It writes!! - Amsterdam, NL

[Fox]

Hi Guys,

I was wondering about the following:
I have a big CNC cutting assignment coming up which was confirmed today. Exciting.
It is big enough to return the costs a Mechmate build - from this assignment alone, if I will do the cutting on the Mechmate instead of outsourcing it.

So (as a long time lurker) I am now very tempted to pull the trigger on a build, since the thing will pay itself instantly ( if I see the majority of the hours I need to build it as an investment :D). The main thing holding me back is I have about an 7-8 weeks deadline on this assignment. I think the cutting would occupy about 2 weeks of machine time ( 10 x 8 hours days - rough calculation after running my files through Mach 3 and adding time for mounting boards changing bits etc. )

Which would leave me a rough 4 weeks to build and 1-2 weeks to debug the machine, and then it has to be in full reliable production for 2 solid weeks, to allow me to be able to deliver 'just' in time. Obviously I am putting a lot of stress on myself having such a deadline, which can be a good ( it HAS to be finished) and a bad thing ( if it is not finished ) at the same time. I could devote my time and that of an employee to the build for about 4 hours everyday + evenings & weekends. So....

1) Could this be done at all, or am I asking for it ?

2) Whats is the standing build record for a full size MechMate and what's a rough average build time so far, if you do not count the very highs (multiple years) and very lows builds ( no idea what the record is )....anyone got any idea or even experience ?

2) Any other builders build their machine specifically with their first assignment already waiting/paying for it?


Ps. To give an idea where I start: I have good computer skills, am a experienced cad user, but have only got simulation experience with Mach3. I got Basic skills in electronics, but a friend who's electrician, I can weld (mig/tig) and have a workshop with most common tools, drill press etc. I outsource a lot of CNC work so I have a good understanding of the process, but never owned/build a machine before.

[darren salyer]

Far be it from me to be a naysayer, but I would hate to build machine #2 up against a timeline that tight, much less my first one.
Can it be done? Sure.
I'm sure its already been done.
Is it a good idea to have the deadline of a paying gig looming on the horizon while attempting a build in such a short time? Not for me.

[Fox]

Hi Darren,

thanks for the quick reply.
I value your words as you have build 101, (great number btw), so your talking from experience.
What were/are the biggest hurdles/issues you encountered during your build, to base your advise on ?

[smreish]

With all your parts in a pile: Steel, laser cut parts, motors, etc....it can be done in a long 2 weeks. BradM and Seattle Rep Theater built 5x10 machines in just about a few weeks. My best record is 3.5 weeks (I've built a few). Most people take 8-16 weeks in the spare time. this is for a direct drive or Gearbox design. If you want belt transmission, add a few days.

Key to a quick build is limit your "out of the box" thinking. Use a dedicated computer for Mach and purchase a simple 2D toolpath creator like "cut2D".
Make your machine as simple as needed:
- simple 3 button stations on z-car and gantry
- no laser finder, no dust collection, no auto-z zero (all can be added later)
- Order the right cable chain - don't try to get fancy
- Use the correct steel shapes
- Use the grinder skate method with Sanding disks 36/50/80 grits in succession for a quick profile.
- Use common sense and good craftsmanship.....

building quickly is easy with a good set up.

My timeline went like this with 2 people:
1 person - only focus on electrics, control and wire for a whole 2 weeks, help lift, drag and clean when needed.

1 person fabricate:
- Cut steel (all of it) 1 day
- Drill all the steel 1 day
- build base table 2.5 days
- cut and profile rails 2 days
- build gantry 1/2 day
- build y-car 1/2 day
- build z - spider and z-slide 1 day
- build cable tray and track holder 1 day
- test fit all 2 days
- paint 2 days
- reassemble 1 day
- square and shim
- wire and terminate 3 days
- test
- cut parts

Elapsed time was right about 180 manhours

This was on machine #3 with a lot of experience (and a few jigs left over) - The first one (#5) took about 3.5 months in my free time

[Fox]

damn you guys are quick :) but that's great, as I need to kinda make the decision this weekend I guess !

Thanks Sean,

Really helpful post, KISS method sound a smart plan. Add the fancy stuff later when not really needed/whenever possible. The only thing I think that might be NEEDED though in my build is a vacuum as each 2 x 3 m sheet ( 60 or so in total ) contains many (80 or so) smaller parts. I think with the number of sheet it might be smart to invest the extra time in that ?? Screwing and and unscrewing all those quickly adds up time wise as well ?
Cut2d sounds like a good plan as well, downloading trial as we speak.

Ps. Already got the dedicated computer, screen etc.

[smreish]

A quick vacuum setup is as easy as a few shop vacs and a "wafflegrid" etched in the top sacrificial spoil board. I made mine in a day on my old machine. Plenty of examples here on the forum or CNC zone or SHOPBOT forum. It's not hard to make a simple vac setup for use....the challenge is your best bet depending on part size is TAB your parts AND use vac.

[smreish]

I built my first table with simple tools:
- sawmax steel cutting circular saw
- 10" abrasive cut off saw for simple cuts
- 3.5/4" grinder (three of them set up differently to keep the tool change down.) One for skate, one for weld grinding (rough) one for sanding and fine grinding (fine) Its amazing how much time an extra 20 dollar grinder will save you!
- drill press with VERY Good lighting.
- cordless drill for tapping with pierce point spiral bit
- used only 3 sizes of holes/taps In the US It was 1/4" 5/16" and 3/8" - 12mm for carriage - limits the needs for many drill index's

[Fox]

yeah I though so, I have to surface the spoil board anyway and milling the waffle grid could be my first programming after that. Seal the edges will take a day or so to dry out I guess, but then the basics should be there. I already have my eye on a industrial vacuum pump at a local surplus that could work. As I am going to be cutting clear PMMA and polycarbonate for this assignment, I would like try to avoid the tabs all together, so I have good edges straight off the table, so I will need a proper vacuum.

I already have dust collection running at my workshop, for starters I think I could just hack a hose from the ceiling and ducktape it to the spindle for getting out the most mess, and afterwards design a proper dustfoot. I guess two weeks of cutting it's nice that at least the basic stuff get out, instead of doing nothing at all.

We think alike; I already use 3 grinders in my workshop,3 cordless and 3 wired drills for exactly that reason, time saving... I have always 1 good 'expensive' one, and the other 2 tired/cheap ones, but good enough for rougher work. Also avoids people watching each other while they work ;-)

I am trying cut2d as we speak, easy enough it seems. I have just generated the first few tool paths with it, and fairly quick as well. Was curious if it would work well with .ai, and .dxf files Rhino3d ( my cad) put's out, but no problems I see so far.

[danilom]

So now the Mechmate edge quality will suit the job, Fox

Hurry up to cut the steel and order electronics your wasting time :D
as Sean said with everything on the pile you need 2 weeks minimum if you know every step, on one machine it took me a week 24/7 work time to put it to work after the metalwork has been completed.

[rischoof]

fox,
welkom op het forum.
Thats a tight deadline.
I lose, or say, spent a lot of time to figure out how I want to my machine to be and which on the forum available design / solution is available and if not, design myself (z axis , gearbox, desc cnc instead of mach 3, diagnostic in he machine) find a previous proven build and copy as much as possible

going to the different shops for getting the materials is costing a lot of time.
if you miss a part you have to go back to the shop, that is costing a lot of time. be aware that some parts are not on stock and have to be delivered.
Order on line, will be delivered at home.

In my opinion it is very important for you have to find a complete as possible part list and try to order as much as possible, as soon as possible, from as less as possible suppliers

some parts you will have to buy from us or azie. (bob, spindle, stepper driver v-bearings)

buy some extra bolds and nuts so you have spare.
use a cable overview and figure out which cables and connectors you have to use or better to be quick, wire first quick and as direct as possible.
I spent a week with 2 people for my wiring and testing (longgggg days)

I learned that the machine doesn't need to look fancy to do the job.
skip painting
skip homing and use end stops (aanslagen) for zero positioning. move your machine by hand to this position
use only emo's on the button box.

on the shopbot forum (yes I look sometimes there for things things they are making) there is a topic were people with a project can look for machine owners who are willing to do the job. there are some mechmates in Holland and Belgium

when I start making products as a "fresh" operator I run into all kind of problems. it's cost me a lot of time to get used of the way of working in cad / cam programs and debug my cutting file.

[bradm]

Sean's estimates are pretty good. I was one of the faster builds, but I already had all the electrics and computer pieces nailed from prior experience in CNC control. My steel work was close enough to Sean's daily estimates.

The red flag I will throw is that 1-2 weeks to gain enough experience to run the machine flat out for 80 hours of cutting time without errors and failed parts is aggressive. It takes a number of repetitions of running the design -> CAM -> cut -> examine -> adjust cycles until you'll know how to approach various types of cuts in various materials.

If this is a simple, outline style job with consistent depths and pieces about the same size, you can probably pull it off, especially if you're repeating the same thing on a series of sheets. If it's complicated, or the material is particularly expensive, or every sheet is different, it might be tougher to do while you're still on the learning curve.

With that said, I hope you do decide to go for it.

[Fox]

@ Danilom, I guess we'll see about that,
But remember I never said it would not do the job, I was just ASKING if it would, and being cautious and prepared before spending hard earned cash on it :). Edge quality is luckily a bit less important on this specific job, but still important, because I work mainly on plastics and aluminum composites ( I see you/your friend have experience with that) where edges are in sight. I did some research the past months on the forum especially on the work people with Mechmates are making, and I think with some improvements ( belt drives, welding some parts etc) gathered from this forum, and reducing the speed where/if necessary I should be able to get the edge quality the job demands, otherwise I would not be in this predicament and asking you questions ;-) Thanks for a heads up it is possible. I find your build inspiring, and have seen you come up with some good work and solutions. A lot of guys do not visit the forum as frequent after their build ( They are probably too busy making money with it), but I really appreciate when I see experienced builders helping others out with their experience.

@ Rik
Hoi Rik, goede adviezen, precies dezelfde overwegingen maken nu ook dat ik een beetje op een tweesprong zit, waar ik dit weekend een knoop in door ga moeten hakken.
I already have a rough idea on what the machine specs will be like, and have studied the plans and such for which components I need, but I am going to decide this weekend if I am going for it. If I do, next week will be 'lost' in my planning for decision making and ordering all parts.

Which machine is actually working in the Netherlands ? I only know of 1 machine, that is still being built but not finished ? I know of two machines in Belgium. I am in contact by PM with the builder in NL but his project is sleeping/delayed for the past few years. The guys in Belgium I have tried to contact them for a visit but I don't think they are open for that ( I respect that). But it would be a massive help to be able to talk to a local Dutch guy, in terms of where ordering what and such and see the machine first hand ( mayeb even run some test). Please tip me if you know of a working machine in NL !

@ Brad, thanks for the confirmation on Sean, that's reaffirming it can be done. It's the unknown problems ( I have build some custom machines before) that eat up time. I take you warning on the 2 weeks learning experience very serious, I guess that's the second hurdle. Once the machine is build in time, it instantly needs to perform. If all goes well it should, but what if... I feel confident about the cad drawing (already there), and setting up for the material ( only one material ). What I am worried about would if you get not so easy to diagnose problems that prevent the machine from running properly/reliable.. like interference for example which I sometimes see eating up weeks on the forum..

I have a big decision to make this weekend.One of those decisions you can not say up front if it will be a smart or stupid one :confused:.. If I do jump into this... you going to be seeing a lot of me over the next few months ;-)

[darren salyer]

I'm sure we'll be watching closely, and when the inevitable head scratcher comes up, you probably won't wait long for an opinion on solving it. Lots of very helpful people here.

[Fox]

Hi Darren, thanks for the supporting words. I feel much more confident it can be done then a few days ago, but it's still quite an undertaking. It's only me and 1 co-worker here, so it's going to be a very intense first 2 months of 2013 if we attempt to pull this off. Going to do some proper calculations over the weekend for the time, with shipping delays leadtimes etc, and see if we can make it happen.

Ps. Is there an interactive version of the builders log available ( meaning if you click the name of the builder it shows their build thread ? ) Right now I am checking it by entering the name in the search box and then look for post of this person. Would be helpful to do a quick filter on the quick build threads, or those with similar specs to what I need...

[bradm]

A couple of thoughts on the topic of interference. These do stray a little from hard proven fact and into my own personal superstitions based on past "unpleasant experiences":

A) Good quality, name brand VFD if you're going for a spindle. Absolutely NOT an HY bundled in with the spindle.

B) Ground all parts of the machine.

C) Don't skimp on the wire quality. Good shielding cabling is expensive, but worth it.

D) How high is your ceiling? If it's high enough to do a simple single hook swag for the dust collection hose I'd run the spindle line that way as well, and just keep it away from everything else in the short term - you can come back and pretty things up in the cable chain later. Bonus: time savings on your dust collection engineering for the short term.

E) Don't bother with spindle or dust collection integration into your controls, keep 'em manual to start. So just the computer, BOB, drivers, and motors as one subsystem, the spindle control as a separate subsystem, and the dust collection as a separate system. Sure, it'll cost you an extra 30 seconds a sheet to turn them on and off, but you'll avoid (or isolate) your gremlins. It'll also make it easier to diagnose when you integrate them one at a time later.

[danilom]

For quick access to the parts look for Damen CNC and Stappenmotor.nl they are in Netherlands and have quality parts.
Stappenmotor works with Deitech and can source same motors and drives along with spindle and other stuff.

[Fox]

Hi danilom, thank you.
I already was aware of these two suppliers, but thanks anyway, wondering how you know them though !? Are you related to any Dutch or ?!

[rischoof]

for the machines in Holland, if I remember well, Hans in badhoevendorp has build one and there has been a build in antwerpen.
For shops in Holland.
for lasercutting and metal, try "the cromvoirste" in oisterwijk. you can order lasercut parts online I think they also can deliver the other steel frame parts, but each local steel construction shop should be able to do that. (I used to live near Tilburg)

I ordered my spindle from deitech (order also a range of er20 router bit mouning collets) 4, 6,8,12 mm is most common.
try to order motors, driver (leadshine) gears, cable chain + mountings, spindle
from them. If you want to order other parts, sent Claudia from Deitech a mail she's able to deliver more than they list. I know they ship also by container to holland, but I think this will be to slow for you.

for frequency controller try marktplaats, there are always some listed. get your frequency controller programmed correctly can be tricky, buy one at a place where they can help you. bring your spindle and some hook up cables and let it run

for water cooling for your spindle, use tapwater and let it finish in the drain.
check if it is not freezing in your workshop during winter.

for electrical stuff, try conrad or rs-catalog,
If I go shopping for parts, i have always a picture downloaded from the part and the specs listed at a piece of paper to make clear what I want.

for v bearings, superior bearing in us, was not able to find something cheaper in asia,

make a bold and nut list, buy as much in one time put everything in boxes so you have clear overview of the parts you have.

and last but not least: GET A GOOD COFFEE MACHINE to survive the coming weeks if you decide to build one

[danilom]

No, I have no relatives there. I am just well informed :)

[alan254]

took me 6months to read and understand toe posts

al

[ger21]

I don't own a mechmate, but have about 15 years programing and running big industrial CNC routers.

With a lead time that short, I'd recommend outsourcing the cutting, and invest the profits into a mechmate, that you can build at a more leisurely pace.

As Brad says, the learning curve will take considerable time. I also have a feeling that the vacuum table will not be as simple as you think. Especially when cutting small parts. Small parts can be difficult even with a very good vacuum system (25HP or more, $10,000+ pumps).
You might be better off onion skinning the entire sheet, and then using a small trim router to cut them free. But it's very difficult to give an accurate answer.

[Fox]

Hi Ger,

Normally I would totally agree with this way of sensible thinking ( I am still racking my brain over all the details calculations and possible issues as we speak). And I still might agree and not do it - because this is betting that everything works my way & not Murphies Law...)

I have also contemplated buying a machine and invest more money and less time. But.... there's no comparable machine out there that I can afford (table size 2x3 mtr) and going to the bank is out, so that option is off the table. Outsourcing or DYI it is.

The catch is of course, this outsourcing (no matter where I go in the Netherlands) cost me a lot of money; I calculated it roughly ( with material prices-machine time-bits-handling, the whole deal) and it would at least double my profit if we do the cutting ourselves (If I don't count my hours to build the Mechmate, that is). That EXTRA profit would pay for the good MechMate. If I would out source and build the Mechmate later, it would eat up almost all my profit, plus the risk that the Mechmate build will be postponed/spread out over a longer period, because of all kinds of other deadlines that get in the way-it's never quiet time in my shop.

So the goal being: investing a lot of hours over two months, and end up with making a profit on it to pay the bills, and get a nice spec mechmate as a bonus for all the hard work and extra hours. The Mechmate will serve me many years to come, and make us a lot less dependent on that cutting part when it coems to tight deadlines, prototyping etc.

Catch being:
A) I will have to work my ass off for the next two month (story of my life)
B) I am doing 'risky business' of course.... once I commit that's kind of a point of no return, and I have to live with whatever difficulties comes on our path. After I buy all the mechmate parts it would hurt financially to not continue but I could still make the decision to finish it later, but after I buy the material (50+ sheets) it becomes imperative I finish the Mechmate to do this job.


VACUUM
The small parts are not THAT small, I guess ??. Around 20-25cm diameter ?! or is that still too small to hold with an affordable vacuum ( I Can't spend 10 G's on that !) I am cutting entire sheets that cover the whole table with these small parts + bigger parts.. so I would only loose vacuum in the cuts ? It might even be possible to trow over a piece of thin pvc sheet over half the table when that side is done ?

[parrulho]

I believe, if you can get all parts on time, you can do it. My table is not finished and I can not speak by myself but is was done in the past:

http://www.mechmate.com/forums/showthread.php?t=1568

http://www.mechmate.com/forums/showthread.php?t=2451

http://www.mechmate.com/forums/showthread.php?t=2180


Go for it. Good luck!!!

[Fox]

Hi Paulo,

thanks for the encouraging words. I am crazy busy with running my shop and in the mean time figuring out all kinds of details to be able to pull the trigger.
I decided this weekend...CNC is coming to my shop FIRST HALF OF THIS year... but I still did not decide a 100% on doing it before this pending deadline (2 months). But figured I should at least waste no time and commence planning/gain order info right now and see if that goes wel.. if not I still have some time to back out, if it goes well it makes it easier to shell out the cash ( the first difficult thing..... building will be the next)

What I did the past 2 days/nights:

1) Researching all over the forum + cnczone for my best configuration choices, like spindles drivers guides etc.
2) Asked quote for steel + laser cutting with a friend of mine ( he might even let me use his metal shop to build the frame ). Frame will have to be 2 x 3 meters effective space, to be able to cut our plastic board sizes in one go (2050x3050mm most often)
3) Visiting hans (http://www.mechmate.com/forums/showthread.php?t=2371) tomorrow evening, The only other builder in the Netherlands I know of... He might not be the best to talk to about my crazy plan ( he started in 2009 and is not finished yet ;-) But he can sure help me with a lot of questions I imagine on parts and such.
4) Visiting Damen CNC (http://www.damencnc.com) tomorrow afternoon, after speaking with them on the phone today on parts and such.
5) Tested and decided I am going to use USBCNC (http://www.usbcnc.com) as my controller software. Was in doubt with MACH3+smoothstepper as they have a huge userbase obviously, so more 'support' to be found online in the future. But I voted for USBCNC (http://www.edingcnc.com/) because of their nice interface (no issues related to serial port) and very good reviews I have read on the Dutch CNCzone and heard from DamenCNC which were all in favour of USBCNC (So I dearly hope I am one of the early adopters of a great program here, and not hit myself on the head later). Good thing is also it has a built in limited CAM, so for my first assignments ( tested with the dxf i need to cut) I might not even need to buy a CAM like cut2d, ( but might still do to be safe).
6) Taken an old XP system with TFT and started to upgrade and convert it to a dedicated CNC control system.


I will update this list BELOW here as I gain new info, and ANY help/advises/sharing of exprerience is much appreciated as I have limited time on my hands to make good decisions:

SPECS:
bolting table (future location uncertain )
2x3m cutting surface/table size (cost prices on steel+lasercutting this from my friend with metalshop, I can invest savings this in upgrading other parts)
Affordable but noisy vacuum by means of roots blower style, or maybe is I can buy a big becker pump at a very good price at an surplus ( currently negotiating).
Watercooled Chinese Spindle DTS-23 (maybe DTS-27): 2,2 KW with Delta or other good VFD (not chinese anyway)
Maybe linear guides (supported round ones or Hiwin -> need: to mill the machine then) on all axis, but definitively on Z. ( maybe...because of accuracy, and avoid time consuming grinding, costs are making me think twice )
rack and pinion driven
Z will be custom with ballscrew (and hwin linears) ( metal shop friend has CNC mill).
Leadshine AM882 Drives
4:1 belt driven gearbox
Nema 34 PK296A2A-SG3.6 or 7.2 motors ?? I must say I am not sure about this; is this a good choice with the 4:1 reduction or should I opt for PK296-F4.5A ?

(LATER ADDITIONS/UPGRADES)
USB pendant
Bill Pentz Cyclone dust extraction

[darren salyer]

Ummm yeah, linear guides and ball screw z slide...... reinventing the wheel definitely won't help you meet the deadline if you go that route.

[Fox]

Thanks for the constructive criticism Darren. You are quick:)
Might I counter .. that I am not reinventing the wheel, these are all Mechmate improvements I have noticed here on this forum ( see below) that could suit my needs for this machine. Just as I need to decide what size of table I need, and adapt the plans to my needs, I see these points similair. But nothing has been bought/decided yet, that's why I am asking you all for your opinion, so you're advising against linear, noted.

Linear Slides on table: http://www.mechmate.com/forums/showthread.php?t=1148&highlight=linear+slides

Linear slides/ballscrew on Z: http://www.mechmate.com/forums/showthread.php?t=2135&highlight=designer&page=3

Bolting table: http://www.mechmate.com/forums/showthread.php?t=2359 & http://www.mechmate.com/forums/showthread.php?t=1329&highlight=roll+pin

Vacuum: http://www.mechmate.com/forums/showthread.php?t=849&page=3 (http://www.mechmate.com/forums/showthread.php?t=849&page=3)

To all....: I am merely making a list of options to contemplate if i will
A) include or not
B) Now
c) Later

so.... please chip in !

[darren salyer]

I"m not here to debate whether something is an improvement or not.

Best of luck to you sir.

[Fox]

Hi Darren,

Ok, this subject is sensitive, not sure why :confused: but if for a moment you would forget that I mentioned these improvements/modifications/whatever you want to call it..

Would this reduction/motor/driver combo be any good ? I read good comments about both the Geckos and the Leadshines, and I think this digital leadshine should be able to drive this motor form their datasheet. But I am not sure if this kind of motor matches a 4:1 reduction ( even though I researched a lot of threads around here ? )

Darren salyer
I'm sure we'll be watching closely, and when the inevitable head scratcher comes up, you probably won't wait long for an opinion on solving it. Lots of very helpful people here.

[domino11]

The SG3.6 or SG7.2 already have a 3.6 or 7.2 gear reduction built in to the motors. If you want the same motor to build a belt drive around, the F4.5 is the one to go for in the oriental motor line.

[Fox]

Thanks domino ! Makes sense if you know what the numbers are for...

[Surfcnc]

Hi Fox

I have been reading your thread and wish you the best of luck.

With respect to the modifications you would like to make, from my perspective no problems at all.
What I would be very disappointed about is if you mine the forum for information and knowledge then do not document your mods for us all to see.
It is a tough call especially when you are under a time constraint but please make it a priority to give back and share what you learn.

Just a quickie on your USB based controller. Don't do it !!
Recently I have just read the Shopbot forum from top to bottom about USB communications issues and if you have any, significant pain is heading your way.
The board you are using may be good but many other variables such as the PC controller itself, cable quality, ground loops etc are also factors that you will not easily make allowance for.
I will not hand you the answer but my firmest recommendation is to use Ethernet communication for your controller if you wish to move away from the safety of the parallel port !!

regards
Ross

[Fox]

After a very long day (18 hours) literally just arrived back from my visit to Hans and Damen CNC. Kevin, owner of DamenCNC was a very nice and knowledgeable guy, learned a lot.

After that visit I went to Hans...he was VERY kind, invited me to his house and workshop and showed me all the parts he gathered, and what he build so far.
He got a big treasure chest at his house with a complete Mechmate in parts in it ... has even got an supercool ATC from Blurry Customs; NICE !

He helped me tremendously with all kinds of tips and info and warnings.. THANK SO MUCH HANS !

He also warned me 4-5 weeks in his eyes is too little time to do it ( like many others are sceptical ). But he provided me with a list of suppliers to save time.
First I am getting some shut eye, tomorrow I got A LOT to think about...


PS.
Hi Ross, if I am doing the 1 month build I won't have much time left to keep you all up to date in detail, besides from asking a lot of questions when in doubt/troubles, (I have a young family and a business to run at the same time, (so imagine my predicament), but will do a detailed write-up afterwards, I've learned a lot from reading others !

[bradm]

You can count me on the side of the folks that say you need to build a stock, unmodified MM (belt drives if you wish) if you're on a tight schedule. If you want to explore the other versions, it will cost you more, not less time. Of course, if you decide to take the slower path, then you can explore as many changes as you have time and money to handle.

[Surfcnc]

Thanks for that Fox, it is quite clear that you have set yourself a very steep hill to climb.
A few strategic pics along with way will also serve to remind yourself what you did and how you did it.

If you want to build the best machine you can, it is a mistake to use an already superseded and troublesome communications bus on the machine controller.
Ethernet is a high bandwidth bus and has superior noise rejection as it uses twisted cores.
The newer Cat 6 cables also have a thicker wire and can be purchased with an additional mylar shield to further eliminate interference.

If it seems like I'm harping on about this - that is because I am :).

regards
Ross

[Fox]

Hi surf,

I am thinking of purchasing CPU5A4E Economy 4 axes with Ethernet. (http://www.edingcnc.com/index.php?pagina=producten)

I will be using Vectric Cut2d for the first job(s) later upgrade it or use something else when I know what I like/don't like about it. So far I like it, it's dead easy. Already have run simulations on my all my cutting files with milling pocketing and drilling included so I do not have to worry anymore if I can control the machine on the software side, that's now sorted.

I am now super busy on my shopping list for steel and electronics.

(@HANS: I have found a postrpocessor for all VECTRIC software on USBCN on CNCzone (http://www.cnczone.com/forums/post_processors/62323-vectric-pps_usbcnc.html))

[Surfcnc]

Hi Fox

I have had a quick read up on the CPU5A4E.
You should be able to install a wireless adapter (USB even) into the controller PC and have it networked to your LAN quite safely.
The idea of having a separate subnet for the CPU5A4E and the machine controller is similar to the other ethernet controllers.
The configurator set up is also similar and from experience I would suggest switching off the firewall until the communications is established.

The pulsing is adequate at 125K and faster than the Mach3 100K using the parallel port.
The smoothstepper runs at up to 4000K but in practice this is not often used.
I am wondering about expansion though as the smoothstepper has options for a further two emulated parallel ports.

The price for the CPU5A4E board and software combination is certainly competitive against say a PMDX 126, SS and Mach3 combo by a significant margin.
In the end if it fits your needs and it works well, it does not matter what is locked away in the control box and running on the PC.
You have already stated that you understand that the user base for CNC4PC is smaller than the Mach3 install base.
As long as you have established a pathway to some quality help with the hardware/software combination that is all that matters.
The only downside is that this wonderful forum will stay mute if you run into problems as Mach3 and Linux CNC pretty much rule here.

Regards
Ross

[Fox]

Hi Surf,

that's the only reasons why it did not buy it yet, I'm gonna be the first with a Mechmate and USBCNC... In general I am not the one to shy away of being the first, if it seems to make sense why not ? Somebody got to do it :) But of course..in this case I have a tight deadline. Luckily it's Dutch and on the Dutch CNC zone there is a userbase and the guy who invented it is also fast in replying to problems. And in then end a CNC big or small is still a CNC. Either way both versions of setup work in test setup, and these are the least of my worries right now as there's not much delay in the ordering process for these ( inquired already). Now grinding my teeth at more urgent stuff; all electronics parts and the cutting list...

[Fox]

I found a nice affordable vacuumbed option I think I will go for (Hans tipped me):

http://www.shopbotblog.com/index.php/2010/07/the-k-i-s-s-vacuum-manifold/

with

http://www.talkshopbot.com/forum/showthread.php?t=11813

The plumbing seems straight forward, and the box can be cut to learn my first steps on the machine. Any opinions on these vacuumcleaner engines ? Any Mechmates with this system ( I've seen Ken uses them I believe) ?

[Fox]

Revised planning on the Motor/driving side of things:

starting with:

Gecko 203V (http://www.geckodrive.com/geckodrive-step-motor-drives/g203v.html)

wired by means of Bipolar parallel wiring to a:


oriental motors PK299-F4.5A (http://catalog.orientalmotor.com/item/all-categories/pk-series-stepping-motors/pk299-f4-5a)

In combination with a 4:1 belt reduction

Using XL belt, pulley 72 & 18 tooth, 20 tooth pinion

to

standard spec rack.

[Fox]

Update:

Mechanical: 20 tooth is for inch based machines so I should use a 24 tooth as a minimum for my metric based machine, correct !?


Onto the electronics; can anybody confirm my calculations below ?!

PK299-F4.5A (http://catalog.orientalmotor.com/item/stepping-motors--1068/pk-series-stepping-motors/pk299-f4-5a?&plpver=11&origin=keyword&by=prod&filter=0)

Inductance: 2.5
Voltage: 1.9 [Bipolar (Parallel)]
Current per Phase: 6.3 [Bipolar (Parallel)]

Gecko 203V (http://www.geckodrive.com/geckodrive-step-motor-drives/g203v.html)

Min volts: 18
Max volts: 80


My calculations for the power supply:

So, where do we go in the range 24 to 80V?
[B]Volts:
32 x √2.5 (Motor Inductance) = 50,59644 Volts
VA/Watts
6,3 (Motor Amps) x 4 = 25,2 Amps and 67 % of that =16,884 Amps
16,884 (Total Amps) x 50,59644 (Volts) = 854,27029 VA/WATT

So I need an 50 volts 850 watt Power Supply.
Is this assumption correct ?

According to the forum I can go up in voltage about 5-10% and the VA could be a bit bigger or smaller without much problems, to find the proper power supply, but not go as high as a 1000 watt as it might cause problems with the inrush.

At Antek, they have no 50 + which matches 800+watts, until I go over 56 volts but that is more then the 10 % added, so fall back to 48 volts ?

Should I order:

just the 48 volts: http://www.antekinc.com/details.php?p=411

Or 48 and 24: http://www.antekinc.com/details.php?p=413

I am asking cause I thought I have read somewhere people uses the 24 volts for optional stuff like lights ? or am I mixing things here ?

[danilom]

If module 1 rack then what I found is only 20, 26 and 30 tooth pinions , the lower number of teeth slows you down but increases resolution,
What I seen almost everyone use 30 tooth module 1 (3.14 or PI, tooth distance) with belt reduction 3:1

you can use 12 or 24VDC depending on the relays, contactors and other stuff you plan to use on the machine

[Fox]

Thanks Danilom !

Yes Module 1. Then I order 20 and 26th or 20 and 30 tooth- so I have two options in the end to experiment if necessary.

I am willing to sacrifice speed for increased resolution but only if necessary
I am doing the 4:1 ( I know about the theoretical wrap issue - ) though.
I thought I read somewhere here that 20 tooth works only with inch based racks, and that 20 tooth on module 1 would be too small to use grub screws ?
( But I read so much the past week, and slept so little that I sometime think I hallucinate)

Ok, I will decide the second output to be 12 or 24 based on the rest of my system, so besides that my calculations/assumptions are correct ? (I need to start ordering..)

[danilom]

48V should be ok for motors supply

if a pinion has a too small collar for grub screw, same as 15 tooth pulley which I had to bore to suit 14mm motor shaft,
what I do is drill two-three holes (depends on space availabe) and tap a thread to connect it to a larger piece (if possible I welded it couple of times). I think it could be the term "hub" to call that piece which is added to the pinion or pulley.

[MetalHead]

You need to make sure your voltage matches your motors faily close so that 48 should be fine. You may want to explore other motors as a cost savings if they are available. The price point in the States for that motor is almost 3x a comperable equivalent. Also I have seen in a lot of the newer builds not having secondary output DC voltages in the box. This keeps the costs down on the PS. Unless you have specific designs in your cabinet for the lower voltages, they are really not needed with the newer BOB's. You just need to make sure your BOB can run off of 110 volts like the PMDX-126 can. If I am remembering correctly the older setups used the secondary DC voltages to power the older PMDX-122 BOB's.

[Fox]

Thanks Mike,

I'll keep that in mind. I was looking into Motion King and Leadshine for savings (saw this savings story pop up somewhere else). But... I am very limited in time, so I might just cave in and stay with the current choice, my head is hurting and time is running out quickly. Just finished calculating all my cutting list and lasercutting for the steel, mailed it out to 3 vendors, hope to have answer today/tomorrow, going to bed now (3PM )...


Danilom; found this on the pinions again: is this old or still active in terms of suggestion for my pinions in metric with 4:1 drive ?

http://www.mechmate.com/forums/showthread.php?p=12605&postcount=39

[Gerald D]

Update: Mechanical: 20 tooth is for inch based machines so I should use a 24 tooth as a minimum for my metric based machine, correct !?

24 tooth is the minimum for "module 1" gears . Correct.

(if you use module 1.5, which has bigger teeth, you might be able to go to 20 tooth, but I have not studied it. Module 1.5 is not so common as module 1. The best smoothness will be with module 1)

[Fox]

And...awake again (since 1,5 hours).
Luckily I am "combat trained" by my 2 years old :)...but that was a short night again.

Thnx Gerald !
I will order the 24 tooth and the 36tooth option then. I need smooth, so Module 1 will be it.

I've spoken with my friend ( in the steel trade) about your plans, and he was impressed with the thorough and clean job you did on the plans. So am I, now I am diving into all the little details. Even though I was already up to speed with the general plan, I must admit it's quite a time-consuming (with my deadline) read to find/figure out all the details needed on a build ( I know why you choose this strategy ). But the good thing is almost always when I start looking I discover that somewhere on this forum you typed the answer.

@ Mike can I also buy a .dxf for the proxy switches time saver of you ? Would be more convenient/economic for me if I can order them with the lasercutting parts.

[MetalHead]

I can ship a set to you in the post. I can send you an invoice

or

Look here

http://www.mechmate.com/forums/showthread.php?p=32914&postcount=23

[Fox]

Ah, thanks added them immediately to the lasercut files !
Normally I would buy them from you but with this deadline it's quicker to cut them with spiders etc !

Can anybody confirm what my total width would be of the table with effective cutting surface of 3100x2100 mm ? Maybe measure your own table ? I know I would need to add 500 mm to the Y, but I think the motors are sticking out over that ?

I need to know how much room I will have left in my preferred designated area which is a bit tight (8000 x 3450 mm wall to wall ) to be able to walk around ( on one side ?) the machine when it's operating.

[rischoof]

fox,
For my machine add 36cm to your y axis (measured over the outerside from the C profiles. my motor+ gearbox is 14.5 cm outside the c profile
total to your y is 36 + (2x14.5) =65 cm

[Fox]

Thank you Ric,

don't know where I would be without you guys !

[bradm]

The absolute total width is from the back of one of the X axis stepper motors to the back of the other, so it varies a small amount depending on the size of the stepper. Normally, the front of the stepper is 20 mm or so beyond the length of the gantry steel. Belt drives push things out another 30 mm. So, a working estimate is that you'll have 2100 + 500 + 2 * ( 20 + 30 + 100 ) = 2900 mm in your 3450 mm space.

Presumably you'll do full sheet loads into the end of the table, and perhaps place one long side against a wall. That would leave you with a small area in the middle at the back that you would not be able to reach without climbing on the table. My space is constrained on two sides, and I have to load over the rail and climb to get anything at the back rail; I tend to layout my sheets so that I don't have small pieces at the back side.

You won't need 850 watts worth of power - all four motors will never draw their maximum at the same time. There are plenty of successful reports at 300 watts; there is no risk in going over, it just costs money you don't need to spend. Since most folks aren't comfortable when they hear this, go ahead and get a larger supply, but you might consider something in the 450 watt range instead.

You may also want to consider using "half coil" wiring instead of bipolar parallel. You can read up on the pros and cons elsewhere on the forum. This needn't affect your ordering,

[Fox]

Hi Brad,

valuable insights. Stupid but I did not think about the engines not all being loaded at the full at the same time, but hey... it's my first CNC.
I will downscale the power supply, saving money is good.

Space is a bit tight indeed, my workshop size is comfortable, but divided into many different rooms ( not just drywall- they use real stone here in the Netherlands :) ) So this room the best/quickest to clean out place I have available right now ( now storage/junk room) My workshop itself is bigger, but don't want to take up all the floorspace there with this big machine obstructing other work/machines.

Your tips on parts planning will be taken into consideration. I was also contemplating if after cutting your could just pull the cut sheet towards you to slide the cut pieces with it, being able to pick them out one by one ( drop them on a cloth on the floor ), a bit like a conveyer belt. ( with 20 mm mdf this become heavy work - but with a sheet of 4 mm plastic it might work, if the spoilboard has not yet seen too many cuts ? we'll see.

[Fox]

Hi Gerald & others;

Asking for a second opinion:

For a MechMate with bolt together table frame with flanges like seen on the forum for instance by BuiBui.

Option A) the left side
Option B) the right side

http://s13.postimage.org/jrf5wz8pf/mechmate.jpg (http://postimage.org/image/jrf5wz8pf/)
I am using an Wide flange 200 x 100 mm I profile for the main beam.
I am using 50x50x5 mm beams for the frame.

Option A allows me for regaining some more of the lost space under the table for storage of boxes etc.
Option B is true to the original design, but using square box profile instead of tubes.
I would choose Option A, but only if I am not comprising the performance/strength of the frame in a significant way. Then I would fall back to option B) and triangular boxes :-)

[Surfcnc]

Option B is more rigid.
Option A is still more than adequate.

All those options limit direct access to the underside of the table as you have to lift objects over the lower bar.
Please note that storage of items / sheets would normally be slid in from the short side or what some might call the ends of the table.
The table is also typically loaded from the ends, so having your storage strategy consistent with that is quite important.

A better option is to diagonally brace the long side of the table as per the plans.
For the table ends if you wish to use the underside as sheet storage see the pic.
13687

There is a strong focus on the table design in the forum with lots of variations on the basic theme.
Typically these variations are mostly cosmetic / vanity modifications as is quite typical of the average DIY fan.
The real issue is reading the plans in fine detail and understanding how the table goes together.
The business end is the upper table, the electronics and the software control and tuning.
Set your goals for your storage requirements and quickly move on to the main game if you are on a short time frame for the build.

If you intend to bolt the table as myself and Bui Bui have done, expect this to take more time to build.
To make this process as quick as possible you should draw your own CAD lower table drawings to highlight any clashes with the bolt connections.
The CAD drawings will also give you the ability to make mild steel drilling templates out of flat bar for your hole spacing's.

Not all measurements are directly provided in the plans and they will need to be inferred from your chosen table size.
Until you can calculate these measurements on your own, you can assume you do not know the plans well enough to proceed and get it right first time with no redo's.

Regards
Ross

[Fox]

Hi Ross,

Thanks for all the suggestions.

I already did the drawing I show myself in CAD, luckily I am fairly quick with CAD.
Even though the deadline I wanted so spend one day and night on checking the thing in CAD, for issues. Found some and corrected. Also understand the dimensioning much better now. Was just wondering about the compromise in strength between these two designs for the side, I am not an engineer, and have not run a CNC before. I do understand we want the main beam as stiff as possible, and a diagonal is generally stronger, as they triangulate the hell out of cranes and such I was just wondering if this solution would be "good enough". I figured I loose some strength, and maybe gain a bit by using the I beam for the main beam so I would be ok, but I just am not sure about that. if there would be thousand KG on the middle of the BIG beam, i would for sure take the diagonal route, but with the router most forces are sideways, wanting to rack the thing, which I hope to catch with the ladder construction ?

I do have left a little space to slide some stock 2050x3050plates in from the front, but I designed it so it is only small, as I think adding a lot of boards will quickly add many KG in weight, which will start pulling on the structure and that can't be good. Other stock will be slid, under the machines lower beam, or elsewhere.
I also asked BuiBui about this design and he has replied me he has not found any movement issues other then he would replace the top beams with the C profile like in the plans ( he used box ). So I guess I should be fine. If it really tuns out an issue I can always add an diagonal later in the left and right square of the ladders.

Ps. I have to send out the cutting list tonight, so I just need a second opinion before I make a stupid mistake in search of some storage space :)

"Standard"
http://s2.postimage.org/reny7vxyd/mechmate_standard_big.jpg (http://postimage.org/image/reny7vxyd/)

"Adapted design"
http://s7.postimage.org/cw6q8zjbr/mechmate_mydesign_big.jpg (http://postimage.org/image/cw6q8zjbr/)

[zumergido]

you can start building and order the laser parts while you asemble the base table.
remember you have to asemble the table and paint it. once you have the base table you can cut the rails on it at least one day cutting.
if you are alone there are some things you will need help like moving the gantry..

[KenC]

Hi Fox,
1 month is "enuf" (not plenty) time for this build. The main trick is to have all the necessary parts & material in hand & a pair of helping hands for heavy lifting & assemble the table.

Of all the works, the rail grinding is the sucker, not difficult, just time consuming. I would strongly advice you start grinding the rails once you have the skate in hand. AND DO IT OUT DOORS!

DO NOT skip the Kitchen table project before any electrical installations, you had been warned!.

A bolt on table will take at least double the time to build then a weld-on, because you will need to prepare the brackets (unless you have a pair of helping hands) & most importantly it is harder to align everything together (even if you have a pair of helping hands). I know because I'd built both.

Prime your steel before you do any cutting will save plenty time at the end of the build.

You only need 300VA transformers for the PSU, just take our word for it if you have no time to read through the forum. & yes, 48Vac is a good choice. Don't forget that you need at least 20,000uF capacitor with at least48V rating, 65V is my choice.

If you need DC for your bob or other light load electronics, just grab a wal-mart wall plug or any cellphone charger for the job. will save you plenty of time & money..

For direct drive, 24 teeth is the smallest pinion physically possible to fit on to your stepper motor shaft.

Take down a wall if you can, collecting the cut parts is PITA when you can not access from the side especially when they are small & plentiful.

Take time to level the rails & do better than your best to ensure every part of the table is square & through.

[Fox]

STEEL IS ORDERED

@ Fernando
painting afterwards, not now. Table building will be finished next week with help of my friend in the steel company.
@ Ken.
Rails are indeed an issue. Asked for quotes on V rails to cut time short. If in US I would have ordered from Superior bearings. Quotes in NL for V rails + wheels are crazy, unhardened rails for my machine about 2000 euros (!)... I can buy Hiwin linear guides for that ! If time schedule permits it will be the DYI grinding, will perform test with skate in hand next week to calculate time, already ordered enough L profiles just in case.

I am on a 4:1 gearbox with belts, but on geralds advise will use 24 tooth for gear.

I will square everything with micrometer. Table is built in professional workshop with all tools/weldingtables/clamps available, I hope we can get it into 1 mm accuracy on frame. I hope I beams for main will help reduce warp on the mainbeams.

Bolt on is necessary for transport movement in and out of my shop. I know welding is quicker and stronger, but it is like it is... So bolted it will be.

First assignment has to be in small room, later we will see.
I can walk around it see image.

http://s7.postimage.org/ouof4io4n/mechmate_in_shop.jpg (http://postimage.org/image/ouof4io4n/)

[Fox]

I know a H500 x W400 x D150 is minimum requirement for a control box. (http://www.mechmate.com/forums/showthread.php?p=2898&postcount=10)
I want to mount my box in between my side frame, where I have a space of W825xH400 mm.
Looking for boxes I can place a W600 x H400x D210 box in there so doors open standard. So basically a bit bigger then the minimum "portait" version, but then laying in "landscape" modus.

I wonder: could I manage fitting my electronics horizontal in this box ? Or should I stick with standard postion ( remember my box is laying down, so standard horizontal rows will become vertical in my box ! )

red box is electronics box, green one will be VFD box.

http://s1.postimage.org/fiun1w7hn/mechmate_control.jpg (http://postimage.org/image/fiun1w7hn/)

[bradm]

You can also mount the electronics on the wall above the machine. If you do re-orient a box, think carefully about the hinges.

[smreish]

I want to make a quick comment on the prefabricated V-rails.

They do not save you time.

To properly mount the V-rails you have to cut the support material - aluminum architectural angle - to height. Then drill A LOT of holes - precise holes.

After assisting Nils during his build and watching the effort, it was clear that the amount of labor hours was equal, if not longer with the prefabricated V-rail.

Just my view from doing both processes. Don't interpret this as a negative for the prefabricated rails - I love them.

But from a purely time sensitive view, it's not a time saver. With the proper set up on the skate with an agressive 36 grit disk to do the major profiling, I bet I could do all of the rail fabrication in 2 days flat by myself.

[rischoof]

Fox,
orientation for your electrical components should not be an issue. if i see the picture of your available space I should also consider wall mounting. Do you see yourself laying on the floor clamped between the mechmate and the wall doing some trouble shouting? if mounting under the machine, check if you can open the door. place the machine that you can just walk between, and chose which side you mount the cable gutter ( I decided my one wall side)
Working outside at this moment is not an option If I hear how the weather is in holland. I received snow pictures today

[Fox]

@ Brad; I can't really do that: in my room it will: obstruct my walking, or has to be mounted too far away, or above the table. Hence this location ( good tip on the hinges, I already looked into the spec for some data on this, could not find it. Will now order big box first so I change size on vfd box if needed because of the hinges, leadtime is short luckily.

@ SM;
noted. I might use linear rails, but go for grinding first.

by pure luck I got offered lightly used second hand linear SKF rails the other day while talking to someone. And 2x3,5 meters of 25 mm linear rail for a few hundred euro total ! I thought SCORE, could not believe my luck.
But then found out:
the two rails are no longer produced by SKF, and they have only two carts with them. No newer carts will fit acccording to SKF, their advise was... buy new ones ( @4000 euro). Then I thought... hey... I buy one new rail of the same dimensions, it will still be a good deal. And then I calculated the X amount to buy the new rail and it is 3850 mm.... so that stings a bit ( I thought it would be around 3,50 meters or a bit less with my X of 3050, dunno why I had that number).

Still contemplating on buying them just for the Y, but a bit hesitant, because of matching two different brand rails then ( friction?!)+ if I have to machine v rails anyway for the X the Y is relatively easy work.

@ Ric
Great ! Strip that off the list. Ordering all electronics etc. this weekend. Wall mounting is not really viable...see above.
I was contemplating on mounting the box after all initial troubleshooting is done, just have it losse in the room on some trestles. I hope I do not have to lay under the table under daily basis once its up and running ??. And then maybe still mount with quick releases, and have some "loose" cabling so I could again put it on a stool, for more convenience if/when needed. I see a lot of commercial CNC have the boozes on the table, so I guess it'll work somewhat.

Dutch: Sneeuwt en vriest idd lekker, dat zal wel tegenvallen straks bij terugkeer:). De elfstedentocht waanzin "het-gaat-echt-wel/niet-gebeuren-dit-jaar" is al weer losgebarsten ;)

[Fox]

Can any body point me towards the best compatible stepper from www.motionking.com (http://www.motionking.com/Products/index.htm)
as an alternative for my current choice oriental PK299-F4.5A (http://catalog.orientalmotor.com/item/all-categories/pk-series-stepping-motors/pk299-f4-5a) ?


I am a bit lost, so may specs so many options, don't want to make the wrong decision.

http://www.motionking.com/Products/Hybrid_Stepper_Motors/2-phase/34H2A_Stepper_Motor_86mm_1.8degree.htm

Would this be a good option: 34H2B9850 ? (I see they only make Bipolar on request)

[danilom]

Those are like I use, my friend Milan and Pajka from Macedonia, they are Deitech 86HS9802 same spec. 5Amps and around 45V is ok for them, others used 9801 on Mechmate its a lower Amps but higher voltage motor but with same torque. Depends on your drives.
If you use DM856 from Leadshine or Gecko then you can use the 9801 one.

[Fox]

I don't know if I understand the last sentence correctly.
I see your friend uses DM856 from Leadshine to drive his Deitech 86HS9802 motors.
This turned out to be a wrong choice andhas been replaced ?
Or do you mean if you use this driver or the Gecko 203V you can choose between the 9801 or 9802 bases on personal preference ?

I do notice it saves a considerable amopunt of money to go with the leadshine vs gecko and motionking vs Oriental, so I might take that route and spend it on some otehr parts like linear drive.
But then if I understand correctly you did not wire them Bipolar parallel ? Or did you order that custom, in short: how did you hook them up ?

[domino11]

I believe he was referring to the fact that some drives are capable
of using higher rail voltages than others. Some motors cannot be used with some drives,
as the require a rail voltage higher than some driver can support. The gecko 540 is limited
to 50V rails for the motors, so you have to pick motors that work with up to 50V and no higher for example.

[danilom]

Thats right Heath, I meant that with 80V capable drive you have more motor choices.

DM856 is THE most silent and smoother drive I have seen, we accomplished feedrates of 42000 mm/min

Wire bipolar parallel and have no worries. It works best for me.

[Fox]

Cool.. I save almost a 500 euro on that combo; no-brainer, ordering from Deitech !

[Fox]

Ok, working things out with Deitech at the moment, want within teh next two days... for engines, controllers, e-chain etc.

QUESTION: Should I also order some milling bits from them while I am at it ? Or should better stay away from 'Chinese' cutters ? My experiences with drill bits on this topic.

If so; anybody got some tips on what an absolute beginner should order for starters ( some basic cutting mdf, doing the spoilboard leveling, some bits for testing alu, and cutting polycarbonate acrylates ? ) THNX !

[danilom]

For basic cutting with carbide spiral up-cut bits look for carbidechiu on ebay, I use bits from him all the time more than a year. Even some dia 10mm hrc55 labeled I use for hardwood cutting, guitars and stuff, it cuts like hot knife trough butter and stays sharp for ages.

For spoil board there are large diameter bits for leveling.

For aluminium look for single flute bits but not all work nice, it depends on the geometry of the tool.
Same thing for polycarbonate or acrylic plexiglass cutting, single flute 4-6mm carbide.

Best thing is to find one from a well known manufacturer and try it, if it works ok but the price is a bit high look for a Chinese equivalent, for important jobs use real quality and Chinese for every day cutting.

In the end everything depends from proper feeds and rpm.

[Fox]

Again asking for urgent help:

86HS9801 engine
Leadshine DM856

engine spec: mH 4.1, 4A
I want to wire it Bipolar parallel.
I do not see different specs on the Deitech site, but it does state 8 wires.
Do I need to multiply the amp value ?

If not

32 x √4.1 = 64,79 Volt
5x4=20x0.67=13,4x64,79=868.18 VA/Watt
( taking one extra engine into account for future 4th axis here).

Watt can be around 300-500, and still be fine everybody says ( engines not asking full power all at once ).

Then I wanted to built my own powersupply, cause Antek will be to much delay, and no local alternative:
DYI powersupply (http://www.mechmate.com/forums/showthread.php?t=350)

toroid, check: use a 64 volts, 400 va

rectifier:Needed to convert the 50Volt AC of the example to DC (Direct Current). The DC voltage is 1.414 times higher, giving just over 70V. (The Gecko's limit is 80V).
Ehm...64 volts x 1,41= 90,24 volts ?!? Will that mean I kill my drivers (also 80 volt max) the instant I put power on it ? What am I missing here ? Should you have taken this multiplying factor into account when talking the DIY route; choosing the toroid lower then the calculation suggests ? - not mentioned anywhere- (http://www.mechmate.com/forums/showthread.php?t=56) I am lost in this electrical mambo-jumbo.. and time is running out quickly. Please help !!

I continue the math for the capacitor just in case...
Capacitor=mF=(8000*I)/V= (80000*20)/64=25000mH= 24000uF (only one available), 100V ( this also needs to be able to take the 1.41 higher voltage, I've read ) or I could/should I just hook up 3 x 10000 uF x 100 volts ?

[danilom]

no you divide by 1.41 , that is the voltage output of the transformer, after the capacitors do their work it will be 64V

you can parallel 3x10.000uF

bipolar parallel current for 9801 is 4Amps , 8 wires gives you 4 separate winding which you can parallel to get bipolar PARALLEL connection.

This and other information given here are very valuable and should not be taken for granted, what you need if you get into building this machine is to understand all this and its done with reading build logs and discussions about these information to get a better insight into it.

Regards

[Fox]

Danilom, believe me I have read/still am reading a lot on this forum. Metalhead can probably see that by my user statistics over the past few weeks. And I've read a lot over the past years ( I am member since 2008) I am not being lazy. I thought I understood it, but electronics are not my strong point. The information is often so splintered, often a thread starts out with a conclusion which at the end has 10 other conclusions/suggestions. I am not moaning about that, just asking for help and explaining why, casue I want to understand every bit on this machine.

I thank you VERY much for your quick reply, but I am very sorry to say I do not completely understand the answer:

no you divide by 1.41 , that is the voltage output of the transformer, after the capacitors do their work it will be 64V

What do I divide by 1.41 ? Are you saying that I should have divided the outcome of the voltage calculation for the toroid in this post (http://www.mechmate.com/forums/showthread.php?p=449&postcount=1) as I suggest above ? Do I completely missed reading that in that thread or is it elsewhere ?

[danilom]

If you need 64 Volts to power a stepper drive, then a pure transformer output should be 45V AC !

After bridge rectifier and capacitors you get 64V DC

[domino11]

Your 64 volts you calculated is DCV. Therefore you divide the 64 by 1.414 to get the ACV of the transformer you need. In this case it would be a 45V AC transformer secondary. When rectified and filtered it would give you the 64 VDC you calculated. :)

[domino11]

Sorry Danilo, guess we were typing at the same time. :)

[Fox]

THANK YOU ! for putting up with me-helping me out. Appreciated.

Ps. So, my hunch/back reasoning after reading this figure in the rectifier thread and getting all puzzled was correct, and I do understand this stuff a tiny little bit, or at least now I do.

But.... how ever did you guys figure that out when you were first building and knowing nothing ?! Same way ? Or am I really the first one ever to be so stupid ?

Imo it woudl be a good idea to put a small tip "for DYI powersupply toroids divide this number by 1.41" on this in post 1 here: Calculate the Voltage and VA size of the DC power supply for stepper motors/geckos (http://www.mechmate.com/forums/showthread.php?p=449&postcount=1).

Or at least fix the link in that post; part of my confusion was fueled by the fact that the link Or consider to design & build your own. in post 1 links to the same thread/calculation, instead of this (http://www.mechmate.com/forums/showthread.php?t=350) as i think it should ?

[KenC]

No offence, conversion from AC to DC is really a very elementary, I'm very sure its been mentioned many times. If you had googled, you will find enough information to fill a library.
I remembered I have a few very comprehensive calculation in one of my building threads.
The forum is a treasure of info on CNC building. but sometimes, you gotta look else where suit your learning level.

[Fox]

No offense taken, everybody asks questions on different subjects/is skilled in different subjects. I am sure you have asked a "stupid" questions in your life to someone else (maybe even on this forum). Google and the search on this forum can be very helpful, but you've got to know how to ask the "right" questions, to get the right answer. Human interactions prevails imo when really stuck.
At least we got it sorted in a very quick manner, thanks to all the help here on the forum. In 2 years time, I might also think it's "easy", right now it was darn difficult :) So I will try to remember that when in two years time I need to answer a newbie here.

Anyway, let's agree I am totally stupid and all that, let's move onto more important stuff. Electrics orderlist now sorted. Ordering today. Hope to have steel for table arriving tomorrow, and the lasercut steel will be done by the end of this week if things work out.
So we are moving onto actually building stuff; looking forward to that. Brace yourself for some questions here and there along the way :D

[KenC]

While you are waiting for your electronics parts, I strongly suggest you focus on reading up on "Kitchen Table Project" & start asking questions on the subject.

[Fox]

Will do ( was already reading it to see what elektro/mechanical parts were needed, beside the big ones)
but I just hit a big issue. Deitech can't supply the engines 901 or 902 untill after Chinese new year= two weeks from now:-(

[danilom]

If in a hurry ... sorotec in germany has some nice motors which I also had used

http://www.sorotec.de/shop/product_info.php/info/p2310_schrittmotor-5-9a-bipolar---5-5nm-ds.html

[Fox]

Thnx Danilo,

mailed them ! Also talking to stappenmotor.nl for alternatives.

[Surfcnc]

That's motors Fox - are you sure your reading ?

regards
Ross

[Fox]

Huh ?
Motors are pretty essential for a Mechmate I would reckon, I don't want to build the Fred Flintstone version... :-)
So when I received a message from Claudia I can't have them in time, I need to look for alternatives immediately, right ?
Or what are you aiming at ?

Btw. I e-mailed you with a question if your beltdrives .dxf files are Inches or Metric based ?
I thought AU was metric, but I noticed the Flange bearing you suggest is an Inch based model ??

[Surfcnc]

Hi Fox

It was just a hint that "stepper motors" are not called "engines".
The use of the correct terms really helps as servos motors, stepper motors and even hybrid steppers with encoders all power CNC machines.
Never seen anyone use the term engines so thought I would pull you back on track.

I saw that post but could not find it once I had the time to go back and think about it.
Not ignoring you on that one, just disorganised, so my apologies.
The files are metric in MM.
The flange bearing may be a mistake as I ordered some and noticed the inner shaft diameter was larger than what I had intended to use.
To make it work check that the outer diameter fits the hole size in the plate and that the center bearing hole suits your chosen shaft size.
VXB bearing on the web are a great source for cheap flange bearings.

Regards
Ross

[Fox]

Hi Ross,

ahaaaaa :) I see, "Engines" is of course Dunglish (http://en.wikipedia.org/wiki/Dunglish)because I work too many hour days lately, type in a rush on the forum, and English is not my native language. I thought you were after me because of what Ken said (http://www.mechmate.com/forums/showthread.php?p=60416&postcount=82) and felt a little offended...

"Motor" is used for "engine", "motor" and even "motorbike" in Dutch. Hence my mix-up.

Thnx on the metric, I can source the flange bearing you mentioned but have also found a metric alternative. Was just wondering if I woudl be doing it all wrong again, ending up with metric stuff on inch based designs, making everything not fit as it should. Your bearing had 12,7 inside, no problem if you axle is 12,7 mm.

[Surfcnc]

Hi Fox you are spot on with the design there.
It is hard to read what the intent of some posts are at times so don't sweat the little things.
Everyone helps here but sometimes that help might extend to a little tough love !
Even the most negative of comments will often have benefit after the initial ouch wears off a little :)

[Fox]

True that. You've already proven you can built one. I'm still trying ;)

[KenC]

Fox, Don't get too exited about your motor delivery. your 1 month is almost up...
This is my suggestion.
While you are waiting, you can start work on the Kitchen Table Project. There are actually a lot of work if you are doing it for the 1st time, many of the builders with your similar electrical & electronics skill (me included) know it very well. Figure out how to power up the BOB, the cooling fan & other bells & whistle. The E-stop + Mains contactor/switch circuit & the DC Power supply. that will eat up some of your idle time. As for the testing, you can grab any stepper motors for the purpose. Stepper motors can be salvage from old printers, air conditioner ... etc use your imagination.
If you still has time, start stuffing up the control panels & wire up.
Even if the motors are not delivered when you are done with your control panel. You should be working on your rail grinding & mechanical installations & more....

[Fox]

Motors: decided.
I have decided my back up on the motors ( Danilo's tip has 29 pieces in stock, and can be delivered in a few days),so have asked a pretty please just in case to Claudia if she can make a miracle happen, and if the motors still could still ship with all the other stuff, if not, fall back on that.

My control box models is decided/ordered. Will use two boxes on table; one with everything besides PC and VFD which will be in the second one ( is that an issue? pc and vfd ? I know some do not recommend vfd inside controlbox because of interference with bob, will it also cause trouble for the pc ? ). Will use a pivoting monitor and keyboard ( maybe touchscreen later on) like Liaoh did fixed pivoting on corner machine.

Controlbox
My powersupply parts decided/ordered.
My drivers are decided/ordered.
Motors: see above

Am looking at the other electrical parts for the kitchen table project.
Cables decided form Gerards advises, need to calculate length as I am much bigger then Doug Fords table - example lengths.

Going to use BIG surplus heatsink similair to pmdx, in side of box for drivers etc, no fan needed I hope.

Bob ( USBCNC ) need 9-24 volts powersupply ( am looking at second hand 24 v 5,0 amps din rail transformer offered for cheap at local ebay ) Otherwise use Geralds example of transformer from his box or cheap plug in. Run e-stops with 24 volt relais as well.

Will built two sockets for mains in side of box like Gerald.

Thermistor relays not needed as I think my cheap chinese 2,2 KW ordered from lovehappyshopping (http://stores.ebay.com/love-happyshopping/_i.html?_nkw=2%2C2kw&submit=Search) does not have this option ? Could I instead protect it from overheating by limiting current with a fuse ?

Dinrail terminals blocks you always but as a single row (I see at farnell) ? So I need to buy 30 single rows with double contacts, right ?

Mains contacter deciding on it( how many amps will be enough, should I just take the same like mains ( 16A or 32A for 3 phase ?)

In doubt if if input with 3 phase or single phase power ( have both at shop, but 380 not yet in the room I am planning to place the mechmate ). Any advandge beside the higher load on using 3 phase ?

E- stops buttons etc decided.
Want to have ordered all parts latest next Saturday ( saturday I can talk to a friend who's electrician, just to make sure I am not blowing myself up).

[KenC]

Control panel space is never big enough :)

I reckon the VFD deserve a panel box all by itself. I like my PC away from everything else. Especially when I'm in a hurry to build.

If you can afford it. 3kW spindle will be nicer to own & use.

Since you are in a hurry, I would say leave the thermistor for later. BTW, I'd not known of anyone burning their spindle because of overheating during operation but premature bearing worn out is fairly common when you plunge too hard at the beginning of your CNC cutting career.

You can never have too many Din rail terminals blocks.

While you are at it. Look at dust collecting as well. This is something you really need when you are sweeping the wood chips later. Vacuum table can wait, for now... ;)

[Fox]

Hi Ken,

Correction: Vacuum table is needed. Dust can wait. Cutting PMMA and Polycarbonate sheets into many individual parts for first job !
I think I will still try the PC with the VFD or PC with BOB in one box thingie, but for my first assignment I'll just put the PC on a chair somewhere, away from all interference.

[KenC]

Anything that make you happy ;)
You might be right. a simple vacuum table will suck any thin plastic sheet snugly.
Anyway, dust collector work badly on some plastic chips.
One last attempt to state my point on separate PC. There will only be one or maybe 2 very long cable (which is the printer cable & the HID console). A lot easier handle & aesthetically pleasing than having a big bundle of cables running all over the place. & long cables = $$$

[Fox]

Hi Ken,
Thank you :) right now I am not so happy, encountered some nasty setbacks (steelorder was F*d up by the steelcompany so I am 4 days behind on my schedule) + I made a calc error.

Points taken on the cables, I would need to run 3 cables: 1 USB ( for USB hub in monitor ), 1 monitor cable, 1 monitor signal cable. ( mouse and keyboard are usb and can plug into monitor ), but Liaoh I noticed put his pc in his monitor housing I might follow, just means I need to build a housing/buy one.

[KenC]

In my pass life as a project manager, I always say " You can only plan...What ever will be will be..."

You forgot keyboard cable, mouse cable & maybe a USB Hub cable. Don't forget the power cord for the Monitor...

If you go liaoh's way, Just buy a panel PC & mount it on a casing. that will save you tons of work & time. & the casing can be made slick & streamline. I am dreaming of a ATM kiosk look-a-like panel...

[Fox]

I was forced to decide to outsource the job.
I feel the pain of this decisions both financially and personal... as I am not the kind of person to easily give up on a deadline (in fact I never did in my whole life - it's the first time ever I did not achieve the goal I set out to accomplish in the time I intended).

We all knew everything should work with me to be able to make the deadline and due to recent issues which were mostly out of my hands; steel not delivered in time, supplier lead-time issues, Chinese new year, my wife being ill passed week (I had to take care of our 2 year old) and now this weekend my friend whom I was supposed to weld the frame with at his shop struck with illness... I surrendered; it's getting very unlikely/impossible my Mechmate will be completed in time and I'll be able to produce everything needed on it within the next 2 weeks. I was pulling in a lot of favors with friends and family (and you guys) to get it done in time, as I could not do it all on my own in this short time, but sometimes favors can't be had within a deadline.

It was a very hard decision to make.

I will be continuing the build, the steel is already cut to size and piled up in my friends shop, so we'll weld it up as soon as he's feeling better to have it out his shop... but everything will be in a slower pace.
I myself will take a breather for a few days from the forum and the project to cope with the disappointment. Thanks all for your support so far, see you back somewhere later this week.

[southernduckie]

Sorry to hear about your situation i am sure u will look back and realise this is not so much of a bad thing, spend the time and get the MM working right and keep looking for the next job life will reward you.:)

[Fox]

Hi Guys,

I am back ! Thank you for all the kind words on the forum and by PM ( you know who you are ). I think it was the right decision in the end... it has to be now anyway :)

Spoke with my buddy today, he's feeling better but got a lot of catching up to do on his regular work. This week will be spend for both me and him doing just that ( I also had a lot of regular work piling up when focusing on the MechMate the past 3 weeks).

Upcoming weekend I will prepare/double check myself on which-hole-goes-where and how-to-weld-what-to-what, so he does not have to wait for me to figure out stuff while we are actually welding/drilling/fabricating, and it should be pretty smooth sailing.

All steel is already cut to size and all laser cut parts are also ready ( including laser cut and engraved bolt on flanges for easy alignment).

Next week we will spend the evenings to weld and drill the table, gantry... all the major steel assemblies basically, and the goal is to have the thing out his shop and into mine before the second weekend of February.

I will do all remaining ordering/fabricating from this point onwards by myself in the evenings and weekends. I will start with the third week of February on the belt drives, rack and pinion or the kitchen table project...whichever parts are in first.

Keep you posted !

[Fox]

Hi guys ( Ross I e-mailed you as well about this) ,

tomorrow night we start to drill and weld the base table together !
Exciting stuff.:D

One last minute thing keeps nagging me tonight and wondering for a second opinion from (a larger-) bolted table owner:

Did you "oversize" your bolt holes ? If you did... by how much and both holes or only 1 of a pair ?

I contemplate: a bolted connection is already weaker than welding, so a perfect fit would help with strength. But if I drill 'exact' 12 mm holes, I wonder if it'll ever come apart with welding stresses and such ( or can be re-assembled for that matter ) ? I noticed for instance the support board is bolted with M8 but the holes are 9 mm. Did you use the same +1 mm strategy ? I don't want to be too tight-nor unnecessarily loose on the critical joints, it does not have to come apart everyday, and the bolts then will be replaced anyway.

I've searched the forum but can't find anyone mentioning holesizes/tolerances besides Ric, but his table is a fair bit smaller then mine will be.

(I do know about the 4 mm hammering pins trick for re-alignment purposes, but even then I think it might become a big pain in the ass to get bolts back in holes when they are 'exact' to size ?)

Ps. I am going to be using M12 (8.8 or 12.9 HT) bolts, with self locking nuts/locktited. My best guess right now would be 1 hole at 12 mm and 1 corresponding hole at 12.5 mm for all my flanges/main connections (each connection 4x2 holes)???

[smreish]

Bolted table 5' x 10'
Main beams and crossbeam's drilled 1/32" over for a the 1/2" bolt.
Thus - a 17/32"" hole on both holes of joint - for a total of 1/16" movement.
Angled structural washers on tapered flange side of both sides of joint.

I know many others adhered to different methods or standards - for me, it worked well with a very square and parallel table base.

[Fox]

Thanks Sean,

Translated it into metric (Geeeeez I would go crazy working with these fractions in my workshop :D ):

1.524 x 3,048 meter table
bolts 12,7 mm
both bolt holes 0.7938 mm oversize = holes 13,49375 mm
movement for each connection = 1,5875 millimeter

so I might be heading for two holes at 12,5 mm ( instead of one at 12 and one at 12,5 mm ).

Ps. I picked up on the angle washers earlier ... but I only need them on my support board luckily.... as I use I beam and laser cut flanges for my table design.

[smreish]

Fox,
If you layout for the hole drilling process is spot on, then limiting your drill size to one tool is most likely okay.

Some of the builders have "less than precise" layout methods and need that little bit of wiggle room.

The design of the MM is fashioned to accommodate these imperfections in tolerance. Bolt together main table has been done countless times, but not part of the MM original intent. You do sacrifice a bit of rigidity with the bolting, but it is minimized with proper cross bracing and spoilboard attachment.

12,5 mm would be my choice with 12mm fasteners. Less drill bit sizes, less money.

[Surfcnc]

Fox I have answered your question by hotmail as your 'respond by email" is turned on on your forum settings.

Both bolting and welding of tables are very strong.
Most movement is not typically in the bolted joints or the welds, it is along the spans of the metal.
The material itself has elasticity and only good design can minimise that but never fully eliminates it.
Even epoxy granite has some resonant frequency.
It is important to not over think or over design the Mechmate and get on with the business of building a machine with good tight tolerances.

On the question of strength, a poor weld with low penetration is probably weaker than a properly tightened M12 8.8 grade bolted junction.
It is extremely difficult in the real world to make accurate generalisations about strength.
Imagine your self suspended from a beginners weld 20 stories up or from a M12 HT bolt.
Personally I would choose the M12 HT bolt as it is more of a known quantity and that is exactly why critical welds need to be certified.

Regards
Ross

[Fox]

Thanks Ross,

Got your advices !
Based on averaging both your and Seans approach/advice I'll be trying to keep them true to size initially ( your approach), and can always enlarge them a bit more if/when needed (Sean). Making a hole smaller is much more difficult I reckon. I will test it throughout the process and adapt where necessary. Thanks guys !

Into the workshop tonight; will post some pictures if it's not too late when I am back home.

http://s7.postimage.org/3lhe0rc6v/useless_without_pics_2_gif.jpg (http://postimage.org/image/3lhe0rc6v/)

[Fox]

We have lift-off !

We got 3 hours in tonight - not as much as I would have liked - but my friend starts at 6 in the morning so he gets to decide when we go home :). Felt great to have physical parts in my hands.
I managed to check everything against the drawings/dimensions and my cutting list, found a few small 'mistakes' on quantities (most in my favor), and welded small assemblies flanges on main beams, feet on legs etc, numbered all matching parts for ease of welding. To be continued next Thursday night with lots of drilling and welding. For now some pictures:

Some of the lasercut parts:
http://s11.postimage.org/duwqgaqnz/IMAG0667.jpg (http://postimage.org/image/duwqgaqnz/)
Test fitting:
http://s8.postimage.org/5rfhyz9gx/IMAG0659.jpg (http://postimage.org/image/5rfhyz9gx/)
Main beams:
http://s17.postimage.org/flswnkbkb/IMAG0662.jpg (http://postimage.org/image/flswnkbkb/)
Pretty straight:
http://s3.postimage.org/5zke7xeun/IMAG0665.jpg (http://postimage.org/image/5zke7xeun/)

[xraydude]

Okay, I am jealous of that jig/fixture/welding table. Used one in a previous job years ago and there is nothing like having the right tool for the job.

ted

[Fox]

@Xraydude

Indeed, it's really nice to have all those helpful tools and huge level surfaces etc around. I could do it in my own shop/garage just like most of you guys but this was an offer of my friend I could not refuse.

[Fox]

QUESTION:

1030420W Welding the "monocoque" Y-Car

Is this right ?

http://s4.postimage.org/6uzaemdrx/gerald.png (http://postimage.org/)

Mine is like this ? Is that a mistake: should 1 have been cut in mirror ? Or is Geralds example an old drawing that has been modified ?
http://s12.postimage.org/yznuscytp/later.png (http://postimage.org/)

[xraydude]

Yours is correct. The elongated slots should be diagonally across from each other.

Ted

[domino11]

Geralds picture is of the original design. Your is the latest cut which does not have left or right sides. If you check you can only get the tabs to line up one way. :)

[Fox]

I noticed that of the tabs, which made me less doubtful, but still.. good to know guys thnx

[Fox]

Sorry about the Spam links under previous photos, they were added by the free image hosting and I can't remove them because I am not allowed to edit the posts after 30 minutes, and noticed it afterwards

Update:
worked about 8 hours with my friend past week divided over 3 nights and a Saturday morning, and here's what happened in a short photo essay:

Handy tools:

I bought a self-centering center for copying center of holes (recommended ! works a treat)

http://s7.postimage.org/n56sgf5rf/IMAG0687.jpg (http://postimage.org/)

My friend had a self-centering magnetic drill press:

http://s17.postimage.org/6caiek33j/IMAG0680.jpg (http://postimage.org/)

So I drilled a lot of holes 6 mm holes followed by enlarging the same to 12,5 mm holes ( I wished there was a shortcut to that but good holes/good work takes time)...while my friend started on the welding:

http://s11.postimage.org/fpolqwd77/IMAG0681.jpg (http://postimage.org/)

By the time I was finished with drilling the main beams, the legs etc:

http://s17.postimage.org/vbauf30vj/IMAG0686.jpg (http://postimage.org/)

My friend had welded these wheel supports:

http://s3.postimage.org/lcscs6rgz/IMAG0685.jpg (http://postimage.org/)

And these front and back frames:

http://s3.postimage.org/9wqj7w2oj/IMAG0695.jpg (http://postimage.org/)

http://s9.postimage.org/z0fp5wedb/IMAG0700.jpg (http://postimage.org/)

And these side frames:

http://s8.postimage.org/7yawy1atx/IMAG0689.jpg (http://postimage.org/)

So when we bolted all lasercut flanges ( holes 12 mm ) to these frames ( 12,5 mm holes ) we were able to weld up from these parts asemblies like this:

http://s18.postimage.org/q62abpa55/IMAG0699.jpg (http://postimage.org/)

To a table like this (2,5 meter x 3,85 meter):

http://s4.postimage.org/ymyc9nxdp/IMAG0705.jpg (http://postimage.org/)

http://s12.postimage.org/lgpe5ljil/IMAG0710.jpg (http://postimage.org/)

Which to me and my friends utter joy (even he was surprised it was that close) leveled out within 0,25 mm precision on parallel,level and square for the main x Beams, and on 90.00 degrees on all 4 legs in both directions, measured with a digital level (accuracy 0.05 degrees). :D I know this kind of excitement is a bit childish cause it'll move that much just by sitting in the sun, but still... And it's very stout with 50x50x5mm square tubing, 100x100 I-Beam profile legs, and 200x100 I-Beam profile main rails, and the 80x45 UNP for the table.

After we high-fived each other and ate a sandwich for celebration [ 10:00 in the morning after starting at 6:00 AM + at his workplace meant no celebratory beers ;-) ] he then started to clean up all the tools and dress some welds.. and I drilled (by hand) 4 mm holes trough all the way for the flanged connections for ease of re-aligment with 4 mm roll pins in the future.

Then we unbolted the entire frame, and we're able to carry everything out with relative ease (still heavy/big metal).
I did measuring/weighing in my CAD program when designing the table so that everything can be carried by two persons trough a standard door opening.

I think we even could have made our holes 12 mm everywhere, as we had no issues whatsoever on getting the bolts out. It really helps my buddy does metal for a living, so he know a thing or two about welding and distortion + he has all the alignment tools and tables for big steel a M.M.M. (Man-on-a-Mechmate Mission) could wish for.

Then it went in to the back of a 4.3 meter deep loading bay of a rental Extra-Extended-Mercedes-Sprinter-Van ( I made this boltable design so I can still move it around with a 'widely available' rental van, and I do not need to hire a truck driver ) :

http://s12.postimage.org/dcznjtqml/IMAG0711.jpg (http://postimage.org/)

and now sits in my shop... (...disassembled )
I still have to drill the holes for the support board after i rebuild it ( I can borrow my friends magnetic drill for that), and my friend will weld up the gantry and carriage...
Then the main table will be finished and look something like this:

http://s11.postimage.org/vcd5jk9ub/mechmate.png (http://postimage.org/)

I can then move on to the rails, Z-slide and kitchen table. But first need to get the big assignment out the door, and catch up in general on parts of my business I have been slacking so besides some ordering I'll go quiet on the build for a week or 2.

[jessyjames]

Wow. That is just too cool. Thanks for sharing this information and pictorial with us. Sure wish I had friends like this who also do metal work.

Again thanks for the pictures. It's so neat seeing how other people build their tables and techniques.

Keep up the great work,

James

[Fox]

Thanks James,

I know excactly what you mean...as I was hunting for these kind of pics only a few weeks ago ( still am actually ). Matter of fact is, most of my solutions are copies of, or directly inspired on pictures I found here on this forum from other members. Sharing experiences is what makes this forum work, as we can learn from each other, and then pass on the knowledge gained to others.

Since a thread usually starts with a lot of technical words/questions (mine did) I"ll try to keep posting pics on most of my major steps, from now on. Not always easy with the grinder in one hand and the expensive mobile in the other :)

[parrulho]

Nice table, neat and funcional. Congratulations. Good tools make the work more easy.
wish to have some of this at my place...

[Fox]

Hi Guys,

I have a question/poll about the cable management.
I have obviously read the Cable management thread by Gerald (http://www.mechmate.com/forums/showthread.php?t=498), and am leaning on using the 57Wx25H energy chain (inside dimensions) for all three axes ( would this work on the Z - or is that overkill ). I will be designing custom supports for the energy chain, which I want to finish shortly so I can round up the base table. For this I need to decide the size of the energy chain first.

Looking for some opinions from Mechmaters using watercooled spindles + maybe some additional accessories on their machines if that is indeed what their are using also or if 38x25 mm would already suffice for cabling and waterhoses etc ?

For the 38x25 mm I have seen a cheap option on our local craigslist at the moment, and it seems Deitech also only stocks that size

[smreish]

Fox,
The use of the larger chain is not an issue for the Z as long as you have room with your spindle/dust collection set up.
I personally used the same size E-chain for all 3 axis' without any issues. I found it to be more cost effective too - only one purchase.
My extra long 48" z - is shown here with similar size chain your recommending:

http://www.mechmate.com/forums/showthread.php?t=460&highlight=smreish&page=99

[danilom]

Larger inside dimension of energy chain means its overall stronger and can support larger free spans.

My friend put 100 x 50mm inside diameter and it can stand on a 3m span without any problems even with 4 x 8 core cables and two water lines and 4 x 2.5sq mm spindle wire, of-course he found it second hand.

My advice is for large spans, check the documentation about free span lenght

[Fox]

Once again thanks for the quick responses guys, lightspeed !

@ SMreish
Thnx, cost as well as future expansion/add-ons was on my mind indeed ( to use same/bigger size on all three axis).

@ Danilo
Another argument, I am convinced.

Larger chain it is; hunting for second hand versions a we speak, otherwise a new Chinese equivalent.

[Fox]

Danilom,

your friend is that Paijka you are referring to or another friend ? It looks like Paijka is using much smaller in his thread (http://www.mechmate.com/forums/showthread.php?p=58376&postcount=53)?

I just found a 103 x 25 (inside) Igus chain advertised second hand locally for cheap.
More width then I need really, but if the price is right for my meters I might take it (although Z is maybe not fitting ?).

[danilom]

No it's another friend with also a Mechmate, use larger for X Y and get some smaller for Z, but I must say, any large (over sized) piece of equipment looks really good on the machine :) , like a large control cabinet on Pajka's machine. Any person that looked at it was astonished, at some point we even considered a larger one. He wanted not just a machine that does its job, but also a showpiece. That's why the machine was professionally painted with fillers and sanded. The cable chain is I think inside 75 x 35 inside
Second hand parts can be somewhat helpful to lower the budged but get the once very expensive part with real quality, that's my opinion.

[Fox]

but I must say, any large (over sized) piece of equipment looks really good on the machine

Must be a male thing :D :D :D

[Fox]

Ps. If my eyes are not deceiving me I think Gerald used that size ( 103 x 25 (inside) Igus chain)

on his original (http://www.mechmate.com/Forum/messages/329/646.html)?

From that view and from a quick CAD sketch it looks nice, but is fouling the use of full Z height ( cable chain hitting object as it is lower then Y beams because of bending radius of 175 mm)

[Fox]

Hi, I got another question for you guys...

...would it make sense to have the machine ( once it's leveled and squared up -use some clamps for that routine- ) center the holes for the support board in the steel beam, then drill them up to size with the magnetic drill ?
Then drill two diagonal holes in the support board, align it, bolt it down, and run the hole drilling + countersunk screw routine with the machine ?

Not that I mind doing them by hand if really needed, but I was thinking.. hey, I am building a CNC machine here, which are great at repetitive work, and way more accurate then I am by hand...

Makes sense ... or not ?

[normand blais]

I would try it ,I cant see why it would not work. You might have to rig a drill instead of a router to do that .It would save time

[Fox]

I am not going to try to drill the main steel beams, I think that would lift the pinions of the rack ? Just center it ( touch it with a drill, or maybe even fit a automatic spring loaded center punch in the spindle and make the center points that way ). For the mdf I do not need to fit a drill, I'l be fine with the Spindle I imagine.

Dreams for sure, not on that point yet, still need to finish my base table and then move on to the kitchen project ;-)

[Fox]

Hey guys,

been catching up with my other work, cleaned out the destined Mechmate room re-located some stairs that we're in my way... wired some proper lighting in the room and painted all brick walls pure white ( now that makes a huuuuge difference for a pleasant working environment in a no-window room ). Waiting on last coat of paint to dry, and can then re-assemble the table somewhere this week into it's new home. Unfortunately the next week I also have to travel out of the country for business for 3 weeks end of March-April, so it will be untouched for a little longer when it's back on it's feet.

Want to pick up momentum again upon my return and get it running asap, so I am trying to order all big parts/long lead-time parts before I leave. Have spend more time studying plans and parts for the kitchen table project and have cleared most obstacles for that but have a nagging question i hope you can help me out with:

I want to wire Bipolar parallel the Deitech 86HS9801 (http://www.driver-motor.com/ProductContent.aspx?ID=47) ( in know the cons vs pros - so let's assume this is a 'fact' ).
Looking at the data at their site .. I assume this is unipolar data.

Bipolar parallel current = Unipolar current X 1.414 (http://www.mechmate.com/forums/attachment.php?attachmentid=6813&d=1256639835)

Which would mean 4,0 x 1.414 = 5,656 A

The leadshine DM 856 driver (http://www.leadshine.com/Product_Show.aspx?ID=115) offers 5.6 A max (4.0 RMS). Would this driver still be safe ?

- Or should I move a step up to the DM 570 (http://www.leadshine.com/Product_Show.aspx?ID=40) which offers a max of 7.0 @ 5.0 RMS)?

- Or.. would neither be sufficient ?

I am asking, because I can't properly judge ( first CNC ) if the RMS value is that important ( I imagine the motor will not be asking full power most of the time ) and/or if the 0,056 amp underrating of the DM 856 driver is actually an issue to worry about ? ( or can it be easily neglected like with the calculated VA power of a power supply usually being much higher the numbers people actually use... )

Thanks for any insights !

[Fox]

does anybody have an idea about the question posed above ? I am still puzzling with this issue ( electrics are not my strong point ). Thanks in advance !

[danilom]

Fox, 9801 is 4Amp motor, so its ok to use it with DM856
mine 9802 are 5Amp and if I would run them 7Amp they would surely blow up :)
so that "unipolar" in their table is some error as everyone else rates it as bipolar.

also we are running one machine without problems with DM856 and 5Amp 9802 motors

just give them voltage and its fine.

[Fox]

Thanks danilom, another thing off the list. Hope to drill all holes next weekend, and then flip the table. Going to order everything electrical over the next two weeks. Main deadline is to have it up and running before summer holidays, looks a long time away, but man there are a lot of little things to think about...

[Fox]

Superbusy at the dayjob the past few months, so only baby steps, but progress nevertheless....

http://s21.postimg.org/5oci0x4gj/IMAG1014.jpg (http://postimg.org/image/5oci0x4gj/)

http://s16.postimg.org/hxro6cey9/IMAG1019.jpg (http://postimg.org/image/hxro6cey9/)

handy helper ( in red )

http://s21.postimg.org/503nhz5qr/IMAG1018.jpg (http://postimg.org/image/503nhz5qr/)

on it's feet wheels.

http://s16.postimg.org/5ketzfp9t/IMAG1020.jpg (http://postimg.org/image/5ketzfp9t/)

with table frame.

http://s16.postimg.org/cown8gwj5/IMAG1021.jpg (http://postimg.org/image/cown8gwj5/)

detail of bolted connections ( these bolts are temporarily - dismantled later for paint and replaced )

http://s16.postimg.org/nejis7n3l/IMAG1023.jpg (http://postimg.org/image/nejis7n3l/)

wheels for moving it around ( wheels will mount in two positions - this is the high position, so it can load onto trailer and such, the low position will only have 5 mm between feet and floor, just enough for small movement indoor.

http://s16.postimg.org/oe58a6cvl/IMAG1024.jpg (http://postimg.org/image/oe58a6cvl/)

dowel pins ( 5mm ) to align everything nice and properly again worked like a charm..

http://s16.postimg.org/9kqkpf54h/IMAG1026.jpg (http://postimg.org/image/9kqkpf54h/)

UNP profile mounted to I beam.

http://s16.postimg.org/9aj262qi9/IMAG1027.jpg (http://postimg.org/image/9aj262qi9/)

under the beast

http://s16.postimg.org/jjbjbwek1/IMAG1028.jpg (http://postimg.org/image/jjbjbwek1/)

partly welded ends of I beams

http://s16.postimg.org/5s729opm9/IMAG1029.jpg (http://postimg.org/image/5s729opm9/)

holes for support board

http://s16.postimg.org/4gzb17u0h/IMAG1030.jpg (http://postimg.org/image/4gzb17u0h/)

eat your heart out mr. Eifel :)

http://s16.postimg.org/eprs71i29/IMAG1031.jpg (http://postimg.org/image/eprs71i29/)

straight as morningw....:D

http://s16.postimg.org/mtzwbs4hd/IMAG1032.jpg (http://postimg.org/image/mtzwbs4hd/)

like I mentioned before, the room around it leaves to be desired. At least it'll be nice and warm during the cold winters.

http://s16.postimg.org/3sqgi9vap/IMAG1033.jpg (http://postimg.org/image/3sqgi9vap/)

birds-eye trough door opening.

http://s16.postimg.org/bdy2rhfox/IMAG1035.jpg (http://postimg.org/image/bdy2rhfox/)

designed to fit ( and it did, just :) )

http://s16.postimg.org/3mhcsxbjl/IMAG1036.jpg (http://postimg.org/image/3mhcsxbjl/)

Now for the dilemma.... do I go mechmate on my rails or these 2 x 25 mm SKF rails + 4 new carts ....I basically got for next to nothing... but 3,5 meters instead of the 3,85 meter I need.

http://s16.postimg.org/j98m6apbl/IMAG1037.jpg (http://postimg.org/image/j98m6apbl/)

just a tad too short... damnit... they are sooo nice... but will only fit the Mechmate gantry design if I compromise/downsize on my effective cutting size (3050mm x 2050 mm) If I would do a 'commo CNC' single big beam gantry it would fit without compromise. But Mechmate parts are already lasercut and the two big grantry beams are bought, just not welded.

What would you do !?

[swatkins]

I would go with MechMate design.. Why settle for a smaller cut volume when you can have the larger size without compromising the accuracy of the machine.

[Fox]

Actually; mechmate gantry and L rails= full size cutting. OR single beam gantry with SKF 25 mm rails = full size cutting. The whole catch is the SKF rails are 350 cm long NOS (new old stock) models, so I can't buy more wagons then the 4 I have for them. I have already decided I will use a custom z, like many mechmates have with linear rails and ballscrew. So my dillema with these 3,5 meter rails is, if I buy 2 similair hiwin rails for my Y, the whole machine has linear rails and full cutting envelope if I use a single beam gantry. Or stick to the plan... and have L profiles on my x and y and linears on my z and sell the 2x3,5 meter..

[Fox]

Decided to push on MM style a few weeks ago. The cheaply bought 350 cm SKF linears are going to be stored, and I have decided to make some small adjustments to the plans for holes etc ( Hiwin and SKF etc use mostly 60 mm hole spacing) so I made my rail holes 120 mm spacing ). I can always quickly upgrade to them in the future if I want to this way.

Makes the most sense as it will get me everything in size I want, and without any major modifications or custom design ( which always costs more time then anticipated ) nor extra costs.
I think common sense kicked in and under the motto "if it aint broken-don't fix it", I want to experience first hand the results from the ground rails, as I've seen enough MM's online with good accuracy from them-but none in real life, before thinking of any modifications.

Have since then put on the support and spoil board (28 mm MDF), welded my Y car yesterday, and will weld my Gantry later this week. Big order with motors and spindle energy chains is coming in from Deitech beginning of next week. And next week I will also start on the rails. Working on Kitchen table in between. Last thing on the build list is the modified Z slide (screw and linear guides), which I am currently designing, then major constructions blocks are finished, and I need to start assembling everything and I can start on the wiring.

Still a lot to do, but slow progress.

[Fox]

Progress; Order from Deitech came in today !
A lot of money, but happy with the pile of parts.

Description and Model No. | Quantity | Unit | Price | Total Price
-----------------------------------------------------------------------------
Stepper motor 86HS9801 | 5 pcs | USD48.00 | USD240.00
Stepper driver AM882 | 5 pcs | USD65.00 | USD325.00
Timing pulley 18XL | 5 pcs | USD10.00 | USD50.00
Timing pulley 72XL | 5 pcs | USD26.00 | USD130.00
Timing belt 170XL-10mm | 10 pcs | USD3.00 USD30.00
HF spindle 3.0KW DTS-24 | 1 pc | USD476.00 | USD476.00
Spindle bracket for DTS-24 | 1 pc | USD30.00 | USD30.00
ER11 shank 1 pc | USD35.00 | USD35.00
ER11 collect chuck (1-7mm) | 7 pcs | USD6.00 | USD42.00
ER11 collect wrench 2 pcs | USD8.00 | USD16.00
ER20 collect chuck (1-13mm) | 13 pcs | USD6.00 | USD78.00
ER20 collect wrench | 2 pcs | USD8.00 | USD16.00
Energy chains 25x38 in 2 meters | 1 pc | USD26.00 | USD26.00
Energy chains 25x77 in 6 meters | 1 pc USD | 108.00 | USD108.00
FedEx Freight charge for above goods to Holland | USD300.00
----------------------------------------------------------------------------- +
Total (by bank transfer): USD1,902.00

Question on the Energy chain, when locked in it's "straight" position it does not want to lay 100 % flat but has a slight bow to it ( ends touch but middle raises off the floor ) .
Is this normal/to be expected because they pre-load them to counter any 'sag' of the weight of the cables and themselves ?
Or is this the difference in quality between a Chinese brand and for instance IGUS ?

http://s11.postimg.org/3xhrld5wj/IMAG1197.jpg (http://postimage.org/)

http://s11.postimg.org/wpoj4udk3/IMAG1199.jpg (http://postimage.org/)

http://s11.postimg.org/6g3gm1rmr/IMAG1200.jpg (http://postimage.org/)

http://s11.postimg.org/60nlfagbn/IMAG1202.jpg (http://postimage.org/)

[Surfcnc]

Hi Fox

Been waiting for someone to use these drives.

The stall detection is the big thing these bring over a std digital drive.
Can't believe I just wrote that but things do keep moving on when it comes to technology.
The alarm on the drive(s) should be wired back to an input in Mach3 to ensure an orderly stop on all axis's once a stall is detected.
While it may or may not save the piece it is working on it will save any subsequent parts, so that has to be a good thing.

Ross

[Surfcnc]

Fox

Also noticed Claudia has sent you the motors with the key ways rather than the two flats like the original Motion King Motors.
The key ways will knock out easily but remember now to just tap one grub screw into the pulley hubs rather than the normal two to bite into the single keyway slot.

Ross

[Fox]

Hi Ross,

1)
Do you have happen to have any idea/answer on my Energy Chain question ?

2)
that ( and the higher output) is the exact reason I chose those leadshine AM882 drives in the end and at the relatively small higher price offered from Claudia compared to my originals plans for the leadshine DM856 driver... I thought it was worth a shot to try this.
I'll be using EDINGCNC CPU5 - B Professional with Ethernet (http://www.edingcnc.com/index.php?pagina=detail_producten&productenid=51) instead of MACH3 though but same procedure applies.

3)
I specifically ordered the keyways, I actually prefer them over grub screws as there's no risk of slipping pinions/pulleys albeit they require a bit more advanced work to pulleys and pinions.
I will do this in my gearbox to put the 18 tooth pulley on the shaft of the motor:

http://s14.postimg.org/gon2ktgap/solution_small_pulley_on_motor.jpg (http://postimage.org/)

[Fox]

If anybody can tell me more on the energy chain 'issue' ...much appreciated !

-Another question that has been puzzling me a lot lately: I want to order my rack and pinion. I have just found a supplier called maedler (http://www.maedler.nl/Product/1643/2479/2315.aspx)who can provide pinions with induction hardened teeth only, so the rest of the pinion can still be easily modified/worked upon, while the teeth are hardened and more wear resistant, for 9,40 euro each I can't complain ( to harden standard pinions I got quoted 40 euro per pinion ). I will combine these pinions with standard racks. The pinion ( module 1 metric ) :

http://s9.postimg.org/5t7l1s1jj/Stirnzahnrad_induktiv_gehaertet_jpg.jpg (http://postimage.org/)

Research and previous tip from Gerald have learned me that the minimum metric pinion is 24 tooth for mod 1 metric ( smaller would compromise smooth running ).

I am using a 4:1 belt reduction.

My software is EdingCNC > CPU5 - B Professional (http://www.edingcnc.com/index.php?pagina=detail_producten&productenid=51)

My motors are the Deitech 86HS9801 (http://www.driver-motor.com/ProductContent.aspx?ID=47)

My driver are Leadshine AM882
(http://www.leadshine.com/productdetail.aspx?type=products&category=stepper-products&producttype=stepper-drives&series=AM&model=AM882)
SO.... I mix it all up into the spreadsheet Gerald provided on the subject and get this result.

http://s9.postimg.org/3qbeipmtr/Microsoft_Excel_Gear_Speed_Steps_Freq_Calculat.png (http://postimage.org/)

I understand how the pinion size with it's pitch diameter, and the steps from the driver influences the steps needed per mm and the resolution. I also understand the influence a gearbox has ( higher torque, higher res, lower speed ) so that a higher motor speed is needed to have the same machine speed. I understand there's a direct relation between machine speed, frequency, and motor RPM.

My questions on the part I don't understand:

1)
What would be a 'happy' move speed for the MM with these specs ( I mean no use flying the thing around if it's too much stress on the machine-but quicker is better when quality does not suffer and it's not causing too much wear). I have no idea, I found sooo many different speeds on the forum, I am lost. The 9000 was the value that was already in the spreadsheet, but i think it was just there for putting something there,or is there a reasoning behind it ? If I change it to 20.000 then my motor rpm and frequency obviously increase which leads me to :

2)
With my motors; what is the max 'happy' RPM for these (not mentioned on the site - I know many people use them - can anybody enlighten me ??)

3)
what is the max 'happy' allowable Frequency ?


My goal is to find the biggest pinion ( bigger pinion = less wear = smoother running ) that still leaves me with a reliable but also fast ( within MM capabilities) move speeds. Maedler can provide pinions in 24, 25, 26, 28, 30 teeth ( and higher, but I think 30 is kind of the max that fit's nicely with standard plates ??? )



Ps.
Last question ( sorry for all the questions - I spend already 3 weeks looking for answers and can't find them):
I know Gerald based the microsteps in his spreadsheet on the Geckodrives defaults. With my drives I can toggle all kind of microstepping: "Microstep resolutions are programmable, they can be set from full-step to 102,400 steps/rev"
So I could also tweak with this parameter as well, but I assume this is something you do not really want to do too much tweaking with, and first try to get a good set up with the other parameters ?

THNX in advance !

[danilom]

My 9802 three years ago also came with key, its a good thing.

Get 26 tooth, if it proves you need more speed get 30 , if resolution get 24. Keep 26 for spares.

Using pinion harder than rack can maybe ware rack sooner if you cut a lot in one area. Be sure to check on rack hardness.

[Surfcnc]

Key ways are good agreed.

Happy maximum speed just try 10000 or 10 meters per min go even slower (like half) if you like or are feeling cautious.

The Gecko drives are set at 10 microsteps, however you have a wide range of options, so play a bit once it is running.
Initially start with a low value and then try a high value to see (hear) the difference in movement.
The Kitchen table project is a good time to do this.
As the microstepping increases the available torque decreases so there is a cost for increasing smoothness by increasing microstepping.

The minimum allowable frequency is the kernel speed the controller PC will require to reliably pulse the signals.
This is a calculated value that will for example increase as the move speed increases and or the microstep value increases.
Gerald's gear step calculator is awseome :)

10 Microsteps seems to be a pretty good trade off all round but I ran my Z axis at 16 microsteps for ages and nothing bad happened.

The USBCNC software is good stuff, just finished a test of it myself.
Your board is capable of pulsing at 125KHz so you are well within the abilities of the board to provide an adequate pulse stream in your speed range.
Also note that the pulse generator is on your board so you have the benefit of a hardware pulse generator over the Mach3 software pulse generation.
You will notice that the toolpath display is unaffected by playing around with it during the machine operation, unlike Danilo and myself on Mach3 with the smoothstepper.

Energy chain - don't worry those cables are heavy and most of the MM chains end up sagging a little esp on the X axis.

Ross

[Fox]

@ Dainlo; thanks for your advice, beside common logic ( it makes sense to me to do take the middle and work up and down if I can't work it out ) but is there actually a calculated theory behind it ( do you know the max hapy motorspeed ? ), or more a hands-on trail and error approach based on your experience ? Because otherwise you could also take 25 tooth or 28 tooth just the same ( as long as I increase the microstepping to get the resolution I want, and I do not exceed the max motorspeed )?

@ Ross
- ok I will try to stay around the 10 meter per minute for starters.

- ok, will play around a bit with teh micorstepping ( was already planning to do this - that's why I bought them - but your values help for staring me somewhere)

- I now understand the Frequency a bit better, and in USB CNC - now named EdingCNC there's only an option to set MAX Frequency. I read in the manual that this is done, because some drivers will not be able to run higher input frequencies. The Maximum - MaxFrequency is the 125KHz of my board, and my leadshine driver states in the manual it can accept up to 200KHz max input Frequency. The Eding CNC board differs from Mach 3 I think with it's on board processor, so it does not rely on the PC processor connected to it for the pulses ( which can be the limiting factor with the parallel port if I understand correctly).......So........ is my conclusion correct that I can ignore the Minpulse rate from Geralds spreadsheet unless it becomes a value higher/close to the max 125 Khz value ??

Cool to read you also have been playing around with the demo. I have heard/read a lot of good stories on EdingCNC and seen it function on professional machines, but the Mach3 user database is currently a million times bigger. Most people I see using it now are Dutch. But other then that I agree with all the advantages you state, that's why I chose it over Mach3.


Both your advises taken into the spreadsheet I get this result:

http://s21.postimg.org/a5u9jniav/Microsoft_Excel_Gear_Speed_Steps_Freq_Calculat.png (http://postimage.org/)

I think I will start with this situ and then play around a bit, when it's up and running.

[danilom]

Aim for around 20.000 mm /min rapids (thats around 800 IPM)

Microstepping can not improve resolution, just smoothness which will you have enough at 10 microsteps

smaller pinion can affect resolution.

What is it you want to achieve. "max happy motorspeed" what is that? Under no load you should be able to spin at 1000 rpm. Depending on the pinion it will produce certain feedrate.

Question is what FEEDRATE you want? you can use any pinion and limit the feedrate so you stepper spins reliably if you don't care about feedrate.

[Fox]

Hi Danilo,

Thank you for your comments, but I must admit I am getting more confused. :confused:

Simply put: what I am aiming for is my MechMate performing at it's best capacities ( in speed and accuracy ) when cutting plastics and aluminium, without over stressing it (reliability). I assume it's what we all aiming for ( although some might be more into speed then accuracy )

Working hard is ok, that's what it needs to do,without breaking, missing steps or loosing accuracy because components get too much stress. My feedrates will depend on this and the materials imo.

Maybe I am over analyzing, but everything is quickly tunable without cost, except for pinions. These need to be then bought in a different size, machined and replaced on the machine. The total of that will not be the end of the world, I just thought to properly aim before I shoot the first time.

So tried to make a prediction with the calculator on what would work best. Then things came up I did not know: max happy speed for MechMate to work reliable ( because of the design /structure ), max motor speed, max/best micro stepping settings.

So all my questions and then your comments on those followed and here we are with more questions assumptions:D:

- Do I understand your 20.000 mm/2 meter minute in "rapids" = means straight line full speed ? ) Is that a general cutting speed ( I know it also depends on materials and bits) or a Jog Speed ? I see this mechmate (http://www.youtube.com/watch?v=EwO_KL1GRww) is cutting the spoilboard at 1150 IPM = 29210 mm/minute (!) which is extreme I think, but i also see much more considerate speeds.

- Driver Microsteps did affect resolution in the spreadsheet but I did not read correctly it is microstep resolution that is affected, I see now. I understand this is different from the actual resolution ? I thought it meant with 1 microstep the machine would travel 0,006 mm ? And I though the lower the number the better your accuracy becomes, and that one can influence it with microstep settings or pinion size ? Am I making a fundamental thinking mistake here?

- Happy motor speed = I want my motor to be happy at the best max speed of my machine. Meaning a stepper motor in general maxes out at about 1000-1200 rpm I found somewhere on google, because after that torque becomes to low to be useful ? But it was also written that some steppers might make it to 1000 rpm, but will not be useful at all anymore. So my thinking was; does anybody know the maximum RPM for my steppers to be still useful. Because then I could fill in the the other parameters in the spreadsheet based on that ?

- Feedrate I want: the best feedrate possible within the MM capabilities/my config.

Anyway .... some scenarios at 20.000 mm/second with either 10 microsteps or 16 microsteps all give stepper rpm around 1000 rpm, so that would be the max speed If I follow my google assumptions, if I want to go faster I would need bigger pinions ?? :

http://s16.postimg.org/dogikqg7p/Microsoft_Excel_Gear_Speed_Steps_Freq_Calculat.png (http://postimage.org/)

[danilom]

20.000 mm/min rapid positioning G0 or jog at 100%
for cutting you will use much less, leveling the spoilboard is easy on the machine as it only peels a thin skin... just to level it, usualy 0.5-1mm

microstep only increases Steps per mm, that is the number of STEP signals needed for motor to move the machine 1mm, thats theory. In a real world you will have no less than 0.05mm positional accuracy or repeatability. Consider also the non linearity of a rack, even ballscrews have their accuracy for a given lenght.

[KenC]

Nobody know or at least nobody that I know of took the time to plot their own Torque Vs Speed curve at the Voltage & microstepping condition...

you are right in saying its useless knowing the "happy speed" as tuning is free & trivial.

In real life, microsteps doesn't increase resolution as far as the number said. ~8 is about the best before you hit the point of diminishing return. so many just use 10. The real resolution improvement comes in the form of mechanical speed reduction. The million dollar question is: are you sure you realistically expect your wood or plastic's thermal expansion is below 0.05mm?
Than again, when you step your motor speed down, especially with high ratio step-down, you are effectively running your motor into the low torque region of its Torque-Speed graph... & if you are unlucky (which real world happens a lot :( ), you may end up with lesser torque than w/o speed reduction... tough...

[Surfcnc]

Fox

Correct on the step frequency as long as your gear step calculator says you are under 125Hz you are good to go.
In practice you will never get near that frequency with your steppers and belt drives.

Ross

[rischoof]

fox,
I had the same with my energy chain, it was not straight.
I didn't like it, so I made them straigt.
get out the yellow buttons en disassemble the chain.
I removed a little material with my table saw from the black part
it's small ammount I removed and after a copple of parts I re-assembled and looked if it was flat. I had to re-adjust the table saw a little and then the ammount of material I removed was OK to make the chain flat.
finally I didn't install the yellow bottons back

[Fox]

@Danilo, @Ken, @Ross, @ Rik:

THNX guys, much appreciated, between the 4 of you, you have answered all my current doubts/questions enough I think to be moving on to the next steps !

Ordering rack and pinions tomorrow.

Gantry I hope to finish begin of next week.

I try to order my cables and some connectors this weekend ( have to do some research ).

Kitchen table project starts this weekend. I will see if I can fit in the rail grinding first week of July, and have the kitchen table project finished by then. That would mean a final push in the second week of July to get it all assembled and aligned and making the first cut, and then well-deserved holidays with the family ( they are lucky we go abroad so I can't play with the MM - and maybe it's better for me as well :))...

Upon my return play with it for a week or 3, and then disassemble for paint ( so I can still do that outside ).

Wel... that's the plan, let's see if life/work does not get in the way...

[Fox]

Started on my modified-Z 3d drawings today to work out a teahnical drawing and order sheet:

1)2 ground slabs of alu, indexed& milled for parallel linears with tapped holes - check -
2)2 bought 20 or 25 mm linears guides + 4 carriages - check -
3)Ball screw - question below-

1) What have those of you who are using a ballscrew for your Z used ? - Does a 25 mm ISEL with 5mm Pitch sounds like a good plan for a 700 mm extended Z ?

[rischoof]

fox,
25 mm is big, use smaller diameter. the weight of spindle has te be accelerated and slow down, thinner is lighter. will be easyer to turn from left to right for you stepper motor

700 mm? I don't know why you need that length. I had a spindle 50 cm stroke, total 700 mm length, motor on top. the point of gravity is to height. when mounted on the Y car. I can see that it is shaking. i don't see it in my cut quality, but i can see and feel that it is shaking. that's the reason I built a shorter spindle, and keep my point of gravity low

[Fox]

Hi Rik,

I have a Nema 34 ( same like X and Y planned ) for the Z ( a bit overkill, but already got it ) so I think it will work with the 25 mass ? I mean ... I was chosing 25 over 16 because it would be stiffer introducing less whip on the 700 mm Z. Still think I should go smaller ? What are you using for the pitch ? 5 mm or smaller... I was in doubt on that pitch for increasing the resolution on the Z.

I want/would like the 500-700 mm for future projects ( like foam cutting under the current level of the spoil board ). In normal position it behaves the same like a standard length Z, it would just point up in the air. But it's interesting to learn from you that the Y-car is not up to the job of holding it up steady and straight. I would think that it would, when you tighten make the Z slide of massive alu fit exactly in between the mounting points. Are you sure the shaking is not from the Z itself, but the Y car flexing ? That would change my game plan considerably. I thought the only 'weaker' point in the MM Y&Z was the Z slide with the bent spider, whcih can be improved upon with a custom Z.

Was your 700 mm Z an modified Z with linear guides and screwballs ...between two massive 15-20 mm aluminum plates ?

[rischoof]

http://www.mechmate.com/forums/attachment.php?attachmentid=12923&stc=1&d=1327290909
this is my first Z as the slide is fixed to the y cart. and the unit is moving up and down. the motor is mounted on top an massive aluminum mounting block. this makes my spindle top heavy. this is an aluminium frame with single row bearing on each side. the weight the motor has to move up and down is 18 kg and when the motor is powered off, it just will stay in position

with my new spindle I am trying to improve. so the axis is shorter, stroke is still 30 cm, my ball screw spindle was not shorter.
This time the spindle is mount on the slide, so the unit is mounted on the y cart instead of the slide. motor is mounted lower than in my previous one.
the weight the motor has to move up and down should be lower, I expect Half.
I will upload some pictures on my building log from my new spindle

[Pajka]

you don't want to go lower than 5mm pitch as you will loose a lot of speed.

///Danilo

[Fox]

@ rik, interesting to say the least.. i'll email you for some more pics and details. My plans where pretty much in that direction dimension wise and layout wise, as the linear rails will be pointing away from dust, and they stiffen the longer slide...., Although my calc did not accumulate to 18 kg, it still is much higher then moving it the other way round. The nema 34 weighing in at around 4 kg does not help when mounting it on the moving part of the z.

@ danilo with the standard mm layout the z has the same speed as x and y, but with a 5 mm pitch ball screw ( taking into account stepper maxing out around 1000 rpm ) it means a max z speed of around 5000 mm per minute, so what would one actually need as a minimum z speed to be not limited in functionality !?

[rischoof]

fox, in my building log you can see more pictures from my spindle.
for 3 d, you z has to move quickly. but I don't have experience with that. I can play with my speed just changing the ratio from the belt drive between stepper and ball screw spindle. I didn't mount the stepper directly on the ball screw spindle but with belt drive, I used for my previous one 1:3 (motor small, ball screw large) this because I need some torque to hold my 18 kg in position. I have a gas spring laying around but didn't install it.

[danilom]

If you plan to do 3D work then all 3 axis should move in similar speed or you will lose performance because other 2 axis will slowdown for Z. In 2D cutting there is no such problem and 5m/min Z should be ok.

I used 20mm dia, 10mm pitch ballscrew recently... from Hiwin .. must say its fantastic but the ballnut alone was around 180 eur

[Fox]

Hi danilo, makes sense. So again a trade-off between speed and accuracy. If I would take the 5 mm pitch the whole machine would have to slow down to 5 meter speed for 3d work, while doing 2d it is not an issue....

BUT especially with 3d work resolution matters, so I wonder where the compromise lies, don't you sacrifice too much with a 10 mm pitch on your z ?

As with everything I am looking at industries examples ( not that that bif manufacturers are always right but still ) and I notice pitches around 4-5 mm are used on z drives most often !?

[danilom]

If you ever looked at some big iron machines, routers, lathes mills
you will find really large pitch ballscrews. like 20mm pitch

its because of high count encoders on the screw they get the precision not the pitch.
How do you think a router can achieve 80m / min with a 5mm pitch ballscrew? spinning it at 16.000 rpm ?

I always count my expected POSITIONAL ACCURACY by using a 400 (halfstep) resolution. A stepper motor will always tend to get close to a full or half step.
So with a 5mm pitch direct 1:1 belt you get 0.0125 mm accurate positioning , with a 10mm pitch its 0.025mm.
For me a 0.025 positional accuracy is fine even for metalwork. Most machines will flex more.
A backlash on chinese ballscrew is around 0.05 to 0.10mm for comparison on how precisely you can rotate a stepper motor.

[Fox]

Hi Danilo,

Thanks for your fast reply !

1)
I was talking about commercial machines comparable in size and function. Those machines you are talking which about I indeed know ( I work with companies that have 1 or more of those ) ... But they use servo's and ballscrews of 40 mm diameter with rotating ball nuts for X and Y ( you don not want to know the price for those ) and more of these exotics. They need reliable production speeds and accuracies well beyond what the average MM frame is capable off or intended for. So not comparable with the MM imo.

The lower to medium end commercial machines I was talking about, I noticed often using 10-15 mm pitch or higher for X and Y ( when ballscrew) and around 4 or 5 mm pitch for ballscrew Z slides. I think the shopbot for example uses a 5 mm pitch Z, (maybe not the best reference ).

2)
Your last paragraph; calculating with halfstep resolution is a helpful insight;

10 mm calculated at halfstep obviously is 5 mm reolution at fullstep ( which is what I was doing ).

But on the choice you made for 10 mm, are you mostly doing 3d work or 2d work - and did you apply this Z to an ( or your) MM ?

[danilom]

No ballscrew on my z , but I have some 5mm ballnut around which I might mount.

I used 10mm on a cnc drilling machine made recently. Mostly because of its 2500mm table which had to rewind back after drilling, 10m/min .... compared to 5min/min means time=money

They drill around 350 pairs of pipes a day so 350 x 15 sec = 1.5hr lost on rapid positioning more for a machine to get back

[Fox]

Thanks Danilo,

That situ and axis 10 or higher pitch makes sence.
I will give this some more thought, but probably stick with my original plan for 5 mm pitch for z.

[Fox]

Not dead yet - and instead of hijacking Tokamak's thread any longer :D - here is a sneak preview what I have been up to lately ...

The build had been delayed by paid jobs and life in general, and I still don't have enough time to keep the thread updated regularly, cause I am putting all my spare time in the machine again lately... So don't expect much more from me until it's actually running and sorted.

Here we are about 1 month (about 40 hours I think) away from the finish ( my best estimate).

My Z is almost finished ( completely custom design ) bolting together as we speak.
The table is unpainted but finished ( Rail have been cut, gantry has been rolling e chain mounted).
Controlboxes are wired and motors and spindle have been have been turning.

To do: all little things that eat up sooooo much time, the final electronics (proxies etc) and wiring up the entire machine (pull cables).



Ps.
Moved img to next post as forum was not cooperating ( with the 5 min edit limit)...

[Fox]

Grinding L profile to 28 mm +/- 0.1 mm, making good use of flat spoilboard and Geralds built in clamping space.

http://s30.postimg.org/fp5jf3hm9/grind_height.jpg

Grinding the V-angle on the L-profile to +/- 0,1 mm - first with 36 then 80 grit sandpaper stuck on 2,5 mm grinder disc with double sided adhesive tape.

http://s4.postimg.org/a5n9zxf19/grind_V.jpg

Random image before mounting ground rails to machine, (jacked on wooden blocks) with E-stop boxes in view.

http://s16.postimg.org/7l7yby2qd/machine.jpg

Using two control boxes ( power and control hardware divided ) this is PWR control box, using DIN rails to mount all goodies DIY 1000VA powersupply, got lucky on ebay with unused siemens VFD (60 USD !!!)

http://s28.postimg.org/i4laz1ril/control_power.jpg

This is the DRIVE controlbox, using Eding CNC (former USBCNC) CPU 5B ethernet.

http://s30.postimg.org/4x7z0025t/control_drive.jpg

This is the CAd of My 100% custom Z-slide with 55 cm of stroke. 35 kg of pure steel and aluminium (!) All parts have been made, bolting together but no picture at hand.

http://s30.postimg.org/9apckrldd/z_slide.jpg

[Tokamak]

I look forward to your comments on my page each morning. Good to see that you have posted some updates. You are way ahead of me on the electronics. I wish I had my motors turning. Nice compact box layout.

I read somewhere about VFD electrical noise causing issues in the control box. Not sure if there is any proof of that. But after reading it I decided to mount my VFD away from the controls. Still need to find a box for that.

[Fox]

That is why I mounted all power (noisy) away from all control electronics. Each have their own metal cabinet on the machine. My Eding CPU5B with ethernet (http://www.edingcnc.com/products.php?productid=51) is also much less prone then mach3's parallel port to this noise.

The Eding CNC company advertises in their manual with machines with vfd in the same box without problems so it might be overkill, but better safe then sorry imo.

I have two boxes which are each 60x40x25 and my setup is now way fuller then I anticipated when I started... bigger controlboxes are always better :D But this is all I could fit under my table ( they will be on the far end of the X under the spoilbaord mounted in rubber )

[rischoof]

Fox,
Nice to see your progress. nice controlbox. some remarks for your z axis.
Is the green slide including your spindll going up and down, or is the purple plate fixed to the y car. In that case the motor has to pull 33 kg up and down. there will be a chance that if you don´t power the motor controller your z axis will sink down by its weight. can be avoided by choosing an other gear ratio between stepper motor and spindle, or mounting the gas spring
I builded almost the same concept, but when I start running the machine the z axis started vibrating when running curves. it was to heavy in the top. finaly I made a new, shorter one, and this one has no vibrations.
think allready where you want to place your limit switches and home switch, and drill the holes in advanced. save you some time in assembling and disassembling the z axis

[KenC]

Nice to see another ball screw + linear guide Z-axis.
If I were to do mine again, I would couple my motor to the screw with timing belt like yours.
BUT I would have the motor mount at the back to gain 80~90mm of Z-travel.
Mine is direct coupled, the alignment took a bit to time to get right & I feel the motors is to high in the air.

FYI, In a control penal 6" or 150mm is consider "Far Away from". 2"/50mm is marginal.

[Fox]

If I get a few minutes this weekend I might even post a picture IF I finish assembly of my Z.

@ Rik,

I have an electromachanical brake on my ballscrew (blue cirkel topleft).
It is enabled when there's not power to my drives, and when there runs power to my drive, there runs power to my brake, which is then released.

The mass is a worry I had, but I have a backup plan for that already that will eliminate it for such a problem for sure... so we'll see..

The entire Z was drilles and tapped 100% by CNC taken into account all mountings like proxies etc in advance. It was made accurate to 0,01 mm... I was lucky to have some friends in the right places. (the entire assembly still costs me a lot of money though).

It's really lego. But the stupid thing is i have all parts but I am now waiting for all the right bolts to arrive.

@ Ken
the rails on my Z are on the moving part (stiffens) so you can't move the motor to the back and gain travel... you would actually loose travel ( when the motor hits the fixed plate). I had thought about a rotating ballnut, but thought it was too complex for now.

All my alignments of rails and wagons are taken care of according per Hiwin manual, so just bolt it down and lock it against the shoulders should take care of that.

Last note about teh control panel I don't understand

[Fox]

Here's some bigger pics of my Z.

I have even built in the possibility to tilt my Spindle a true indexed 45 and 90 degrees. Not that I need that straight away, but can be helpful to do some crazy stuff in the future ( as I do a lot of crazy stuff :D )
Also I built ins some basic dust protection for ballscrews and wagons. I have the option to go fully enclosed in the future with bellows.

Spindle is my big friend (380 Volt 3,2KW)

http://s15.postimg.org/4q3tt7ow7/Z_big_1.jpg (http://postimg.org/image/4q3tt7ow7/)

http://s15.postimg.org/rp0jc4kw7/Z_big_2.jpg (http://postimg.org/image/rp0jc4kw7/)

http://s15.postimg.org/5455st8zr/Z_big_3.jpg (http://postimg.org/image/5455st8zr/)

http://s15.postimg.org/hg826awuf/Z_big_4.jpg (http://postimg.org/image/hg826awuf/)

[KenC]

I was referring to the spacing of components inside the control penal.
BTW, I don't have any breaks on the screw, it won't go anywhere.... Mine didn't move a bit... and even if it moves, it would be the ball park of 0.1mm when power down. Nothing I can't live with. There will be a handsome saving on whatever breaks you can come up with.

[KenC]

BTW, my ball screw Z-axis. here (http://www.mechmate.com/forums/showthread.php?p=51455&postcount=436)

[Fox]

Hi ken,

brake is for peace of mind, when I hit E-stop I am sure my brake keeps the Z where I want it, for example it won't drop on anyone else his fingers while the spindle slows down.

I still don't understand the control box comment, you mean my spacing is too small ?

[KenC]

2 reasons I went for the ball screw Z were
1) more rigid (& it actually is)
2) no need for gas spring or what ever for holding... (3 years & nothing drop except for loose spindle clamp because I forgot to tighten the screws...)

Think about it, stepper motor is named such because it move a step at a time. & we "perceive" it is moving continuously when actually it is a strings of steps at different intervals. If and when motor power is cut, unlike other motors, stepper motors makes no more steps & stops abruptly...

No, didn't even wipe it clean after use....

I just enjoy simplicity. too lazy to work on extra precautions which won't happen, most importantly, I don't have the budget to over do. :p

As with "bigger is not always better", "more is not better". Its just me, anyone may do whatever they like.

[KenC]

As for the control penal, I'm trying to say politely your spacing distance is enough. Any more and you will be way over the top.

[Fox]

Now I understand, and I totally agree my z is way-over-the-top-future- proof, but I have my reasons :D. As with my control boxes.

Ps1.
Of course your linear z is also already "overdoing it" compared to Gerald's plans... I just took it to the next level. I am happy with it, but its by no means meant as an advice to follow for aspirant builders. Stick to the plans will be good enough for most.

Ps2. when there's no power to your stepper it won't hold, it will have some magnetic resistance in a rotation, but that's it. You rely on the friction in your rails/wagons, ballscrew/ballnut, and drive ratio to hold your z up, which in most cases i agree is indeed enough. I have also never used the airbags in my car, I had to pay extra for them when I bought it:rolleyes:.

[KenC]

More space in the penal is actually a good thing :D
Honestly, My Z-axis is considered minimalist in comparison :D could had go for smaller rails & thinner aluminum base plate. BUT I don't consider it over-doing, it is a rational improvement.
Why external break is not required.
1)The stepper motor do have detention torque & it is fairly significant just have a look at your stepper motor specs., which is another unique stepper motor feature.
2) the ball screw has higher mechanical ratio in-comparison to the original MM rack & pinion setup minus the weight of belt reduction.
3) My slide assembly (alone) has a static friction which require 16kg pull force to overcome. (I'm sure you will measure yours after you assemble yours).

Conclusion
I have one working hassle free for years :D
I'm saying all this in good intention to tame your worries & avoid unnecessary expenditure.... electro-break is additional cost, the aux equipment to make it work are also additional cost. Maybe all this saving will pay for your A axis or a second gantry.
I rest my case.

[Fox]

Appreciated as always, ken.
'Problem' is I already have my brake, and the aux controlling it...as well as an a-axis :D

[Tokamak]

Fox,

Revisiting our conversation about smoothing the V Rail to remove variations I mad a measurement.
Installing a dial indicator to measure the variation of one when I measured + and - .010" (.25mm).

Seems to be worth checking. Easier to check after the Gantry is rolling.

John

[Fox]

Are you referencing to my other topic (http://www.mechmate.com/forums/showthread.php?t=4506) about preloading the y rails, to compensate for the bending of the tubes due to the weight of the y car ? Or what did you actually measure ?

[Tokamak]

Fox,
Sorry for the confusion. This was referenced to post #159 of my build.

http://www.mechmate.com/forums/showthread.php?t=2526&page=6

We were discussing the need to sand the X V-Rails with a longer V-Block to reduce variations from the grinding. The dial indicator in the picture is measuring the variation in the rolling height of one V roller as it moves in X. The change in height is due to the inconsistency in the cut of the bevel on the top of the rail.

John

[Fox]

Ah, yes, I see now. I did use the v block !

[Fox]

Hi guys,

I am very close to firing up the beast to make a first cut, but have a quick question:

I have powered up my steppers and jogged the machine for the first time. Already felt great to see it move under own power !

But amongst some other things i need to sort, I noticed that with the motors enabled, but not moving (sitting idle) they make a hissing/stuttering sound, not really loud, but enough to be heard when standing next to the table. Is this normal ? It seems not ok to me, but it's my first CNC so I can't judge. When jogging/moving it seems to diminish/be overhelmed by the other sounds.

I use 86HS9801 (http://www.driver-motor.com/ProductContent.aspx?ID=47) motors and leadshine AM882 drives (http://www.leadshine.com/productdetail.aspx?type=products&category=stepper-products&producttype=stepper-drives&series=AM&model=AM882).
I feed them with 71 volt.

[danilom]

Tune the drives to motors by Leadshine software and the RS232 cable provided with drives.

Hissing to some extent is ok. it should not be loud just audible when you get very near the motors.

[Fox]

Ok, I will see into this, I autotuned them by flipping the dipswitch (4 I believe) and thought that might do until now, cause I don't know what I need to do to tune them.... but will look into this.

It is audible, but no loud, but also not only when I put my ear near them. I think from about 2 meters away when there's no sound in the room. In this room I can also hear the relais click in my controlbox, when there's no other sound. (it's empty besides the MM with stone walls)

[Fox]

Hey Guys,

another issue:

My Sicce Nova (http://www.sicce.com/prodottiDettaglio_eng.php/prodotto=nova/idprodotto=18) impeller pump, benchtested perfect, but now struggles to pump the coolant round. It barely makes it, and there's not enough flow left at the end of the lines to keep my spindle happy I am sure. I am using 8mm od (6mm id) Pu tubes. Table is 2x3 meter Z = 80 cm.

So what do you use/should I use ? I would like to try and keep the hose size relatively small ( 20 mm hose would fix the problem maybe, but won't fit in my energy chains i think, and there's the restraint at the spindle anyway with those tiny inlet en outlet of 4 mm. Basically I think I need a higher pressure pump to pump coolant trough the system ?

I'd prefer a submersible pump.

[danilom]

This is overhead chart, your pump is declared at maximum 1.5m where it will pump about 50L an hour "maybe" . Comparing it to my atman 104, 38W pump you need at least the "Extrema" model since its 32W pump. and has a maximum overhead at 2.5m

http://www.sicce.com/backend/prodotti_sicce/img_upload/img_tab1_3/1008171155130_nova_curve.jpg

[Fox]

thanks for your quick help, I think I will even go to their syncra line then, because my line restriction will make the flow considerable less I imagine, I see on alibaba a lot of 75 watt pumps are used at 3000 l/m for the spindles, where they screw a copper inlet restircion of 6 mm on:

http://img.alibaba.com/img/pb/125/000/847/847000125_751.jpg

So maybe their syncra 3.0 ?

http://www.sicce.com/backend/prodotti_sicce/img_upload/img_tab1_3/1008171228130_curve_syncras_20-25-30.jpg

[Tom Ayres]

Fox, I mounted my small cheapy pump right on the y car to avoid running into all of the problems you have stated, I really didn't have space on the energy chain to run a line, it hardly weighs anything, I ran 12v to it so far it works great. I don't know if others have done anything similar or if it will last long but it keeps the spindle very cool.

[Fox]

I contemplated that solution ( still am as a makeshift solution - I am sur ethat woudk work, as I beanctested it with 10 meters line or so )

But in then end I would like to get it off the gantry. Reasons are it's already very heavy in my machine, and cluttered, the slushing etc. Plus I already have everything in place. Just need a different pump now :-(.

What kind of reservoir, spindle, and pump are you running on there, and do you use a radiator ( I try to avoid that, os I do not have it laying around here now; another expense ) ?

I have a 3,2 KW 380 volt spindle btw. So above MM average.

[Duds]

Hey Fox, Stick your pump in a bucket of water and hang it from the ceiling already and make that YouTube video ;)

[Fox]

I promise the video and pics will follow soon :D but this is not the only thing I need to check before firing the beast up.

But I am less then a week away for sure. EXCITING

[Tom Ayres]

I do have a fan and a radiator also on the gantry. The pump has a reservoir built-in, not very big something like 150 mL or 250 mL

[KenC]

No radiator, just a industrial coolant pump & reservoir as per any lathe & milling colant set up.
I reckon even with its good heat insulation property the long PU tube (8~9m to-and-fro )already dissipated enough heat from the water even without the heat mass of the reservoir.

[Fox]

Hi Tom, yeah I though so, so the in between is ,without extra expense a no go.
@ Ken, good to know, that's what I was aiming for as well with my intended trough e chain setup.

On the hunt for a pump that can be delivered tomorrow or at least before the weekend as we speak.

[Tokamak]

Fox,

Put the pump and reservoir on a table next to the machine at the same height as the spindle. The pump won't waste pressure lifting the water.
Put a thermocouple in the outlet hose of the spindle. A lower flow will allow the water to heat more on the way through the spindle. A higher flow will not heat the water as much on the way through. In either case the water is cooling the spindle.

More flow decreases the outlet water temperature but doesn't necessarily cool any better.

Heat loss (watts * .239) = cal/sec
Flow (GPM * 3784/60) = grams/sec
Water temperature rise= (cal/sec)/(grams/sec)

I play an engineer on TV.

John

[Fox]

Put the pump and reservoir on a table next to the machine at the same height as the spindle. The pump won't waste pressure lifting the water.

That was when it barely made it ;-) when I put it in position it did not make it go round. New pump already ordered !

Put a thermocouple in the outlet hose of the spindle. A lower flow will allow the water to heat more on the way through the spindle. A higher flow will not heat the water as much on the way through. In either case the water is cooling the spindle. that makes sense, but it was really not flowing enough to cool.

Thanks for the formula's I will run them.

[rischoof]

fox,
had same problem with my pump. do you still have the problem when the complete line is filled with water? I filled the hose wiht coolant with a "trechter" when I closed the end by folding the hose double and close it temporary with a t-rap. if you can place the resorvoir and pomp on a higher postition it should start pomping arround the fluid and get the air out.

[Fox]

Yes, I did a similar thing, but then it barely pumped round, bigger pump is on the way, but will be here somewhere next week. This weekend I hope to test it without spindle but with pen. Will post some progress.

[Fox]

WOHA ... I created a monster ... albeit an intellectual one :D

http://s29.postimg.org/klittb1hj/IMAG3317.jpg

http://s15.postimg.org/ybrsipcff/IMAG3321.jpg

http://s10.postimg.org/ubpgo1cw9/IMAG3316.jpg

[Fox]

Got led on my gantry, mister, traffic light, lasers, Z-brake, got air connections on all legs, plus sockets with 220 volt, all buttons work... basically, I got everything working .... except the spindle (well that runs as well, and I can control RPM etc, but as you know but I have no cooling ( better pump arriving next week) .....

So.... I opted for a pen (flipped the spring - for a cheapskate plotter )

It felt great to see it moving. Need to sort out all kinds of little details, but hey it works !!!

I AM OFF TO A CELEBRATION BEER.

Big thanks to Gerald for the design and to all of you for the advise and support.




Ps.

For an intellectual monster you need:

Brains (Leadshine AM882 drives, CPu5B EdingCNC Ethernet, DX 24 volt (5v ctrl) Relais board, Eding CNC BOB, custom built opto iso sensor interface, DamenCNC mist/coolant bob + Z brake bob)

http://s2.postimg.org/bqp1jpt6x/IMAG3072.jpg

Muscles ( custom 1000 watt 71 V toroid powersupply, 380V Siemens Micromaster, Meanwell 12 + 24 volt powersupplies, 24V contacters, 32A mainswitch, 16A and 1A circuitbrakers)

http://s29.postimg.org/bzbxl9ah3/IMAG3073.jpg

[IMMark]

Happy times Fox...congrats!
Mark

[Duds]

Yay! :d

[Alan_c]

Veels geluk! Number time... :-)

[MetalHead]

Congrats !!! That Z-Slide is interesting. Can't wait to see it doing cool stuff. Your beast have logos on it ?

[racedirector]

Congrats! A stealth build, well done. Dust very very soon by the looks.

[Fox]

Thanks guys, no... it has no logos and no paint yet.
Will look for a sign shop an cut some.
First cutting will happen when the new water pump arrives for spindle coolant, should be here coming week.. In the meantime I will use the pen to get settings and files prepared, and grow some confidence :)

[Tom Ayres]

Congrats Fox, Well Done! By-the-way you can run your spindle for a few minutes safely for a test cut, mach has the 'roadrunner' test file you can run under a minute.

[Fox]

Yeah I know, ( not using mach btw, but edibg has similair files ) but my instincts tell me its not bad to have to wait untill the spindle to double check everything. Already found some loose bolts here and there, adjusted gearboxes, greased my rack and pinion etc. So you all have to be patient with me some more ;)

[rischoof]

Congratulations with your good looking machine. Is ashame that you have to disamble the whole machine for the paint job.

[Fox]

I'll survive, I planned it this way.
When you build a custom motorbike or car it's common practice.

[Tokamak]

Congratulations Fox. Those last few details can be the toughest. Now you can enter phase 2 and figure out how to run the software. Happy cutting!

[darren salyer]

Well Done!!

[Fox]

Now you can enter phase 2 and figure out how to run the software.
Yeah, mostly got that figured out, I have years in CAD under my belt, I just need to figure out some quirks in the Eding software here and there which have me puzzled, but finding correct cutting parameters and hardware for the range of materials will be a challenge for sure. But looking forward to that.

Yeah, it's never finished I know...
Still have to built Dust and Vacuum table ( have motors already in stock ) after it's cutting those will be some of the first parts of the machine ( after a bed for my little son, that will be the first ''real' part after al test have been finished.

Thnx Darren !

[isladelobos]

Congratulations, looks strong.

The motors are hot on 71Volts?

[Fox]

No... motors are 45 degrees ( on the outside, I have no means of temping the inside) after running 3d files for one hour. Plesse note air running - not cutting yet, so teh motors are not really pushed to their limit yet.
I listened to Geralds advice and exprience, because I want max performance out of my machine. There are many running lower voltages. There have been heated (:D) debates on this mattter, search the forum.

[Fox]

I need advice, what would you guys advise on setting my velocity and acceleration to considering my large table and my electronic specs (see thread (http://mechmate.com/forums/showthread.php?p=69437&postcount=186))

[Tom Ayres]

My acceleration is set around 25 but the speed is set above 600 (can't remember exacts). I found the machine will beat itself to dead on direction changes if acceleration is much higher.

[Fox]

25 inches and 600 inches.. I assume ?? then I was already jogging above that :o.
I am in metric = velocity in mm/s, and acceleration in mm/s^2.

[bradm]

Go as fast as you can go :)

Start with 25 / 600 ipm ( 10 / 250 mm/s, right? ). Increase your speed until the motors stall during accel or decel. Tweak your accel down / up until the stall goes away, or drop the speed by 5% and tweak again if necessary.

Then, drop your speed by 15 - 20%, and test violently changing direction in x,y and diagonal and ensure no lost steps. There's your max speed. If you have lost steps, you may need to drop the accel a bit.

Let us know what you end up with. Always useful to have another data point.

[Fox]

Ok, I'll start playing.

[Tom Ayres]

There you go. That's good advise, when initially set up I just stopped adjusting thinking it was good enough for now, I too might go ahead and increase the speed.

[KenC]

Its good to write a short G-code commanding the axis to to go to-and-fro for the tuning
for Z-axis
G1 z0 F(desire speed in mm/min)
G1 z200
G1 Z0
G1 z200
G1 Z0
.....

[Fox]

Ah i was thinking, there must be a test program to do this. I found emc has it ( a program to find max velocity and accelaration).

Is this gcode you write (i am not familiar with programming g codes manually) something it could copy paste in notepad++ and save as, and run ?

[racedirector]

Fox, pity you're not using Mach, there is an add-in called the Break In Wizard, it generates gcode to exercise the machine multiple times....

[KenC]

Fox, I don't write my own G-code per say, got aspire to do the hard labour. I reckon its always good to know the G-code well enough (or cheat buy having a g-code pocketbook) for occasional troubleshooting.
yes, you can use a wordpad or notepad to write one & save it with .tap extension for Mach3 & .ngc for Linuxcnc.

[Fox]

Fox, pity you're not using Mach, there is an add-in called the Break In Wizard, it generates gcode to exercise the machine multiple times....


This one ? (http://www.machsupport.com/forum/index.php/topic,13915.0.html) That would have been nice. Courtesy of a huge user base, Eding does not have that yet. But I will see if I can convert that for use with Eding CNC. I am glad to get rid of the parallel port and be running Ethernet though;).


@ Ken, yeah I know, let I was more asking if:

G1 z0 F500
G1 z200
G1 Z0
G1 z200
G1 Z0

G1 x0 F500
G1 x200
G1 x0
G1 x200
G1 x0

G1 y0 F500
G1 y200
G1 y0
G1 y200
G1 y0

is a valid G code file when saved, or do I need to add headers , closing brackets etc ? I am familiar with coding html, java and css not G code..

[racedirector]

Yup, thats the one. Works pretty well, I used it on my previous machine.

[Fox]

I think I need a good test file. My values seem way too high according to advises ?

I had my machine set at X and Y @ 500mm/s vel and 250 mm/s^2 accel.
The Z is @ 195.3mm/s ( it made that number from 200 as input ?) vel and 150 mm/s^2 accel.

It jogs fine at these speeds, and it 'cuts' (writes) fine as well with some dummy programs.

[ger21]

Are you sure those numbers aren't in mm's/min?

250mm/sec/sec accel is about 1.5 G's, and is about 6-8 times quicker than the quickest home built machines.

[servant74]

https://en.wikipedia.org/wiki/Standard_gravity says 1G is 980 cm/sec^2. So 250mm/sec^2 would be roughly .025 g, but I could be off. (Changing units always got me in physics.)

[ger21]

My mistake. I was looking at his feedrate in mm/second, and multiplied the accel by 60 by mistake.

Let's try this again. I think that 500mm/sec is 30m/min, which is very fast. But an accel of 250mm/s/s is much too slow for 30m/min.

[KenC]

[URL="http://www.machsupport.com/forum/index.php/topic,13915.0.html"]

@ Ken, yeah I know, let I was more asking if:

G1 z0 F500
G1 z200
G1 Z0
G1 z200
G1 Z0

G1 x0 F500
G1 x200
G1 x0
G1 x200
G1 x0

G1 y0 F500
G1 y200
G1 y0
G1 y200
G1 y0

is a valid G code file when saved, or do I need to add headers , closing brackets etc ? I am familiar with coding html, java and css not G code..
You are definitely on the right track.
I don't quite get what you meant completely. its an ASCII file if I remember correctly. damn, its been so long ago when I last did this., :p
I suggest you open a simple G-code for Mach3 & look for your answer.

Also,
1)its better to split the above mentioned G-codes into 3 separate files; one for X axis, one for y-axis & one for Z-axis. Its always advisable to tune one axis at a time, will save your sanity. :D
2) Should add plenty reciprocating strokes, the above codes only entails 2 strokes. IMHO, 20 strokes is barely enough.
3)Also, have a lot more strokes length in each test file, since you are building a 1.2mx2.4m table, you should have the X & Y strokes a lot longer, at least 1m & 2m strokes respectively.
4) have one hand on the E-stop when you start the G-code. :eek:

Its a great time for you to have an hands-on feel of what the speed & acceleration numbers entails. While you are at it, keep an eye on the raping time too (0 to full velocity) IMHO, 0.1sec should be fine, but don't cast this number on stone....

[Fox]

Thanks to all, will built me some proper test file, and slow things down before I try to built speed again..

[Fox]

Update:
Got my new coolant pump Sicce Syncra 3.0. It does the job perfectly. Shame it took over a week to get here, cause the place where I ordered it f*cked up big time on their side.

Then time ran out ... and I had to move my entire life (planned move to another nicer but old house) in 1 week.
Meaning absolutely no MM time (12 hour days painting, fixing etc.). Got the move that done (sortof, still surrounded by my life in boxes).

FFWD, to tonight. Pump mounted, tested everything, need some last (wiring) fixes and then..hopefully first cut this weekend.

[Fox]

http://s28.postimg.org/i70bsk5h9/start.jpg

http://s28.postimg.org/z3pekzd19/machine.jpg

http://s28.postimg.org/5pto5eabh/milling.jpg

http://s28.postimg.org/v9vybtvp9/result_1.jpg

http://s28.postimg.org/ua5ng4gjh/result2.jpg

Fixed my pump situ, considered waiting untill my Dust management was working for 1 second..nahhhh.

Just a piece of grabage mdf ( some holes we're already in it as you can see ), and my 6 mm two flute hss endmill could not pocket the entire letters (see H) cause they we're to small, but who cares.. IT CUTS.
Happy with the quality of the cut, have no idea if my feeds and speeds were any good.

Surface of the mdf (the top of the pockets) needed a light sand to get rid of fibers, but the inside edges of the pockets are like this straight of the machine. Each letter is approx 25 mm high and a little over 6 mm wide.

Before I do any more cutting my priority is now with cutting a working dustfoot :D

[KenC]

Well done!
Prioritize a working dust foot is very wise.

[racedirector]

Excellent work my friend, looks fantastic. Congrats on your first cut!

[Fox]

Thanks Ken/Bruce,
It feels great that it's more then just a really big printer :D
I already got a million projects waiting that do not include corny sentences,but first the dust foot !

[Duds]

Great work, I love the message and modern font, looks great!

[Tokamak]

Great job Fox. Nice looking cuts. Now fun begins!

[Fox]

Thnx guys !

Yay ... FUN ! Waited a year or more on that. Almost there. CAD Drawing a dust foot right now...

[Fox]

Mike, do I get a number if I slap a logo on the bare metal ? Or does it need to be painted ?

[Gerald D]

The old rules were:
- painted
- logo
- proof of cutting, writing or drawing pictures

[Fox]

Thanks gerald,

Will honor that/you. Means no number anytime soon, but that's ok.

[MetalHead]

Hey I painted a table before with a 4 inch roller.

[pblackburn]

Gerald and Mike,
I think if the build is out of stainless there should be an exception. That would be a nice looking machine. And expensive.