Experimental hybrid MechMate – Tehran, Iran

[Atifeh]

Hello Gerald and all,

In my opinion MechMate is not a mere CNC machine. It is a way of thinking, solving problems and widening the horizons of learning. It is a concept with a design philosophy behind it.

Many have build the MechMate and enjoyed its building and working. My perspective differs though. I enjoyed the design. I tampered with the design to fit to my shortages and found out that I was altering and redesigning it, with inspiration from its own design.

I was here 3 years ago when a machine took almost a year to build, but now I am sure we can arrange a contest for who could make a MechMate faster? And you will see it will be operative within 6 weeks. My point is: this design is proven, is working and running business, therefore there is no need to alter it or to introduce a new design for it. And if there is, no one better than Gerald himself can do it. Alterations that I have made were only out of necessity.

With this prologue I would like to share my experience on building an altered or “hybrid” MechMate, as suggested by Brian and John with thanks to John, Irfan and Brian for their encouragement.

Originally, it was my log of the work done. I have made some changes. You might find some descriptions redundant, outdated or even irrelevant. That’s because they slipped my eye during editing. There are a lot of pictures which I did my best to reduce their size to an average of 60-70kb. I wonder if my remaining quota will suffice for all of them.
.
.
.

Although it has been 3 years since I registered in the Mechmate forum, I did not have the opportunity to make one. The reason for such a long period of delay were numerous. There were a multitude of limitation, acquiring parts was the major problem, specially the laser cutting parts and a precision bending shop to provide me with the main parts and a CNC machine shop to provide me with the v-grooved rollers. These took me a long time to survey without any result.

It seems to me that finishing this project might bring some confidence to those who are more or less in my situation, i.e. limited access to financial resources, technologies or parts and equipment. Over 11000 people have joined the MechMate community, while about 40 machines have been numbered.

My aim was to make a 4 axis machine capable of 1250x2200x250 mm working envelope and 500 mm diameter for circular objects. I intended to make it to produce: foundry patterns, PU and fiber glass moulds, small wooden wind turbine blades, model aircrafts and similar products.

Limitations:
Without access to laser cut parts I had to redesign the gantry and the carriage. Fortunately for the carriage linear motion part, there were two other designs in the forum which I could use, although it was also designed to sit on MechMate’s carriage design. Therefore the Z axis was based upon size 45 linear guide and rail ( ABBA BRH45AL).

Luckily, Rick from Superior Bearing was very cooperative and I could get the bearings through a friend in the US.

Oriental Motors are not available here but there is a variety of Chinese and Taiwanese stepper motors. I have to build timing belt reducers as well.

There is yet another limitation. My workshop is small and I can not move heavy objects like a 180mm U channel around it, therefore I had to improvise another method for the beams.

The spindle will be a 220V, 3ph., 2kW, ER20 collets, Turkish made (Arel), which is less than half the price of ELTE. For VFD, a DELTA type E was recommended.

The controls will be remained to be decided. I am still chewing on the choice of bipolar or unipolar connection. Either I will use my own made unipolar half stepping boards, or I might get available Chinese microstep drivers. The breakout board I shall make myself.

The structure will be built in modules which would be assembled with fasteners. The structure comprises 4 segments of frame which will be bolted together.
The machine was designed and drawn in MDT6. Here are some images of it.

6459

6460

Below pictures show the design of the gantry and the carriage. I would appreciate any comment or suggestion about this design.

6461
6462

The structural steels used in this project are: U-channels 80mm, square tube 60x40x3 mm, 100x50x2.5 mm and 40x40x2.5 mm and iron angle 63x63x6 mm and 50x50x5 mm.

I started off by making other assemblies rather than the frame.

The reducers:
Timing pulleys are 25mm wide M5 type, 24 and 72 teeth. The reducers does not have a swing hole but the pivot is made of bronze bushing. There are two makes, one for X and Y axes and the other for the Z axis.
Made some patterns with a small CNC that I made 4 years ago:
6463
Took them to an Aluminum foundry:
6464
Machined them on my drill press equipped with a XY table:
6465
Here are the machined parts
6466
And here are the assembly stages:
6467
6468
6469
6470
6471
6472
6473
6475
6476

And the finished reducers:

3 used for X and Y axes.
6477
6478
6480

1 used for Z axis.
6481
6482
6483

[Atifeh]

Cutting the rails:
This is the setup I used for cutting the rails. I used my 800W Makita mini grinder and the mounting plate was made to its proportions.
6484

Setting up for cutting:

6485
6486

Adjusting the height:
6488


With this rail cutting fixture each disk (Chinese made), cuts about 5520 mm of rails and shrinks to around 65 mm in diameter.
6489
6490

I had to make some fixture for grinding the rails, so I thought of giving plasma cutting a try. I edited Gerald’s drawings in order to make the holes 2 mm tighter so that I could drill through the holes to their nominal sizes. The picture shows as-cut quality of plasma cutting. As mentioned minor through hole drillings and filing were required to clear the holes but the set works really fine.
6491

Now I have bought a dedicated 850W Chinese grinder for this job.

This is the grinding arrangement:
6492

I did not have a successful experience using grinding disks. It might be the quality of the disk I use or my faulty way of grinding. After a few minutes of grinding the disk face became smooth and shiny and metal removal dropped to nil, while the grinding surface of the rail became discoloured due to excessive heat. Any idea why?
I found out that using grit 36 sanding disks were more efficient. The noise level is much less and the grinder works cooler.
6493

This is the sanding disk after successfully grinding 8800 mm of rails, both sides.
6494

An annoying problem: is the development of a groove on the contact surface of the height adjustment screw with the rail (shown below) which forces one to change these screws and readjust the height periodically. It gets worse since I have to index it by 90 degrees which on a M10 screw will be 1.5/4=0.375mm. I replaced the screws with high strength steel screws. The grooves were somehow less but they still exist.
6495

Has anybody came to a solution for this? I thought of threading an oil hardening steel rod and then heat treating it to about 40 Rc but I suspect the other way around might happen, i.e. the screw scratches the rail down. Maybe phosphorous bronze or copper beryllium could be used?

Here is the quality of the surface using grit 36.
6496

I also used grit 80 sanding disk to obtain a smoother finish. The finish was remarkable but due to the fact that minute feeds should be imposed on the grinding set, the process is long.
Manual sanding with Grit 220 ordinary sandpaper and kerosene as surfactant, virtually yielded a mirror like surface which reduces the rolling friction and I believe would suffice, therefore I dropped the grit 80 process. Of course the deep lines created by grit 36 sanding disk will not be removed by this process but I believe they are of no importance.

Everything gets magnetic.
6497

[Atifeh]

The carriage:
The design is simple and self explanatory.

6502
6501
6498
6499
6500
6503
6504
6505

Making the gantry:
I was wondering how could I bolt the rails on the beam while both ends of the beam are sealed? So I came up with this idea. I drilled through 3 of the holes on the gantry beam. Took 3 M8 rod screws about 150 mm length and passed it through the nut plate. Whenever I want to assemble the rails, I would insert the rail in the corresponding holes, lift up the nut plate, secure other screw, take the rod screws and then bolt these three. Detaching the rail will be in the opposite sequence.
6506
6507
6510
6508
In order to get suitable parallelism between the end plates of the gantry I used a method which you have all seen in http://oneoceankayaks.com/madvac. I named it “resin padding” (as I don’t know the exact name of the process). The compound used for this process was self made and is very reliable and cheap. If there is any interest, I can describe the whole process and materials used. Of course various brands are available. For example I found Devcon liquid steel B, the price of which was out of my league.
6509
6511
6512


This is the gantry and the carriage mounted (without the Z assembly)

6513

[Atifeh]

IT IS GOING TO BE COMPLICATED.

Today is March 20th, 2009. I was not satisfied with the design of the gantry. There is nothing wrong with it, but it didn’t capture me.

I TEMPORARYLY SUSPENDED THIS PROJECT AND STARTED ANOTHER ONE.

From now on you will see the second machine which is more like a MechMate except for the carriage. Afterwards I will return back to this machine.

[Atifeh]

As I said before I was not happy with the previous gantry, I was wondering if I could use the original MechMate gantry but with the modified carriage.

We have 2 weeks of our new year holiday, so I converted Gerald’s drawings to 3D and played around with it to see how it could be managed.
I did some changes on the parts in order to produce them with plasma cutting.

The product will be a 3 axis machine with 1900x3680 mm envelope to cope a full board of MDF and with a vertical movement of about 100-140 mm.

The gantry is the same as MechMate design but the carriage is still different, hence laser cut parts were unavoidable to use in the original MechMate carriage design. The carriage will be identical to the previous make.

Motors are PK296A2A-sg7.2. I got them through someone visiting Malaysia.

The structure will be built in modules which would be assembled with fasteners. Aside the main and table beams, the structure comprises 4 segments of frame which will be bolted together.

6514
6515

Below pictures show the design of the carriage. I would appreciate any comment or suggestion about this design.

6516
6517

Structural materials:
The main structural steels used in this project are: U-channels 140 (for the main beam) and 80x40 mm for the table, square tube 40x40x2.5, 60x40x3 mm and 100x50x2.5 mm and iron angle 63x63x6 mm for the rails and 50x50x5. I bought the steel for both machines.

My intention was to sandblast all the structural steels even before cutting and sizing them, so that I could get a good surface for scribing and later painting, which turned out to be time consuming and excessively expensive, therefore I stuck with wire brushing and grit 36 sanding disks. Small parts and also the fasteners are phosphatized in the workshop (it’s easy) for preservation and better surface for painting. Here are the parts stacked in trolleys (my current workshop is pretty small and a lot of displacement is inevitable).

6522

Plasma cut parts:
So far I have provided the motor plates, stiffener closures and the carrier plates by plasma cutting. In some cases I had to change the drawings to undersize the holes so that they will be drilled through later to the actual size. Deburring and finishing, whenever necessary, were done with grit 50 sanding disks.


Printed 1:1 drawings are bonded to the parts, center pinned and drilled on a drill press.
6520

A blank and other drilled parts.
6519

Quality and defects of plasma cuts. It is more profound in thin sheets.
6521
6523

Other plasma cut parts.
6524
6525


A REMINDER: if you do have access to laser cutting, do NOT try plasma cutting or even waterjet. It is not worth the time consumed and the quality attained. Remember: under no circumstances plasma cutting is a cheaper alternative or a solution or a substitute for laser cutting. It is like Charlie Chaplin’s gold rush movies in which they were short of food and cooked and ate their leather shoes and laces. Now, the leather shoe is the waterjet. For plasma, try plastic slippers!
My approach to plasma cutting was only an act of desperation!

[Atifeh]

The rails:
63x63x6 mm angle iron were used for the rails. I managed to cut both X and Y rails but grind the Y axis rails which are shorter. I will grind the X axis rails when assembled on the main beams.

Markup for drilling the X rails:
6526

A 2 meter steel ruler that I bought recently proved to be a very useful tool and better than tape measures. Markings will only give the longitudinal position of the holes. The lateral position of the holes were scribed for each hole against the reference edge.

Preparing the nutplates for gantry beam and Y rail markup and drilling.
6527
6528

I found this tapping adaptor, lying around in a tool shop and bought it very cheap. Bought a Chinese chargeable drill, removed the chuck and equipped it with this tapping head. It is powerful enough to tap up to M8, pass 2 and 3 taps. Pass 1, it cannot manage.
6529

Transferring nutplate holes to Y rails:
6530
6531

Transferring nutplate holes to gantry beam.
6532
6533

The previous cutting fixture was not comfortable to work with, mostly due to the nasty fume of burning MDF resulting from the sparks. I made another one which now its holes match Gerald’s cutting/grinding fixture. Also lined the circular part with a piece of Aluminum strip. No more fumes.

I tested the upper part of Gerald’s rail grinder for cutting This setup only allowed me to use the cutting disk up to around 85 mm in diameter since I used 63 mm angle irons. I made another plate for the new fixture, without the 45 degrees flap so that I can go down to 65 mm.

Grinding (actually sanding) the surface of the cut face of the rails with the new fixture.
6534
6535

[Atifeh]

The main beam:
I chose U140 over the U160 and bigger sizes for the main beam, the reason was its weight and complexity of movement and displacement in my workshop. The drawings show that I still have about 140 mm of Z travel but I have to lower the position of the X axis e-chain supports.

The lower side of the beam which will sit and bolt to the frame was scribed and drilled before cutting the 45 degrees end cuts.

End cuts done with the remainders of the cutting disks from rail cutting.
6549

Beam end plates cut and bent.
6550

Stages of transferring the X rail holes to the main beam:
6551
6552
6553
6554
6555

The center finders are made by myself. They are plastic mould ejector pins which are precise, extremely hard and durable, comes in different lengths and cheap. I grip it in a drill press chuck and grind it from below to leave a 1mm cone on its flat surface. Have never re-sharpened them in the last 4 years and are still working.

6556
6557
6558
6559

Drilling the beam
6562

Here are the beams drilled, deburred and finished.
6561
6560

[Robert M]

Nader, I’m Speechless…. :)
Maybe took 3 yrs to get there, but NICE prep & work so far !
We should open a sub-group for those of us taking +/- 3yrs to start a built :o

Many thanks for sharing you zillions photos, it is inspiring for me, a guy who as a very hard time with steel, welding….darn, I’m a woodworker who is getting a new passion for steel too :D

Well, can only add congratulation and can’t wait to see your hybrid finished.
Amicalement, Robert ;)

[jehayes]

Nader: Great job! Thanks for the excellent photos. Can't wait to see the machine assembly! Joe

[Gerald D]

Nader, great history! Your attachment quota has been increased.

[riesvantwisk]

Great job indeed and VERY inspiring and some good ideas aswell.. I just realized I could have done some things better, I think I am fine still.. but with your pictures and ideas I think I should have done the same..

great job!

[Polder48]

Nader, thanks for an inspiring story of your build. Great pictures and a lot of inginuaty involved. Keep on going!

marveless job!

cheese Polder

[isladelobos]

Very accurate work
Well taking pictures
Nice ideas.
Thanks !!!!

6568
A lot of bandwich charging this web thread :D
Max:5mb/ps

[MetalHead]

Great work - I am amazed everytime I read this forum. You all are great. I'm a slow starter also and still collecting parts and dividing time on other projects. But I will build a MM for sure !!!

Again great build!! And Great work.

I think steel is a great medium and those who can even scratch the surface (no pun intended) building with it always amazes others !!!

[jhiggins7]

Nader,

Great work, comprehensive build history, inspirational build!

I think you have captured the very essence of the MechMate way.

Good luck to you as you finish your MechMates. Keep up the exceptional work. Gerald has increased you allocation, so please keep documenting as you go.

If we can help you please don't hesitate to ask.

Thanks for sharing your inspirational build.

[Atifeh]

Guys, thanks all for your kind words and support, glad to tighten a small M2 screw in this forum. After all it was most your ideas.

Robert, Thanks for your interest and kind words. This might not sound so good, but I have to admit that we are all drifting from wood working to machine building! I for one!

Nader

[Atifeh]

Table cross beams:
A drilling template is made so that the hole pattern on all the table U channels be reasonably identical.
6569

The indents of the center finder (as seen in the picture below) will give the longitudinal position of the holes. The lateral position is scribed for each hole cross a reference edge. After I drilled all the holes I realized that this was a mistake. I should have drilled the holes all along the template to attain a straight line.
6570


This is a simple template I use to mark the 60 degrees cutting of the table U beams:
6571


Here are the table channels bored, deburred, cut and finished. The holes next to the outer holes are drilled through both upper and lower wings of the U80 beams. I am using the two guide through holes concept on the edge side of the spoil board as previously devised by Sean (SMReish). I intend to tap the other 4 middle holes right on the channels so that fixing the spoil board will be done from above. I assume the other way around would be that someone lies under the table and secures the wedge washer and nut, while another one fastens the bolts from above.

6572
6573

I will FAST FORWARD to the assembly stage of the finished machine so that you can see how the drilling template is used to locate the matching holes on the MDF board. Please note that the outer holes on both sides were marked from below with an 8 mm center finder and drilled in order to obtain the fixed points. Fortunately, the total job was a success, although a number of holes were about 1-3 mm off their centers, which was due to the U80’s cambers resulting from the above mistake, which were corrected. Countersunk 8x30 mm screws were used to secure the lower MDF board to the table beam..

6574
6575
6576

[Atifeh]

Frame:
Making the legs:
The vibration dampers had only one nut. I had to buy a M16x1.5 mm tap to provide a positive locking for the nut.
6577
6578
I bought a lot of cheap HSS tool steel strips and blocks, only to use them as shims while levelling the frame. They are 200m long and when halfed provided a lot of shim plates with a lot of combinations. Their tolerences are within 0.1 mm. These were the sizes I bought: 2x10, 3x10, 4x10, 5x10, 6x10, 8x10, 10x10, 20x20 and 25x25 mm. you can see their use in the below picture.

6579

Here are the stages for mounting,tack welding and final welding of the frame:
6580
6581
6582
6583
6584
6585
6586
6587
6588
6589
6596
6590
6591
6592
6593
6594
6595

[Atifeh]

Gantry

The carrier plates were cut blank by plasma cutting, the holes were transferred by scribing and then drilled. The bean shaped or elongated holes were shaped by filing.
6597

Preparation for assembly:
6598

I checked the parallelism with a well levelled table and laser square.

6599
6600

Here are the sequences of welding the gantry:

6601
6602
6603
6604
6605
6606

The finished gantry:

6608

This was my first try on weld straightening. There was some 15 mm of camber. Went off real good. I hope it is not the beginners luck!

6609
6610

[jhiggins7]

Nader,

Nice work, great pictures. Love the castor idea...portable but stable once it is relocated.

[Atifeh]

The carriage:
Here are the stages of making the carriage:
6611

6612

6613

6614

6615

6616

6617

6618

6619


A proof of negligence:
While in MDT, I did not account for the thickness of the double sided tape. Also I should have mounted the 30 tooth pinion instead of the 20 tooth. The result was that the motor assembly did not have enough room to slip inside the rack. I had to re-drill and re-file the bores.

6620

I used KR-32 idler needle bearings instead.

6621

Another hole placement miscalculation.

6622

The pinions:
There were only module 1 pinions with 22 and 26 teeth available, both of which had a hub thin to accommodate the 12.7 mm motor shafts. I ordered a machine shop to make 4 each of, 20, 22 and 30 teeth pinions with enough hub to be drilled. He made them on a Chinese milling machine with teilekopf, amazingly precise. Cost me around $10 (equivalent) each.

[Atifeh]

Microswitch supports:
Mechanical microswitches are a lot cheaper than proxy switches. In addition I could not make the carriage arrangement of MechMate which had built in locations for the proxy switches, so I chose to use mechanical microswitches. Improvisations were required for their arrangement and room for adjustment. The following pictures show these arrangements:

Limit switch for marking the end of table in X direction. This was a bad choice. I should have placed it on the other side. I wanted the Home side to be uncluttered with these devices. Will review this on the first machine.
6623

Parts required for the Y axis Home/Limit pair of switches assembly.
6632

Marking the place for the Y axis Home/Limit pair of switches.
6624

Y axis Home/Limit pair of switches arrangement secured.
6625

The adjustment distance:
6626

Closeup of how it works:
6627
6628
6629

Y axis Home/Limit pair of switches and the other end limit switch.
6630
6631

Parts required for the X axis Home switch and the autosquare end assembly.
6637
6638

Z axis has also a Home/Limit pair of switches with an adjustable height slide.
6633
6634
6635

This was the last part made. What would happen if the end limit switch malfunctioned and there were no soft limit lamp on?
6636


Echains:

Details of Y axis Echain:
6639
6640
6641
6642

Details of X axis Echain:
6643
6644


A design problem:
I think there is a glitch with the arrangement of the X Echain. You see, it is convenient to park the gantry during idle times at HOME position. One reason would be the loading of the machine which is from the front. When parked at home, the Echain rolls up from its support and gets suspended. I noticed that my Echain sags under the weight of the cables. If it is indexed 180 degrees, then while gantry is at rest or idle position, the Echain will rest on the support, rather than be suspended. Of course it is not an easy task. A lot of cosiderations should be reviewed, for example the cable routes from Y Echain and the sweeping characteristic of the X2 motor. A lot of laser parts should be revised then. Please comment on this.

[Atifeh]

Next will be: The Assembly, Control Panel, Cabling and the final Touch Ups.
They I shall return to the First machine.

[Atifeh]

The assembly:
Painting:
Two part Epoxy primer and paint were used. I gave 48 hours interval before applying the next layer and then tested the quality of the paint adhesion with the scratch and tape method. The adhesion quality was outstanding, but the quality of the paint not so. Mainly due to the local weather phenomenon of suspended dust which took about a week or so to clear.
Each individual part was primered and painted.

6646


Squaring the machine when you are alone and nobody is to help.
6647
6648

Aligning the rails:
6649
6650

These are the stages of assembly:
6651
6652
6653
6654
6655
6656
6657

Installing the rack:
The rack was almost 4000 mm. twice it broke from the weld joint. At last I preheated the joining parts with a torch to about 400 degrees C, checked it with thermochalks, and then commenced welding. This time it held.
I suspended the rack by the two end screws and some suspended wires, washed the contact surface with aceton and placed the tape, lifted the top paper cover and carefully raised the rack.

[Atifeh]

Control:
I finally settled on unipolar wiring. I opted for the speed. Bought 4 microstep 0.5-6A max.80V unipolar drivers from local suppliers (originally from: www.bsjd.com). I tested them and seems fine for the price. They are opto isolated (something I don’t like) but they claim it works up to 200kHz.

Ordered a 220/28 VAC transformer to get around 38 VDC. The capacitor is a 20,000 uf 80V.

The breakout board, I made myself. It is a buffered board using 74HC541 (around 35mA buffered output). I just read the logic diagrams of PMDX-122 in their manual and mimicked its logic. The circuit is simple and no ENABLE used for charge pump. Instead I allocated 4 pins for relay devices which I don’t know what I’m going to do with them?

Honestly, I don’t see any virtue with this charge pump idea. Do you really need to start your machine prior to the computer? I always start the computer, then the Mach and then start the machine. What’s wrong with this procedure? And just for the kick of it, I once started the machine, then the computer. Nothing except for some grumps and growls happened. They are exaggerating this feature that without it your gantry might rampantly ramble around and such, I suppose.

The breakout board was designed as a single sided board with PCBExpress (www.pcbexpress.com). I then printed it in pdf format and had it printed on film. Then used laminate sheets on ordinary PCB copper boards, exposed it to two cheap 4W UV lamps that are used to identify fake bills, developed it in Na2CO3,acid wash it and made it. The board totally populated cost me around 5$ and works flawlessly.

I also used a small 220/12V 20VA to provide 12 VAC to the Estop and limit circuit and a small 5/12 VDC 1A board which would feed the breakout. The breakout feeds the drivers. The 5/12 VDC could also provide power for future relays.

Here is my wiring diagram for the live part. Please note the use of 12VAC for the Estop and limit circuits. It is far safer than 220V.

The brown/blue lines are the 220VAC PH. and NULL and the green/orange are the 12VAC live and neutral lines.
6658

I would suggest visiting this site: http://www.veppa.com/ekts/
There is a free program that lets you play with different relays, contactors, motors, switches and other electrical parts. You can then simulate your circuit and every fault will be revealed. It is designed by a university professor in Turkey as an educational aid. The interface is not much of a fun (as Livewire) and needs some patience, but everything practical I know about designing electrical circuits owes to this programme.

I believe the next stop for us wood workers who are becoming machine builders, would be migration into the realm of electronics! So shape up!

The box stages:
I bought a panel which had only 3 holes at the bottom side, before drilling holes in it I decided on my priliminary layout.
6659

Next I drilled holes with Morse round cutters:
6660
6661
6662

This idea of auto air filter by Rad racer was very nice.

Stages of populating:
6663
6664

Please note the grounding hub next to the cooling fan.
6665
6666

The driver and breakout board section.
6667

Here is the control panel:
6668

I took extra care to divide the control panel to different zones to minimize the noise of AC and high current DC voltages on the control 5VDC’s. Here are the different sections.
6669


What that is not shown are the groundings. Every individual cable shield was grounded to a main hub inside the panel which then left the box to join the main earth cable. 6mm cables were used for grounding.

Here are some pictures of the panel connectors:
6670
6671
6672
6673

The VFD runs manually. For the first machine I will use a 0-10V board (remember? This is the second machine).

And for the last, here is the total cicuitary of the panel, except for the interconnections to drivers and breakout board which differs for various makers. Unfortunately the pdf version is only in black and white, but the edrawing copy is in colour (in rar format).

6674
6675

Grounding:
Dug a hole on a corner of the new workshop which is a basement of a building, about 700 mm deep. The soil was moist in that depth. Picked up a copper strip 6x60x500 mm, drilled at one end and screwed two cable shoed 6mm copper wire to it. Hammered it in the hole so that it went down another 400 mm into the ground. Layer by layer filled it so that in the middle, where the copper strip was buried, I put a mixture of coal and salt, and around it the sieved soil. Then a nylon sheet on top of that so that it will not loose the moisture. I will then occasionally pour some water in the hole to maintain its conductivity. The potential, relative to the NULL line was 6.1 Volts. It should have been less than 5 Volts but to get that, I had to dig deeper which was not possible. I believe that will do.
6676

For collecting the static electricity that accumulates in the dust collector duct, I use a bare braided 14x1.5 mm copper wire and after isolating it with shrink tubes, screwed one end to the steel structure of the building.

Here are some pictures of the cabling:
6677
6678
6679
6680
6681
6682

Arrangement of the panel:
6683
6684
6685

[Atifeh]

Different views of the machine.
6686

6687

6688

6689

6690

6691



Trial run:
I bought an ER20 collet, 6 mm and tried some patterns. At 10th microstep there were no stepped cuts in a curved line. The squareness was bad on a single sheet of MDF, about 4mm in 4000 mm which was corrected. I set the speed at 2500mm/min, the Mach frequency is 25kHz. Unfortunately all the pictures of the works were on a camera chip which I have misplaced it somewhere. I will post the works in a short time.



Last words on this machine:

I really had a lot of exciting and good times, specially with works that I have never done before. A list of my favourite operations were:

Straightening the beams and rails by welding.
Cutting the rails.
Grinding the rails.
Finding that there is only one shop in Tehran that has wedged washers sizes 8 and 10 mm, and now I know it.
Paying the cost of a one week leisure tour to Malaysia in a 5 star hotel for the wife, which turned out to be much cheaper than having TNT bring me the PK motors from there!
Mounting a thin 4 meter object with double sided adhesive tape.
Painting with 2 part epoxy paints. 4 years ago I bought some epoxy paints for a job and was wondering why is it named epoxy when it is only in one container and there is no hardener and is no different that other paints?!
Finding out that after 55 years of age, a broken toe could mend in 2 weeks, just by ignoring it and a little rest!
Understanding the correct way of shielding and grounding.

Now, the excitement still continues since I will be completing the first machine which has a fourth axis and since I have 4 more output pins there are a lot of possibilities to use.


Now back to completing the first machine.

[Gerald D]

Honestly, I don’t see any virtue with this charge pump idea. Do you really need to start your machine prior to the computer? I always start the computer, then the Mach and then start the machine. What’s wrong with this procedure? And just for the kick of it, I once started the machine, then the computer. Nothing except for some grumps and growls happened. They are exaggerating this feature that without it your gantry might rampantly ramble around and such, I suppose.

A real story: We have a general power failure in the town. 20 minutes later one of our staff starts changing the cutter in the spindle. The power comes back on and another guy switches on the control panel again. But the spindle also starts by itself! (Luckily nobody touching it right then, but the spanner/wrench goes flying).

What happened was that the PC took its power from the main supply (not the control cabinet), the PC did an automatic reboot when the mains power came back and this did a false signal to the PMDX before Mach started running.

[jessyjames]

Nader great job on the build thus far. I really appreciate the many pictures that you have allowed us to view through your process of building a nice looking mechmate.

I'm one of those builders who don't really post anything and just gets to building. Maybe those days are over..?

Keep up the hard work. Gerald's design is just outstanding and the knowledge on the forum is just short of amazing..

Keep the pic's coming Nader.

Respectfully,

James

[Alan_c]

Nader

Very nice work and well documented as well - interesting solutions to some of your challenges. In South Africa we have a saying: "'n Boer Maak 'n Plan" wich roughly translates to a farmer always has a solution to solve problems - I think you qualify as an honounary Boer :) well done.

Alan

[domino11]

Nader,
Thanks for sharing those pictures and the story. Very well done and documented. :)

[cncb]

Are the photos not showing up for you guys?

[cncb]

Thats weird, they show up in IE but not Firefox..

[domino11]

Brian, try clearing your cache? OR do a reboot?

[javeria]

WoW nader, I was away for a while - and looks like you posted soo much man

good for you - and good for the so many people with lesser resources who can still build a cnc by using ideas you have used

RGDS
IRfan

[Atifeh]

Thank you all for your support and kind words. I am glad this thread proved to be useful.

Once again back to the first machine. This is still the history of the late works. The machine is undergoing painting and from that point the progress is up to date.


Preparing the legs:
I moved the leg dampers outside of the legs for ease of adjustment.
6720

6721

The previous method that I devised for welding the transverse frames was complicated and errative due to the fact that the main frames were not guaranteed to be quite parallel and in the same plane. This time I used this method. Measured and welded the frames on a flat surface and then matching the main frames to it. It proved to be better and more precise.
6722

6723

The end pads of the frames are tack welded in this position:
6724

6725

Straightening of the beams and rails by welding.
6726

6727

A lesson learned: when straightening a U channel, do NOT weld on the web. It will bend outside, disregarding welding it on the inside or or on the outside. The reason is the low mass of the web relative to the high mass of the wings which holds more internal stress. Even grinding will not return it to the initial state. Instead only work on the wings.

My parameters besides Gerald’s recommendations were:
Use as much amp as you can to concentrate a lot of heat.
Maintain a constant width of weld. This is very important. When I welded wider, more lateral movement was made, the heat concentration was quite high and the bending was more profound.
Start with 600 mm apart. Wait after completely cool. This is also equally important. Then if necessary, make corrections within this distance.
In some cases grinding reversed the action of straightening, in some cases it did not.

Marking and transferring rails holes to the main beam after straightening.
6728

6729

Table beams:
As before a drill mark template was made. I avoided the not so important mistake in the previous make. This time a straight line was drawn along the length of the beams and the template was matched with that line and the center and center pinned. Now, when drilling the table board, whatever longitudinal mis-shape the beam might have, will not interfere with the location of the middle tapped holes, once the first and last holes are matched with the template. Below pictures show this sequence.
6730

6731

6732

6733

Beams being drilled:
6734

[Atifeh]

6735

6736

6737

6738

6739

6740

6741

6742

6743

6744

6745

6746

6747

6748

Please note, the name of the pictures represent the action done.

[Atifeh]

Arrangement of the microswitches supports and pylons:

X axis end limit switch and the previously forgotten end stop tab. Now both positioned on the other side.
6749

X axis home switch and end stop.
6750

Y axis limits and home switches.
6751

Supports for the button box.
A hole was drilled to let the cables pass through from the X1 side to the X2 side. Cables will go through a plastic cable duct which will be mounted inside the gantry beam. (details will come later).
6752

Details of the X and Y echain assembly.
6753

6754

6755

6756

6757

[Atifeh]

6758

6759

Please note: the 4th axis pictures are in the related topic in this forum and will not be discussed here.

[wiwatto]

Absolutely, Well done "Nader"

Thank you for you best idea, i like the way you worked very much.

[boaterri]

Well done, sir. I also like the adjustable work supports. (the saw horses)

Rick

[Atifeh]

Thank you Wiwatto.

Thank you Rick. I will post some details of the sawhorse.

[Atifeh]

All fasteners were put in place loose. The wheels and dampers were installed and the set was levelled. I squared the assembly diagonally and tightened the first and last table beams. There was about 8 mm variation in the width of the machine on the rail rest beam. With flat belt tightener I narrowed the distance and tightened all the table beams while measuring the diagonals on the frame. There remains about +/- 2 mm variation which will be adjusted during the rail adjustment. Then I tightened the transverse frame bolts.

6786

Placed the rails and X axis echain support in place and scraped the paints on the contacting face of the rails.

6787

This is the detail of the wheel and the damper.

6788

[buibui]

Great job, Nader. I really enjoy reading your build and photos!

[wiwatto]

I waiting your progress

For me, Absolute confirm " Your are my inspiration " MechMate

Thank you again all done in this forum.

[Gerald D]

Wiwatto, you have told us enough of your inspiration, now we want to see your progress! :)

[Atifeh]

Thanks john.

Rick, here is the details for the sawhorse. The original credit goes to Fabrica. His posted pictures inspires lots of ideas for sawhorses.
6806

[Alan_c]

See here (http://www.mechmate.com/forums/showthread.php?p=1833&postcount=14) for more sawhorse detail.

[Atifeh]

Alan,

Sorry for the mix up. I stand corrected. The idea was yours but it was in Fabrica's thread. My MechMate picture files are sorted by the name of the initiator, e.g. Yuri, Irfan, AlanC... therefore I only referenced from the folder. I should be more careful when referencing.

[Alan_c]

No problem, at least my wacky idea was able to help someone :)

[Alan_c]

I am not an engineer, but I am a bit worried about your placement of your damper/foot, is there not a danger of that plate bending once the gantry, support board and spoil board are in position (I also tend to climb up on my machine quite a bit, but then again I am on the wrong side of 100kg's)

Maybe you could think about adding a brace as shown below, that should not hinder your adjustment of the dampers/feet.

6826

[wiwatto]

Wiwatto, you have told us enough of your inspiration, now we want to see your progress! :)

Hi Gerald

After "ATIFEH ( Nader )" finish his job i will post some picture in 2 years ago, that started building 1st MM. Nowday i have plans to rebuilding it again. Also i have something i never told your are "YOU ARE MY "HERO" since i start to interesting CNC Machine.:):):D:D:):)

Yes "ATIFEH" you please post your progressing. :):):)
I like the way your work

[Atifeh]

Alan,

Thanks for the idea. I should have done that before but I have to figure out a way to add this brace without dismantling it, since I am not good at welding vertical joints. The plate is 10 mm thick and only 32 mm projected.

If Gerald and everybody else agree and feel suitable, lets allocate a thread for different jigs, fixtures, clamping and work holding systems, intermediate storage systems and likewise that the members have made .

[KenC]

Nader,
Alan is right to point the need of reinforcing the legs. If you are not confident with your vertical weld, I would suggest you "joint-up" many spot welds to form a strong weld. That what I'll do ;)

Cheers

Ken

[Atifeh]

Bending stress = Bending Moment (N.mm) / Section modulus (mm3)

For cantilever beams:
Bending Moment (N.mm) = concentrated load (N) x distance (mm)

Assuming that the total weight of the machine+mdf boards+workpiece+myself =12kN, each leg will take more or less around 4000N of load.

With 18mm distance from the cantilever fixed point, the bending moment will be:
4000N x 18 mm = 72000 N.mm

The section modulus on X-X axis for a 40x10 mm hot rolled strip will be:
(40x10^2)/6 = 667 mm3

Therefore:
Bending stress = 72000/667 = 108 N/mm2

For mild steel (st37) the median bending stress limit is 290 N/mm2 > 108N/mm2

Alan, Ken,

These calculations were meant for only one purpose. To ease my conscious, refraining from a vertical welding with unsuitable electrodes, 100 mm above the ground, a lot of paint scraping and paintworks to be mended later!

[Atifeh]

I had a cold this week, so I stayed at home and worked on the electric wiring. Here is a zip file containing: the wiring diagram (dxf) and the simulation (ekts).

6907

Earlier this week I bought a Q2HB68MG bipolar driver (www.bsjd.com) for testing. These drivers are microstep, 24-80V and 0.5-6A. I have never before used bipolar motors, so this was an exciting experience.

I had 2 types of motors available:
4 wired 4.2 N.m, 6A/phase, 2.4V, 0.4Ohm/phase, 3.4 mH/phase
8 wired 2.2 N.m, 4.5A/phase, 2V, 0.45Ohm/phase, 1.7 mH/phase

You can see the pulse generator I built with a NE555 with jumper selectable capacitor rows and a pot. It gives 0.8Hz up to 10.9kHz pulses.

Also note the power supply unit I use for bench testing. Tied to a 24VAC, 250VA transformer which gives approximately 36VDC.
6908

The driver was set to 1/5 step. The 6A motor which has higher inductance, topped up the frequency and the rotor could not be stopped by hand. The lower inductance motor (4.5A) would not turn beyond 7.76kHz in serial wiring but did the full 10.9kHz in parallel mode.
Since maximum 3 motor will be working simultaneously, I think Mach could manage turning these three.

[wiwatto]

You have done some amazing things.
A picture is worth a thousand words

[Gerald D]

. . . . For mild steel (st37) the median bending stress limit is 290 N/mm2. . . . .

Just a casual note: Mild steel and ST37 are not quite the same thing. I agree with about 290 N/mm2 for mild steel, but you may be interested to know that ST37 was a German abbreviation for "Steel 370 N/mm2" :)

We could have a long discussion on stress calculations, but just some quick observations:
- Most times, in the frame of the CNC machine, we are concerned for stiffness, and not so concerned for risk of failure.
- But, at the support feet, often a spring element (low stiffness) is used for damping/isolation - you can claim that you have an isolator :)
- But, if it were a true isolator, then you need to design for fatigue. The fatigue limit of mild steel is about 50% of the upper stress limit. (Plus you have a weld at the highest stress area) :(
- But, the loading for the fatigue condition will not be 12kN - subtract your body mass :)
- etc.

Engineers sometimes calculate, sometimes use instinct . . . . .:D

[Atifeh]

Gerald,

Unfortunately the term mild steel has found a very wide definition by people who are paid to complicate things! Long ago, it was only meant for the cheapest structural steel. Now there is a whole range of disputes around its chemical composition and/or its strenght. To make it even more fun, mild steel is now even graded! like Grade 250, Grade 350, etc.
These standards and there desparate attempt to adapt to each other make me crazy. To tap a M6 thread, DIN says drill 5.0mm, ANSI says drill 5.15 mm (or at least this the the preference for my MDT6). Low quality drill bits have some wobble, so I drill 4.9 mm!

You and Alan were right. It is the stiffness concern. I agree and believe that this little brace would give the construct a lot more rigourous look. And about the failure, actually I was thinking "how would a 36mm tab be bent when it is supported on some sort of polymer?"

[Leko]

Just a thought. If you have difficulties with vertical welds, a brace like this could alleviate them. Drill for the two bolt holes and then you only have to horizontal weld it.

[Gerald D]

Or this:

6926

. . . . but in real life I would leave it as it is now! :)

[Atifeh]

Thanks all for the ideas. For now I will leave it as it is, as Gerald suggested, but in the future I will use Allan's Idea which is the most simple and yet does not confine accessibility.

[Atifeh]

Greetings to all,

These past weeks, besides a hectic period of work, mostly I managed some other activities.

Bought the whole lot of electrical parts and cables for the next stage.

Bought the computer but have not set it up yet.

Designed a whole new set of power board, 12/5 V board, pwm to 0-10 V and a more compact breakout board (since I have to use two boards). Had them printed on film and made the bare boards.

Attended a one day course for programming the DELTA VFD-E, which I learned how to interface it with external signals. These VFD’s have a lot of potential and I think it is worth learning the plc section of it.

I will be revising the electric schematic which will be posted after completion. Now the VFD will work in two modes, 1- which receives the start and pwm from MACH. 2- a switchover which overrides MACH and gets the RUN and rpm manually, through its detachable mini panel.
I decided on this switch in order to execute the warm up sequence as devised by Mr. Richards without entering MACH. I would not be using the reverse turning, as my spindle is an ER20 collet. Reversing will probably loosen the nut and be quite dangerous. Please tell me of any possible uses for reversing the spindle. The only use I could imagine is for tapping.


Besides, these were the progress made on the machine:

Mounting the racks:
As you have noticed, I use 35mm, M6 screws to hold the rack at a distance from its final position. This provides enough room to strip the double sided tape. For the X axis a middle support is used as well. Then the screws are turned and the rack is elevated to about 2mm below its final position. This facilitates the positioning of the rack and provides a means to once again check the linearity of the rack against the rail. After binding the rails, the screws will be replaced with M6x25mm.
7115

7114

7113

Z plate:
These were the original plates for mounting the spindle.
7116

I revised them to incorporate several holes for: attaching a Z echain, bracket for home and limit switch, a moveable dust collector attachment, additional holes for relocating the spindle plate:

7117

7118

Arrangement for the tension springs.
7119

7120

[jehayes]

Nader: Very nice! Thanks for the great and instructive pictures.

[Atifeh]

Thanks Joe,
Hope they are helpful.

Besides the VFD and the drivers which I bought, I made two buffered breakout boards and a 12/5 V regulated power supply.

7205

I wonder if I could use the 12V for the PWM to 0-10V board for the VFD, as the 12V and the 5V have common grounds. Any comments?


I decided to install an echain for the Z axis which might prove useful in the future; an air hose for a high speed turbine spindle, a water colled spindle or … The size is 25x30mm.
7206

7207


This is the configuration that I use in my gearboxes. A M12 bolt through a bronze bushing, 13mm high. The gearbox stands clear from the support arm by about 1mm which could be less if i grind the bushing.
7208


I tried the standard MechMate 35 turns, 2mm wire, 16mm OD, with elongated length of 128mm, which happened to be unsuitable. While cranking the Z plate down manually, at a point the pinion slipped off the rack. I made a spring with 16 turn, which is very hard to expand and the slipping was eliminated. Now I have to optimize the length of the spring. I suspect this problem is due to slope of the spring and the horizontal vector of the force.
7209

7210

Here is the detail of the spring assembly:
7211

[liaoh75]

Nader, very interesting build. Please keep the pics coming!

[Atifeh]

I changed the Z motor spring to 30 turns of 2.2 mm wire, 16 mmOD. Still slipping. Reduced the number of turns to 23 and it works fine.
I guess the below picture shows the fault. The point of effect of the spring is very much near the pinion axis, not providing enough leverage. The spring should be attached farther up.
7254

The components and arrangement for the Z axis home and limit switch.
7255

7256

7257

Details of Y axis echain:
7258

7259

7260

[Gerald D]

Nader, that y-car design (bolted rectangle) was used by ShopBot and it was not a success. The major cure for the weak rectangle was to add a second y-motor. Another cure was to add a plate over the rectangle - link (http://www.mechmate.com/Forum/show.cgi?tpc=10&post=1033#POST1033)

[Atifeh]

Gerald,
Thanks for the hint and the link. I studied it but could not comprehend the problem. What was the problem with the ShopBot? What stress condition made it faulty?
Thanks.

[sailfl]

Nader,

Your machine is looking really good.

[Gerald D]

A 5-minute sketch:

7277

[Atifeh]

Thanks Nils,

Thanks Gerald,
I got the problem. Do you think that welding the rectangle structure members together will help eliminate this problem? I have to say that the square tubes were quite fit to the angles and bolting them was actually a bit hard.

[Gerald D]

Nader, I would rather not discuss that further. I am trying very hard to concentrate on the MechMate design only.

[wiwatto]

Dear Gerald

Take it easy SIR, Don't worry about your MechMate design, even you got more questions or 10,000 photographs. Since June 2007, I know what is the Real MechMate Design. And i hopefully everyone also.

Thank you very much for your hard work on the MechMate.

Nobody, will change my mind " Gerald D is my hero"

[Gerald D]

I not worried about the MM design (and it is NOT perfect), I am only worried about what will happen to my free time and this forum if I start discussing other designs.

[Atifeh]

I personally agree with Gerald. If I were in his place my answer would be the same. It is not fair to introduce various designs and ask him to comment. Gerald was very considerate to note a fault in my machine. I myself, have to improve it with my mindset on my own design (which somehow I did).

[Atifeh]

Details of X axis echain: after the echain a sheet metal duct is installed. A perforated plate has been devised to carry one end of the flexible cable hose.
7494

Spindle and its mounting plate installed.
7495

[Atifeh]

The cables I use. From top to bottom:
Common ground signals cable (except for the motors), i.e. Estop signal (pin10), Pause, Resume, Zset, Home X1,X2,Y,Z,B and GND.
Spindle cable.
Motors cables.
Microswitches routing and Limit Switches/Estop 12VAC cables.
7496

I started with the cabling of the machine, soon to notice I have to clear up the matter of the control panel, regarding its entries and exits. So I left the cabling and will attend to that after completing the panel.

As you may have noticed, there are no free walls inside this workshop in order to hang the control panel. Also I might be moving this machine to another workshop. So some flexibility should be considered and I have to make a portable one.

Ordered a 500x700x220 mm control panel. Upon collecting it, I noticed he made a mistake and made it to 500x800x220 mm. He gave me a discount to the initial price for his error. All for the better! Now I can install the VFD inside the box.

These are the panel and the support structure.
7497

Here are the sequences for its preparation.

The first step was to locate various components inside the box with minimum voltage interference, considering the routes for the cables, interfacing connectors and input/output devices, and also their location outside the panel walls and door. Also the path of the cooling air to the VFD, the drivers and the transformer should be considered. Metallic plates might be even necessary to channel the air flow to the desired path. I tried different layouts and this was chosen:
7498

The next step was to mark the location of the ducts and other components, inside the box.
Next I measured different components dimensions and draw and printed them on paper. Glued them on each panel surface, center pinned and drilled them. Holes larger than 16 mm were cut by hole saw. Sizes: 19, 22, 25 and 43 mm are usually the most common for preparing a panel. For 120 mm fan a 114 mm hole saw should be used. All cuts were done by hand drill.
7499

All holes were deburred both from inside and outside, as this is really crucial. This is one side completed:
7500

Two rectangular openings should be made. 1- for VFD’s mini keyboard. 2- for the parallel ports.
Here is my method for making rectangular opening on the control panel.
Cutting is done with the remainder of the cutting disks used for cutting the rails. These depleted disks have reduced diameter and work better in short distances. A diagonal cut helps to separate each tab easily. The rest will be finished by filing.
7501

7502

This is the double LPT port interface:
7503

Here are the views of the completed bare box.
7504

7505

The transformer is quite heavy and will be backed by this U channel. The angle will support the other end of the flexible cable hose. Therefore, for moving the machine, the only tasks to be done are to disconnect the connectors, the earth wire and this angle iron. Then the control panel can be moved independently.
7506

The auto air filter and its assembly components. Two off will be installed. One just above the VFD and the other one to provide air for the drivers.
7507

[Atifeh]

The following changes have been reflected in the new diagram:
Some errative colour coded wires are corrected.
The preliminary wiring of the VFD and the 0-10V board are added.
Another Estop switch added on the panel.
There amy be other revisions.

For DELTA type-E VFD the following programs are to be done:
MI1 -- 02.01 = 2 (external terminal, keypad start/stop disabled)
MI3 -- 04.05 = 8 (operation command selector = external terminal)
MI4 -- 04.06 = 19 (operation command selector = keypad)
MI5 -- 04.07 = 22 (source of second frequency command)
DCM -- 02.09 = 4 (source of 2nd frequency command = digital keypad potentiometer)

Also.

AVI -- 02.00 = 1 (source of 1st master frequency)

The attached file is a zipped .dxf file

7536

[Atifeh]

Here will be the stages of wiring the control panel as I progress.

Prior to starting, the box was vacuum cleaned and cleaned with a damp cloth. No traces of drilling chips.
Placing and securing the cable ducts inside the panel:
7537

These are the consumables normally used for wiring.
7538

The wiring operation will be divided into 4 stages:
The 220VAC stage, the VFD stage, the drivers stage and the breakout and interface stage.

The 220 VAC wiring stage:
The relay and terminal blocks mounted on the DIN rail and tag numbered. Please note the spacer under the rail for ease of wiring (idea from Gerald).
7539

All other components related to this stage are mounted in their respective places.
7540

This is the space I made myself to start with the wiring.
7541

The 220VAC stage completed. All wires tagged according to the drawing:
7542

Mounting the VFD and connecting the feed wires. Mounting the transformers:
7543

7544

Fastening the transformer from behind.
7545

Mounting the 39VDC module:
7546

[javeria]

thats actually good organized work - I am so ashamed :(

[Zanchet]

Nader

You´re the MAN!

[Atifeh]

Connector side of motors wiring:
7725

Driver side with separate grounds for motors wiring:
7726

Arrangement of the motor/driver wires:
7727

Preparation of wires between the driver and the DC power supply.
7728

7729

Placing the drivers and its wiring. Duct covers removed to show the routes better.
7730

7731

Details of power supply connection. To avoid clutter around the driver, the second set of ground wires are connected on the power supply side, rather than the driver side.
7732

7733

Spindle UVW connection:
7734

Connection route of the VFD switchover switch and the mini keypad to the VFD. Another Estop is also routed through this cable duct.
7735

Mounting the 5/12 VDC power suppy:
7736

The PWM to 0-10V board is still under development, therefore I have to run the spindle in manual mode. Upon completion, I will start a separate thread for it.

[Atifeh]

Here are a set of carbide bits, newly bought. Although I was told that they are usually used by stone cutters but they work outstanding on MDF. On the other machine, a 6mm diameter tool cuts 6 mm depth per pass @1700 mm/min without any deflection in the tool. I have not tested them beyond this limits but I suppose they have the potential for more. Wall faces are smooth with acceptable tool marks.
These are 5,6 and 8 mm shanks.
7737

These I have bought to test the 4th axis. 6mm shanks:
7738

[Alan_c]

Nader, nice panel layout. Ideally those drivers should be mounted vertically so that the heat can dissipate off the fins of the heatsinks. as they are now you might get hot air pockets collecting under each heatsink. I know you are going to have fans on the case but I doubt much air will move through there. Of course it also depends on how much heat those drivers will generate - it might be of no concern - perhaps Mike Richards could weigh in here, he has far mor experience with these kind of layouts.

Never mind, I see above you have considered even putting in baffles if required - it would help if I read properly before commenting. :o

[KenC]

Nadar,
I second Alan's concern. The drivers might be just a bit too closely spaced to encourage air flow. I not an expert in this as well, it might not be a concern if the drivers never heat up.

[Atifeh]

Alan, Ken,

Thanks for your comments. You are both right. To tell you the truth, my first layout was that the drivers were arranged vertically, drawing air from above, but in order to have isolation of high voltage to control voltage, and also to have access to the microstep setting dip switches, I had to place them upside down. I just didn't like the idea to read the texts on the driver upside down!!

In the previous machine, the motors were wired unipolar. They actually worked with very little heat generated. The motors were 3 Amps. Now they will be 4.5 Amps and wired bipolar serial (except for the 4th axis which will be wired bipolar parallel) and I am not aware of their behaviour, but we'll see how it goes.

After finishing the panel (which you will see in the next post), I noticed that baffling might not be a feasible idea, but as you may have noticed there is a space between the VFD and the drivers, just above the 5/12 volts power supply. I could add a couple of 12V computer fans if the temperature builds up.

[Atifeh]

This is a single connector that incorporates: Estop signal, Home (X1,Y,Z), Home X2, Home B, Pause, Resume, Zset and a common GND.
7751

All the wires that will lead to the breakout boards:
7752

Breakout boards are installed and wired.
7753

Connectors between the parallel ports and the breakout boards are attached.
7754

Den of snakes, i.e. the grounding hub. Two more wires will be passed. 1- the 6mm wire from the machine, 2- the wire that goes into the earth.
7755

A thorough vacuum cleaning was in order to remove any residue of snapped wires. This is the panel almost finished.
Please ignore the yellow/green wire that goes from the transformer secondary to the ground. It was an aweful mistake and will be replaced by the negative line, just after the bridge to the grounding hub.
7756

Air filters installed:
7757

Pictures of the work. Nameplates will be attached later.
7758

7759

A 60 mm cable duct attached to the panel chassis (with double sided tape) which will carry the cables to the other side of the panel.
7760

Nail varnish on M3 nuts and screw ends on the back of the panel to restrict their loosening.
7761


Next step will be wiring the machine.

[Robert M]

Nader, As clean as one can do…. you’re the guy:)
Very inspiring for many !
Thanks for sharing those. Amicalement, Robert ;)

[Claudiu]

Nader, that came out professional! Good Job.

[Atifeh]

Robert, Claus,
Thanks. These works are inspired by your and others works in this forum.

[Atifeh]

I devised a sequence for cabling the machine. I made a lot of mistakes on the other machine and this procedure is the result of that experience. It is only a methodology which you can devise and adapt one for your MechMate. It is done in two stages and a number of steps:

Stage one: easy part
1-the 6mm yellow/green ground wire. It is the easiest but gives a measure of length for some other cables.
2-The Spindle cable. The length of this cable is more or less identical to Z and Y motors. The cables that I bought are marked with their length every meter. I cut a cable adjacent to its length mark to keep a record of the length used. Just at this stage I realized that the arrangement of the control panel is crucial for continuation of the wiring. Now that the control panel is ready, other wirings will be easy.
3-The 12x20AWG cable to the button box is the next. This cable is almost equal to the X1 motor length.
4-Routing the X2 motor cable is the next easy task.

Stage two: hard part
1-This stage is a bit hard since the limit switches and the estops should be connected together in series and for this purpose, suitable junction locations should be devised, also consideration for their shield should be regarded. The result would be 2 pairs of wires. One pair will carry the 12VAC, which is a separate cable, which upon activation will de-activate the main contactor. The other pair will go into the button box and joins the 12x20AWG wire to signal pin 10. this pair is only attached to the Estop buttons. A note: I personally prefered not to include the 12VAC cable in the control signal cable, although there was enough wire to do this. I thought AC voltage might noise the control signals.
2-A 2x24AWG cable will be allocated for the X2 home which will also end up in the button box.
3-A series connection for X1, Y and Z home switches with the same considerations as stated for limit switches, which again will lead to the button box. Or as described below, they can enter independently.

The 12x20AWG can be used in two configurations. A 10 wire configuration which has independent home switches, therefore enabling the use of “REF ALL” in Mach. The pins and port will consist of:
Estop signal , port #1 pin 10
X1 Home , port #1 pin 11
Y Home , port #1 pin 12
Z Home , port #1 pin 13
X2 Home , port #1 pin 15
B Home , port #2 pin 11
Z setter , port #2 pin 12
Pause , port #2 pin 13
Resume , port#2 pin 15
GND

An 8 wire configuration which uses X1,Y and Z home switches as one entry (the normal MechMate preference).
Estop signal , port #1 pin 10
X1,Y,Z Home , port #1 pin 11
X2 Home , port #1 pin 12
B Home , port #1 pin 13
Z setter , port #1 pin 15
Pause , port #2 pin 11
Resume , port#2 pin 12
GND

I decided to wire the 12x20AWG cable, from the button box up to its connector, as with the 10 wire configuration, but to connect the wires on the other half of the connector (i.e. the side that is connected to the control panel) as 8 wire configuration, therefore reserving the possibility of changing the configuration only by changing the wiring inside the panel. 12x20AWG was the only cable available in the market which would suit the purpose.

In addition to these stages there will be one motor and one home switch cable for the 4th axis which will be routed separately.

Routing the ground, the spindle and the Z home and limit switches:
7776

7777

Although it would be very neat to change the wires from the inside of the motor, I am not practicing it, mainly because I read somewhere that opening a stepper motor will ruin it.
I use in between connectors. In my opinion its advantage is that one can change the serial/parallel bipolar wiring only by cutting the motor side connector and rewiring it in a new connector, without changing the motor cabling. The shown connectors are 6 pin.
7778

7779

A plastic cable duct is attached to the inside of the gantry beams to transfer home side cables to the echain side of the machine.
7780

Details of series connection for limits and home switches:
7781

[Gerald D]

Second picture: I don't see a connection between the ground wire and the gantry?

[Robert M]

I'd say at the 4th pictur, it's there ;)

[Claudiu]

Rob, I guess its the Y Car but not Gantry!;)
It should have been visible in pic 2 where it comes out of the cable chain meeting gantry.

[Gerald D]

4th picture is connection to y-car (see idler roller and spring below)

In principle, there must be a ground connection between the 2 cable chains.

[Robert M]

Like this one (http://www.mechmate.com/forums/showthread.php?p=31&postcount=1) , looking at the 3rd picture top t bottom (“ from the table to the gantry….” )

[Atifeh]

Gerald, thanks for your keen eyes. It is apparent that I have forgotten it. It will be fixed by tomorrow.

[Atifeh]

Mounting the Z axis motor. I changed the spring to 2.5 mm wire. The 2.2 and 2 mm wire went into plastic deformation and the spring did not return to its original length.
7788

Correcting the missed gantry ground terminal and the arrangement of the cables for the gantry and the carriage:
7789

7790

7791

Mounting the Y axiz motor and spring:
7792

Mounting the X2 axis motor. The X1 axis motor will be installed after doing the button box.
7793

The Echain and the duct (open):
7794

End distributor of the cables equipped with flexible hose type cable glands. These will lead into the control panel.
7795

7796

Covering the duct and isolating its sharp edge against the cables:
7798

7799

Cables for the 4th axis:
7797


So far the easy part (stage one) is complete. Now to the hard part:

Before starting, I made this correction. The ground wire was removed from the secondary of the transformer and attached to the negative of main power supply.
7803

All the dangling wires are guided through a flexible hose. Well almost all.
7800

7801

The 12VAC (live) wires of the combined limitswitches and the Estop (in series):
7802

The cable was then checked for circuit continuity with a multimeter, each time activating one switch and of course checking it for a possible short with the shield. The two cable shoes will go into the Estop button of the box.

[jehayes]

Nader: This information and the photos of the wiring are extremely helpful. Thanks for your hard work in making them available. This is just where I am in building so the timing couldn't be better. Thanks again, Joe

[Atifeh]

Joe,
Glad they are helpful. If you need any details for the photos I can post it in here.

[Atifeh]

The 12VAC circuit of the Estop is shown in the upper picture, connected in series, cableshoe attached to end in two of the NC terminals of the Estop.
It might be worth noting that the gray cable is a 2x2x24AWG cable that brings both the home and limit switch of the Z axis. The home switches are tagged inside the box.
In the lower part the wires are identified; some will connect to the buttons, others connected to the 12x20AWG cable.
7848

A common ground is bound to the switches and the shield wires connected together and isolated.
Two wires were redundant, but I kept them just in case.
7849

The button box installed:
7850

Each and every button and microswitch checked by a multimeter against possible wrong signals, shorts with each other, or with the common or with the shield.

The X1 axis motor mounted:
7851

[Atifeh]

The motor cables through the connectors. They are tagged on the cable and on the panel for ease of tracing.
7877

Motor cables inside the flexible cable hoses bundled and passed through the glands, off the top of the panel, to provide a clear path of air flow through the fan.
The 12VAC limits/Estop cable and the 12x20AWG cable run through the cable duct to the other side of the panel.
7878

Motors tagged too:
7879

The next step which is very important is to:
1- check the continuity of the motor wires.
2- check for short of each wire with the shield (it appeared that I had one case short, which was corrected).
3- read one by one the resistanse of the coils from the motor to the drivers, while checking any possible short between the two coils. If you are wiring unipolar, this step is very crucial. On the other machine I had one wire mix up which was corrected before starting it up.

The exit side of the cable duct for the 12VAC limits/Estop cable and the 12x20AWG cable:
7880

Again each wire was checked with the related button and microswitch. This is also very important. One might easily mix up the left and right hand glove nature of these connectors and the colours of the wires. I did in one instance.
7881

The general view of the control panel, finished.
7882


Next, I will be configuring the computer.

Today I had an alarming suggestion. I was told that with a PCI parallel board, it is likely that the two ports (on-board port and the PCI port) do not send signal simultaneously. I don’t know whether it is right or wrong, but I hope he is wrong.

[Atifeh]

First of all, I wish you all a very happy and prosperous year ahead.
There has been some mixed activities within the last weeks which I will organize and post.

The main (under) board is placed on the machine and set for its position. With a centerfinder, the two outer sets of the holes are marked.
8024

The markings are indented with a center punch and drilled 6.5mm. the previously mentioned template that was used for marking the holes on the table beams is used to determine the respective matching holes on the board. A 6mm spring pin will fit inside the hole and through the holes on the board, fixing it tightly in place.
8025

8026

By using a piece of carbon copy paper, the indent marks of the centerfinder will be more visible.
8027

8028

The indents of the centerfinder are not deep enough for a reliable drilling, therefore a center punch is necessary to provide deeper indents for a better drilling. The 3 rows in the middle of the board were drilled 6mm.
8029

8030

A countersink tool with 6mm drill is used to provide the countersinks for the M8 screws.
8031

Since this machine will be used for jobs other than cutting, I need some work holding facilities.
The space between the table beams will be marked for holes which will be later used for different kind of clamps. These holes were marked with pencil and drilled 5.3 mm.
8032

Next they are drilled with a stepped drill for making countersinks for socket head screws. The depth, enough to hold M8 nut.
8033

The next step will be to drill the through holes to 9mm so that the M8 thread could pass through it.
The nuts,along with a washer are coated with epoxy resin and inserted inside the holes. Additional resin is applied to the exposed part of the nut.
While it is setting, a screw should be fastened from the other side so that the nut stands vertical in position. This is very imporatant.
8034

8035

The main board in place:
8036

The second board in place. Again screws must be fastened from below to keep the nuts in their vertical position. They should remain until the resin is completely set.
8037

[Robert M]

Darn Nader,
this simplistic but soooo effective approach is indeed nice & fresh to see. I like very much the way you approach you work & methods to execute !
To each it’s own, but for me,….yep …I like what I see….. :)
Thanks for sharing, it’s keeps my inspiration up !
Amicalement, Robert ;)

[jehayes]

Nader: As always, very nice work. I will use your technique when I do mine. Joe

[normand blais]

Hi Nader really proffessional build. To bad your MM is not working it would have zip trough those holes in no time.

[Atifeh]

Thanks Robert and Joe.
Thanks Normand for reminding me of the problem. I have to find a solution for it.

[normand blais]

Looking at your build I cant wait to see what you will be making with the machine.

[Atifeh]

I made a few clamps. Some from the tailings of the rails and some from 50x50x5 angle iron. Pieces cut, deburred, phosphatized and drilled a 8.5mm hole in the middle and 2 or 3 holes drilled and tapped for M5 screws. These screws are necessary to bite into the wood they are to hold in order to secure them on the table.
8058

8059

M8 Grip knobs are not available for longer than 60 mm. I bought some grips with internal M8 thread and thread rolled screws which are available in 1 meter lengths. The screw is cut to 120 mm and was fixed inside the knob with Loctite
8060

8061

If you intend to use your machine for mostly cutting jobs, then Normand’s note is quite valid, but if pocketing and 3D carving is what you mostly do, then the arrangement that I chose would be practical. But since profiling and cut-out is almost inevitable in most jobs, I devised another spoil board. It is cut from the scrap boards laying around the workshop, in different sizes, and drilled with a template, again made from scrap boards. It will sit on the main spoilboard and will be scratched, leaving the main spoilboard intact (I’m really spoiling the spoilboard!).
8062

Below pictures show the configuration of the clamps in action. Please note that the knobs are not fully tightened to show the arrangement.
8063

8064

[m_leblanc]

Looking at your build I cant wait to see what you will be making with the machine.

I wouldn't do anything with it, I would rent a showroom and have body guards to watch it while I would do the entertaining and answering questions.

It's almost a shame to make dust with that kind of tool

[Claudiu]

Absolutely True, Mars

[javeria]

hello atif - miss you - why did you stop posting?

[Atifeh]

Hello all,

Sorry for the delay. The machine is running but I have been through some adjustments, fine tunings, last minute works and also a lot of work on Artcam, DeskProto and PowerMill, CNCToolkit, and finding software for 4th axis. As before, all works are documented and I will tidy them up and post them in a short period.

[Atifeh]

In one of his posts, Gerald suggested to isolate the motor’s shaft end opening, to be sealed with some adhesive tape.

8340

Some other modifications were done prior to starting. I replaced the springs that hold the motors to the racks with the same number of turns, but with 2.2 mm wire. Now the gantry is really hard to sway. The reducers that I made are quite heavy and the current calculations for the spring tension don’t apply.

I also relocated the spring anchor as was shown in picture “Zgearbox05”, a little further up in order to provide more leverage.

The grub screws were another problem. Using Loctite would solve the problem forever, and I mean forever. That would include no possibility for the replacement of the pinions in the future. I tightened them, injected a couple of drops of aceton into the hole to reduce the surface tension and then some nail varnish around the screws. Later, I can loosen them by using some aceton.

[Atifeh]

Installation is quite easy. You open up the case, put it in place, start the computer and install the provided software. Mine is a Wipro (around $11) which works fine. You then open the device manager and note the port number from its properties which you will need later to enter in Mach.

I see a delicate problem though, which might be nothing more than an obssesion, but I appreciate any comment on it.
The parallel port inlets on my control panel are isolated from the common ground and the earthing (which are virtually the same).
I ran a ground wire from the computer case to the “earth hub” inside the control panel. From which a single 6 mm ground wire leaves for the earth well.

Now the tricky part:
The PCI parallel board is screwed to the case itself, so this one is actually attached to the common ground.

What is happening:
When the panel is off and the computer is on --- I see the LED’s on the breakout board having a dim glow.
More strangely: the LED’s on the 12/5 VDC power suppy also have a dim glow.

Now:
When I disconnect the power supply wires from the breakout board, the glow on all LED’s are gone, which indicates that there is a current flow from the computer to the breakout and to the power supply and back.

I wonder why is this happening? What is astonishing is the fact that all LED’s glow, disregarding the fact that some pins source current, and some sink.

So far there has been no problem regarding its operation, but this phenomenon annoys me. I hope someone could help me to figure this out.

[Atifeh]

I do not have the Mach3, but I have an old version of Mach2 which served me fine through years. I suspect I could activate the Z setting function with Mach2.

Posting the xml files will do no good since you all work with Mach3 and I suppose the xml configurations should be different. But I have provided the parameters for my tuning which might prove to be a useful guideline to set the machines.

Three configs were devised:
1- XYZ config
The B axis was used to slave the X2 axis for X axis.

2- XYZ-A config
The B axis was used to slave the X2 axis for X axis.
The A axis was used for rotary motion.

3- XYZ-B config
This config could be used with RhinoCAM, since you can define 4 axis milling strategies both in XYZA and XYZB.
The B axis was used to slave the X2 axis for X axis.
The A axis was used for rotary motion.

Since I had extra input pins, I assigned home switches separately to each pin, therefore accidental clicking on the “ref all home” will do no harm.

The following pdf files are the first two configs. Also a blank sheets which you might find useful.

8341

8342

8343

[Atifeh]

During setting up the Mach, I had the VFD switched off. As you have noticed, an automatic fuse is installed inside the panel which isolates the VFD, monitored by the “VFD Online” signal lamp. This has been implemented to protect the VFD, while going down its shut down sequence (which is quite long when you are testing the integrity of the panel). When I first switched on the VFD, and started the machine, the motors started vibrating and moving back and forth in an uncontrollable manner, when I turned the knob to increase the rpm. I shut down the system and tried to figure out what was the problem. Certainly it should be a noise, but where? To bring a long story short, the problem was in the DELTA’s “original part”, EMI filtered, and shielded cable that connected the mini keypad to the VFD.

You could see the initial arrangement of that cable in picture “panelwiring28”. It exerted a profound noise on the ribbon cables that connected the breakout board to the DB25’s mounted on the panel. By relocating it according to the below picture, keeping it out of the ribbon cables, all troubles were gone.
8344

Lessons learned:
1- When installing the VFD inside a panel, take extra care in isolate the wiring.
2- Never trust that decorative glands named noise filters on any connecting cable, disregarding of who is the manufacturer (period). It is an aesthetic something, only to cost me $25, which I could do it myself with a tenth of that price and wasting a whole afternoon to track a defect.

[Gerald D]

The grub screws were another problem. Using Loctite would solve the problem forever, and I mean forever. That would include no possibility for the replacement of the pinions in the future.

The big enemy of LocTite is heat. To loosen a LocTite joint, just apply the tip of a blue flame for a short period.

I see a delicate problem though, which might be nothing more than an obssesion, but I appreciate any comment on it.
The parallel port inlets on my control panel are isolated from the common ground and the earthing (which are virtually the same).

See this post (http://www.mechmate.com/forums/showthread.php?p=4373&postcount=25).

I do not have the Mach3, but I have an old version of Mach2 which served me fine through years. . . . Posting the xml files will do no good since you all work with Mach3 and I suppose the xml configurations should be different.

We changed from Mach2 to Mach3 some years ago. There wasn't a big change between XML's. But I do remember that the step and direction pins became reversed with each other.


Lessons learned:
1- When installing the VFD inside a panel, take extra care in isolate the wiring.
2- Never trust that decorative glands named noise filters on any connecting cable. . . . .

When we built our machines, we were concerned about the type of issues you faced, so we adopted the following:
- no connectors at the control panel. Run cables through glands (cord grippers) or just holes directly to BOB and drives
- mount the spindle VFD in its own enclosure away from the control panel

[Atifeh]

Gerald, thanks for all your helpful notes.

Actually I made the parallel port connections from Alan's thread. The metal part on the DB25 does not touch the control panel. So even if the parallel port cables should be internally shielded to those metal parts, there will be a discontinuity. It seems that there must be a problem with this isolation. I will check that out and post.

From what you mentioned about mounting the VFD inside the control panel, I assume the whole control panel has been turned into a noise trap due to VFD's function. Is that right?

I will also try some arrangements for future to route the cables directly to the drivers.

[kostas5]

nader
very good joop ..bravo and good luck
kostas

[Gerald D]

Hi Nader, we havn't heard from you for over 2 years - how is it going with you and your machine?

[Giants]

Amazing...