PDA

View Full Version : Max. speed of the MechMate - moving, cutting, jogging . . . .


sprayhead
Thu 08 May 2008, 02:13
Although this thread started on 3D machining, later posts discuss standard cutting & jogging speeds . . . .

Hi Guys,


For 3D contouring, what would be the the fastest feed possible with the Beast?

Francis

Gerald D
Thu 08 May 2008, 02:28
I have no experience of this, but I believe that the limiting factor is your CAM software (and the computer processor power) that is controlling the motors.

sprayhead
Tue 13 May 2008, 03:43
Hi Gerald

Apart from the frequency limit from the computer/CAM software I believe that there would be a limit established by the actual milling operation? Say... until the motors start loosing steps or heating up too much, I suppose.

Here comes a new thought... What kind of maximum cutting force can the mechmate operate until it gets to a point where it starts to loose steps or forcing the motors too much? And actually, how do you know that you are loosing steps? :confused:

Doug_Ford
Tue 13 May 2008, 16:47
Sprayhead,

I've lost steps a couple of times when I was doing stupid/crazy stuff. When I was jogging the gantry at 24 inches per second, I lost steps. The way I knew was that it made a LOUD noise and locked up. The other time I lost steps, I was using a slightly dull 1/4" carbide endmill, cutting 3/4" mdf at full depth and the feed speed was 6 inches per second. That time, I didn't hear anything and it didn't lock up but the design was distorted.

dmoore
Sat 17 May 2008, 23:39
I just completed my MechMate enough to get it cutting (still have to do eletrical, dust management, painting, etc) and without any tuning or adjustment, I was able to get it to cut at 210 inchs/min. Here is the link to the YouTube video:

http://www.youtube.com/v/jGmz7CSeJqc


(sorry for the quality - YouTube cuts it down)

Robert M
Sun 18 May 2008, 01:08
Very exiting & motivating… Thanks for sharing
Keep those videos & more pics in !
Congrad. Robert

dmoore
Tue 20 May 2008, 06:36
I was able to get my MechMate up to over 800 inches/min but it appears that steps are being dropped. I suspect that the problem is within Mach3 since I can pass the 45kHz output test and the failures occur when running two axises at the same time. I guess I now see the reason for the SmoothStepper. It seems that with the 7.2:1 with the 30 tooth pinion, the "real" limit is around 450-500 IPM - anyone have experience at these speeds?

I'm also curious what other people have set for acceleration for the X and Y in Mach3. The gantry isn't light weight but I am impressed with how well those geared steppers can do to control it.

Here is a video of a cutting at 450 IPM, then it switches to 150 IPM for the lettering and then at the end, it switches back to 450 IPM to do the final cutout:

http://www.youtube.com/v/4DbFRGkc084

isladelobos
Tue 20 May 2008, 09:20
Yess David.

Your are a very good beta tester. Thanks !!!

This is a good velocity for cut thin sheets of iron whith plasma.

We need 16000mm/min = 630 inches/min (aprox) for a hypertherm cutter and 0.5mm (26GA)of sheet thin.

I think it is possible with 32T.

Your work is really good.

ekdenton
Tue 20 May 2008, 11:23
:eek::eek: I haven't been above 100ipm yet, and I thought I was cutting fast. Is it safe for the motors to be starting and stoping so fast? Given the weight of the gantry ?

dmoore
Tue 20 May 2008, 11:44
Wow, nice speed....

Maybe Gerald can comment on this... It seems that theoreticly the MechMate should be able to do 1,178 IPM (7.2:1 with 30 tooth pinion).

Looking at this speed/torque chart from OM:
http://catalog.orientalmotor.com/Asset/pk296b2a-sg_stc.jpg

They list the 7.2:1 motor as falling off at 250 RPM (depending on the wiring and voltage settings) which must be the final output, reduced speed. So if I multiply 250*7.2, I get a max motor speed of 1,800 RPM. Punching that number into the spreadsheet (with DP of 20 and Pinion of 30) we get 1,178 IPM on the table (1,375 IPM with a pinion of 35 or 786 IPM with a pinion of 20 - all with the 7.2:1 OM motor).

Of course, it just pushes the problem from one place to the next - the minimum frequency for Mach3 is now 60kHz - pretty much well beyond anything a parallel port is going to put out with more than a single axis going. So, it "seems" that you can get the ~1,200 IPM if you use something like the SmoothStepper.

Of course this totally disregards acceleration and torque issues (though at 800 IPM on the X axis, I wasn't able to extert enough force to stop it).

Is anyone actually running these types of speeds (1000 IPM +) on their MechMate?

isladelobos
Tue 20 May 2008, 13:28
Most of the plasma cutters or drill, how come this far?

I do not think so.

but i need know the MM posibilities and improve


an example : http://www.farleylaserlab.com.au/performance_details_5218_5140.html

dmoore
Tue 20 May 2008, 21:35
I haven't been above 100ipm yet, and I thought I was cutting fast.

The geared shotbots do cutting at 600IPM and jogging at 1,500IPM:

http://www.shopbottools.com/PRSalphaspecs.htm

Is it safe for the motors to be starting and stopping so fast? Given the weight of the gantry ?

Hence the question about the acceleration. I'm still impressed how fast these geared motors can stop a few hundred pounds of gantry and y-car.

Crank up the speed - with the 7.2:1's, 400 IPM shouldn't be any problem - material and bit allowing.

Does anyone know what interface the RPSAlpha's use to get to that 1,500 IPM?

smreish
Wed 21 May 2008, 06:57
These speeds are very achievable, you will just have to size the drive train to react accordingly.

The PRS Alpha has one very important difference from the MM. It uses DC servos and we use steppers. The servo motor rotates up to 3000RPM, the steppers are only good for up to ~1500 RPM, which quickly illustrates why the PRS alpha can achieve speeds double the MM designed speed. If you choose to use Servos, you too can achieve these speeds. Just realize your going to be spending an additional 3 to 4 thousand dollars for the motor package.

Sean

D. van Randen
Thu 22 May 2008, 02:55
Hi Sean

Thats incorrect the PRS Alpha derives the alpha part of its name from this stepper series from vexta. Its a hybrid that uses feedback similar to servo, but is a stepper at heart.

http://catalog.orientalmotor.com/viewitems/stepping-motors-motor-driver-packages-/as-series-stepping-motors?&plpver=11&forward=1&backtoname=&pane=

D (a bot owner)

smreish
Thu 22 May 2008, 05:19
Van,
Thank you for the correction. I am not a bot owner, nor have been. My experience is from trade show attendance. The Buddy demonstration I attended appeared to express that it was using closed loop servo. Clearly, it is comparable to servo, but not.

Thank you again for the clarification.
Sean

isladelobos
Thu 22 May 2008, 07:35
cutting the mamba parts in alluminium and soldering it entire y-axis, reduce the Weight and torque needed for move this in horizontal. but also reduces the vertical weight.

dmoore
Thu 22 May 2008, 07:43
These speeds are very achievable, you will just have to size the drive train to react accordingly.

The PRS Alpha has one very important difference from the MM. It uses DC servos and we use steppers. The servo motor rotates up to 3000RPM, the steppers are only good for up to ~1500 RPM, which quickly illustrates why the PRS alpha can achieve speeds double the MM designed speed. If you choose to use Servos, you too can achieve these speeds. Just realize your going to be spending an additional 3 to 4 thousand dollars for the motor package.

Sean

I'm sorry - when I went back and re-read my question it didn't clarify which area of the ShopBot Alpha I was wondering about - the interface to the drivers. I'm already sure that on jogging and cutting that the MechMate can meet the speeds of the ShopBot Alpha from a purely mechanical point (as mentioned in posts above.) I'm curious as to what interface they are using - because clearly the parallel port can't cut it at those speeds (~1000 IPM). Are they using something like the SmoothStepper that bypasses the parallel port?

dmoore
Thu 22 May 2008, 07:49
cutting the mamba parts in alluminium and soldering it entire y-axis, reduce the Weight and torque needed for move this in horizontal. but also reduces the vertical weight.

This is a really interesting question - clearly something like the (current) shopbot has an advantage due to it's lightweight aluminum gantry. I can't imagine why someone couldn't either braze the steel gantry carriers to aluminum 2"x4" gantry tubes. Of course I think the Y-car and it's parts would likely need to still be steel. The other downside to lightening the gantry would be it's greater susceptibility to vibration due to the weight loss.

smreish
Thu 22 May 2008, 08:42
David.

Your assumptions are correct. The mass of the gantry is a great dampening tool to counteract the forces imposed by the spindle and cutting tool. The actual weight of the gantry is really nominal. As Greg has clinically shown the spring scale method of force needed to move the gantry is only a few pounds. In Mechanical world this is referred to the amount of "drawbar" needed to move an object taking into account all forces in the move. (rolling friction, durometer of wheel material, # of tires/wheels, rotational diameter, surface area contacting rolling surface, etc.)

*note, I'm not an engineer, and the opinions stated are clearly my own* :confused:

isladelobos
Thu 22 May 2008, 10:03
This is other cuestion.

When it becomes the engine and gear with the spring, there is already a kind of clip that does not allow lifting the gantry with his hand.

If raising the strength of the spring, maybe you do not need so much steel?

dmoore
Thu 22 May 2008, 10:31
David.

Your assumptions are correct. The mass of the gantry is a great dampening tool to counteract the forces imposed by the spindle and cutting tool. The actual weight of the gantry is really nominal. As Greg has clinically shown the spring scale method of force needed to move the gantry is only a few pounds. In Mechanical world this is referred to the amount of "drawbar" needed to move an object taking into account all forces in the move. (rolling friction, durometer of wheel material, # of tires/wheels, rotational diameter, surface area contacting rolling surface, etc.)

Good point - I was barely able to lift the completed gantry onto the rails (nearly 200 pounds) but I can push the entire gantry WITH the y-Car and all the motors with one finger (without motors engaged on the rack).

D. van Randen
Thu 22 May 2008, 17:27
Hi David

To answer your question about interface used on shopbots.

They use a proprietary control board, it has inputs / outputs , relay interlocks for spindle/router control , emergency stop, etc. It uses a small microcontroller interfaced via a rs232 serial port with a USB 2 convertor attached to it . Through a suitable USB cable it connects off to the PC and communicates via a USB 2 port mapped as a serial port that the Shopbot control software uses.

Using a 1 gig or more PC with win98 or using a 1.8G or > XP Pro will allow good control of the bot providing you use USB 2.0 to communicate with the bot. ( You may want to look into the smoothstep controller for the higher speeds.)

The advantage of the small microcontroller is that it allows a shopbot "G code" to run. The code is more like early basic and has move commands like g code that can be programmed as well as GOTO , GOSUB, INPUT, etc. Very easy to program .

D

Gerald D
Mon 09 June 2008, 07:05
. . am not a bot owner, nor have been. My experience is from trade show attendance. The Buddy demonstration I attended appeared to express that it was using closed loop servo. Clearly, it is comparable to servo, but not.

ShopBot have two ranges, Alpha & Standard

- Standard uses the same style/type of geared stepper motors that we talk of here at MechMate. Their performance spec is published here (http://www.shopbottools.com/PRSstandardspecs.htm).

- Alpha uses a geared stepper motor very similar to what we have here, but with an important addition.....an encoder on the back of the motor to give feedback to the drive. That's where the "servo-like" closed loop control comes from. The Alpha performance spec is published here (http://www.shopbottools.com/PRSalphaspecs.htm).

Gecko is (was?) developing a drive that can talk to an encoder on the back of a standard stepper motor. This will be much more affordable than Oriental Motors Alpha motors and drives (http://catalog.orientalmotor.com/viewitems/stepping-motors-motor-driver-packages-/asc-series-stepping-motors?&plpver=11&forward=1&backtoname=&pane=). The performance should be on par.

J.R. Hatcher
Mon 03 November 2008, 18:49
Resurfaced the table today, and as promised here is the video. Before I ran the machine 1000 ipm. This time I decided not to run it at that speed. I am cutting .050" deep, .75 path. Sorry about the picture quality but it's all I got. :o

http://www.youtube.com/watch?v=EwO_KL1GRww

Gerald D
Mon 03 November 2008, 22:13
JR mentioned before that he is using "soft limits" for that surfacing program, and one can see how Mach ramps the speed down before the gantry reaches the end of the table.

The first thing I did with watching that video was to time the "strokes" against my watch and verify the 1150 ipm . . . . it looks right if the table is 8' long. Not that I doubted JR's word, but because we have had some funnies with indicated speed vs true speed. (In Mach3 r1.84 we had our y-axis running at half indicated speed, In ShopBot we had all axes running slower than indicated).

Another thing to remember is that JR is using "big" motors and belt-drives. Can be considered as "non-standard" for the typical MechMate builder.

And then it must also be realised that 1150 ipm is on the limits. For table surfacing, a lost step is not the end of the world (except for hitting the hard stop on the backstroke). Not a recommended speed for general pupose work. But nice to know that Mach can get it there, and that mechanically it is okay without jumping off the rails or pinions jumping out or anything else nasty like that.

cncb
Tue 04 November 2008, 04:41
Love the video JR your machine is outstanding. ANd to think most commercial high end machines for the most part cut at or below that speed!

J.R. Hatcher
Tue 04 November 2008, 05:30
My spoil board is 49" X 97". The jog speed G00 on my machine is 1175 (so I can go faster if I need to):D:p. I have soft limits set at G53 X0 X101 and Y0 Y53. The soft limit ramp up and down is 11.5". It was only running 1150 ipm for 6' the rest of the time it was either ramping up or down :D.
The machine gave no indications at all that it was under the least amount of strain, everything sounded normal, smooth and under control. :cool: I don't think it missed any steps at all. ;) Next time I'll check for that.
The dust collector was on, it comes on with G-code/software. I had been working on the pipe system and did not have the MechMate reconnected, human error :o. I thought the last job (1/8" bit 1/4" material) was a little dusty :confused:. After connecting the pipe it did an excellent job even at that speed and volume.

J.R. Hatcher
Wed 05 November 2008, 14:16
I made a mistake in my last post. The jog speed on my machine is 1375 ipm not 1175 ipm. :o

domino11
Wed 05 November 2008, 15:02
Show off. ;)

jrabeneck
Mon 10 November 2008, 08:02
I downloaded these files and opened in Inventor 2009. They opened perfectly. Good job.

noel76
Thu 18 February 2010, 20:58
Mr. J:R#4, what motors and reductions you have on your machine? it looks amazing runing at that speed.

bradm
Fri 19 February 2010, 05:28
You can find out in J.R.'s build thread (http://www.mechmate.com/forums/showthread.php?t=329&page=2).

PK299-01AA Direct Drive. I think he may have changed to belt drives later. It's worth looking through that whole thread, though, because J.R. did a lot of impressive innovation in many areas.