[binfordw]

I'm having an issue when cutting larger decorative parts, my table is losing its zero or however I should describe it. After I dry-run a fancy part, my gantry is out of square to the table, up to an inch. When it cuts inside patterns, and then goes back to cut the outside, it is off and ruins the part.

I marked 0,0, then dry-ran the part. When it returned to zero, it was .5" to .75" off in both x and y. I slowed the program down and ran again thinking high feedrates caused it, but it did this at lower feeds too. I've checked for loose bolts or pulleys, but everything seems good.

I'm not sure what is wrong, I assume its a motor setting in MAch3?

I recently just set the steps to give me correct travel, could this still be off a little and causing this? What throws me is the gantry going out of square though, both the X steppers should be doing the same movements but apparently they arent.


Any ideas?

[binfordw]

I did more testing last night, best I can figure is the two X axis motors are not staying in sync somehow. The mechanical setup seems fine, pulleys are the same and tight, same timing belts in good shape, nothing loose that I can find.

Since the one X axis motor isnt staying in perfect sync with the other, the gantry goes out of square.

[binfordw]

I noticed of the 4 stepper motors that came with my Bladerunner kit, 3 are labeled with an "S" piece of tape, one is not. Does the S have any meaning? Could I be using the "wrong" stepper for the X slave somehow?


I ran the table from X0 to X40" about 30 times and checked, the gantry was still square. It seems it has no trouble with straight moves, but it rather quickly becomes out of square when cutting designs/lots of arcs.

So far Ive tried feedrates from 90ipm to 350ipm, with the exact same results. I also upped the acceleration on X and Y to around 60, it was at 30. Motors respond quicker now but that didnt seem to correct the problem.



What in Mach3 could be causing my Slaved X motor to not stay in sync with the other?

[Torchhead]

Swap the motor with no "S" sticker to the Z and make sure both of the slaved motors are S motors and that both are tuned the same in MACH. Different acceleration and velocity settings (or steps per unit) on the slave motor will make it run differently. Sometimes slowing things down will let you see the problem easier (visually). Make sure the Step Active low on ALL the motors is the same in Motor Outputs. DIR active low can be different depending on which way you want then motor to spin.

The "S" sticker on the motor is for our in-house assemblers. The S means wired serial. That 8 wire motor can be wired parallel and draws 7A and we use it with as a Z motor in the larger Plazpaks with the larger Geckos. Because it;s missing does not mean it;s wire wrong but you can check it with an DVM with a low ohms scale (like 20 ohms full scale). Before measuring the coils short the meter tips together to get the resistance of the leads. A series wound motor will have about 1.5 ohms of DC resistance. A parallel will have about 1/4 of that.

All the units are put through final test with all the motors hooked up and they would have caught a parallel wired motor because the low speed resonance setting would have been different on the Gecko. Can't say they are perfect but out of hundreds of motors we have not seen one get sent wrong. Contact me on the CandCNCSupport form for more timely response.

TOM caudle
www.CandCNC.com

[Torchhead]

Also if it continues on the same axis after the motor swap and test THEN swap the slave Gecko with the Z (two mounting screws and the 12 edge screws on the Gecko). Through measurements determine WHICH motor is the suspect and swap it to Z and use the Z drive for substitution.

[binfordw]

I posted on CandCNC forum, thanks.

I did what you suggested and there was no change. The gantry still gets out of square when I run my program.

I've checked everything I know to check and cant find anything that could be causing this. Everything is tight mechanically, anyone have any ideas on what could cause my X motors to not stay in sync?

Mach3 motor settings are identical for X,Y and A

Also, if I just jog the table from X0 to X40"+ and back over and over, the gantry stays square..

This is driving me nuts, if anyone has any ideas please let me know!

Thanks

[binfordw]

Well,

I tested a program of radom sized squares to see if arcs were cauing the problem, but it still moved out of square.

I can jog the table all over the place and it stays square... But I cant run a program without it going out of square. Could this be a problem with sheetcam settings somehow?

Can someone send me a program to dryrun that was made with a different cam software?- or perhaps even sheetcam, maybe something is setup wrong on mine.


Edit- I tried a diff post processor with the same results. I looked through sheetcam options but didnt see anything that might cause problems.

[binfordw]

Uh..


Well anyway, I think Ive gotten somewhere so far, Now the gantry seems to be staying straight, but my table is still "losing zero". I thought this was because the gantry was getting out of square, but its still off when the gantry seems to stay square.

I marked the sheet and ran my program with the torch off, when it finishes and returns to 0,0 on the part, its off about .450" on x and y..


Any ideas for this? Ive tried various step settings in the motor tuning section of Mach, with no real change in the endpoint.

[binfordw]

After more testing,

The stepper motors arent returning to zero.

I removed Y, and marked the pulley. The longer I run it from "0" to 2" and back, the farther it gets away from the "0" mark.

I pulled the X motor, and it did much the same.

This was at low speed settings. (170ipm feed, 15 acceleration in motor settings in Mach3)


Any ideas?

[Kipper]

Is the pulley/gear loose on the shaft? and can you set the steps per unit by using more or less all the table travel?

Or you're losing steps somewhere.....might have a runt stepper motor...

Can we see some pictures of the rack and pinion/belt set-up?

[binfordw]

Thanks Kipper,

I tested the stepper motors, while I did leave the pulleys on them, they are tight. I just left the pulleys on to make it easier to mark and visibly see if the motor was not returning to "zero"

Since the motor doesnt return to zero that rules out any issues with the pulleys or belt setup, I guess its in the motors somehow, or in the bladerunner controlbox or Mach3. Im just not sure what to test or change next :/ Waiting on a reply from the manufacturer for suggestions.

[MrWild]

slow down your rapids to 60ipm. and your feeds to 40ipm. Sometimes even seemingly slow speeds of 90ipm have problems with lost steps. Are the cables shielded, especially the motor cables in the control box? I bought a project and the motor cables from the drives shielding wasn't grounded and they caused problems in with random steps being generated.

[binfordw]

Well I have tried from very fast to very slow, both in feedrates and acceleration values, and the results seem to stay constant.

Also, Ive checked X,A and Y motors now and all 3 seem identical in how they react. This explains why my gantry was going cock-eyed on my, since A run in reverse as a slave to X, the error is reversed.

As far as shielding, Im not sure. It is a Bladerunner kit, so it was all pre-wired. It does state clearly on the case "no shielding required".


It doesnt seem random at all, I havent checked for exact amounts of error, as in, how far off after a set amount of moves, but all 3 motors seem identical in amount of error after being tested. Faster feedrates and accelerations dont seem to change the error at all.

[millman52]

Well I have tried from very fast to very slow, both in feedrates and acceleration values, and the results seem to stay constant.

Also, Ive checked X,A and Y motors now and all 3 seem identical in how they react. This explains why my gantry was going cock-eyed on my, since A run in reverse as a slave to X, the error is reversed.

As far as shielding, Im not sure. It is a Bladerunner kit, so it was all pre-wired. It does state clearly on the case "no shielding required".


It doesnt seem random at all, I havent checked for exact amounts of error, as in, how far off after a set amount of moves, but all 3 motors seem identical in amount of error after being tested. Faster feedrates and accelerations dont seem to change the error at all.

Generally shielding isn't needed with a stepper system. If it was some sort of "noise" causing the problem(s) it would be a 1 in 1,000,000 chance that all 3 motors would end up off the same amount after each test run.

If you mark the amount of error after test running your part. Does the machine repeat the same amount of error every time?

I assume if you do not re "0" and run the same program again you will end up apx .900 off?? Or if you run a shorter or longer program from 0.00,0.00 you end up with less or more error respectively?

[binfordw]

If you mark the amount of error after test running your part. Does the machine repeat the same amount of error every time?

I assume if you do not re "0" and run the same program again you will end up apx .900 off?? Or if you run a shorter or longer program from 0.00,0.00 you end up with less or more error respectively?

Pretty much exactly. I didnt count the number of times I ran the program to see, but it is very repeatable in the amount of error.

Before I removed the motors, I set a dial on my Y axis, and ran the test. From 0 to 5, then 0. Each time it returned to zero it added .005" on the money. So it went from 0 to .005, .010, .015, etc. Im certain its behaving the same way with the bare motors, adding the same amount of error upon returning to "zero"- its just a little harder to measure.