Losing My Way

I am not sure if 40 is excessive for Y or not. I kind of settled on these values experimentally. I should be lower than your 4x4 since my mass is going to be greater (I presume).

I am preparing to run some tests with marks on the shaft to see if that's the root cause.

Gantry weights should be similar. If anything mine might be heavier, I have some heavier motors and some dust collection brackets mounted.
maybe we can compare configs at some point, i arrived at all my values experimentally also.

Ok, the pen definitely reveals a problem with one of the two motors on the gantry. I ran the very same job that I posted the picture of. I ran it twice. Each time the one stepper was dead on and the other was off an 1/8 of a revolution. Obviously I can't tell if that's 1 - 1/8, 2 - 1/8...

I am disassembling the drive right now and the pulley seems to be on there solid. I had loc-tite on the set screws but I'll pull it and put it on the shaft as well.

Ok, I think this is pretty telling but I would like someone else's input as well.

I disassembled the motor assembly for the side that is losing steps. It looked good. The pulley was solid on the shaft and there was no evidence of slippage. I re-assembled with loc-tite on the shaft as well as the set screws. I re-ran the program and the side is still losing steps.

The next thing I did was go to the driver and swap the cables that go to each of the motors on the gantry. So the machine gantry is now flip flopped. The problem follows the driver to the opposite side. That side is now the side that loses steps. It must be the driver right? I should also note that the input signals are still coming from one set of pins on the lpt port so both drives should see the same signal.

did your mark with the pen between the pulley and motor shaft, or between the pulley and something stationary?
If the marks between the pulley and shaft don't line up anymore then it is slipping.
If you marked between the pulley and something else, I', not sure I guess it just tells you it didn't return to the same position.


It seems possible your driver is bad, but I'd rule out the power supply first.
with everything powered down, unplug your X motor, to reduce current draw, then power up and re run your tests.

I seem to remember reading about the g540 doing strange things when the power supply was barely adequate.
did you get it from cncrouterparts also? is it the 7.3a or the 12.5a supply?
they used to sell the 4 axis g540 kit with the 7.3a supply, but now include the 12.5a supply, presumably because of problems.

thinking about it, if you need to actually cut something for the problem to show up, you obviously cant do that without the X motor working.
hmm.

I was only looking at the shaft in relation to a fixed/static location.

The power supply is indeed the 7.5A supply. Perhaps that's the problem. I don't actually need to be cutting anything to get the problem to appear. I can cut air and it shows up. I am going to do as you suggest and disable the Z and X to see if that eliminates/attenuates the problem.

It appears the power supply is not the issue. With the Z and X unplugged I still get missed steps. I have lowered the acceleration down to 25 on the Y axis but that doesn't seem to help matters either.

If you decide to replace the driver, you can get a g250x by itself, they are the single axis module inside the g540 case.
It's still $80, but beats $300 to replace the whole g540.

You may want to email gecko and tell them the problem, they repair the boards and may have other suggestions.

Hi,

I haven't been following this thread, was just about to suggest you posted your config when I saw you had at the beginning.

Why would you expect what appears to be a Nema 23 powered, big router gantry, to be able to move at 120 inches per second?
10 feet per second, really!?

This with MAX_ACCELERATION = 40 inches per second per second and STEPGEN_MAXACCEL = 80 instead of the 10 - 20% headroom recommended.

Any mechanical and electrical issues aside, I would suggest you drastically reduce your figures.

By drastic I mean divide your max velocity by 100 and acceleration by 10, with the stepgen max at plus 20%

You also have a very low base-thread figure at 30K. Make that 50K for now, irrespective of any optimistic latency results you may have had.

THEN do tests to see if one driver is U/S, which is doubtful, then slowly increase the figures until you hit problems again.

regards