X3 Lost Steps

[pzzamakr1980]

I am having some trouble with lost steps. I am running 276 nema 23 steppers on the x and y and a having a lot of trouble with lost steps on those axis. The acceleration is set to 1 with rapids at 60. I turned those down in an attempt to fix the lost steps. What is everyone else using for their steppers and acceleration settings. I have gecko 203s and a big 72volt 12amp PS.

I was debating getting larger steppers, but know that that may not fix the problem, since it might just be my settings.

[cadmonkey]

I haven't done any actual cutting but I have 330's @ 50V/15A and I dropped my rapids down to what you're running while I was setting up the X. I'm taking it slow sorry I can't give real data. I can't physically make the axis loose steps at 50ipm with any force my body can exert. I hope to get all my limits and Y preload taken care of in the next month and then can actually make the chips fly.

Interested to hear what others are running. I'm already considering stepping up (pun intended) to 400+...

[pzzamakr1980]

I cant cause it to lost steps by pushing on it either. Its when its making multiple complex moves, such as pocketing, contouring, etc. What do you have your acceleration set at?

[cadmonkey]

5 in/s, but I expect that will change too. It was an arbitrary starting point. I'm running EMC2 - what are you running?

[pzzamakr1980]

Mach. That seems a good bit high.

[rc-monster mike]

I had a similar issue with the G210 drivers on my mill. No visible lost steps, and I really couldn't stall the motor on my own, but it missed steps as noticed in the finished part. It turned out to be a communication issue of sorts. I am using a Camtronics controller, and the BOB used seemingly could not spit out the info fast enough for the 210 drivers, and it created this problem. I also had another Gecko controller with 201 drivers and a pmdx122 BOB, so while troubleshooting the problem, I swapped drivers. The 210 drivers worked fine in the controller using the pmdx BOB, perhaps due to the buffered signal (BOB buffers the signal before, and the 201 drivers haven't missed a beat in my Camtronics controller. I had several discussions with Dan at Camtronics while trying to work through it, and it seems that the 210 drivers actually require a longer pulse width(longer than the 40ms Mach can deliver) , so the buffered signal used in the other controller likely provides this. The 201 drivers don't require as long a signal and have worked perfectly in the Camtronics board (Dan tried talking me into the 201 from the start, but I wanted the extra step multiplication options that the 210 drivers offered at the time).
I am not sure if my explanation is helpful or not - I generally understand the issue I had, but I am not up on the lingo so much. Basically, I had a communication problem related to the signal pulse width. The drivers simply required a longer signal, which Mach alone could not provide. If you have the same problem, it may be fixed by adjusting the signal length(in motor tuning I think), or by using a buffered BOB (seemed to work for me, as the 210 drivers work perfect with the Pmdx122 BOB, and the 201 drivers work perfect with the un-buffered signal on my Camtronics controller).

[S_J_H]

Are you using ballscrews?
I have nema 23 425oz/in and only a 28 volt xylotex system. No problems with lost steps at 60IPM and 5 for acell using Mach.

Steve

[pzzamakr1980]

Yeah, I'm using the cncfusion deluxe kit.

The BOB might be something to consider. I am using a straight plain ole BOB with a SIIG card in the computer. I wonder if its not receiving a strong enough signal to push the motors properly. I will also check the motor signal length too.

What is everybodys acceleration set at. Thx

[rc-monster mike]

My machine is a Minitech MiniMill 4, so it isn't apple to apples(linear bearings, ball screws), but I run it at 330 IPM rapid speeds and the acceleration setting at 10 (about .5 second ramp up). I am using 254 oz/in nema 23 steppers with I think a 42 volt +/- power supply. When I was having the problem, I went all the way down to 40-50 IPM with the same results (no noticed missed steps, but chamfers and like features were offset from each other). It will easily go to 350 ipm now.

[pzzamakr1980]

So you are saying you adjusted your acceleration to fix it?

[Eddiep]

See his first post above - not an motor tuning issue, but a communications issue with the drivers.

Originally Posted by pzzamakr1980 So you are saying you adjusted your acceleration to fix it?

[rc-monster mike]

Originally Posted by Eddiep See his first post above - not an motor tuning issue, but a communications issue with the drivers.

Exactly. I was trying to say that motor tuning DIDN"T fix it - same problem was present regardless of motor tuning. One thing I did notice while I was having the problem - it didn't miss steps when I ran it in single line mode (constant velocity turned off). The motor tuning never changed the error. It was fine in single step mode, but always was "off" in CV mode, regardless of motor tuning. This is when I started farting around and discovered the communication error - the 2 controllers I have were very similar, but with different drivers and a different BOB. The 210 drivers work fine with the buffered BOB - I suspect this is due to their longer required signal pulse, which Mach alone could not deliver. Try running in single line mode and see if the error is still present. If it is, you may need a buffered BOB like the PMDX 122 to accurately run in CV mode. You may try adjusting the signal length as well - according to Gecko, the G210 drivers need a 50ms signal, but Mach only goes to 40ms, so without the buffered signal, I was getting error - the buffered signal seemed to fix it. I am not familiar with the G203 drivers you have, but it is reasonable to assume your issue is similar to mine, given the similarity of the actual results in use.