Mesa 5i25/7i76. Looking for some help.

Hi,
How have you got on, |I have just setup my ClearPath SD servos and they seem to work really well. Only working on the bench at the moment will be fitting them to a machine in about three week (just waiting for ballscrews).

Still getting my head around the HAL stuff. The Clearpath have some great feature like hard home ie you drive it into a hard stop, and it will record it as home. just hooking up the HLFB wire, so if they fault it will stop the axis.

Seems a shame to ditch the ClearPath technology and go back to std steppers. If I can be of any help let me know.

I've switched to NEMA 23 stepper motors now and they are working good. I couldn't get the ClearPath to track properly and couldn't figure out what the problem was. It could be that my step drive settings are off but I haven't played with them since then.

Hi usert_I, I am not a Linux specialist but I work for Teknic, the manufacturer of the ClearPath integrated servo motor. Using motors from different lot numbers or purchased at different times should have no affect on the performance. You mentioned that you switched motors to different axes and that the "slow motion" problem followed the motor. You also mentioned that both motors are set up as 800 counts per rev but that one motor seems to run about 1/10th as fast as the other. It sound to me like one of your motors is an -ELN model (6,400 cnts/rev) and one of your motors is an -RLN model (800 cnts/rev). This would explain the difference in behavior. It may not be the problem but you should check this out first.

Hi, set my step pulse timing to a little more than minimum Teknic spec, that gives a little and enusre pulse are alwasy within spec. Step Pulse 1.2uS, Step|Cycle 2.2uS, DirChnage and Step 50nS, Dir hold time 1.2uS. I looked at the pulse the MESA card puts out a scope and found them to be quite accurate, but at 1uS (default I think) they are very tight. The edges on the pulse are not square and there can be some jitter.

If you want to sell you ClearPath, let me know what size they are and PM me.

Cheers

TeknicTom, they were both -RQN models (1/4" shaft).

When I set my step pulse timing I followed the minimum Teknic spec as a guide. Mine was as follows:
Minimum step pulse width = 1 uS (or 1000 nS).
Minimum step cycle time = 2 uS (or 2000 nS).
Minimum time between direction change and step input change = 100 nS
Minimum direction hold time = 1uS (or 1000 nS).

I was not aware that it can be a bit tight and did experience some jitter. So that could be the issue and using more than the minimum may solve that.

soehlrich, I sent you an email.

I generally try to run timings about double what minimum timings are for most drives I set up (1.5 if the times are long). Usually the max feeds the machines are capable of don't come anywhere close to pushing even those limits.

In addition there is no gain to be had by making the setup and hold times short.
That is, setting them to 5 usec will cause no performance loss at all, even on a high
performance system. This is because this delay is only invoked when reversing
where the pulse rate is very low.