Vectrac CNC Knee mill refit from Fanuc 0M to Camsoft

may I am full of beans, but I am reading those bumps in the bottom graph as a following error, not an overshoot. Wrong???

Getting tired of this for today. I will put T14 into the control tomorrow and try to run some parts. See if it throws any tolerance and or position errors.
 
They are both, I think. :cautious:

I need to clarify a term here Following Error is a deviation from 0, can be lead or lag.

In this case, where the move is in the positive direction, any deviation above 0 is overshoot, or lead error.
When moving in the positive direction:
Error >0 is lead error
Error <0 is lag error

I hope I have the signs correct in the above explanation.

TE is the difference between commanded (RP) and actual (TP) position, but I don't know for sure if it is calculated by RP-TP or TP-RP internally in the Galil. That would change the sign. I need to run a test to confirm this. I'll do this later and report back.

A small amount of lag error is not a bad thing. Lead error is not good, that means the actual position is ahead of the commanded position.

In any case, T14 still looked the best.

Have a relaxing evening. We can play tomorrow if you have time. :)
 
I will make time to run the control. But I got a real job to attend to. two apple deliveries to grocery, and a ton of picking - all day long.

OK lead and lag make more sense for definition. I am reading this as a lag error. When I collected actual data last week, it was lagging errors.

BTW, with these huge 1.8 KW motors, the table mass is negligible. I can use same settings on all three axis.
 
Ahhhh, no wonder you are only reaching 2 volts. Those motors have a lot of torque. They accelerate my 2000 lb lathe carriage to 400 IPM in about 100 ms, and that's only because I have it turned down a bit. There is more there but I don't want to beat up the machine too bad. It already bounces a 10,200 lb machine around a bit, you can feel it in the floor.
 
OK, I moved the carriage while watching the TE and going in the positive direction, if the following error is positive then it is lag error.

TE=RP-TP

So, moving in the Positive direction:
Following error >0 = Lag error, a little bit is good.
Following error <0 = Lead error, bad. Also lead error on stopping = overshoot.

and, moving in the Negative direction:
Following error >0 = Lead error, bad. Also lead error on stopping = overshoot.
Following error <0 = Lag error, a little bit is good.

Moving faster will normally increase the magnitude of lag error.

In either case, when properly tuned, once stopped, TE should = 0 +/- 1 count.

So in your graphs above, following error excursions into negative territory = Bad. But you may not to be able to get rid of all of it.
 
IMHO, the signs also change when decel vs accel. Yep, the counts are getting ahead on decel; because its a tad behind decelerating. Not such a bad thing.
 
ran a few programs... she purrs like a kitten. time to make some serious chips.

FWIW, I wrote and debugged a generic trocoidal mill slot program this morning. It makes a series of "D" shape moves as it works its way down the slot. I'll test it with an endmill when I catch up on my real job. I am "old fashioned", prefer macros of cutting ops over trying to get CAM generated gcode running.
 
Congratulations :clapping:

Going to need a video of chips flying :)
 
email from DMM tech service:


Hey Jim,

In torque servo mode, the 3 Gain parameters have no effect in the drive's control whatsoever,

Only the Torque Filter Constant is used to apply a smoothing filter to the torque delivery. The lower the setting, the heavier the filter. Setting to 127 will turn off the filter completely.

So in torque mode, the drive just follows command analog voltage and outputs that torque to the motor immediately. The analog input and torque command to the motor is updated every 100us.

So pretty much all of the tuning is done in the controller for this mode. Initially, set Torque Filter Constant to 127 to turn off the filter so the motor response is as quick as possible. Then when your PID tuning is done, you can apply a bit of torque filter to smooth the motion a bit.

Thanks,

Stephen

DMM
 
You must log in or register to reply here.
Back
Top