SCM retrofit - servodrives throw a fault, when velocity exceeds 2,5 m/min

Hello!

I am doing a retrofit of a cnc machine, made by SCM. Basically it is very similar to Biesse CNC machine, even the servodrives are the same - Yaskawa Servopack SGDH - 50DE for X axis, SGDH - 20DE for Y and SGDH - 10DE for Z axis.

I have made it so far that I have connected enable signals and e-stop loops and servodrives to analog outputs of 7i77 board as well as encoder feedback is connected, so I got the motors to move.

I started to tune feedback loops based on this tutorial:
forum.linuxcnc.org/10-advanced-configura...etailed-how-to#88210

My problem is that as I reach velocity of approximately 2500 mm/min, the machine stops as the enable loop is interrupted. And I have no firm clue, what is going on. I can only suspect that it is because servodrives go to fault, but I do not see a reason for that.
X axis is really really hard to move by hand - I would understand that servodrive would fault due to excessive load (it easily could be some mechanical problem with linear bearing or whatever), but the same happens also if I try to move Y axis faster than approximately 2500 mm/min, but mechanics are good there - I can easily rotate the ballscrew with hand.

I would appreciate any ideas, where to start from. After the machine has stopped, there are 2 messages on the drives - "pot" and "not". Basically they show that motion is not allowed in positive and negative directions, but that is only because of default wiring scheme that "servo ready" signal is wired also to POT and NOT inputs of servodrive. So I cannot directly see, if there was an error.
Can I access the drives' history of faults? Anyone has experienced something similar?

Thanks in advance!
Viesturs

If this speed is equal on all axis, maybe theres some safty speed-monitoring which is implemented in hardware in the cabinet.
So that if something is not ready in an multi-stage saftycircuit (light grid, step mat) the drives not allowed to go upper then 2500mm/min

db1981 wrote:

If this speed is equal on all axis, maybe theres some safty speed-monitoring which is implemented in hardware in the cabinet.
So that if something is not ready in an multi-stage saftycircuit (light grid, step mat) the drives not allowed to go upper then 2500mm/min

Thank you for the clue!!! I had noticed "manual speed restriction" signal in wiring plans, but servo drive input signals did not have anything related to that. Upon closer inspection it turned out to be output signal from servo drive - basically servodrive has an output, which goes "true" when servo motor speed exceeds predefined value (there is a special setting for that). And this output interrupts "servo ready" loop...
So I tuned it up to 36 m/min. Faster seems a bit dangerous, because the travel for Y axis is too short to tune it. I noticed that at 36m/min analog output value is 3V, which indicates that machine is running at 30% of its intended max speed. 36/0,3 = 120 m/min. I have NO idea, how can they tune feedback loop for such a speed, when there is less than 2 m of travel. Do they create a settings profile on test machine (with much longer travels) and then simly copy-paste the PID values?

120m/min is the theoretical speed at max servo rpm. I don't think that the mechanics can take that.
At my weeke cnc factory sets velocitys to x 65m/min, Y+Z to 40m/min and that is horrible fast...... at 10V x would be 130m/min.

Tuning can also be done at a time window of 200ms(if this is enough for accel to max speed setpoint), there is no need for big travel distance. Most torque and velocity tuning is done with an symetric step or sin sequence as source.