I'm using Mesa 5i20/7i33/7i37 and Axis display to run analog servo cards. I wanted to be able to switch the AC side of the servo amps DC supply so that one could do other controller/ PC tasks when not running and not have to listen to the Servos sing.
I realize I could probably tune the amps better to reduce the sing, add it to my to do list. Going with the shutting off the amps route, I trial tested by connecting the amp's ssr to the Axis manual "FloodOn" hal pin. This seemed to work well. Wanting to leave the flood output available for flood, I added some buttons via pyvcp. Turning the amps on via the pyvcp button seemed to make the steady state/ no motion loop unable to maintain the 0.0001 position that it was able to when fired from the "FloodOn" pin. Other routes to fire the output seemed to have the same problem. If it were physical wiring difference I would suspect signal noise. But this occurs when just changing the software link /'wiring.' The amp power is a discrete output, so I wouldn't think this would make any difference as far as threads and or realtime. But if that's it, what are my options besides 'Mist' and 'Flood' to link to my amp power?
I agree with PeterW. Disable the drive and it should quit singin'. Don't drop power, especially to the encoder, and you won't loose position. Careful if you have a vertical axis. Vertical axes should have holding brakes.
I tried the enable output. However, I currently power the 7i33t via the pc 50 pin cable, so the output goes high as soon as I power on the PC. Would rather not add another power supply to 'external' power of the 7i33, but will consider it.
I was hoping that there was some easy software fix to power my amps on and off. I'm just a little surprised and baffled why the built in pin/signals from 'flood or mist' verses a new link via pyvcp is so different or would even effect either the control loop or the encoder/error feedback.
These signals (4 = 1 per 7I33 analog out) are intended for just this purpose: deferring enabling amps until the software has initialized the interface hardware. They should alo work fine for your purpose (disabling your AMPs when EMC is not used)
With the new 5I20 Firmware (HostMot2), these enable outputs will also go low on watchdog timeout