Recommendation for Max ripple voltage for DC servo drives?

I just got a CHNC with GN6 controls. This has 90v HIAK servo drives. The previous owner "converted" this machine to single phase. I have been chasing quite a few issues down, one of them I think may be due to the extra ripple voltage I would be getting from having a rectified single phase vs 3 phase coming to the servo drives.

To size a suitable cap for this I would need to know what is a reasonable level of ripple voltage for input power to DC drives such as this.

The problem I think may be a result of this is the velocity feedback I am getting from the revolvers looks like the sound wave of an instrument while the axis is in motion.... Also to be noted, is that this setup has the resolver and tach (I dont think I use) on the ballscrew and there is no feedback on the servo itself.

How are you reading the resolver velocity?

MESA 7I49, I have not had a chance yet to dig in and to verify this was actually setup correctly.

Servos drives are generally pretty resilient to DC bus voltage variation. I doubt that is your problem.

The resolver excitation is via a low-voltage supply on the 7i49 so is isolated from the servo drive bus.

So I assume that you are trying to work out if the motor speed is changing with the bus voltage? I would guess not. What is the frequency of the variation? Does it seem plausible that the motor speed could change that fast?

Thanks Andy. This video is of the x axis before I started messing with anything... The increments of motion seen here can be felt when the drives are off, I'm guessing the permanent magnets give it this feel. When I removed the X belt, it feels smooth as butter.

Not sure this is related at all to the waveform I have seen with the resolver feedback. I haven't measured the frequency of the feedback "ripple" from resolver velocity, but I think it would be within a rough order of magnitude of the ripple frequency from the rectified single phase.

Just to illustrate the reason I suspect ripple voltage may be an issue, especially if the old cap is toast.