Warning to users of servomotors!
-
I am referring to AC servomotors used in place of steppers, not hobby servos used to change tools, etc.
I have been using iHSV servomotors in my new sand table build and was forced to make some mods to the mechanism which reduced its size in X and Y slightly. When I put it back together and ran a test file, not even thinking about the smaller dimensions, the mechanism ran into the end of an axis and promptly blew a power supply and Duet board up.
The servomotors will suck a lot of current to keep up with the input signal and in this case I think that when the mechanism hit the end of the axis, there was probably a large current spike as the motor tried to keep moving, which would have shut down the power supply, but it was moving very fast at the time, and shutting down the power supply would maybe have generated a large voltage spike on the supply line due to back EMF from the motor. I think... Whatever happened, it blew both the Duet board and power supply.
I tried replacing the voltage regulator chip but the Duet board was still dead, so I assume that a lot more than just the regulator chip bit the dust. I ordered a replacement board.
The 24V power supply is also dead- the output voltage has dropped to around 12V and bounces up and down a couple volts. I ordered a replacement.
I will be powering the Duet board from its own supply going forward. That should isolate it from the motor power lines and protect against similar mishaps.
-
Yikes! That sounds exciting in a bad way.
Dealing with overvoltage during deceleration can be very tricky.
One product I wrote firmware for ended up with something like a dozen Zener diodes in parallel to dump the extra voltage generated when a customer stopped moving their bed in an emergency situation and we dumped all the motor EMF back into the power supply.
-
@mrehorstdmd Ouch! Genuinely sorry to hear of your mishap (and thanks for the warning).
-
@mrehorstdmd, what power supply model? Reputable?
-
@zapta Meanwell LRS-200-24
-
-
Ouch!
My guess is that when the servo motor hit the hard stop, it first increased the current to maximum to try to overcome the resistance. When that failed, it turned the current off, at which time all the energy stored in the motor inductance was dumped into the power supply rails, causing a voltage spike that blew both the PSU and the Duet. The stepper driver chips and the fan mosfets on the Duet are rated at 30V, although if the drivers are not energised then in theory they can take 60V. The capacitors on the VIN line are rated at either 35V or 50V. Your PSU most likely had output capacitors rated at 35V.
Using a separate PSU for the servo motors is certainly a good idea. You might also consider connecting a large (e.g. 10000uF) capacitor between the power rails, to help absorb the surge if this happens again. More complex schemes are possible, e.g. turning on a mosfet to provide additional load when the power rail voltage exceeds a limit.
-
Back in the days of GeckoDrive, they documented a simple circuit to clamp the supply in situations like this:
https://www.geckodrive.com/support/returned-energy-dump.htmlI'm thinking that a sufficiently sized TVS diode would would resolve the issue, although given that the JMC iHSV servos are rated up to 50V with 36V nominal, you're probably better running them at 36V separate from the Duet anyway as suggested above.
-
@yngndrw Excellent! Thanks! I will put a couple of those together when I rewire everything.
-
Some power supply have 'over voltage protection'. Will this solve the problem?
E.g.
-
@zapta The supply I used had overvoltage protection. good for 28.8-33.6 V. Apparently the voltage spike exceeded 33.6...
