Pinch Valve Servo Overheating?

[Maxwell S Raderstorf]

Hey All,

We are having an issue with our pinch valve servos becoming very hot and after a certain amount of time running, stopping functioning properly. We have not definitively found that the overheating is the thing causing the servo dysfunction but it seems the most likely cause. The dysfunction ranges anywhere from only being able to get to the first one or two openings on the valve to completely ceasing to take commands. Any insight on this would be appreciated. Has anyone else run into similar issues around the pinch valve servos ceasing to function or getting very very hot while running?

Best,

Max Raderstorf

NYU

[anmo...@colorado.edu]

I've had similar issues. Several times I've even blown the servo completely. What I think is happening is there is too much stress on the servo caused by the pinch being too tight. Thus when the servo tries to turn, it struggles to fully complete it's turn and overheats, eventually breaking the motor. Obviously there is a balance here: you want the pinch to be tight enough to prevent leakage, but loose enough so the servo can turn easily.

I've had some success relieving stress while maintaining a tight seal by (1) sanding down the center part of the 3D-printed pinch clip and (2) swapping out the pinchers for slightly smaller ones. Regarding solution #2, what I recommend doing is taking out a handful of your pinchers and lining them up next to one another. You'll notice that even though these parts are laser cut, there are noticeable differences in their heights. Use the shortest ones and see if that relieves stress on the servo.

Another thing I've found useful is to always label the pinch clips and pinchers that go with each chamber. For whatever reason (probably due to the variability in 3D printed parts), different lines have a different optimal amount of sanding and sized pinchers. Through trial and error I've nailed down the perfect combo for most of my chambers, but I'm having to constantly adjust them.

All of this being said, I just had another issue of overheating of the pinch valve servo this week. So hopefully one of these two solutions will fix it, but if you come up with anything better, let me know.

-Andrew Morgenthaler

University of Colorado Boulder

[Chris Takahashi]

Unfortunately this can be a problem, which is why the newer versions I recommend using all solenoid pinch valves.  One thing I found that helps is greasing the cam to reduce friction.  The heat comes from there being enough friction that the servo never makes it to it's target.  You can actually hear the servo buz in between motions when this happens.  

There are a couple other solutions that may work.  

1) a servo that uses PI control or some other method with zero steady state error.  (I've noticed most appear to use proportional control only).

2) modify the code to turn off servo PWM a second or two after a location change.  This would keep the (insignificant) error but power down the servo between motions.

I've wanted to do (2) for a while but haven't had time to test my patch (testing is the time consuming part).  If someone wants to attempt a firmware fix and send a push request that would be cool.  Otherwise since I see more people than I expected still using servos I may try and do it sooner.

[Chris Takahashi]

This is great advice Andrew.  It's actually what I did in my first versions to get everything "tuned".  Once tuned the same servos lasted for years until I replaced them with solenoid pinch valves.  

On Friday, March 31, 2017 at 2:13:05 PM UTC-7,  

[anmo...@colorado.edu]

I totally forgot about the solenoid pinch valve option! I'm definitely going to switch over to those. Considering 95% of the issues I have during runs are with the pinch valves, the extra ~$500 cost totally justifies the hassle.

[Chris Takahashi]

Yep.  Plugging solenoid valves into positions 1-8 should "just work" (for 12v solenoid pinch valves).  They're expensive but really worth it.