Strange Issue. Ideas on cause?

Right so I’ve got an issue. Our flywheel is a dual flywheel geared 25:1 with a separate left hand side and right hand side (not mechanically linked). It uses 6 motors and incorporates PID control with an exponentially moving average.

The problem I am now having is that the recovery time has dropped significantly. It is not to do with the PID as far as I can tell. When I run balls through at full power the recovery rate is still slow (roughly a second). It used to be 0.34sec. I haven’t touched the PID at all. Since it was working we have done a cortex replacement because we destroyed the old one (My personal thinking is that the replacement cortex is also broken). The vexnet connection also resets now when i am running 8 or more motors ( 2 intake and flywheel/ 4 drive and flywheel). There is also now a low drone sound when I am running the flywheel. I was thinking friction problem but everything seems to be in order. None of the shafts are bent and there is no metal on metal rotation.
For relatively old pictures please see here:
I will upload more recent pictures tomorrow if needed.

I’m going to replace this cortex with another one, but I was wondering if there could be any other issues. My exponentially moving average takes the last 10 values, could this be too many?

Anyway thanks in advance - Alex (Error 404)

VEXNet is likely disconnecting because something is drawing a lot of current and voltage drops a lot.

I wonder if your quad encoders are getting slowly destroyed - it looks from the picture like they are sitting directly on the flywheel axles and it may be too much for them.

Another thing to check is that your PID is not forcing very aggressive slew rate on the motors. You may need to protect them from having commanded power doing large jumps, like from 1 to 127. @OverlyOptimisticProgramer asked in a read-only thread what could happen in such case. It is ok for power to go from 127 to 0, but if you go from 127 to -127 or from 1 to 127 in one jump, bad things could happen. Right gear was almost new before the competition and this is what we discovered the next day:

Ideally, if you hear any noise, you need to start pulling one component at a time to see if it is a culprit of the problem. You need to check that all motors still have good gears and are matched. You can remove motor covers and pull the gears without removing entire motor. This way you could both inspect the gears and disconnect specific motor from the geartrain. If gears are still good, then make sure they are well lubricated.

Sorry I should have mentioned changes we have made since those pictures were taken. We moved the quad encoders to the next axle along, so that they weren’t being driven too fast, could they be damaged?They are still reading values. As you can see it is quite difficult for us to deal with the insides of those particular motors.

You could just unscrew the green covers on top of the motors. This will give you access to the gear (in the picture) that gets destroyed first. You definitely want to check those gears and see how each motor behaves in isolation.

Also, what ports are your motors connected to?

Ports 1 and 10 have slightly different power control profile from ports 2-9 (see May 25 post in this thread) and it could be that some motors are slacking while others are doing most of the work.

But sound should be the best indication of what is going on. How does it change when you run all motors at some fixed power, without any PID? How is it different at various fixed power levels?

Sound increases with volume not pitch as i increase motor power. each side is run by two ports between 2-9 (Y-cable for two of them). I will check the gears at our session later, but I haven’t heard any clicking or grinding from the motors. (The ear test: Put your ear right up against the motors while spinning the flywheel. Definitely a safety goggles moment)

If the encoder is reading a value such as 250-260, it might be broken. I think that it is the default value or something.

One issue could be a blown out or accidentally reversed motor. I would highly recommend unplugging all the motors but 1 and turning the flywheel on. Obviously there will be lots of resistance, but you should hear some straining, or the flywheel may turn slowly. If this doesn’t happen, the motor/ internal gearing is broken and needs to be replaced. Hope this helps. :slight_smile:

Seen this exact same thing much more this year with flywheels. It used to be the “speed” gears up front but now it seems to be this gear more and more. Luckily these are in the motor repair kit but depending upon how and where your motors are, the fix could be painful.

One of our teams thought they blew their cortex but it was a motor shorting. Check current draw too if you have the tools and suspect similar issues.

Actually, you need slew rate both ways to protect the motors. I have to have it otherwise we get some loud clicks even spinning down from the lowest power setting (1700RPM) It doesn’t have to be super slow, but it needs to at least be enough that the motors aren’t straining terribly. You could get away with 1 to 127 however, it just may stall the motors too easily.

You need to also check for cold solder joints in your motors. Where you replace the motor gearings (like speed or torque) you can check to see if there are gaps in the solder joints which hold the PTC, Voltage and Ground wires in place for both the external connections and the motor itself. We had this issue numerous times with joints that would heat up, break off, and cause the flywheel to draw almost 5A of current. We would destroy batteries and couldn’t shoot full court until we replaced the joints (only because we couldn’t reach the required 2400+ RPM for a single flywheel).