Cartridges Breaking

This is an issue that has been plaguing us for the last few days. When we were using a beta cartridge, we could apply negative power to our flywheel to slow it down extremely quickly and effectively (for double shot). But with the production cartridges, when we do the exact same thing, the cartridge breaks after just a single double shot attempt. Anyone have any idea what we could be doing wrong, or is the production quality just really low?

https://lh3.googleusercontent.com/GBf3CN3cH9NXOzdjMc29jeDwh0HtJkFe0cEZES6Esxvmtw9ecCBGC7K0Fyj3HU6F96TdwrelMIJlnAZ-ZgTBHlMMAuh1Adc_rSs3k9Og5Chu2RfpI2WswcDzecXDTzGEbJqp_k3mZMJa13aAYdSOjgh-7xqcchfqUVmb4Y2LFPiCQwCaUS5kyOyn_JePoIq80G9XjaPnJWt-XvYnamiMowEHhi7x1RIJfYsl3z89AWdPd7h_xbec5h_KEwhTDwMs8QLqJmedrgnsxrsm6-5qhyQVm69Pl5CUWgCrU34sycriVPY_bzomExvstXGIfDBqJSP4QvOCpHTZzlxE7Aw2n4xkbvxLgo21XugPMzVtgIhZH1VHJLhuJ13VJ07oe3W5sEs6nosC4mKCXKOCouMDZ8ySsA2k62zvCawwk3i5AABy0tvyeYOQEsSU0ax-1L_TdDAJhezaNSNA7KTRfgM0TfrWq7MT-9I_JrSac0RmcdOW2Zt17ojSoCKsUZmmhoSeUb37XtT_cItyRpIH14QhEauc8LohJrecIDEtXVftea94M2hMVokzOm3yVZ084Vrh9cE5uGG_RpEtyyTScLHdI22bE_qjDCLi0SHMSbRWW2yJRqZhVoooSy4zRC6EFBoRQdHMLaNXB2FNaR1RkrJn2lIsLA=w1376-h1834-no

You are allowed to use beta cartridges in competitions I think, so that should fix your issue. Emphasis on “I think”

^The official Q&A post went something like “You shouldn’t use beta cartridges, but since they are practically identical, EPs shouldn’t have to check them”.

So, if I understand this right, more or less, you guys were running the flywheel at full power, then set the power to full negative until it reached a certain RPM, which you would then spin the other way to keep a lower RPM? Were you automating the task, or just the operator eyeballing the speed?

It’s automated, and the negative power used is -45, applied for 180 milliseconds. It’s really surprising that it’s breaking so quickly. Beta cartridge started clicking (different issue), so we can’t use that anymore.

Could you use the brake function? Or maybe use a negative acceleration curve to slow it down more gradually.

I’m surprised that the production cartridges are weaker than the beta ones. However it is not surprising that something breaks when you immediately reverse the direction of a high-inertia object. The energy of the motor will be directly acting against the stored energy of the flywheel which puts ridiculously high forces on the motor and gearing. Something has to give, and usually there is a (cheap) part designed to fail before anything else (so that an expensive part doesn’t die).

This is the best answer. We already know these motors can pretty easily twist axles, and the cartridges are most likely a mechanical fuse that will break before the motor breaks when under with extreme stress.

If you are giving it a power of -45 that’s not good. Why not do a coast until the flywheel lowers to the designated speed, better yet you could try using the brake method of the brake class for the motor,

Well, yea, I understand how much stress the cartridge is going through. It’s just annoying that there seem to be differences in behavior between the beta & production cartridges when they’re supposed to be mechanically identical according to VEX.

At this point, it seems like the only option is to lower the moment of inertia of the flywheel so the braking required is minimal.

No, the motor isn’t SUPPOSED to be under that much stress, it’s like trying to get a car to a certain speed, and then throwing it in reverse with the pedal to the metal when it gets above the speed. You shouldn’t do this. Please just use motor.stop(braketype::coast) so that the motor isn’t placed under the amount of stress your asking of it, which is an unreasonable amount. Lowering the inertia of the flywheel will help, but its not a good solution.

Sure, I could make it coast down to a lower velocity (which will happen after 5 seconds or so) but that wasn’t my original intention. The goal was to slow it down as fast as possible for double shot, and it was working very well with beta cartridges. My question is why production cartridges fail like this and beta cartridges don’t.

Beta was to test out ideas for the final product run. In production they may have used different materials.

You have the wrong focus here. The question should be how can I design my system with production gears.

But yea, I’m going to start focusing on alternative solutions.

Does this issue occur with all production cartridges or has it only happened once?

4 production cartridges broken in the same way so far. Each time, we thought we had found the issue and addressed it, but the cartridges kept on breaking.

Maybe you can use a ratchet to allow the wheel to slow down slowly and not break the internals of your motors. I apologize if what I said earlier came across as blunt, but the expectation you’re placing on the internals is a tad unreasonable, as the weight of 3 traction wheels at a 25:1 ratio is almost 5 kilograms, trying to stop that suddenly won’t work with the small amount of force the internals can absorb.

Don’t rely on the beta hardware acting the same as production. VEX even outlined that during the program.

What you’re doing is physically taxing on the motors and isn’t a good use.

Out of interest, how are they breaking? As in what part of the cartridge gives out?

There could be myriad of reasons why production cartridges are weaker than the beta hardware.

It may be a different weaker material that couldn’t absorb shocks as well. Or it could be injection process that is not well tuned or uses less precise molds, which results in worse tolerances, and/or allow voids or fractures to form if process is rushed and optimal cooling profile is not followed.

https://www.google.com/search?q=injection+moulding+defects

Regardless of the reason if there is any free play between gears and you slam them with reverse force they will eventually break. 393 motors had enough power to shred metal gears, which happened a lot with NbN flywheels. V5 motors use much better balanced planetary gear train than 393s and nylon gears should absorb shocks better, but there simply is far more power that is flowing through…

+1

As many people had said the best strategy is to decelerate more gently to allow any slack in the geartrain to be picked up over longer time when you switch in reverse and avoid large shocks.

I would, first, use brakeType::coast,
then 20msec later brakeType::brake,
then 20msec later start sending negative power levels in small increments with some conservative slew rate.

Below is the data from flywheel spinup (back from NbN season). Notice the violent spike and high force deceleration in the beginning of the graph - this is what happens when you pick up the slack between the gear teeth.

https://vexforum.com/index.php/attachment/56e8292d989e6_recovery_datalog.png

And this is what happens to the gears if you don’t program adequate slew rate to protect them from aggressive acceleration / deceleration:

https://vexforum.com/index.php/attachment/5670d633a8e0c_flywheel_damaged_gear.jpg

Would it be faster to take the low shot first and spin up for the high one second?

1 Like