Drive Overheating

[murdomeek]

if u are using chain, make sure its not too tight
just add a link even if it seems too “loose”. chain too loose is better than chain to tight

[Telemascope]

Heh, sorry to turn this thread into a strategy discussion, but we are going 2:1 on 4 inch wheels :P. It’s especially beneficial when the doublers and negators are introduced and the entire field is open – if it takes you a majority of the thirty seconds to score a doubler, you close a lot of possibilities with negation and last-second strategy.

Yeah, whatever works for you is fine. We just found that 2:1 introduced a lot of mechanical problems to deal with, and for our style of play, wasn’t particularly beneficial over 1.6:1 for our style of play. However, whatever works for your style of play is fine, and good luck :D.

[JesseCRN]

if u are using chain, make sure its not too tight
just add a link even if it seems too “loose”. chain too loose is better than chain to tight

I might try that. It is pretty tight.

Thanks for all the options to try yall’!

[Thorondor]

Our robot also encountered overheating issues last year at world’s. This was obviously quite upsetting as it caused us to perform rather worse than we had hoped. In order to prevent it from happening I would suggest trying to reduce any amount of friction that is occurring, and make sure the motors have some airflow. Another thing that really helped us was to improve our driving. Try to be as fluid as possible in your movements, in other words don’t jerk it around too much. This reduces stress on the motors and helped us quite a bit. If all else fails, torque the drive train more, but based on your gear ratio you shouldn’t have that much trouble unless your robot is extremely heavy. Another thing worth mentioning is that if you encounter the overheating problem in a match, the best way to deal with it is let the robot sit still for about 10 seconds. It may seem painful to do nothing for that amount of time, but afterwards you should be able to move freely until they overheat again. If you attempt to continue moving after it overheats it just doesn’t work.

[jpearman]

I think it’s worth understanding a little about what happens when the motors “overheat”. The wiki page for the 393 motor show that the over current protection device is a HR30-090. The datasheet for this device can be found here.

There are a number of things to learn from the datasheet. The first is the current that may flow that will not trip the device (IH), this is 0.90A for the HR30-90 (hence the part numbering). The next important number is the minimum current at which the device will always trip, for the HR30-90 this is shown as 1.80A. The third thing of interest is the time to trip at 5xIH (or 4.5A) this is shown as 5.9S. These numbers can also be found by looking at the graph of current Vs trip time for a typical device.

https://vexforum.com/attachment.php?attachmentid=4930&stc=1&d=1323039742

I have shown the HR30-90 curve in Red (ignore all the others, they are for different parts) as well as indicating the stall current for the 393 motor( 3.6A ) and time in yellow. What this shows is that at an ambient temperature of 25 deg C the time to trip the PTC is somewhere around 8 Seconds, however, even at half the stall current the PTC will eventually trip.

Referring back to the datasheet we can see how the trip currents change with temperature. At 40 deg C the max current guaranteed not to trip has decreased to 0.75A, at 50 deg C it’s fallen even further. This is affecting the red curve in the graph above and effectively moving it to the left (I think) so now the time to trip at stall current will have dropped to perhaps 3-5 seconds. PTC’s are not very accurate devices and, along with the tolerances in the motor (20%), mean that some will trip sooner, some later.

It’s not that the motor is overheating and tripping the device but rather that the PTC over current protection device will trip sooner. I guess you could argue that that is the same thing but I think it’s subtly different.

[jgraber]

What is the axle stackup on the left side 269?
Looks like motor, bearing, metal, spacers, 12t HS sprocket, spacer, encoder, metal.

It looks like you are missing an outside bearing on the shaft through the encoder.
Given your setup, likely a pillow block bearing on 1/2" standoffs from the outside rails would fit the best.

[jumper11]

Check a few things:

1. Make sure you have bearing blocks on every piece of metal that has an axle passing through it.

2. Make sure the chain has the correct amount of tension. (i.e. not too tight or loose)

3. Check for friction in the system by removing the motors.

4. Connection between the chain and the front 12-tooth sprocket may not be meshing well enough given its angle, so you may not be getting the full power from that motor.

5. Weigh your robot. Since you are using steel, 15 pounds seems a bit on the light side. (however given the gear ratio this probably isn’t the problem)

Good luck!

[JesseCRN]

I will try all of those. Also our drive base and tower are steel while our arm and intake is aluminum

[NHRC-Vin]

Does your robot’s wheelie bar drag on the ground? From the video, it looks like a plate of metal is dragging on the back, center part of your robot. My team’s robot always used to overheat, and when we removed our dragging wheelie bar the problem completely disappeared.

[vamfun]

I tried to look at the video but it had been removed by the poster.

How sure are you that the motor PTC’s are overheating and not the Cortex PTC’s?

Are you using a power expander?

If you can tell us how you have distributed your motor loads among the motor ports maybe we can comment on whether you might have a problem there.

I.e. if your lift/intake was active during or before you used your drive motors you can pre-dispose your Cortex PTCs to tripping prior to seeing big motor loads. You can trip one half of the Cortex while other functions may still be running.

Jgraber made a good point about the apparent missing bearing block on the encoder. You essentially had a cantilever load on that axle which can double the axle friction loads on the bearing block and also put side loads on the motor bearings.

[JesseCRN]

I can answer one of those questions right now.

The reason we dont have a bearing block on that certian poin is because the encoder is
on the orher side.

I will get back to you with the ports later.

I do not have the robot with me

[mediumdave]

Encoders aren’t made for bearing loads. Perhaps you could install a bearing and put the encoder on standoffs?

[jgraber]

The reason we dont have a bearing block on that certian poin is because the encoder is on the orher side.

Well there is your problem, right there.
Given your setup, likely a pillow block bearing (http://www.vexrobotics.com/276-2016.html) on 1/2" standoffs from the outside rails would fit the best, and it doesn’t even require any axle de-stacking.

You should NEVER have to gear a drive further down than that, (1:1.5 for torque)

“NEVER”, in this context, is a shortcut version of “Never, in my experience,”; the longer version makes it clear that other people may have other experiences…
Roundup quad-carrier prototype needed more torque than that with 4x 393 motors, 8x 4" omni-holo drive, due to usual 45 lbs gross weight with cargo.

[JesseCRN]

I want to first thank you all for comments. This is a great
community.

Here are the things we are going to check, or do.

1. Put bearing block or something on the other side of our c channel
   set up so the axle will spin freely without extra friction.
2. Change the wheelie bar on the back to some sort of omnis
3. Make sure there is NO excess friction.

I think thats all.

Thanks