Side of Drives Giving Out
 

Our robot has been having a bit of a problem as we have been practicing for the Rah Cha Cha Ruckus in a few weeks. We will go to drive up the bridge, get up a little ways, and a side of the drives (commonly the left, occasionally the right, sometimes both) will just give out. They stop working and the robot spins backwards with only one side of the drives working. I'm looking for possible causes, as we have not clue where to look.

Here is the set-up. Each side of the drive is powered by 1 CIM motor, through a CIMple box. We have a wide orientation robot with two wheels on each side, so this set-up is chained to both wheels. My hitch with giving details is that I did not help design or build this robot. This is the 2012 robot of a team I joined in July, so I wasn't part of the design process.

I speculate (and I am by no means a mechanical engineer or an expect on motors) that the CIMs can't handle the load of going up the bridge and just shut off temporarily. We maintain communication with the robot, and after a few seconds can continue driving around. I am lead to this conclusion by that fact, and that the robot has no issue driving around on flat group. Since there is only one CIM per side, it's a hefty load, even for such a tough motor.

If there are any details that might be pertinent to the discussion, let me know. I know this has been a persistent problem (it occured during testing over the Summer). The next time I will be able to see the robot work will be Thursday morning, so I will be able to gather data and look for anything else based on feedback.

Any help would be appreciated. Due to lack of mechanical mentor support, I don't think any changes can be made. But I would like to know what the issue is. It will certainly influence how we play at Ruckus. Thank you in advance to any who are willing to offer advice.

-Leeland

Re: Side of Drives Giving Out
 

This is probably what's happening. You should be able to confirm this by listening to the PDB for a tripping breaker. Is there any reason you can't add 2 more CIMs to the drive?

Re: Side of Drives Giving Out
 

The other two CIMs are currently being used in the shooter and bridge arm. Definitely wouldn't have been my choice, but that's what we're working with now. We were originally going to make the necessary changes to move all the CIMs to the drives and put different motors in the other parts of the robot (we have already recieved the new motors and everything), but I have been told by my team leader that due to lack of support by Mechanical mentors, all mechanical upgrades are off.

If this really is the issue, I'm going to try and make a second push to make the necessary mechanical upgrades.

Re: Side of Drives Giving Out
 

If your team is hesitant about replacing both CIMs, you should be able to stick some FPs through AM Planetaries for a bit of a boost in your drive. A bit pricey if you don't have the parts lying around though.

Other factors to consider with this: Where is your CG? If it's too high or off to one side this problem might present itself, but you'd probably notice the robot being quite tippy already.

Re: Side of Drives Giving Out
 

CIMs don't have internal breakers, so they will keep going until they are dead. Once dead, they are DEAD. They can be worn from hard use and simply swaping in a fresh pair might help, but not likely.

It sounds like you are either tripping the breaker, or resetting a Jag. You didn't mention which speed controllers you were using, but by the sound of it, I'm guessing they are Jags. Either way, you are over currenting your drive.

Check for binding in your drive train. Check for chain misalignment. Lubricate your transmissions and chains. Possibly, change your output sprocket to a smaller one to gain more torque out of the transmission. If all else fails, swap the Jags for Victors.

Re: Side of Drives Giving Out
 

I fully agree with Chris. To check whether that's the reason, check if the CIMs heat-up more than they usually do. If they don't heat up, it won't say much for or against the theory (because from what I understand the situation only lasts a few seconds), but If they do, overload is almost definitely your problem. I want to say definitely, but I've seen enough strange things to not commit to it :rolleyes:

Also, if the CG is a bit closer to the left (and the overload theory is correct), it might explain why the left gives out more often than the right.
A related long-shot: do you have a turret?

Re: Side of Drives Giving Out
 

FRC motor controllers can easily deliver current well above 40 amps, which in short bursts wont trip the pdb breaker immediately, but if your cims are sitting stalled on the bridge, it is likely that this is your problem. This could be due to the gearing ratio also, and the motors not having enough torque to keep driving up the bridge. Maybe consider a change in your gearing ratio.

Re: Side of Drives Giving Out
 

If you are using Jags/PWN switch the jags and see if the stall is opposite. Could be you are resetting your Jags dues to improper gearing and stalling the CIM. Makes sure your firmware on your jags are up to date.

If you can switch to Victors try that.

Re: Side of Drives Giving Out
 

Leeland, I also agree with chris that you probably don't have the torque. If you don't want to run the ruckus with 6 CIMS, I would suggest you swap out the CIMple box with the tougbox from the previous years. There is a larger reduction in that gearbox for more torque and shouldn't be too hard to make this change.

It would also help if you know what the reduction is after the cimple box.

Re: Side of Drives Giving Out
 

It sure sounds like an overload tripping the breakers.

Also check the gear ratio in the gear boxes. If you have the 4?? to 1 it would cause the problem you are having.

Re: Side of Drives Giving Out
 

Was the robot ever able to climb the ramp? If this is new then look for something wrong. Cims are very reliable, but they can be burned at by extended use at high loads. (Which is beyond their design spec)

Otherwise it sounds like you do not have enough torque. If adding motors in not an option, you can lower the drive ratio by changing the chain sprockets on the wheels. You will lose speed but gain acceleration, turning & pushing ability along with climbing ability. Sprockets are available from AndyMark.

Our locomotion chassis is geared for about 10 ft/s & has no problem climbing the bridge. (long chassis, 6 wheel drive, Simple boxes, 8 in wheels, 1 cim per side.)

Re: Side of Drives Giving Out
 

I appreciate the advice everyone has given so far.

My first course of action is going to be to try and rally the mechanical mentors on the team and make the motor changes we had planned on earlier, including moving 2 CIMs to the drive base.

If that doesn't work or I can't get enough support, we will try changing the sprockets on the wheel to larger ones to lower the ratio. Since that seems like a substantially less time consuming job, this seems like a likely route if I can't find the support to change the motors.

Again, thank you everyone for all the help. I will be doing more research Thursday, as well as talking with mentors about getting things in gear for upgrades.

The robot used to be able to climb the bridge reliably. It can still climb it on occasions, but the failure is more frequent.

We do not have a turret. Part of the issue I think we are having is that the shooter is largely integrated with the body of the robot (see picture), and most of the shooter hardware is over the left side. I don't really think it's a coincidence.

Re: Side of Drives Giving Out
 

are you running jags? you might be popping the over current protection when you climb the ramp.

Re: Side of Drives Giving Out
 

We are indeed using jaguars on our robot.

Re: Side of Drives Giving Out
 

I would never advocate doing this in-season, however it's an offseason, why not just run the 6 CIMS? A number of teams did this in the offseason after the whole BB775 case short issues. I'd just contact the Ruckus planning people to double check.

I'll even come over and help you out in the morning when I get there if you need it.

Re: Side of Drives Giving Out
 

Leeland,
The Jags have an over current protection circuit that will remove motor drive for 3-4 seconds and flash the red LED on the face of the Jag. When a breaker trips open it resets almost immediately and there is a audible click. When repeatedly tripped it can cause a buzzing sound.
It is quite possible that over the months since build, hardware has loosened and mechanical parts have moved. This could cause friction in the drive that was not present earlier in the season.
Since you state that the robot can drive after a few seconds, my money is on the Jag over current protection.

Note: We had some CIM failures earlier this season. Upon inspecting the interior of the motors, we found that some of the winding wires had actually broken away from the commutator. These were older CIMs and had been put through some excessive driving. I believe that the overheated motors weakened the copper at stress points and caused the cracks. I posted pictures in a different thread earlier this year.

Re: Side of Drives Giving Out
 

I'll be available to help too, if you need it.

-Nick

Re: Side of Drives Giving Out
 

So if we are activating the current protection circuit in the jaguar, is that an issue that can be resolved by adding a second CIM to the drives?

I did notice a Jaguar flashing red when we lost driving ability, but I'm not sure if it was the Jaguar responsible for the drive motors. I'll check that Thursday. When testing, I'll also have someone watching the Jaguar for the right drive motor, to see what happens when that side fails.

I will also see if we can pull together and get an effort to perform maintenance on the mechanical parts to help adjust the friction in the drives.

Also, thank you very much, Justin and Nick! I really appreciate the offer. Our goal right now is to be able to show up to Ruckus and not need extensive repairs, so that's what we're working towards. I really do appreciate the offer, and I will let you know if we will need to take you up on it :)

Re: Side of Drives Giving Out
 

Leeland,
Adding a second CIM would help with sharing the load. The flashing LED on the Jag is the fault LED. It can be caused by a few items like over temp and under voltage as well as over current. Start by tipping the robot to one side and without power turned on see if you can easily rotate the wheels on one side and then the other. High friction would show up in this test. It is also possible that once you have the weight of the robot on one or more of the wheels, high friction results due to wheel bearing misalignment or failure.
I already described the over current condition. The under voltage condition can be caused by loose wire connections, bad crimp terminals or loose hardware on the main breaker, battery, or PD terminals. Generally, those last few would also show up as radio or Crio resets. In programming you can try adjusting motor speeds to a lower value to minimize over current.

Re: Side of Drives Giving Out
 

On top of what Al said - It's dirty, but switching to Victors might do the trick if your drive is such an edge case that it started failing after a season of wear. The Jaguar over current time out is way too lenient.

Another benefit of gearing down the drive a little bit would be that if you're hitting the bridge at less than full speed, a geared down drive would go the same speed at a higher PWM command, resulting in more available motor power in addition to the increased torque of a higher gear reduction.

I'll echo Foss and Nick in offering you help at Ruckus if you need anything. We could probably spare a few CIM-U-LATORs and 775s if you're willing to take that gamble.

Re: Side of Drives Giving Out
 

The jaguar will flash red quickly when you're going in reverse (normal) and flash red slowly in the fault condition

Also check the electrical connections directly on the jaguar and after the jaguar.

Re: Side of Drives Giving Out
 

In the past we have run extra CIM at the ruckus. They're usually pretty accepting of that.