12V Voltage on Robot Frame (Perhaps because of a BaneBot)

[Dan252]

Hey, we need your help.

About a day before we bagged and tagged we noticed that the DS lost communication every couple on seconds and everything on the robot flickered (we also heard it because of the solenoids). When we checked for the source of the problem we noticed that the frame of the robot had 12V on it. We checked with a multimeter by putting the (+) on battery + and (-) on the aluminium frame.

We found that the problem root is a BaneBot motor whose common is shorted to its cover (which touches the robot frame). What was weird is that the multimeter showed resistance between the (-) wire and the motor cover only when the motor was at a certain orientation. When I removed the motor to test it alone, not embedded in the system, I did not see any resistance (and I varied it's orientation between each test).

Do you think the motor is really the component that is causing the problem or might there be another cause?

Also, our team had built 2 robots, one for each idea we had (so they are completely different, also their electrical). And in that second robot the multimeter also showed that there is 12V on the frame but on the other hand, communication worked fine and was solid. (So might the problem be in the software?)

Thanks for any help,
Dan #3075

[electroken]

Based on your description, it sounds like your frame is actually connected to battery negative. The polarity is irrelevant though.

Some RS-775 motors have been known to get internal shorts to the housing. In some cases these connections are intermittent and orientation dependent. I haven't seen this with any of the RS-550 motors with my own eyes, but the designs of the two motors are very similar.

While this condition will cause your robot to fail inspection, I doubt it is the reason for your communication problem. I would check for metallic debris in the Digital Sidecar and perhaps try another ribbon cable. Our Digital Sidecar can attract metal chips from distant planets.

Of course code is a possible culprit, but I'm a hardware guy so...

[Dan252]

Originally Posted by electroken

Based on your description, it sounds like your frame is actually connected to battery negative. The polarity is irrelevant though.

Some RS-775 motors have been known to get internal shorts to the housing. In some cases these connections are intermittent and orientation dependent. I haven't seen this with any of the RS-550 motors with my own eyes, but the designs of the two motors are very similar.

While this condition will cause your robot to fail inspection, I doubt it is the reason for your communication problem. I would check for metallic debris in the Digital Sidecar and perhaps try another ribbon cable. Our Digital Sidecar can attract metal chips from distant planets.

Of course code is a possible culprit, but I'm a hardware guy so...

Thank you for your comment. We tried taking out our electrical board (made of poly-carbonate) and disconnect all the motors from the speed controllers and testing it outside the robot where there is surely no short. In this condition everything worked fine and the communication was solid. We didn't change anything in the hardware nor in the software. How can you explain this?

[Alan Anderson]

It might not be the motor's case that's shorted to the power, but the output shaft. When installed on your robot, the rest of the assembly probably has metal that could provide a current path to the frame from there.

[tkell274]

Our team had the same exact issue with the bane bot motors we were using to run our rollers. It turns out there was a bad batch of them last year that all shorted out. Cost us two jags before we figured that out but now we have switched to a cim for the time being because it has the right speed and does not short.

[fox46]

This is how we fix case short issues with the RS775:

http://www.youtube.com/watch?v=zpxolE76bZw

These motors have cost us failed inspections in the past. Usually the windings short to the armature which grounds to the frame through the bushings at either end or the interface with the output shaft. Your only solution is to get a new one and hope it isn't case shorted or mount them in a plastic to electrically isolate them.

[NotInControl]

With the motor removed, do you still have a short to the frame?

How is the dlink wired? Is it going to the proper port on the power distro board and through the 12V-5V regulator?

Is the communication intermittant only when the robot is running its motors, or if it is just on, you notice comm dropouts.

Posting your DS logs might help.

Definetly, get the short to frame fixed, because you could potentially be providing 12 volts to other things inadvertly mounted to the frame.

Note, you don't need to worry about resistance or voltage as much as simple continuity. If you have a digital multimeter with continuity mode on it, you can check to see if there is any continuity between the + or negative side of the PD board, and the frame. And you can do all of this without the bot being powered, so you don't need to worry about frying stuff if there are shorts.

You can also do the same thing with the motors you suspect are shorted (checking for continuity between either motor terminal and the motor housing).

Any continuity is an issue and needs to be rectified.


-Kevin

[Dan252]

Originally Posted by NotInControl

With the motor removed, do you still have a short to the frame?

I don't remember :/ I'll check it first thing tomorrow

Originally Posted by NotInControl

How is the dlink wired? Is it going to the proper port on the power distro board and through the 12V-5V regulator?

Yes

Originally Posted by NotInControl

Is the communication intermittant only when the robot is running its motors, or if it is just on, you notice comm dropouts.

Only when enabled (even if no motors are actually running). - Which may hint that there is another short somewhere?

Originally Posted by NotInControl

Posting your DS logs might help.

Tomorrow i'll be in the workshop and post those.

Thanks for helping

[NotInControl]

Originally Posted by Dan252

I don't remember :/ I'll check it first thing tomorrow


Yes


Only when enabled (even if no motors are actually running). - Which may hint that there is another short somewhere?


Tomorrow i'll be in the workshop and post those.

Thanks for helping

Not a problem.... three more question to check once you get back to your robot.

When you have comm issues, are the lights on the dlink do anything (turn off, go red, or are they always blue?

Are you running tethered, AP, or Bridge?

On the driverstaion, are you noticing any errors in the user message box? (watchdog timers etc.)

As for diagnosing your short:
If without the motor connected to the robot, you still have intermittant comms, or a short to the frame.

The fastest way to run through the circuits, is to pull all of the breakers on the power distro board. Also unplug power to your digital side card, analog, and solenoid card. Turn the robot on, and check continuity between the frame and the PD board.

Plug in each breaker one by one and run the check until you find the circuit that is causing the short. After running through the breakers, add power to the analog, digital, and solenoid cards one by one, and perform the continurity checks again.

If you notice the short after plugging in one of the CRIO cards, then that narrows your search down, to the electrical components connected to that card.

If even with all breakers and cards unplugged you still have a short, then only what is still being powered is causing your short. your PD board, CRIO, or Dlink, main breaker etc.

Hope this helps,
Kevin

[artdutra04]

Originally Posted by NotInControl

The fastest way to run through the circuits, is to pull all of the breakers on the power distro board. Also unplug power to your digital side card, analog, and solenoid card. Turn the robot on, and check continuity between the frame and the PD board.

Removing the breakers probably won't work in this situation. The breakers are single-pole on (AFAIK) the high side of the circuit. Since they are reading a delta 12v between battery positive terminal and frame, it means the robot chassis is shorted to battery negative terminal somewhere. If a RS775 is case shorted, removing the breaker won't show the problem.

I would check first to make sure cRIO is isolated, then check to make sure the camera is isolated, then physically disconnect both wires leading to every RS775 (one at a time) from the M+/M- side of the speed controller, then check any custom sensors (such as Sharp IR sensors), as these in that order are the most likely causes of grounded frames.

[Dan252]

Originally Posted by artdutra04

Removing the breakers probably won't work in this situation. The breakers are single-pole on (AFAIK) the high side of the circuit. Since they are reading a delta 12v between battery positive terminal and frame, it means the robot chassis is shorted to battery negative terminal somewhere. If a RS775 is case shorted, removing the breaker won't show the problem.

I would check first to make sure cRIO is isolated, then check to make sure the camera is isolated, then physically disconnect both wires leading to every RS775 (one at a time) from the M+/M- side of the speed controller, then check any custom sensors (such as Sharp IR sensors), as these in that order are the most likely causes of grounded frames.

What can be insulated in the camera? It's case is made of plastic, or you ment it's LEDs?

Anyway, our cRio is insulated and I'll have to wait until the competition day to check how is it without the motors connected.

Thanks
-Dan

[Mark McLeod]

Originally Posted by Dan252

What can be insulated in the camera? It's case is made of plastic, or you ment it's LEDs?

If you are using an old Axis 206 model camera, then the case is grounded and must be isolated from the robot frame. The mounting bolt will conduct.

If you are using a newer Axis M10 model camera, then the case is isolated and you don't have to worry about that conducting to the robot frame.

[Answer42]

Originally Posted by artdutra04

I would check first to make sure cRIO is isolated, then check to make sure the camera is isolated, then physically disconnect both wires leading to every RS775 (one at a time) from the M+/M- side of the speed controller, then check any custom sensors (such as Sharp IR sensors), as these in that order are the most likely causes of grounded frames.

This. My team had the same issue during 2012. It was traced back to the C-Rio forming an intermittent connection with our frame. Your symptoms sound very similar to what was happening with our bot.

[Al Skierkiewicz]

If you have a frame short with the motor installed and it is gone when you take the motor out of the robot, it just might be you had screws that were too long and were touching wiring inside the motor. It is unusual to measure +12 volts on the frame when nothing is running. If you indeed had a positive 12 volts while the meter's other lead was on the +12 terminal of the PD, you were measuring the 24 volt supply that feeds the PD less the 12 volts of the battery. Check the insulation on the cRio.

[mickztlin]

Everyone else probably gave the right answers, but If it werent for the short, I would check to see that it was plugged into a 20 amp breaker. Without power it wont be able to send out signal to more than a few motors.

EDIT: Im reffering to digital sidecar