12v12 chip getting SUPER HOT upon powering

I have a 12v12 that has been working fine. I am powering from a 12v source. When plugged into computer everything seems to be in working order…input and feedback working fine, no strange errors. However, when I plug into my 12v power source as of yesterday the yellow light comes in and quickly alternates with the red and then red goes solid. Even from being plugged in for only 2-3 seconds the biggest chip on the board gets SCALDING hot. I have never had this problem with the board and have even had individual heatsinks on the chips on board to help in cooling. What could the problem be?

Is this the jrk 12v12 that has a burned out VIN trace as we discussed in this thread? Is this the 12V power supply that was involved in burning out a switch and a fan?

You said it was working previously. Did you change anything about your system recently?

I think you should check for shorts on the board, especially near the VNH2SP30 (the largest chip), near the motor outputs, and near the motor power inputs.

If you haven’t already, try disconnecting everything from the Jrk except USB, and see what happens when you apply power to GND and VIN (of course disconnect it quickly if the chip gets hot).

Do you have a multimeter that can measure resistance? We can’t trust the exact numbers very much, but if you measure the resistances between all possible pairs of GND, VIN, A, and B while the board is unpowered, then all those resistances should be at least several kilo-ohms (I just tested this myself on a working jrk 12v12).

–David

Yes, this is the very same one that I had that trouble with. However, it was determined those problems were due to a secondary grounding of an exposed wire.

I have desoldered the blue connections from the VIN GND and Motor A and B inputs to see if somehow something there had become shorted together. Upon retesting with nothing else plugged in except the 12v input to VIN GND the same result occurred…that chip becoming instantly hot upon the yellow light flashing real fast a few times followed by the error light coming on. I also removed the two heatsinks that were installed on the board with no change in the result. I also removed the bridged wire for the 5v VIN output trace wire that had burned out. No change there either.

When I connect a multimeter I am getting the results you expect…a random reading of a few kiliohms.

I don’t believe anything has changed in my configuration. I had the controller working flawlessly for several weeks with a 10k joystick POT on input. The only change I made prior to this issue starting was switching out the joystick POT for a touchPOT strip for input and I changed to a larger spur gear on my motor, but don’t see how these should have affected anything.

As for shorts I don’t see any apparent short points around the chip or the motor/power inputs.

I have attached links to pics of the front and back of the controller. The area around the GND isn’t black though pic may seem that way. It looks a little rough back there due to the desoldering, but did not look at all like that prior to.

http://www.flickr.com/photos/42522757@N02/5071613286/

http://www.flickr.com/photos/42522757@N02/5071004853/

Unfortunately, I can’t see anything obviously wrong with your jrk from those pictures, but from your descriptions it sounds like the jrk has been through some rough times. If you still want to try to figure out what went wrong, try to remember what you were doing right before it stopped working and let us know as many details as you can.

–David

Well, I am still not 100% sure what caused the short. I think may have been a piece of conductive debris between a power and ground point somewhere that was causing the short or more likely a little ball of solder sitting inside a few of the holes under the main chip on the board bottom as I now am getting power without error light or chip getting hot. The other possibility is a bit of solder may have overflowed from that bridge spliced vin trace. Either way, with nothing else plugged in I just get yellow status solid light which is what I’m looking for.

David, you are right, this controller has been through hell and back being a small PCB in the hands of a VERY inexperienced, green, wet behind the ears, noob like me. But, I must say now, that will all my trial and error and lack of caution around short/unintentional grounding type scenarios…this little guy just keeps on ticking =)…KUDOS to Pololu for making a controller that can take a beating and keep on ticking =).

Only downside (depending on how you look at it) is that you guys are so fast on shipping that I can’t go back and cancel the “replacement” I ordered yesterday cause it’s already half way to me ;)…oh well, I guess I’m gonna be able to make two of my contraptions now…hehehehe.

Thanks again for the continued help. Keep it up cause the support you provide is one of my many reasons for sticking with Pololu =).

Ok David and others…The previous HOT chip issue was apparently not being caused by a short. I thought I had it back up and working cause I was able to get the chip to stop overheating. I spoke to soon.

Now I have discovered that what is causing the error light to come on is a motor drive error of some sort. In serial or analog input mode if I set the input target to anything above 2048 the motor will run. If it is set to anything below 2048 the error light will come on, motor will stop, and chip will get super hot instantly. Also, I noticed during the testing to discover this that any changes to the target input will not result in a change of speed in the motor. The only thing that will change motor speed is dropping target below 2048 which of course stops motor and causes board to overheat. I notice that the instant this starts to happen that current spikes then quickly drops to zero as motor stops and board starts to overheat.

I’m at a loss on what would cause this. Prompt help…as I usually receive here…will be GREATLY appreciated.

Hello, please do not start a new topic each time you discover something new about your problem. I merged your new topic in to the old one.

Are you running the jrk in Feeedback mode None? In that case, targets above 2048 mean “drive the motor forward” and targets below 2048 mean “drive the motor in reverse”. So what you’re saying is that when the jrk tries to drive in reverse, it gets really hot, and when it tries to drive forward, it can’t control the speed. It sounds like one of the low-side MOSFETs inside the motor driver chip is permanently turned on; that would explain both effects that you are seeing. Unfortunately, that would be permanent damage that is not easily fixable.

–David

No, I am running in analog feedback mode. Uncontrolled full speed constant when target set above 2048. If say I power on with input target above that motor will spin full speed. If I power up with feedback and targt both sitting at zero is instant error and overheat. If start with feedback high and input zero then motor will power in reverse till feedback reaches zero basically then promptly overheat.

Likely this is probably caused by same solution u just gave or is different?

If you are running in Analog feedback mode then a target of 2048 should not have any special significance, and you are giving me an incomplete picture by telling me what the Target it is, but not telling me what the Scaled Feedback, Duty Cycle, or PID parameters are. The most important thing in this case is the Duty Cycle. Please use the Jrk Configuration Utility to look at the Duty Cycle and tell us how the jrk & motor behave for all duty cycles from -600 to 600.

–David