Chip Burns Out

[John Wells]

I've been using this stepper motor control on a circuit board I've designed. It uses a pic controller to drive it and I've produced quite a few and they've worked pretty well. The chip is fed 12vdc with a 5v voltage reg to run the pic. While it runs I will run my finger on the board to check the temperature of the components to make sure there aren't any issues and apparently this causes the motor control chip to short and overheat. I was looking at the datasheet trying to see if there are pins on this chip that may be sensitive to that. I could see that I could potentially short 12v to the pic but that's not what's burning out. It may need some kind of protection. I would appreciate any help.

https://www.mouser.com/datasheet/2/408/TB6560AHQ_datasheet_en_20141001-771503.pdf

 

[KMoffett]

You might want to post your schematic.

Ken

 

[(*steve*)]

Are you saying that running your finger across the board causes the device to lock up and fail?

I would be looking for floating input pins first, but that's just a guess.

 

[John Wells]

See attached - I do have unterminated pins and it functions well like that but it appeared that default for these pins didn't require termination. The pics seemed to have internal pull up/ down so it hasn't been an issue on other boards but I didn't realize it could actually destroy this chip. I believe one of the times it burned out I wasn't touching any of the floating pins. I'm going back into the data sheet in detail again. Could it also be possible that getting the right combination of input pins high/ low that it could cause the chip to short internally? Doesn't seem like it should do that.

Update - the pins I thought were floating have an internal pull down. Also the datasheet says the chip is highly susceptible to static discharge. We handle the boards a lot but not much when power is applied which is when it will cause problems.

 

[poor mystic]

Static discharge damage is permanentl and is usually caused by unprotected handling when the board is not connected.
When handling sensitive boards and components special precautions need to be taken which I’m sure are described elsewhere.
Others are better qualified than I on the matter of circuit design but I would say it is not conservative practice to rely on internal pull-downs... use a ground.

 

[John Wells]

Hmmm. We don't seem to have a lot static here. Humid in Florida right now. Only had a couple failures. Only with the power on and only after I've been touching the board for several seconds which I would have thought the discharge would have already happened. Something tells me it's not static. Maybe some other transient voltage?

I'll take your advice about tying the pins to ground - thanks.

 

[John Wells]

There is both 12v to drive the load and 5 volts for the logic. Could it be there is enough conductivity through my finger to drive a high impedance input to 12v that may cause the chip to self destruct?

 

[dave9]

Rubbing your fingers around on live circuits is improper procedure. Use an infrared thermometer. E L E V E N D O L L A R S !

 

[John Wells]

I have one. There will be people servicing the appliance it's in and I'm sure putting their hands on the board so I would like to bullet proof it as much as possible. If I've designed the circuit with a flaw I want to identify it now.

 

[gorgon]

What driver chip do you use? Can't see any type info on the drawing. However you have several unconnected pins, like the cw/ccw. I would think a floating pin here could make a mess for the driver.

 

[Bluejets]

However you have several unconnected pins, like the cw/ccw. I would think a floating pin here could make a mess for the driver.

Exactly.......Especially with the Op's finger poking around introducing all sorts of noise.

TB6560 driver...??
Where did you get the connection /operation diagram?

Might be better to show what you have in front of you.

 

[gorgon]

Sorry, didn't see the driver reference.
If you read the datasheet you can see that the signals not connected has 100k internal pull down resistors. 100k is too high to give any proper protection if you finger around in the circuit. Depending on the resistance and charge of your fingers you may have both shorts and electrostatic charges going into the circuit. Not very healthy in most cases, and terminal for the driver if you are unlucky.
I've been into electronics R&D for 45 years, and have never used fingering around in the powered PCBs as some kind of test. When not powered we take care of not introducing ESD to the boards, with earthed straps and table cover, and antistatic bags.
You should be aware of the fact that the boards doesn't need to fail now if you zap it, but it could work for a time and then fail due to the same zap. There are countless pictures around showing partly burned through insulation layers inside chips, due to ESD.