Smoke from Motor Control

[wyemine]

Our first post on here - so 'Hello Everyone'.

Having had our Gertboard unused on the shelf for ages, we are now enjoying setting it up to control a small model railway project.

First stage using 11 channels of buffered I/O to operate point/switch motors and isolate track sections. Written an interface with Python and Tkinter. All working well and lots learned along the way.

Next stage to use the Gertboard motor controller to power the (12V 1A DC brushed motor) model locomotive.

Connected MOTA & B at J5 to GP 17 & 18 as per the user manual. Output (MOTA & MOTB) at screw terminal block is currently n/c. Connect 12V DC supply (measures 17V no load) Polarity checked. Motor control starts smoking!

Fuse is still intact. ?? However, input now looks like short circuit to a test meter.

So, can anyone tell me what I did wrong?

What's my best option to rescue the situation - purchase complete new board or try and unsolder/replace the controller chip?

For now I'm going to press on and look at the code examples. But I don't fancy trying to re-connect power to the motor controller until I know where I went wrong the first time.

I'm assuming our board is wrecked - is there anything I can test without connecting the external MOT power supply.

Very grateful to anyone who can point us in the right direction.

Rob
(Hereford UK)

[Wizzabeck]

As a matter of interest were you using the -rg or -wp versions? And did you connect a battery to check the program before using a power supply? I presume by the latter you mean a model railway power unit.

Ken

[wyemine]

Hi Ken and thanks for the reply.

We have built our interface using the RPi.GPIO Python code samples.

We did not do any initial testing with a battery, the PSU is one salvaged from a failed network switch rated 12V at 1A which I hoped would be safely within spec for the controller.

Having now read the manufacturer's datasheet for the IC, I note they are sensitive to back EMF. That gave me an idea as to why ours failed. All elements of the system have dual manual/RPi control. There is a possibility that the Gertboard motor control output was connected to the rail track when it was being operated manually (i.e. from a stand-alone analogue model rail controller). That may have cooked it?

I have a new IC on order and will improve the isolation between the Gertboard and the manual system. (A physical changeover switch - not elegant, but effective.) Just need to search for techniques for unsoldering SMD as I have not done that before.

Will post back on here of we are successful in repairing our board.

Rob.

[Wizzabeck]

Hi Rob,

As a mod. rail. buff it wasn't clear what motor you were driving in your initial post but I suspected you had launched into a railway set-up. I've only been playing with the gertboard for a few weeks but I read the manual as not only an intro. but also a check of what was originally a self-build kit. My "years ago" experience of building pwm controllers with sealing wax and transistors left me with great respect for back EMF )

Ken

[wyemine]

Quite a few folks watching this thread - so here is an update. IT'S WORKING AGAIN

Managed to replace the chip. £2.80 inc next day delivery from RS. Their part No. 6996752.

Unsolder and removed the 2x 2-way terminal blocks. Then remove to motor control chip. Unsoldering to tab takes a LOT of heat. Some slight damage to tracks but nothing too bad. Then carefully, with minimum heat, replace chip and made good damaged tracks as required. Replace terminal blocks and test. Seems OK so far. No sure how much solder I got under the tab so have fitted small heatsink to compensate.

'Plan B' was to replicate the motor control circuitry using the breadboard area adjacent to the failed chip.

Gertboard now fully operational again. So if anyone else out there has a burned out motor control. It can be swapped with basic tools.

Rob

[Wizzabeck]

Thanks for that. Copied to me RaspberryPi directory. I'm bound to need it!

Ken