GCRP
Posts: 3
Joined: Wed Nov 14, 2012 4:19 pm

Troubleshooting a new output board design

Wed Nov 14, 2012 4:43 pm

I have designed and built a GPIO breakout board that has 5 double pole 9V (max 500ma per pole) outputs, 5 single pole 9V outputs and 4 9v DC motor outputs (intended for low power motors). I have used a couple of octal darlington arrays which are working very well for the outputs, but I am struggling with my motor control. I have used a PCA9685 16 channel PWM generator with a couple of quad half H bridge ICs. I am using the Adafruit Python scripts designed for their servo driver with the same chip. I can detect the chip fine using the i2cdetect program, and the python scripts are running without an error, but I am not getting any outputs on the PWM driver (and therefore H-Bridges). It could be a software issue, but I was wondering if any of you electronics gurus out there fancy looking at my circuit design to see if you can spot a hardware design fault. I have tested all my tracks on my home made PCB using a multimeter and can't find any issues.

The design is a PDF file which I have had to put on dropbox because the forum won't accept PDFs or a png/jpeg/gif large enough to show the detail. I have also included a pdf of the trace just in case anybody in future fancies printing their own version!

https://www.dropbox.com/sh/kkzirmw271eckkq/AUW5m2sobB

This is my second version of the board (the first didn't have the motor drivers) and I have already designed a third in my head but not until I get this one working!

Many thanks,

Guy

GCRP
Posts: 3
Joined: Wed Nov 14, 2012 4:19 pm

Re: Troubleshooting a new output board design

Sat Nov 24, 2012 8:39 pm

Further to my last post, I have managed to localise the problem. I purchased an Adafruit PWM driver board and hooked it up to my Pi on a breadboard to test the code. I produced a bit of working code then plugged in my board. Using a multimeter, I measured 3.3v out on each of the PCA9685 pins to show that the PWM driver chip is working. Therefore I think that my H bridge driver chip circuits are wrong. The output of the H bridge out pins goes to around 0.5V when it is supposed to be at 9V and drops proportionately when the duty cycle is reduced at the enable pin. Clearly this is where I need to focus my attentions, so more breadboarding required, I think!

I hope this is of interest to other users working on similar projects.

GCRP
Posts: 3
Joined: Wed Nov 14, 2012 4:19 pm

Problem Solved!

Thu Nov 29, 2012 6:53 pm

Following some intensive breadboarding, I have found the error in my circuit. As a newcomer to electronics, my lack of understanding was key to the problem! The input pins on the quad half bridge ICs (in this case SN754410NE) must either be set high or low. My circuit only allowed these pins to be set high or be left effectively unconnected. What is needed is a pull down resistor on each input pin on the H-Bridge so when no signal is being sent to them by the PCA9685 PWM IC, they are connected to ground, but when output from the PCA9685 is set high this sets the h-bridge input pin high. I have used a relatively high value resistor (47kohm) on my breadboard - I'm not sure if this is optimum, but it works.

I hope that helps other people designing similar circuits. If anybody is interested, my output board will be used to control a Marklin Z gauge model railway layout.

I will probably post a revised PCB design here once I have designed, built and tested it.

Return to “HATs and other add-ons”