Repairing zodiac cx20 pool robot

Thread Starter

smallsam

Joined May 4, 2020
4
I have a zodiac cx20 pool robot, which has started malfunctioning. By my knowledge of it's algorithm, effectively one of the 2 motors in the unit, the drive motor which drives it forward and backwards on its tracks now only appears to operate driving forwards. Notably it also appears to drive forwards when I would expect it to reverse direction.

Any advice on where to start looking? I only have access to a soldering station and basic multimeter.
The motor output going to the drive motor is bottom right of the board, viewed from the IC top side.

IMG_20200504_141258158.jpgIMG_20200504_142429354_HDR.jpgIMG_20200504_142216628_HDR.jpg
 

Thread Starter

smallsam

Joined May 4, 2020
4
Here are some clearer photos of the board. I did pay some attention to the FETs and the electrolytics when I dismantled it earlier but didn't see any obvious bulging or burnt parts. That said I've had very little experience of seeing failures in real life so could easily have missed something.

IMG_20200504_142216628_HDR.jpgIMG_20200504_141411287.jpgIMG_20200504_141333731.jpg
 

Thread Starter

smallsam

Joined May 4, 2020
4
Hi Duke,
Not yet, I'm suspicious of a few things but haven't had time to get it onto the bench and work on it. I'll update the thread if I find anything.
 
I've check the FET's and the transistors the drive the FET's and they seem ok. What are your suspicions? As I have it on the bench at work, but not sure what else to check.
 

Thread Starter

smallsam

Joined May 4, 2020
4
That's where I was going to go first. As the behaviour is that the motor drives all the times I expect it to but only fails to change direction, my belief is that assuming it's an H-bridge or similar setup that there's a failure in there somewhere.

My hope is that the microcontroller is still providing the output to switch the direction but the power stage (motor driver) isn't responding. Rather than invest lots of time trying to reverse engineer the board or attempt to get it into some sort of test rig, my first though was to order the $15-20 of parts to replace the electrolytics and FETs and blindly give it a try.
However if you've gone through the FETs and transisitors that might be a waste of time.

I'd be keen to power it up on the bench to see if you can get it to drive the motor and somehow work out how to force the direction switch (maybe through tilting the board), I assume there's a tilt sensor on there somewhere.

Worst case scenario would be that the firmware is corrupt and it's simply no longer doing the right thing. To prove this point (because fixing that is way outside of my ability), would be to locate the logic level direction coming from the microcontroller and see if it changes direction. The pre-req for proving this out I think is to somehow rig it up and power it on the bench.
 
Top