Design and production of an efficient PC board for a Whirlpool free-standing fridge using an ATTiny
I am deeply apologetic that I am only writing this log entry now. The last year been a whirlwind for me. I moved to the other side of South Africa and started working as an engineer in Condition and Performance Monitoring for our local power utility. The time between being called to hear the offer, being interviewed and then starting work was literally less than 10 days.
Suffice to say, I moved 1,600 km away from the fridge and actually haven't done any Hacking since. I was, though, gifted a Dremel 4000 for Christmas so I know I must start again. Back to the fridge controller though, I did have enough time to build and blow up two iterative controller boards :-/...
The first one blew up due to the switching current, I tried to use a 220V MOSFET but it's current limitation was 10A, the fridge draws about 22A when engaging the compressor. Realising my mistake, I decided to use the original controller's relay. I made the mistake of using veroboard. I, sadly have a severe distrust and ignorance of electricity above 5VDC. The new board was fried by an arc of current over its board, ruining every component. I had had the presence of mind to use optocoupler isolation but that obviously wasn't enough.
I did try a redesign, using a larger veroboard with a broad separation (grinding away 20 or so rather than 1 or 2 strips of copper) between the DC and AC areas and Live and Neutral lines but I have not found a reasonably priced 5VDC-220VAC (30A) relay to complete the board.
So in summary, I blew $30 of components in an attempt to play with voltages I normally avoid like the plague and then had to go off to the coal field for work. My parents have said they're going to scrap the fridge too, so I think I must close this project. But I know that if you have the right components and know how for 220VAC PCB production, this design should have worked.
So I made my boards and wired everything up and... learned that 220VAC live and neutral can be nowhere near each other on protoboard (even if you made sure to remove strips too near each other). Luckily I had decided to use two 5VDC supplies though, so I only blew 2 optocouplers, my control board is still OK.
I am now waiting for a separate 5VDC actuated 220VAC relay board by SainSmart which I can then theoretically just wire to my control board.
Regarding my control board, I am using an ATtiny85 to maintain 3degC using a DS18820 oneWire temperature sensor. I have 2 pins actuating the fan and compressor relays. Another attached to a speaker and LED for the open door alarm and finally an input pin that I plan to sense the door's state (when the door is open, a 220VAC switch is closed turning on the light).
I should be getting the relays tomorrow, so will hopefully finish this build soon.
I have spent the last few days trying to finish this build.
I had to try learn the requirements of components I couldn't replace, as well as try glean how this fridge's controller is meant to work.
In the end, I am in the process of designing 2 separate protoboard circuits to run the fridge. The reason I am making 2, is since one only requires 5VDC and I think separate boards will make installation easier. I am using 2 cheap ($2.50) USB chargers for my control and switching DC circuits and otherwise, as many components reclaimed from the original PCB.
I should finish the designs and first build today.
EDIT: Circuit designs are complete, now to build them :-)
Besides the usual lack of time and focus on my part, this fridge regularly shocks me with some design obfuscation. I have found (by testing the controller's speaker with a 9V battery, how ever that worked) that the logic and memory on the board is still functional.There is an issue with the mains and compressor signal leads it seems.
I made a parallel connection for the compressor and it worked momentarily before the fridge tripped the circuit breaker. I then discovered the board shorted badly but no components I could find were damaged somehow so I cleaned up the solder.
Since then the PCB has ceased operating, I think due to a short within the fridge with the live wire (as I mentioned). When I get a chance I'm going to see if I can make another parallel for this.
At this stage though, I don't think I'm going to need to make a controller, I just need to stop procrastinating :-P