Project: replacing control module with Raspberry Pi (Hotpoint FZA54)

I wondered if anyone else has replaced a ‘black-box’ (proprietary) control module with their own microcontroller? I’ve just begun to do this for an old but perfect Hotpoint FZA54 frost-free auto-defrost freezer due to what seems to be a bezerk control module. Reasons for doing this instead of replacing with a new freezer are various but the main two are: height fits perfectly into a limited space and cannot find new freezers that fit so well with same capacity, and being fed up with no insight or control over what is very very simple operating procedures (no service manual available).

With another unit, an FF92, I have the service manual which describes the operating modes with all the specific details required, and I’ve successfully diagnosed and solved issues on that over the years (and published findings in these forums).

Problems:
Back to the FZA54 (short version of the FZA84). Problems first appeared about two years ago with high temperature LED indicating a problem. Initially tracked that down to a faulty fan. The fan design is rather unique and it took some months before I found an exact physical-size-and-shape replacement and fitted it. During diagnosis, due to lack of a service manual, I’d drawn up a basic circuit diagram, measured resistances on the three thermistors (evaporator, lower compartment, heater), and figured out most of the control module schematic. The freezer was powered off for several months before the new fan was fitted. After being powered on the FZA54 couldn’t reach regular operating temperature. At some point I must have thought I’d figured that out because I wrote on the circuit diagram:

NOTE: Must have door open when powering on else it goes into TEST mode!

I’m not sure what I was getting at with that note since, now, there doesn’t seem to be any difference between starting with the door switch open or closed (note: door is actually closed – the control module is now external and I use a crocodile-style clip to close the switch)

However, I didn’t fully resolve the issues and the freezer was left powered off for a long time (over a year) until now when it will soon be required to do long-term storage for the fruit crop so I’ve returned to it, having forgotten most of what I figured out originally!

I’ve taped two thermocouples to the thermistors in the evaporator and lower compartment so I can accurately monitor internal temperatures. I’ve also got a radio-transmitter clipped to an upper-middle drawer front to monitor ambient air temperature in the compartment. Tests are all from an ambient room temperature (~23°C) start with evaporator free of ice and frost.

I’ve monitored thermistor resistances and operation of each of the devices the control module … controls.

Symptoms:
After power-on, with door-switch closed, the compressor and fan never operate at the same time, and they cycle (compressor, fan, compressor, fan, …) with each device operating for about two minutes before switching to the next. Temperature readings show a minimum of -11.5°C at evaporator and -9.0°C at lower compartment.

With door-switch open compressor will operator continuously (or, rather, for long periods as would be expected) and temperature readings show minimums of -32.3°C at evaporator and -23.1°C at lower compartment, with drawer around -3.6°C. With door switch set to closed compressor no longer operates, fan operates and temperatures on both thermistors equalise around -10.3°C (depends on how long the compressor operated before fan was engaged) and then with the frequent cycling between compressor and fan temperatures rise to around -6°C.

User manual states “It is normal to hear the fan when the freezer is cooling.”

Heater element is not operating during these periods.

By connecting my multimeter in current-sensing mode between PCB pin 4 (240V AC) and pin 3 (fan) of the primary edge connector I can cause the fan to run at the same time as the compressor. Temperatures at the two thermistors stabilise around -25.2°C and the drawer temperature drops at approximately 0.1°C per minute starting around -3.6°C.

I’ve decided the control module has got its knickers in a twist, and that the other components all seem to operate correctly, and that when I manually operate the fan at the same time as the compressor (so warm air is being passed over the evaporator) the freezer will, eventually, reach correct operating temperature.

Project:
So, considering it isn’t worthwhile spending £112 to replace one black-box with another, I’ve decided to use a Raspberry Pi instead. There’s three stages to developing this project:

1. Operating software (easy part to codify the known operating procedures and provide a user interface)
2. Prototype on a currently unused Raspberry Pi model B+ (original version 1.2 with no Wifi, and arm6v architecture)
3. Move to a Raspberry Pi Zero W (wifi enabled small form factor @ £7 when stock eventually becomes available)

Software:
Configuration and twiddling with operating values (such as associating thermistor resistance to actual temperature) will be via web server interface. Basic ‘dashboard’ showing current operating conditions plus options to set target temperature (control module has a physical variable potentiometer for this), enable fast-freeze, and showing measured temperatures, status of evaporator, fan, and heater. Configuration to vary the thermistor resistance-to-temperature ratios, vary minimum and maximum operating cycles for compressor, fan, heater, and anything else I can think of!

I shall publish this as a generic open-source freezer control project that others can adopt. I’ll design-in hooks to control fridge operations too.

Development:
Raspberry Pi Model B+ with a spare 4 inch touch-sensitive LCD attached (WaveShare SpotPear) since this RasPi doesn’t have WiFi; move to a Pi Zero W once they become available. LEDs to provide the visual high-temperature and fast-freeze indication. Possibly a PC-style internal speaker/buzzer to make audio feedback. Momentary push-button to activate fast-freeze.

To switch the 240V AC for the compressor, fan, and heater, a four-channel solid-state relay with GPIO (General Purpose Input/Output) control. An un-shelled un-pinned 240V AC to 5V USB power module to provide the 5V to the Pi and solid-state relay. I may add an additional ADC (analogue to digital converter) module to provide additional temperature monitoring via additional thermistors for e.g; ambient room temperature, top-of-compartment, compressor.

The Raspberry Pi Model B+ onwards has a forty-pin GPIO header. On that there are three ADCs (GP26, GP27, GP28) where voltage measurements can be read. By connecting those to the centre of a voltage divider network with a thermistor on one side and a fixed resistance on the other (e.g. 1kΩ) the resistance in the thermistor can be calculated and therefore its associated temperature.

Three other GPIOs can be used to switch the solid-state relays on-off under program control.

Using the small amount of operating detail gleaned from the user manual a few specifications are known. Target operating temperature is -18°C and lower, fast freeze auto-disables after 48 hours. Ambient room temperature should not exceed 2°C or 38°C for “prolonged periods”.

From other service manuals it is also known that the compressors have recommended minimum and maximum running times.