Dual purposing a refrigerator

[Erroneous]

While side by side fridges are nice and have 1 less problem for this, I'm wondering about a fridge with a freezer on top (or bottom) because they are normally cheaper. And for dual purpose I mean putting a temp controller for freezer (like in this post only with 2 controllers), and another for fridge to use the freezer as a kegerator and fridge as a fermentation chamber.

Possible issues I see with top/bottom freezers are keeping the cornies on their side, which was discussed on this post. Also if the thermostat for most combo fridge/freezers don't allow you to easily control 1 side and not the other, this would be very difficult.

I plan on using an arduino with multiple temp sensors and relays since it would be cheaper and offer more expandability, plus I already have one working with 1 sensor/relay right now.

 

[edgeofblade]

I'm in the same boat. I'm contemplating converting the garage French door fridge into a kegerator/ferm chamber. I have this wonderful plastic fermenter with a big wide mouth... no funnel required. But it's too wide to fit on the freezer side.

I need to test if a carboy or a bucket would fit there. If that's the case, I can put the temp gauge on the freezer side for ferming, and convert the fridge side for kegeration operations. By my measurements, she'll fit two 9 in diameter cornies. If I crank the fridge side to max and the freezer side to minimum, I think I can still strike a nice temp in the fridge side while regulating the freezer side.

 

[MalFet]

Arduinos and relays are effective for many applications, but what are you going to hook the relays up to? It's not just that the thermostat doesn't allow you to control one side and not the other, but that there is only one compressor coil and thus only one thing to control with sensor logic. It can be done, but this kind of conversion usually requires some more serious modification involving controllable baffles and vent fans.

@edgeofblade: I don't know the design of your fridge of course, but in many (most? all?) consumer fridges you can't keep the fridge colder than the freezer. You might want to test this out before you invest too much money into controllers.

 

[certaut]

most refrigerators have the evaporator in the freezer part and control the refrigerator temp with a fan and damper. I tried to do the same thing a while back by controlling the fan with a controller and the only issue was the refer side would get to cold sometimes, so I planed to install some flexwatt http://www.reptilebasics.com/flexwatt-heat-tape/ but procrastination set in and I was given a upright freezer that the conical fit in

 

[FishFace]

The freezer is main cooling side. The temp control on the fridge controls a door and fan that blow the freezer air into the fridge. Get your one temp control on the freezer main control unit. Then get etc unit in line with fridge temp that controls door. The main controls should be where the dials are. You might have to do some stripping.

Theoretically you can do it. The trick is getting etc in the right place. The etc used for the freezer should have a cycling unit for the compressor. The etc for the fridge, since it only actuates a door/(fan) can be of any style.

Would like to hear how it works out.

 

[Erroneous]

Thanks. I figured it wouldn't be like controlling 2 compressors, but more like controlling a compressor to get the freezer down to temp, and then the door/fan (I'm guessing a relay would work for this as well). Of course this means the freezer would have to be colder, so the fridge would have to be the ferm chamber.

I tried to do the same thing a while back by controlling the fan with a controller and the only issue was the refer side would get to cold sometimes, so I planed to install some flexwatt http://www.reptilebasics.com/flexwatt-heat-tape/ but procrastination set in and I was given a upright freezer that the conical fit in

Were you using a love or similar controller for the fridge temp? I'm thinking that since I'd be running on arduino, I can minimize the time between relay switches, which should make it more stable with temperature. Most controllers should have a delay between turning on and off the compressor so it won't damage it, but with a fan and door, I would be safe to do that every 5 seconds.

Also since the freezer would be the kegerator side, I wouldn't have to worry as much about fluctuating temperatures. I'm thinking I'll abandon the top/bottom freezers and keep an eye on CL for a cheap side by side.

 

[MalFet]

Thanks. I figured it wouldn't be like controlling 2 compressors, but more like controlling a compressor to get the freezer down to temp, and then the door/fan (I'm guessing a relay would work for this as well). Of course this means the freezer would have to be colder, so the fridge would have to be the ferm chamber.

Yep. Presumably you wouldn't even need a relay because the thermostat likely has its own, but this would require reverse engineering the thermostat a bit to know for certain.

 

[edgeofblade]

@edgeofblade: I don't know the design of your fridge of course, but in many (most? all?) consumer fridges you can't keep the fridge colder than the freezer. You might want to test this out before you invest too much money into controllers.

I understand. The whole shindig may be a lost cause and that garage fridge will only be good as a ferm chamber, but I'm ok with that. I'll just save a bit more and buy a cheap chest and go keezer instead of keggerator.

 

[MalFet]

I understand. The whole shindig may be a lost cause and that garage fridge will only be good as a ferm chamber, but I'm ok with that. I'll just save a bit more and buy a cheap chest and go keezer instead of keggerator.

Take this unsolicited advice as nothing more than that, but to my mind the easiest solution to this problem is to get a single vessel and to cool it to the coldest temperature you need (i.e. kegerator temps). Then, wrap the fermentation vessels in a fermwrap or the flexwatt tape someone else mentioned, and stick the whole thing in a neoprene carboy jacket. It seems strange to run a heater inside a fridge, but with good insulation on the carboy it's not really that inefficient.

 

[Erroneous]

Yep. Presumably you wouldn't even need a relay because the thermostat likely has its own, but this would require reverse engineering the thermostat a bit to know for certain.

I'm unsure what you mean by this. I'm planning on controlling the door/fan with the arduino, not the built in thermostat. I'm planning on circumventing the fridge's thermostat and control the compressor and door/fan manually.

 

[MalFet]

I'm unsure what you mean by this. I'm planning on controlling the door/fan with the arduino, not the built in thermostat. I'm planning on circumventing the fridge's thermostat and control the compressor and door/fan manually.

Weellll...sorta. It's kind of just semantics at this point, but it depends on what consider to be part of the thermostat. If you use "thermostat" to mean only the temperature probe, then you are certainly replacing that completely.

But, more generally defined, a thermostat includes its switching mechanism (which might be mechanical, or a transistor or relay) and perhaps even its actuator (in this case, a fan and damper). Plus, of course, all the caps, diodes, and whatwhat that make everything play nicely together. You're not so much replacing all of that as splicing into it.

Where exactly you splice in depends on the system's design, but in an ideal world you would recreate as little circuitry as possible, if only because you might need to reverse engineer unlabeled parts. In a lot of fridge designs you wouldn't need (or want) to provide your own relay but rather to use the one already built into the fridge. You could hack all the way in and control the fan directly, but at least on my fridge that was a lot more complicated, a lot more technical, and a lot more fridge-mangling than splicing in further upstream. On yours, though, it could of course be exactly the opposite. Good luck, and let us know how it turns out!

 

[Erroneous]

Ah yes. If the thermostat assembly has the power supply for the fan, and not just a wire coming off of it, I'd have to find that voltage, get a power supply, and do it that way. I'm not sure why using my own relays would be a bad thing, so long as they're rated for the highest wattage found in the unit. Well hopefully I'll be picking up my side by side from craigslist today and can start tinkering with it.

BTW, just modified my arduino's code to accommodate as many sensors and relays as you can fit on the board. It uses OneWire so you can use those nice DS18B20 temp probes, and logs one of them to the arduino's EEPROM every 15 minutes, which can go up to 10.5 days (on 1KiB EEPROM models). You can then connect the arduino via serial/ftdi usb and dump the recorded temps to a log file using a separate program. Anyone interested in the code?

 

[MalFet]

Ah yes. If the thermostat assembly has the power supply for the fan, and not just a wire coming off of it, I'd have to find that voltage, get a power supply, and do it that way. I'm not sure why using my own relays would be a bad thing, so long as they're rated for the highest wattage found in the unit.

So long as your thermostat is a simple two-wire switch? Nothing bad about it at all, except perhaps you'd have more circuitry than you really need (if that counts as something bad). But, all this is speculative until you know what kind of thermostat you are getting. Pray it is a cheap-o model. Those tend to be simpler.

What do you mean when you say you modified the arduino's code to accommodate more sensors and relays? I wasn't aware that any of the libraries had limits on how many sensors or relays you could maintain (beyond IO port limitations, of course). In any case, nice work!

 

[certaut]

Were you using a love or similar controller for the fridge temp? I'm thinking that since I'd be running on arduino, I can minimize the time between relay switches, which should make it more stable with temperature. Most controllers should have a delay between turning on and off the compressor so it won't damage it, but with a fan and door, I would be safe to do that every 5 seconds.

I used a ranco single stage to control the compressor set at 38 degrees and a cheap thermostat from a fried wine refrigerator to control a 80mm fan that blew cold air into the refrigerator (fermenter) side.

 

[Erroneous]

What do you mean when you say you modified the arduino's code to accommodate more sensors and relays? I wasn't aware that any of the libraries had limits on how many sensors or relays you could maintain (beyond IO port limitations, of course). In any case, nice work!

I mean tweaked my own code for the arduino. Originally I simply defined what port was the relay, what was the sensor, etc but only coded for a single sensor and relay. I changed it to an array of values for each thing so that it was (only theoretically) infinitely expandable. This way, if you add sensors and relays, all you have to do is add the right data into the arrays. Sure I could have modified brewtroller's code to my purposes, but for simple things I prefer coding from scratch to modifying existing code. Plus I just love programming.

I used a ranco single stage to control the compressor set at 38 degrees and a cheap thermostat from a fried wine refrigerator to control a 80mm fan that blew cold air into the refrigerator (fermenter) side.

Like I figured. Something like a wine fridge thermostat should still have a safety timer to keep the compressor on for a minimum time before turning off. This would produce more temperature fluctuations that tend to be colder than the desired temp since the blower probably would cool a 60F fridge down pretty quickly. BTW, keeping the fridge filled with as many containers of liquid as possible would help stabilize the temperature.

On the other hand, there is the possibility that cold air is escaping through the door. If you could keep the freezer side warmer, or somehow insulate the door, that might be a solution. Or as previously mentioned, use something to keep the fermenter warm.

 

[Erroneous]

So I took a look into my fridge I got from CL. This one doesn't have a separate setting for the freezer and the refrigerator, rather it has a general warmer/cooler setting and an "air balance" setting. The thermocouple is in the refrigerator, so it turns on the compressor until the fridge is at the desired level. The blower from the compressor goes through a vent in the freezer and a vent at the top of the fridge. That vent has a door that is opened and shut with the "air balance" setting.

This complicates things slightly. Now I'll have to make a new program and get a little more hardware. What I plan on doing is putting a sensor in both compartments, and attach either a servo or an actuator to the vent door. When either side is too warm, I'll turn on the compressor if the time limit has passed. Then, I'll keep the vent door open to a proportion equal to (freezer current temp - freezer desired temp) : (fridge current - fridge desired). So if the current temps are 50 freezer, 62 fridge and the desired are 36, 60, then the math would be 14:2, or 14% open. The compressor would keep going until both sides were cold enough, which would mean the freezer side would tend to get a bit too cold since the blower goes mainly in there even with the door all the way open.

Also, the defrost timer (says it's 10 hours) is connected to the thermostat, so I'd have to figure out how to wire that up correctly, if I'd need to at all. Maybe another relay to begin defrosting if the temp won't come down...

 

[Erroneous]

Just an update. Last night I started controlling the fridge with my arduino controller. One thing I noticed is that (overnight anyway) it kept getting cooler with the compressor turned off. Likely this is because the door for the blower is open and cooler air is leaking in. That'll be remedied after I attach the servo and 2nd thermometer for the freezer section. I didn't yet because I plugged in one of my sensors backwards so now it only reports 127.94 C. I've got 5 more on order so won't be having that problem for too long.

Anyways, figured I would let you guys know how much this setup costs.
1 relay controller: 12.95
1 temp sensor: 4.25 (found a site where I can get them for under $2)
1 arduino w/ usb built on board: 29.95
1 arduino w/out usb: 19.95
USB FTDI board (if no usb on arduino): 14.95
female headers (if no usb on arduino): 1.50

Total cost:
USB arduino: 47.15
FTDI arduino: 53.60
FTDI arduino if you already have the FTDI board: 38.65
Add 14.95 if you want a microSD shield for bigger logs.
Add another 15-16 if you want an LCD.

In short, before shipping about the same price if not a little lower than a love controller, but has the added bonus of data logging and expandability. On the down side, this is without a nice LCD and all the time put into it. Also you can get a love controller working very quickly.

 

[Erroneous]

Just wanted to update. I checked over my 694 KiB log file and did some analysis on the data.

With the set temperature at 68* F (20* C) over a period of 116 days, my refrigerator came on 949 times, had 4 resets (power outs and 1 reset every ~50 days of continuous up time), was running the compressor a total of 7.26 days and was therefore on for 6.2% of the time.