Refrigerator Circulation Fan Hack

Thread Starter

Jim_PDX

Joined Nov 13, 2019
9
TLDR SUMMARY: I want to power a refrigerator circulation fan from two different power sources: the fridge's original source from its motherboard, and also from an MCU that I control. How do I do this?

READ THESE DETAILS IF YOU PLAN TO REPLY:

About two years ago I purchased a new Blomberg kitchen refrigerator. It wasn’t long before we noticed that our fresh greens were freezing regularly unless we placed them on the top shelf. They would freeze in the crisper shelf which was at the bottom of the fridge. During that first year we had a repair person come three times to try to fix it. Once he tried replacing the circulation fan and once replacing the fridge’s motherboard. I got to keep the “broken” parts (I tested the 12V fan he pulled out and it worked fine). Anyway, it’s now been two years and I’m not happy with the temperature control. Veggies keep freezing, but not as frequently, and something is definitely not working right, and I’m no longer under warranty and don’t want to pay the high repair bills.

Here’s what I’ve done so far: Using an MCU (an ESP8266 - Wemos D1 mini), I set up 4 calibrated temperature sensors on 4 different shelves in the fridge. I took readings every 3 minutes and sent them to an IoT cloud service. The results were very useful! (see attached graph)


FridgeTempGraph.jpeg


I concluded the following:

  • the difference between the warmest shelf (top) and coldest shelf (bottom, but above the one crisper drawer) was often around 20 degrees and sometimes more than that
  • the circulation fan blows air from the top shelf down to the bottom shelves behind the back wall where there’s a cooling element.
  • the circulation fan turns on just once every 3-4 hours and is on for approximately 15 minutes

My conclusion from all of this is that the circulation fan is just not staying on long enough.

I’m writing you for a review of my design to force the circulation fan to be on for longer periods, which should keep the fridge from having extreme temp differences between the top and lower shelves.

Here are the parameters I have to work with:

  • I can access the power lines to the the circulation fan. It’s a 12V DC fan using 0.16A
  • I can NOT access any continuous power source within the fridge
  • I want to power the fan from the 5V power source that powers my ESP8266. I’ll run two wires into the fridge with 5V from a wall wart
  • I’ll write an algorithm for the ESP8266 that will attempt to turn the fan on whenever the temperature difference between the top and bottom shelves is over a fixed amount (maybe 12 degrees)
  • I want the fan to stay connected to its existing motherboard connection AND to alternatively power on by my device based on my algorithm whenever it’s not being powered by the motherboard. (i.e. I want the motherboard to think that everything is operating normally, and not throw an error code if it senses the fan isn’t connected). (Is this realistic? I don’t know! Any Blomberg engineers listening in???)

Please review the attached schematic. It uses a DPDT relay “controlled” (the coil) by the motherboard that turns on the circulation fan.

Do you think this will work? A simpler plan is to not use a relay and just disconnect the fan from the motherboard completely and power it only from my own source. Don’t know if this is a good idea or not.

BTW, this is my first post - I have learned so much from following this forum these past months - thank you!!!

Schematic.png
 

MrChips

Joined Oct 2, 2009
22,912
Do you know that circulation is the problem?
Here is what I would do as a test. Take a small muffin fan and run it continuously blowing air from bottom to top.
Continue to take your measurements and see if the fan makes any difference.
 

Thread Starter

Jim_PDX

Joined Nov 13, 2019
9
I feel pretty certain already it’s a circulation issue. Look back at the chart at what happens around 5:30 in the morning. What else would cause a sudden spike in temp in the lower shelves AND an almost as sudden drop in temp in the upper shelves? Then after 20 min or so, it starts to slowly return to the way it was before?

The existing fan is a muffin fan, built into the back wall. There is no good place for me to put an additional fan that will circulate the air well. That’s why I want to force the existing fan to turn on more frequently.
 

Rich2

Joined Mar 3, 2014
175
Why not try turning the thermostat up a bit. It's sometimes useful to have a temperature range. There is no fan in either of my fridges and they are ok.
Beer and raw meat in the bottom butter and eggs in the top etc..
 

Rich2

Joined Mar 3, 2014
175
If it's a fan assisted evap it's a wonder it's not iced up if the fan is off a lot. It's called 'off cycle defrost' where by the fan runs all the time so when the compressor is off the fan defrosts the evap.
 

Reloadron

Joined Jan 15, 2015
5,788
As drawn it will do as you expect. When the relay is deenergized you have the fan control and when the relay is energized it works as advertised.

Ron
 

debe

Joined Sep 21, 2010
1,226
Fan asist evaporators switch the fan off & use a heater to defrost on defrost cycle. Still do on my latest LG inverter fridge.
 

Rich2

Joined Mar 3, 2014
175
Fan asist evaporators switch the fan off & use a heater to defrost on defrost cycle. Still do on my latest LG inverter fridge.
That's how commercial freezers defrost. Never really worked on domestic stuff so you know more than I do then.
 

Thread Starter

Jim_PDX

Joined Nov 13, 2019
9
The thermostat is digital, allowing settings between 33 and 46. It’s already set at 46.

Thanks Ron, for verifying my schematic.

Whether this is a fan assist evaporator or not, it seems that forcing the fan to run longer than it wants to won’t hurt anything. I’ll report back once I get this hack up and running!

Thanks everyone for the input.
 

Rich2

Joined Mar 3, 2014
175
The thermostat is digital, allowing settings between 33 and 46. It’s already set at 46.

Thanks Ron, for verifying my schematic.

Whether this is a fan assist evaporator or not, it seems that forcing the fan to run longer than it wants to won’t hurt anything. I’ll report back once I get this hack up and running!

Thanks everyone for the input.
You could maybe resite the thermistor to somewhere colder but I think you may have nailed it with a better air circulation
 

MisterBill2

Joined Jan 23, 2018
8,024
If the motherboard that you have is the same as the new one then you can trace the power circuit for the fab and add a connection so that it runs constantly.
The daily spike in temperature is probably part of the defrost cycle. But running the fan constantly will not cause any problem. I fixed a commercial cooler that had the fan switch fail by changing it to constant fan on. Less frost buildup and the temperature stays the same all the time.
And if you ever need warranty service just remove that jumper connection before the service person arrives. Simple, cheap, and very reliable.
 

Rich2

Joined Mar 3, 2014
175
If the motherboard that you have is the same as the new one then you can trace the power circuit for the fab and add a connection so that it runs constantly.
The daily spike in temperature is probably part of the defrost cycle. But running the fan constantly will not cause any problem. I fixed a commercial cooler that had the fan switch fail by changing it to constant fan on. Less frost buildup and the temperature stays the same all the time.
And if you ever need warranty service just remove that jumper connection before the service person arrives. Simple, cheap, and very reliable.
If you do this, I would disconnect the defrost heater. Defrost termination should be triggered by time as well as temperature so should work fine, as I said above "off cycle defrost"
 

MisterBill2

Joined Jan 23, 2018
8,024
If you do this, I would disconnect the defrost heater. Defrost termination should be triggered by time as well as temperature so should work fine, as I said above "off cycle defrost"
NO!! The defrost heater is needed, otherwise the end of defrost will not be triggered by the temperature rise in the evaporator area. I doubt that there is a default time limit for the defrost cycle. But there might be.
 

MisterBill2

Joined Jan 23, 2018
8,024
Hi Jim,
The important thing is; Please keep us informed, as to how the experiment goes.
Good luck...……..
And observe that NOT disconnecting the defrost heater is less disrupting to the system and also less work. ALL of them automated defrost systems that I have seen and investigated required a rise in temperature in the frost-forming area to end the cycle.
 

Rich2

Joined Mar 3, 2014
175
All commercial controllers that include defrost have a DET parameter (defrost endurance timeout, usually set at 30 min) it would be crazy not to include it, if the defrost heater fails the defrost wouldn't end until everything in the fridge/freezer is 10°c, or whatever the DST parameter it's set on (defrost stop temperature)

Also what if the defrost thermistor drops off or pulls out of the evaporator? It has been known. Without a defrost time limit The heaters would, and do get very hot after all the ice has gone... Red hot!

Also if the fans are running continuously with the defrost heaters on... Well that's not very good is it, blowing warm air round the fridge.

Anyway, that's commercial systems for you, who knows what hideous penny pinching goes on in the domestic market.

If all else fails you can fit a commercial controller with all the parameters and features.. You could probably pick one up for around £50.
The good ones are made by Eliwell, Corell, Dixel, LAE etc
 

Rich2

Joined Mar 3, 2014
175
I used to work in Supermarket refrigeration and we used to carry the 974 on the van's (ewpc974 at that time) if you like gadgets or are a control freak these are the business :) there are loads and loads of parameters to set like compressor protection, fan stop temperature, probe failure settings, temperature alarm settings.. And many more :)
https://www.ebay.co.uk/itm/123793092394
 

MisterBill2

Joined Jan 23, 2018
8,024
I used to work in Supermarket refrigeration and we used to carry the 974 on the van's (ewpc974 at that time) if you like gadgets or are a control freak these are the business :) there are loads and loads of parameters to set like compressor protection, fan stop temperature, probe failure settings, temperature alarm settings.. And many more :)
https://www.ebay.co.uk/itm/123793092394
OK, certainly the controllers for commercial coolers and freezers have a lot more functions. Residential refrigerators are far more into appearance and features of marginal value. But the circulation fan is vital and the defrost heater is also important because without that ice will block the evaporator and cooling will cease.
And here in the USA, for the $75=50# you might get to look at one of those controllers. Closer to $200 at our Grainger's supply company, and even more at a specialized refrigeration distributor.
 

Rich2

Joined Mar 3, 2014
175
defrost heater is also important because without that ice will block the evaporator
It's not so important in medium temperature applications, believe me I wouldn't lie :) The controllers have a "differential" or "hysteresis" parameter which is the difference between cut in and cut out. If you put that up a little from the usual default 2°c the evaporator has time to defrost in between compressor cycles as long as the fan stays on. As I said a few posts above it's known in the trade as "off cycle defrost"
Wouldn't work in a freezer of course and a meats fridge (-2°c +2°c) might struggle but standard chillers, bottle coolers and veg fridges all off cycle defrost.
Look up Eliwell 961 if you don't believe me, it has defrost parameters, but no defrost probe and no connection for any heaters. It's a medium to high temperature controller with off cycle defrost.

(hope I've remembered the 961 number right :))
 
Top