Compressor refrigerator fan overkill...

[@Michael]

I've had a series of headaches with the compressor fridge in my '19 Coachmen Crossfit. The OEM installation was poorly vented. The dealer that performed the TSB to add insulation and a vent did a really poor job and lied about what they did. I did my own venting and supplemented the vent with powered fans.

The venting and fans worked. I flip the switch on when it's hot in the camper, and off when it's not. With the fans running I've gotten 60F of temperature difference between the fridge and ambient (100% duty cycle). My beer was cold on a 100F day. Life is good.

So instead of leaving well-enough alone, I'm building a temperature sensitive cloud-enabled (IoT) fan control. Overkill, to say the least.

My goal is to make the fan automatic, run only as fast as needed to vent the back of the fridge, and track what it does so I have data on how much vent and fan I need in a particular situation.

The gizmos in the center of the photo are an ESP8266 microcontroller (a small, low power CPU that runs custom software), a PWM motor controller, a 12V NO/NC SPDT relay and a temperature sensor.

The microcontroller has the software and the logic that controls the fans and connects to the cloud.

The relay is a fail-safe. If the microcontroller goes off-line for any reason, the relay closes and the fans run at 100%. The temp sensor (the blue thing dangling below the relay) will be installed above the fridge coils, below the fans.



This is still a desktop mockup. I need a few more doodads before I can get rid of the rats-nest of jumper wires and install it in the camper.

To make it work, I wrote software that reads the temperature sensor, adjusts fan speed by varying a PWM signal (just like your desktop computer) and uploads current temperature and fan speed to a cloud-enabled app. The app lets me adjust temperature set points using my phone from anywhere I have Internet. For example, I can set the controller to run the fans at 10% when it's 80F, 50% at 90F, 100% at 100F, etc. It also lets me override the controller and run the fans at 100%.



If the microcontroller can't attach to Wi-Fi or the Internet, it will continue to control the fans without uploading data.

I don't expect this to improve fridge performance, nor will it work much better than simply buying a temperature sensing PWM motor speed controller from Amazon. But it will give me fancy charts and graphs to look at - something which the simple solutions would not.

I'll report back once it's installed and working.

[rowiebowie]

Your way seems way overkill, but who of us haven't been there.

On my van, Airstream also did a poor job of venting (zero lower venting, when Novacool calls for 30 sq. inches) when they went to compressor fridges. My van was used and out of warranty, so I added my own insulation and fan to vent the cabinet behind my fridge back into the coach living area.

But my simple solution was to wire the fan to a button thermostat that turns on a 100 degrees and goes off at 80 degrees. According to my temperature probe monitoring that area, it is pretty precise.

I also wired it to an off/on switch. But I didn't have the interior cabinet space for a large fan opening, so I used a squirrel-cage fan that vents through a 1-1/2" tube into the coach living area.

https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1

https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1

.

[eric1514]

Originally Posted by @Michael

...
I don't expect this to improve fridge performance, nor will it work much better than simply buying a temperature sensing PWM motor speed controller from Amazon. But it will give me fancy charts and graphs to look at - something which the simple solutions would not. ...

Exactly. I love it.

[mkguitar]

we are still on the Dometic 3 way
it's about removing heat


most recent trip shows that my 3 exhaust fans on the cooling fins are overkill and that the fins are often barely warm to the touch at ambient temps below 80º
( I often remove the cover on the top when camped)


The van came with a popular "kit" installed under the fins on a temp controlled switch-

I had added 2 exhaust fans at my vent- and a tin panel to direct air through the fins- contacting, it too can act a sink

a tiny internal fan to waft the air inside the fridge on it's own switch


I am planning to leave the added fan kit on the temp sensor

and put each of my exhaust fans on a separate switch, the 2nd I'll use when we are at 90º plus


if all 4 fans are running my current draw is about 800mA, Usual use is about 300mA


I look forward to the results with your project- wathcing the thermometer read out is an obsession


Mike

[booster]

I am certainly the outlier in the more fans are better theory, I think.


We have an Isotherm 3cf frig that uses the Danfoss/Secop modular compressor/condenser/fan all in one little rectangular unit that is in the upper right rear of the frig.


I did go through lots of early air routing stuff to get the airflow to the fan in and out smooth and efficient with no looping or such to get the frig to use the same power in the same test procedure in the same conditions in the shop on the bench and in the van using the external venting from Roadtrek. It stays cool at lowest compressor speed in 100*F weather in the sun.



After that I tested fans in the discharge and in the frig and neither saved as much power as they used either by themselves or together. If fact the tiny savings were so small that they could be measurement errors or such.


To put it in perspective, selecting the best compressor speed can save much more energy than fans do by a large margin and depending on the frig it may or may not be the slowest speed. Very surprisingly, just removing the drip tray from ours will save over 10% in energy and only raise the freeze temp a few degrees and not above good freezing.


So for us no need for any extra fans, just a duct and some air bypass sealing. The air that comes out our upper vent is barely warm.

[@Michael]

Originally Posted by rowiebowie

Your way seems way overkill, but who of us haven't been there.

Sometimes overkill is the whole point. I'm also learning a lot about microcontrollers, re-learning my soldering skills last used 40 years ago and my 'C' programming skills last used 20 years ago.

I like your simple solution though.

Originally Posted by mkguitar

it's about removing heat [...] most recent trip shows that my 3 exhaust fans on the cooling fins are overkill

Same in my case. I think that two fans running at low speed will be adequate in most situations.

Originally Posted by booster

So for us no need for any extra fans, just a duct and some air bypass sealing. The air that comes out our upper vent is barely warm.

In my specific case, the solar radiation from the roof heats up the compartment above the fridge and disrupts natural convection, so fans make a measurable difference. Without fans I get about 40-45F cooling under ambient, with fans about 60F. When the sun isn't heating the roof, fans don't make as big a difference.

The adventure continues.

[Davydd]

I have an Isotherm Freeline 115 at 4.2 cf. It sets at 16” high and the inlet vent is at 16” directly underneath the refrigerator and the the outlet vent is at the top of the cabinet at the ceiling 75” above the floor. It is not on an outside wall, gets a good natural draw but there are two fans to help it. So far we have had consistent cooling up to 90 degree ambient temperature with little or no adjustment. Being entirely inside a conditioned space and not on an outside wall helps the consistency. In the past I have had trouble keeping cooling below 40 degrees in both Dometic 3-way absorption outside venting and Nova Kool inside venting compressor refrigerators.

[rogerstg]

I don't think it's overkill. Since you have the skills to do this, why not add the safety measure like Fridge Defend where it shuts down the fridge if the boiler overheats due to not being level for too long?

[@Michael]

Originally Posted by rogerstg

I don't think it's overkill. Since you have the skills to do this, why not add the safety measure like Fridge Defend where it shuts down the fridge if the boiler overheats due to not being level for too long?

It's a compressor fridge, which I believe has its own thermal protection. I'll read up on that though.

As of a couple hours ago it's installed. I modified the software to allows me to set the various thresholds on the fly from my phone, so I can figure out how at what temp to start the fans, and much fan I need to keep the temps within a certain range.





Ideally I'd measure cabin temperature and only run the fans as fast as necessary to keep the coil temperature close to cabin temperature. No point in spinning the fans faster and faster trying to get the coils cooler than the cabin air.

[CaDreamin]

I have two fans for my Dometic 3-way. There is a temp sensor that turns them on and off.

If bright direct sunlight hits the van where the frig is, nothing will help you.

But under other conditions it works pretty well. It comes on very rarely, 90 degrees ambient or higher.

[MarcTremblay]

Originally Posted by @Michael

I've had a series of headaches with the compressor fridge in my '19 Coachmen Crossfit. The OEM installation was poorly vented. The dealer that performed the TSB to add insulation and a vent did a really poor job and lied about what they did. I did my own venting and supplemented the vent with powered fans.

The venting and fans worked. I flip the switch on when it's hot in the camper, and off when it's not. With the fans running I've gotten 60F of temperature difference between the fridge and ambient (100% duty cycle). My beer was cold on a 100F day. Life is good.

So instead of leaving well-enough alone, I'm building a temperature sensitive cloud-enabled (IoT) fan control. Overkill, to say the least.

My goal is to make the fan automatic, run only as fast as needed to vent the back of the fridge, and track what it does so I have data on how much vent and fan I need in a particular situation.

The gizmos in the center of the photo are an ESP8266 microcontroller (a small, low power CPU that runs custom software), a PWM motor controller, a 12V NO/NC SPDT relay and a temperature sensor.

The microcontroller has the software and the logic that controls the fans and connects to the cloud.

The relay is a fail-safe. If the microcontroller goes off-line for any reason, the relay closes and the fans run at 100%. The temp sensor (the blue thing dangling below the relay) will be installed above the fridge coils, below the fans.



This is still a desktop mockup. I need a few more doodads before I can get rid of the rats-nest of jumper wires and install it in the camper.

To make it work, I wrote software that reads the temperature sensor, adjusts fan speed by varying a PWM signal (just like your desktop computer) and uploads current temperature and fan speed to a cloud-enabled app. The app lets me adjust temperature set points using my phone from anywhere I have Internet. For example, I can set the controller to run the fans at 10% when it's 80F, 50% at 90F, 100% at 100F, etc. It also lets me override the controller and run the fans at 100%.



If the microcontroller can't attach to Wi-Fi or the Internet, it will continue to control the fans without uploading data.

I don't expect this to improve fridge performance, nor will it work much better than simply buying a temperature sensing PWM motor speed controller from Amazon. But it will give me fancy charts and graphs to look at - something which the simple solutions would not.

I'll report back once it's installed and working.

I have the same truck, 2018. Same issue.
The end goal is certainly to consume less energy. Aren’t the fans going to consume more energy than the extra consumption of the fridge having to run longer without the fans? Perhaps running the fans only when the fridge runs? I wonder what is the best scenario to minimize batteries consumption. Maybe building an arduino temp sensor that runs the fans only when there is a temp difference more than 2 degrees between below the fridge and above the fridge. Great article. Thanks.

[@Michael]

Originally Posted by MarcTremblay

I have the same truck, 2018. Same issue.
The end goal is certainly to consume less energy. Aren’t the fans going to consume more energy than the extra consumption of the fridge having to run longer without the fans? Perhaps running the fans only when the fridge runs? I wonder what is the best scenario to minimize batteries consumption. Maybe building an arduino temp sensor that runs the fans only when there is a temp difference more than 2 degrees between below the fridge and above the fridge. Great article. Thanks.

I end up with a heat bubble above the fridge/behind the microwave that's hotter than the coils below, so the normal cooling via convection fails and the fridge cannot keep up even with 100% duty cycle. So not only energy use, but spoiled food ... or spoiled beer.

I.E. with no fans on a hot day with the sun is burning up the dark blue roof, I have a fridge with 100% duty cycle at 5.5 amps that's not keeping the fridge cold. My fan solution uses .14A at idle, and each fan uses .1A at 10%, .2A at 100%, so at worst less than one amp.

This is a 'feature' of the way Coachmen designed this unit - the roof heats up, the compartment above the fridge gets hot, and convection stops. It's made worse by Coachmen's poor venting. Even the newer ones have too small of openings above the microwave, and the 2019's and prior had no venting at all.

I've modified my ESP8266 firmware so that it only spins the fans when the temp above the coils is 2 degrees F greater than the cabin temp, and so the fans spin proportionally to the difference between cabin and coil temp.

The Wemos Mini D1's ESP8266's that I'm using are like Arduinos, except smaller, cheaper, and with built-in Wi-Fi, so what I built is roughly what you've suggested. I can even use the Arduino IDE if I want, but prefer VScode with PlatformIO.

I have the microcontroller and fans on an on/off switch, so can further reduce the .14A idle draw by turning off the switch at night when I'm on battery. But then I must remember to turn it on again in the morning. That's a problem.

[Happy Nomads]

My two cents... Stop the problem at the sources.

Paint your roof white, using a modern metal roofing paint formulated to emit in the infrared. White is best, but even a "cool" roofing paint in blue is cooler than a good solid blue in old style heat-retaining automotive paint. Can you buy a gallon? Hmmm can't say, sorry. There is always a way.

Second, add at least a thin layer of high value insulation on the roof over the fridge. You might experiment with a thin 1/4" sheet of aerogel under a protective outer cover. Or mount a sun shading panel 1.5" off the roof to stop direct solar insolation. And mount it well��.

Third, if you are doing as much work as it seems, you must have pulled the fridge out several times. They make a composite aerogel/ aluminum foil / aerogel / aluminum foil sheet you can attach directly to the underside of the roof above the fridge.
You have thoroughly addressed the fans.
Next is the inside of the fridge itself. Tough to do much with a small fridge, but one inch sheets of foil-faced (foil on both sides) styrofoam or polyisocyanurate rigid insulation from Lowes or Home Depot cut to fit against the walls doesn't take much room. I did this with 2" inside and outside a home standing freezer and more than doubled its efficiency. It now cycles less frequently while attaining a temp 10 below zero, about 25 degrees colder than the original product.

Finally, if you love ugly and inconvenient tweaks, cover the outside of the door of your fridge with a full 2" of the foil-backed rigid board. Ugly it is, but cold beer is worth it.

[rowiebowie]

Originally Posted by Happy Nomads

My two cents... Stop the problem at the sources.

Paint your roof white, using a modern metal roofing paint formulated to emit in the infrared. White is best, but even a "cool" roofing paint in blue is cooler than a good solid blue in old style heat-retaining automotive paint. Can you buy a gallon? Hmmm can't say, sorry. There is always a way.

Second, add at least a thin layer of high value insulation on the roof over the fridge. You might experiment with a thin 1/4" sheet of aerogel under a protective outer cover. Or mount a sun shading panel 1.5" off the roof to stop direct solar insolation. And mount it well��.

Third, if you are doing as much work as it seems, you must have pulled the fridge out several times. They make a composite aerogel/ aluminum foil / aerogel / aluminum foil sheet you can attach directly to the underside of the roof above the fridge.
You have thoroughly addressed the fans.
Next is the inside of the fridge itself. Tough to do much with a small fridge, but one inch sheets of foil-faced (foil on both sides) styrofoam or polyisocyanurate rigid insulation from Lowes or Home Depot cut to fit against the walls doesn't take much room. I did this with 2" inside and outside a home standing freezer and more than doubled its efficiency. It now cycles less frequently while attaining a temp 10 below zero, about 25 degrees colder than the original product.

Finally, if you love ugly and inconvenient tweaks, cover the outside of the door of your fridge with a full 2" of the foil-backed rigid board. Ugly it is, but cold beer is worth it.

Welcome to the forum Nomads!

You sound like you have experience in this area and I'm glad your solutions worked for you. However, many of us have rigs with well insulated fiberglass tops, so not much additional is needed there. Most of my heat enters behind the fridge from the outer wall, where I did add additional insulation.

In my case, Airstream failed to install sufficient top and bottom venting (30 sq. inches each) as recommended by Novacool for my size compressor fridge. Plus the top venting they did provide does not have a clear path, but rather must flow behind and then over the top of the microwave.

[booster]

I went back to the other thread where @Michael showed his chimney, sketch of the a sideview and opening sizes.


I wanted to see just how it all fit together and get a closer look at the frig configuration of compressor, factory fan, and coils.


I like the basic design of the chimney around the coils, although I would prefer the chimney part to be a smooth surface like aluminum sheet or foilboard so airflow is smoother. The fan location above them, in that older system would need to be sealed completely horizontally to assure all the air goes up over the coils and not around the edges.


I think the factor fan blows over the compressor to cool it and out a hole in the case or open area to the side. I think a vertical baffle to keep that hot air away from the coils would help. Getting fans as close to, or the exit air ducted to a sealed to upper vent would assure all the air went out the vent and couldn't loop back down and back into the fans. Keeping the volume of the area above the coils small will and uniform will help it not to have hot stagnant areas.



Another posters suggestion of a radiant barrier with airgap at the roof is essential, I think.


If all the air can coming in and leaving is moved by a couple of fans at the exit goes over only the coils and comes in only from the bottom I think it should cool well as the cooling air and coils will not be heated by roof air if it is exhausted with the air off the coils and compressor.


In that discussion @Michael mentioned some redesign was going to be done, so now it all could be somewhat difference.

[avanti]

I may be under-thinking it, but it seems to me that if you have a chimney-style setup opening into the van and with (a) a pinch-point at the coils (b) a thermostatically-modulated exhaust fan at the top, and (c) fan power arranged such that it only runs when the compressor is running, then thing are going to be pretty optimal. You will always have room-temperature air flowing past the coils and the fans will never run any faster than necessary.

Ebay and Amazon are full of cheap fan controllers that are perfect for this.

[@Michael]

Originally Posted by booster

I like the basic design of the chimney around the coils, although I would prefer the chimney part to be a smooth surface like aluminum sheet or foilboard so airflow is smoother. The fan location above them, in that older system would need to be sealed completely horizontally to assure all the air goes up over the coils and not around the edges.


I think the factor fan blows over the compressor to cool it and out a hole in the case or open area to the side. I think a vertical baffle to keep that hot air away from the coils would help. Getting fans as close to, or the exit air ducted to a sealed to upper vent would assure all the air went out the vent and couldn't loop back down and back into the fans. Keeping the volume of the area above the coils small will and uniform will help it not to have hot stagnant areas.

Another posters suggestion of a radiant barrier with airgap at the roof is essential, I think.

If all the air can coming in and leaving is moved by a couple of fans at the exit goes over only the coils and comes in only from the bottom I think it should cool well as the cooling air and coils will not be heated by roof air if it is exhausted with the air off the coils and compressor.

In that discussion @Michael mentioned some redesign was going to be done, so now it all could be somewhat difference.

Here's a new picture of the microwave compartment above the fridge. I have several layers of insulation above the reflective layer, and I've sealed the compartment to prevent convective heat transfer from the sides. The reflective layer is felt backed automotive underhood insulation.



I've also:

• spaced the fridge out into the isle 3/4" to give me more room to insulate behind
• added a 1/2" sheet of polyiso against part of the inside skin of the van (there isn't room for more, and various wire looms and bits of plumbing are in the way. )
• fixed up the chimney area by making sure there were no stray convection paths
• added a bit of fiberglass insulation to the top and sides of the fridge (via a modified water heater blanket)

Here's a sketch of the situation:



Here's a current photo of the chimney, looking down from the microwave compartment.



And a photo of the polyiso behind the fridge, with the 'chimney' insulation removed. To the outside of the polyiso is the factory Ford window. To the inside, the polyiso is covered with a layer of the pink closed-cell foam that Coachmen uses throughout their vans.



And the back of the fridge compartment prior to re-installing the fridge.



Ther might be an airgap between the polyiso and pink closed cell foam, but it'll be inconsistent, depending on plumbing, wiring etc.

--Mike

[booster]

Originally Posted by avanti

I may be under-thinking it, but it seems to me that if you have a chimney-style setup opening into the van and with (a) a pinch-point at the coils (b) a thermostatically-modulated exhaust fan at the top, and (c) fan power arranged such that it only runs when the compressor is running, then thing are going to be pretty optimal. You will always have room-temperature air flowing past the coils and the fans will never run any faster than necessary.

Ebay and Amazon are full of cheap fan controllers that are perfect for this.

I would totally agree with this statement, with the only caveat being the size of the upper cavity and fan location to a lesser degree. In @Michael's case and with some other Novakool frigs with the compressor on the bottom a secondary chimney/baffle to get that hot air to the top probably would help some, too. Compressor on top no problem except getting it cool air.


I am always a bit amazed by the Secop compressor/controller/coils/fan assembly in our 3.0cf frig as it is very small in volume but cools very well with the only airflow a 70cfm computer fan on it going into the unit.

[@Michael]

I'm working on version 2.0 of my refrigerator fan controller overkill. The existing version is WiFi/cloud connected, but I figured it would be more useful if it had Bluetooth capability also. My idea is to have all of the devices in the camper be able to communicate with each other and be controllable whether I have WiFi/Internet or not. I still want WiFi/Internet/Cloud connectivity so that I have visibility and control when I'm away from the camper though.

To enable Bluetooth, I've switched from an ESP8266 to an ESP32-C3 microcontroller and programmed the ESP32 to set and publish various data and parameters via Bluetooth. It's mocked up and ready to install - just waiting for above-freezing temps.



As with version 1, this one has a deadman relay so that if the ESP fails for any reason, the relay closes and the fans run at 100%

I'm also working on adding a display/touch panel to allow for viewing and control of whatever I end up monitoring and/or automating. First try will be to use a Nextion display. The Nextion displays allow for the graphics to be designed with a GUI desktop app, and the variables/parameters/buttons/switches to be controlled via a serial port using simple ASCII commands. One can have full control over what's on the screen, and the screen itself has a microcontroller, so can run simple code and logic.

My plan is to use another ESP microcontroller to act as a bridge between Bluetooth and the Nextion serial interface. This will allow other devices in the camper to update the screen via Bluetooth without having to be wired to the Nextion display, so the display can be mounted wherever convenient, rather than limited to where I can fish wires.

In theory, this could enable monitoring and control of anything that can be remotely monitored or controlled, though I'm not sure how far I want to take this. For now, I just have basic battery status, solar controller status, fridge temperature, and a real-time clock dummied up on the display. I'm working on the Bluetooth interface next.



This type of display is half-way between using a non-intelligent display connected to a microcontroller and using a full featured home automation system/application like @avanti is building.

https://www.classbforum.com/forums/f...ion-12805.html

Using something like Home Assistant will certainly be easier though.

[MarcTremblay]

Originally Posted by @Michael

I'm working on version 2.0 of my refrigerator fan controller overkill. The existing version is WiFi/cloud connected, but I figured it would be more useful if it had Bluetooth capability also. My idea is to have all of the devices in the camper be able to communicate with each other and be controllable whether I have WiFi/Internet or not. I still want WiFi/Internet/Cloud connectivity so that I have visibility and control when I'm away from the camper though.

To enable Bluetooth, I've switched from an ESP8266 to an ESP32-C3 microcontroller and programmed the ESP32 to set and publish various data and parameters via Bluetooth. It's mocked up and ready to install - just waiting for above-freezing temps.



As with version 1, this one has a deadman relay so that if the ESP fails for any reason, the relay closes and the fans run at 100%

I'm also working on adding a display/touch panel to allow for viewing and control of whatever I end up monitoring and/or automating. First try will be to use a Nextion display. The Nextion displays allow for the graphics to be designed with a GUI desktop app, and the variables/parameters/buttons/switches to be controlled via a serial port using simple ASCII commands. One can have full control over what's on the screen, and the screen itself has a microcontroller, so can run simple code and logic.

My plan is to use another ESP microcontroller to act as a bridge between Bluetooth and the Nextion serial interface. This will allow other devices in the camper to update the screen via Bluetooth without having to be wired to the Nextion display, so the display can be mounted wherever convenient, rather than limited to where I can fish wires.

In theory, this could enable monitoring and control of anything that can be remotely monitored or controlled, though I'm not sure how far I want to take this. For now, I just have basic battery status, solar controller status, fridge temperature, and a real-time clock dummied up on the display. I'm working on the Bluetooth interface next.



This type of display is half-way between using a non-intelligent display connected to a microcontroller and using a full featured home automation system/application like @avanti is building.

https://www.classbforum.com/forums/f...ion-12805.html

Using something like Home Assistant will certainly be easier though.

This is very interesting. I’d like to do something like this in the future for my CrossFit. I’m still in the learning process. At the moment I’m displaying data on my android phone from my arduino clone Bluetooth. Android app is created with mit inventor on line. Check this out if you haven’t so far. It’s simple and free. You build a Bluetooth app in no time. No programming.