A different approach to internal refrigerator fans

[ESGWheel]

Originally Posted by jbflag21ds

You just need to be careful about how much velocity you are creating with the airflow. If it is too high, you don't get the maximum heat exchange and thus it takes longer to cool initially and recover from the door being opened.

jbflag:

This is exactly my thinking as well. The specs for the various internal refer fans do not go into detail about airflow that I have seen. Some state 1000s of RPM so I assume significant. Being sensitive to needing only a little flow is the reason I went with the 24VAC mini fans running on only 5VDC. The calculated airflow per fan is less then 2 CFM, but I suspect even less. It really is just a whisper of a ‘breeze’ not unlike when you exhale. But with this gentle flow continuously on I speculate that it will help keep the refer temp homogenous. Some devices advertise 'only on when the refer is in its cool cycle’ but I disagree with that approach. Better to keep circulating to keep temps even then allow one spot to warm up over the others. Will it cycle on more with continuous fans? Maybe, but I think that is a good thing and since no moving parts in an absorption refer, I am not concerned.

And based on all I have read and knowing how these systems work, agree there is more bang for the buck with external cooling which is next project I look forward to.

[ESGWheel]

Been running for a little over a week now with the fans installed and the temp sensors in their named locations (Top, Mid and Bottom). And all the while using the refer (i.e., trailer occupied). Below are the temps look like charted out.

While I expected the Bottom sensor to generally lag the other two and be a little warmer (it’s in the ‘vegetable drawer ‘) I did not expect the top and mid to have a more or less constant 4 degrees difference! I’ll continue to let it play out over the season with it 'as is' and perhaps disconnect the fans to see if any difference. But it looks like I need to crank them up a little!

[NXR]

No way I'd be happy with temps in the mid to upper 40's overnight when the doors are not being opened.

Cold air drops and normally the bottom is colder than the top so something weird is going on. Are you certain you do not have a problem with the bottom seal on the door(s)?

Ray

[ESGWheel]

Originally Posted by NXR

Are you certain you do not have a problem with the bottom seal on the door(s)?

i do not think so but something i will check when back there next week. Thanks for the tip!

[ESGWheel]

Just like Jim Nabors in the TV show “Gomer Pyle” I am surprised. Ok, here goes.

First off, my refer has worked fine, lasts a long time from the beginning. No issues whatsoever. Door seal is all good. And I have never had an issue with food storage. Ice cream does not get solid as typical with these absorption refers but makes ice cubes, no issue.

What surprised me was the deltas in temp throughout the refer. Recall I have 3 blue tooth thermometers (BTT): top, middle and bottom. Initially I compared each BTT to each other and ‘adjusted’ my numbers accordingly, but I was still seeing deltas in operation.

Turns out there is a calibration procedure for these devices, so I did that. And now, at the 32-degrees, are within 0.1 of each other.
Post calibration they are back in the refer these past couple of days and again, I am seeing as much as 8 degrees between the top and the middle (the bottom, in the vegetable compartment also similar to mid only slightly higher). My concern was that the top and middle are really only inches apart, so what is going on?

My first observation was that I had the top one ~6 inches away from the back of the refer. The middle one was next to the back wall. Hmmm…. Perhaps getting some heat from the hot tubes of the refer just on the other side. Move it away and sure enough, temp drops by a few degrees. Wow.

Thus, my first conclusion is that the insulation of these refers is not that great. But it does get hot on the other side where the gas needs to dissipate its heat in the tubes, causing me to again realize that should get more bang for the buck by installing the external fans to help dissipate that heat.

I did swap the top and middle BTTs to check and had same results: about a 5-degree temp difference with about a foot of vertical separation. See pic of placement.



Thus, my second conclusion is I do not have enough fan speed to cause the needed circulation to create a uniform inside temp. Recall these mini fans are running very slow (a 24VDC fan running at 5VDC). So, I will amp it up. Initially I will go to full 12V, which per my pervious calculations, will almost double their CFM. And I just ordered some thin metal plates with sticky on the back. I will use these to mount the BTT on the sides of the refer (they have a magnetic mount). Using these will keep them the same distance from the rear and not occupy shelf space.

One thing I did notice: no ice build-up on the refer fins. Previously I did have this but as noted above by NXR, there is the benefit of no icing by using properly placed internal fans.

More to follow with the 12V speed.

[Triple C]

Originally Posted by ESGWheel

Just like Jim Nabors in the TV show “Gomer Pyle” I am surprised. Ok, here goes.



First off, my refer has worked fine, lasts a long time from the beginning. No issues whatsoever. Door seal is all good. And I have never had an issue with food storage. Ice cream does not get solid as typical with these absorption refers but makes ice cubes, no issue.



What surprised me was the deltas in temp throughout the refer. Recall I have 3 blue tooth thermometers (BTT): top, middle and bottom. Initially I compared each BTT to each other and ‘adjusted’ my numbers accordingly, but I was still seeing deltas in operation.



Turns out there is a calibration procedure for these devices, so I did that. And now, at the 32-degrees, are within 0.1 of each other.

Post calibration they are back in the refer these past couple of days and again, I am seeing as much as 8 degrees between the top and the middle (the bottom, in the vegetable compartment also similar to mid only slightly higher). My concern was that the top and middle are really only inches apart, so what is going on?



My first observation was that I had the top one ~6 inches away from the back of the refer. The middle one was next to the back wall. Hmmm…. Perhaps getting some heat from the hot tubes of the refer just on the other side. Move it away and sure enough, temp drops by a few degrees. Wow.



Thus, my first conclusion is that the insulation of these refers is not that great. But it does get hot on the other side where the gas needs to dissipate its heat in the tubes, causing me to again realize that should get more bang for the buck by installing the external fans to help dissipate that heat.



I did swap the top and middle BTTs to check and had same results: about a 5-degree temp difference with about a foot of vertical separation. See pic of placement.



Attachment 277159



Thus, my second conclusion is I do not have enough fan speed to cause the needed circulation to create a uniform inside temp. Recall these mini fans are running very slow (a 24VDC fan running at 5VDC). So, I will amp it up. Initially I will go to full 12V, which per my pervious calculations, will almost double their CFM. And I just ordered some thin metal plates with sticky on the back. I will use these to mount the BTT on the sides of the refer (they have a magnetic mount). Using these will keep them the same distance from the rear and not occupy shelf space.



One thing I did notice: no ice build-up on the refer fins. Previously I did have this but as noted above by NXR, there is the benefit of no icing by using properly placed internal fans.



More to follow with the 12V speed.

I have an Everchill and I had a similar problem. After some trouble shooting I discovered that my fan’s movement was being restricted due to the way it was installed. After taking the panel out of the freezer, that held the fan, then taking that apart to look at the fan, thats when I discovered that the fan was being overly compressed and could not spin freely when activated. Not sure if yours is the same set up, but I read tons of threads trying to figure out the problem and never saw anything mentioning the fan’s movement being restricted. I hope this helps you, good luck!

[ESGWheel]

Triple C: Thanks for the tip > that would be a frustrating issue. However, before I set up the relay to turn off the fans when the door is open, I did check that the fans were indeed running as installed.

I am now believing that these refers have significant stratified temps without sufficient forced circulation. In other words, the natural convection of the internal air is very limited. Add to that the heating that happens to the back wall from the refrigerant coils just behind it and I think it’s all over the place.

To test this, I have bought another set of 4 of these BTT devices and will set them throughout the refer and measure both with fans on (at a higher speed) and off. I’ll calibrate them this week and get them installed next weekend. Testing may take some time as I am only there for a brief period.

[MacGumbo]

Originally had trouble with our Dometic keep cool and was going to go the fan route. I thought I would pull it out and see what I could do other than the fans. Basically I added insulator board and created a better channel for air flow. After that it worked perfectly and I never put the fans in. Air flow across the coils is key and sometimes you don't need the fans if it is designed right.


Post on my work.
https://www.forestriverforums.com/fo...ml#post1590153

[ESGWheel]

This has been a long haul to gather evidence that my original hypothesis was incorrect. The theory was that a constant but gentle laminar type airflow would even out the temps. Part of this thinking was based on the refer door being shut most of the time. In other words, in a few hours or overnight the gentle breeze would create a homogenous temp. But with the 7 remote temp devices throughout the refer, the data showed not the case as noted in my last post. So, amp up the speeds. First to 12VDC (which put these 24V fans about ½ speed) and then to their full 24V. All the while recording and plotting the data at 15-minute intervals over weeks and then studying the patterns.

The data at 12V for a couple of weeks showed better results but not as homogenous as I would have expected. However, it was enough of a difference to be able to lower the overall refer temp setting. In other words, the refer was not working as hard to maintain the ~ 40-degree temp throughout. This was a pleasant surprise. Finally, using a 12V to 24V buck device, I went right to 24V full fan speed. And the temps evened out even more. Again, I was able to lower the overall temp setting and still maintain the ~ 40-degree temp throughout.

Upon reflection I realize that in my original hypothesis I considered the refer to be uniformly insulated > that no one area would gain or loose ‘cool’ more than another. But I did not take my thinking to the next level. While the refer may be uniformly insulated the fact that there are drastically different external temps translates into internal areas that are impacted more then others. And I did not consider that. Think about it: absorption refers use heat to work. And those hot coils, while external to the refer, are right next to some areas and not others which impact the localized internal temps.

I also did not consider blockage. Just under the cooling fins is the drip tray. And my fans blow down thru the fins. Thus, my gentle breeze is immediately blocked.

All this translated into the need for turbulence and lots of it. And now I have it but with the benefit of not losing any storage space. Also recall these turn off when the refer light comes on - and they completely spool down even before I have the door fully open - so I am not spilling out the cool any more than usual.

Final thoughts: in terms of refer efficiency (running less for same cooling) there is more bang for the buck by installing external fans as noted by the many posts about this. That said I am glad I took the internal fan approach first as I was surprised at the internal temp variations. Now when I do install the external fans, I can take comfort in knowing I am not going to freeze the eggs while spoiling the milk.

[5picker]

Interest data and I sure thought the results would be different and said so in post #8.

Thanks for sticking with this and sharing here.
Nice work!

[Ballistic]

So if I’m understanding your data, a smaller fan with a higher static pressure would perform better than a larger fan with a lower static pressure, or as you put it, ‘…a gentle breeze’.

For example, many might be able to relate this, but look at server racks. They don’t use fans with low static pressure and they sure don’t have laminar flow.

Also, laminar flow and obstructions don’t coexist.

[ESGWheel]

Ballistic,

My first objective was to not use up valuable refer space with an internal fan. A lot of the aftermarket ones are full size computer fans mounted on the horizontal face of the cooling fins essentially consuming that small top shelf. And being familiar with computer fans with their high output I figured it was excessive. Instead, I imagined that a series of tiny fans gently blowing down thru the fins would do the trick. In my mind’s eye I envisioned the ‘gentle breeze’ being created akin to a fog bank slowly rolling down the mountains. Not exactly laminar but not a storm either.

Turns out that the obstructions, which include the items in the refer, coupled with the surprisingly large temperature variations inside (due primarily to outside influence) led me to conclude a storm was needed.

So no, I am not trying to suggest a gentle breeze or use this fan over that. High turbulence is now my (new) conclusion of what is needed. The air inside needs to be constantly stirred and vigorously so.

While I have not done any study, I would assume the aftermarket devices with their high output do create that needed storm.

It’s been suggested in an above post that high velocity fans would create an overpressure in the refer which is not the case and not a concern. However, I would suggest that if using fans that create significant turbulent air flow its better to have them shut off when the door is opened else you could push out the cool air.

[clhjsh]

As a retired engineer that designed hvac and refrigeration systems I can tell you that laminar air flow is not conducive to heat transfer. If you look at cooling coils or even an engine radiator they are designed to produce turbulence which increases heat transfer because air does not want to touch surfaces when it’s moving. Also laminar air flow does not provide for even temperature within a space you need the air to throughly mix.

[ESGWheel]

Clhjsh,

Your point is exactly what the extensive data collection showed: need to mix it up! I did not start with that thinking but that is excitement of gathering empirical data. While not in a certified laboratory, I suspect there has not been a more extensive study of temps throughout one of these absorption refers under different air circulation profiles. So those little battery powered fans that I have used for years: I now think the agitation is not sufficient, so don’t bother. But those high-powered aftermarket fans attachedon to the cooling fins: go for it!

[5picker]

Originally Posted by ESGWheel

Clhjsh,

Your point is exactly what the extensive data collection showed: need to mix it up! I did not start with that thinking but that is excitement of gathering empirical data. While not in a certified laboratory, I suspect there has not been a more extensive study of temps throughout one of these absorption refers under different air circulation profiles. So those little battery powered fans that I have used for years: I now think the agitation is not sufficient, so don’t bother. But those high-powered aftermarket fans attachedon to the cooling fins: go for it!

I said this a few posts up... I'm surprised by the results.

But in response to your statement... that little battery operated fan that moves very little air and wouldn't make what I call agitation inside the box, DOES IN FACT make a difference in my 4 door refrigerator.

I have repeatedly tried it with and without over 4 years and with it running, I consistently maintain 37º temps in the box. I even bought a set of the fin attached fans but never installed them because the little guy does the job just fine. (albeit needing 2 D batteries about every 6 months)

I guess that's the reason for my surprise in the OP's findings.

[ESGWheel]

Picker,

Apologies, I somehow missed your earlier post and thanks for the encouragement! I am sure everyone’s mileage varies but I respectfully wonder: how you know you have a consistent 37 throughout?

Over the years I have used a couple of different monitoring methods (along with the D battery fan) and adjusted the refers ‘thermostat’ accordingly. And always towards to the ‘coldest’ to keep my temp sensor, which was in the middle, ~ 40 degrees. But it was always just one device and it never occurred to me to move it around to see if the temperature varies. My initial data, with the 5V fan speed and 3 sensors, showed large variations that was only reinforced with all 7 sensors installed. Initially I was convinced it was ‘bad data’ > different sensors seeing the same temp but reading differently, so I kept switching them around (No. 3 where 4 was and vice versa). Same results.

Perhaps with a lager volume of the 4 door there is greater opportunity for natural circulation to occur. Clearly not the case in my smallish Dometic DM 2652 CuFt refer.

[Riccochet]

Originally Posted by 5picker

I said this a few posts up... I'm surprised by the results.

But in response to your statement... that little battery operated fan that moves very little air and wouldn't make what I call agitation inside the box, DOES IN FACT make a difference in my 4 door refrigerator.

I have repeatedly tried it with and without over 4 years and with it running, I consistently maintain 37º temps in the box. I even bought a set of the fin attached fans but never installed them because the little guy does the job just fine. (albeit needing 2 D batteries about every 6 months)

I guess that's the reason for my surprise in the OP's findings.

I bought the little Camco fridge fan a couple years ago. It barely blows any air, but that's the point. You want a small amount of air current blowing across the fins. It made a drastic improvement in fridge temps and maintaining those temps. Before I'd have to run the fridge on high constantly to maintain 40-44F, now I turn it down to middle setting and it maintains 33-36F consistently, and gets there quickly.