Efficient Refrigerator DIY

Van Living Forum

Help Support Van Living Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Trebor English

Well-known member
Joined
Jul 20, 2016
Messages
1,401
Reaction score
16
Location
Melbourne, FL
My efficient refrigerator uses very little electricity.  During the period of 10/1/17 through 11/2/17 the average consumption was 10.2 amp hours per day.  The highest 7 day period average was 12.13 amp hours per day with temperatures in the low to mid 90s.  The lowest 7 day period average was 6.78 amp hours per day with temperatures in low 70s to mid 80s.  This is actual in a van use, not an always cool place.  

I do three things to make this happen.  It has an electronic thermostat, it has additional insulation, and I use it as a self icing ice chest.  The freezer compartment contains 5 pounds of water in a couple of zip lock bags and I only allow the compressor to run when the sun is shining.  It freezes at a speed of 5 pounds in 3.5 hours and melts at a speed of 2.5 pounds per day depending on temperature.  

Since I use it like an ice chest I don't put hot left over food in it and I don't use it to chill drinks.  I prefer water at room temperature anyway.  

This is a link to a Bob Wells video about my ice chest from the 2016 RTR.  

The ice cost is not why I did this fridge.  The problem with this ice chest is the small food space capacity.  The volume it can hold is about 3 quarts.  That is enough to upgrade eating from the no refrigeration level but I wanted a bit more space.  Now I have solar electricity so the fridge is possible.  

The efficient refrigerator has insulation similar to the ice chest.  I started with a 1.7 cubic foot Haier fridge.  It had 1 inch insulation all around with the hot condenser tubes heating the outside skin.  I added 2 inches (4 half inch layers) of polyiso to the inside surfaces.  Since the back and bottom don't have condenser tubes those parts got insulation on the outside.  The insulation I added inside is about 1 cubic foot leaving 0.7 cubic feet of space.  That's about 20 quarts.  

I power this with a 100 watt panel, a $10 PWM controller, a $10 fuse box, a Harbor Freight 400/800 ($20 with coupon) inverter and a 75 amp hour group size 24 "deep cycle" battery that is also the engine starting battery.  This is what I had been using to charge my laptop and cell phone, and run led lights, MaxxFan, and my electric shaver.  The main reason for the solar power system was to have the MaxxFan.  I modified the inverter by connecting a pair of wires to the back side of the on/off switch so that the thermostat can turn off the inverter when the fridge has frozen all the water.  

It is my opinion based on my experience that regular starter batteries can regularly give 5% and occasionally 10% of the capacity according to the reserve capacity minutes without seriously shortening the battery life.  With this group size 24 battery I think it is good for 10% and occasionally 20% of the 75 amp hour 20 hour rate specification, not 50% like golf cart batteries.  That means I can start with the ice all frozen and have one no solar rained out day using ice and the next no sun day using the battery.  So, 2 days of no sun and then I have to have good sun or start the engine for the third sunless day.  If that happens I'll add another 2.5 pounds of water or maybe switch to salt water.  

The refrigerator was $80 at Walmart.  The 4 x 8 sheet of polyiso was $10 at Home Depot.  I used 2 cans of Loctite spray foam for $12 also from Home Depot.  The electronic 12 volt thermostat was $5 from eBay.  The fridge total is $107.  The solar power system was about $150 plus the $80 battery that I needed to get anyway because the prior starter battery got to the end of the warranty period and died.  

Some details:  When the inverter turns on the surge current is about 55 amps.  When the inverter is up to speed and running it takes about 0.2 amps.  When the compressor starts the surge is about 75 amps.  While the compressor is running the inverter takes about 6.5 amps.  To measure these currents I use a 0.001 ohm store bought shunt resistor and a Fluke model 189 meter.  The meter has a handy max / min / average feature.  You set up what you want to test, start the meter then turn on the load.  Measuring 1 millivolt = 1 amp with the shunt, 75 millivolts max is 75 amps max.  The average feature is what I used to calculate amp hours.  I have an old time electric clock.  When the compressor runs, the clock runs.  So average amps times elapsed hours gives amp hours.  

While the compressor is running the inverter takes about 6.5 amps.  The solar panel produces 3 to 5 amps.  After the fridge has run for an hour and a half and shut off the battery gets charged to put back what was taken while the compressor was running.  The battery ends up getting cycled twice each day.  In the morning the fridge runs.  Over night I use led lights and the MaxxFan.  The next morning I don't start the fridge till the battery is up to 14.5 volts.
 
That is really impressive.
I dislike the consumerism advice to just buy as much as you can afford, that dummies like me follow without thought.
And, even with the added insulation, you have more fridge space than me.
 
UptownSport said:
even with the added insulation, you have more fridge space than me.

A little bit of fridge makes eating much better than shelf stable only.  It doesn't take much.  

Mine is about 20 quarts minus 2.5 quarts of water/ice so 17.5.  What size is yours?
 
so much good stuff in this post!  Do you have any pics of the setup?


During the period of 10/1/17 through 11/2/17 the average consumption was 10.2 amp hours per day.  The highest 7 day period average was 12.13 amp hours per day with temperatures in the low to mid 90s.  The lowest 7 day period average was 6.78 amp hours per day with temperatures in low 70s to mid 80s. 

Impressive.  I think the average daily consumption for 12v compressor fridges is something like 20Ah.


The efficient refrigerator has insulation similar to the ice chest.  I started with a 1.7 cubic foot Haier fridge.  It had 1 inch insulation all around with the hot condenser tubes heating the outside skin.

Is there any space between the fridge and the coils?  I've got some 1/4" XPS that might slip between them and keep coil heat off the fridge surface.  Of course, it might decrease coil cooling.


I power this with a 100 watt panel, a $10 PWM controller, a $10 fuse box, a Harbor Freight 400/800 ($20 with coupon) inverter and a 75 amp hour group size 24 "deep cycle" battery that is also the engine starting battery.

I love that you are doing this and documenting.  I think carfolk could benefit greatly from starter/house combos like that.  How often are you having to water the battery?  How long has it been in double-duty service?


It is my opinion based on my experience that regular starter batteries can regularly give 5% and occasionally 10% of the capacity according to the reserve capacity minutes without seriously shortening the battery life.  With this group size 24 battery I think it is good for 10% and occasionally 20% of the 75 amp hour 20 hour rate specification...


I've been doing some independant reading and have come up with similar numbers.  And for onlookers:  I am currently testing a setup like this using a similar $10 shunt charger with a raggedy old Group 24 in my office



That means I can start with the ice all frozen and have one no solar rained out day using ice and the next no sun day using the battery.

It'd be cheap and easy to test salt water and see if it made a difference.  Go, Trebor, Go!  :)


The battery ends up getting cycled twice each day.  In the morning the fridge runs.  Over night I use led lights and the MaxxFan.  The next morning I don't start the fridge till the battery is up to 14.5 volts.


How deep is each of the two cycles?
 
There is no space between the fridge and the coils. The outer skin on the left, top, and right gets hot when the compressor runs. The condenser is not a grille across the back but rather the steel skin. Inside there is the little freezer compartment.

When I first got the battery I checked the water and specific gravity real often. That dropped off to weekly quickly and then monthly. I still check it monthly and keep a record. There is one cell that uses a bit more water than the rest. Each cell takes 20 to 30 ml per month. That lets me know that it is charging to the level where electrolysis is disassembling water after the charging finishes. The specific gravity confirms that.

The battery is a year and a half old with the double duty use. I have been running the refrigerator for two months.

The two cycles are each not very deep. During the day when the compressor is running it takes 6.5 amps and the solar panel generates 3 to 5 amps. That means 1.5 to 3.5 amps for an hour and a half is used from the battery on a regular day. If there was no solar and no compressor one day then the next day gets up to 3 hours of running. So the daytime cycle is 2 to 10 amp hours. Overnight 5 amp hours is about it if the MaxxFan gets used a lot.
 
A small computer fan moving air across the steel cabinet would help the heat transfer from the metal skin. The use of 50% glycol instead of water as a heat sink would maybe last longer.
 
Do you turn on the inverter manually every morning, and off at night?
 
Weight said:
A small computer fan moving air across the steel cabinet would help the heat transfer from the metal skin. The use of 50% glycol instead of water as a heat sink would maybe last longer.

I have a small computer fan blowing on the compressor.  The end of the compressor near the outlet tube was the hottest part so the fan is positioned to blow there.  In normal use the compressor runs a few minutes at a time and then shuts off.  It never gets really hot.  The way I use it the compressor runs for up to three hours non stop and it gets hot.  I don't think it gets hot enough to have a lubrication failure but the fan was easy to install, uses very little power, and lets me think that it will last a long time running cooler.  

To add a fan to cool the sides and top would require a shroud and fan like a Corvair engine.  A cardboard box to control the air flow would work.  If the fan failed then the box would be bad.  What I do is use some wood pieces that are 3/4 inch thick to keep stuff from being directly on the top.  The sides are open and work ok as designed.  I do keep a piece of cardboard handy.  If the door is open and the sun is shining on the black steel I cover it.

If I decide to switch from plain water I think my first attempt will be Atlantic Ocean water.  It is very conveniently located and salty enough to make a difference.  The price is right too.
 
skyl4rk said:
Do you turn on the inverter manually every morning, and off at night?

Basically yes.  What I want is a low voltage disconnect to turn it off at night when the sun goes down and the surface charge dissipates and the voltage goes down to 13 volts.  Then in the morning I want it to wait until the battery voltage is up to 14.6 volts before turning it on.  Ideally it should be at 14.6 for long enough for the battery to be well charged.  

My solar charge controller has a "load" output with programmable voltages just like what I want.  It doesn't work because the start up surge lowers the battery voltage to about 11 at the minimum point which shuts off the low voltage disconnect.  That turns off the thermostat.  The thermostat relay turns off.  That turns off the inverter eliminating the startup surge.  What I need is a low voltage disconnect with a time delay.  It would be ideal for it to be a 10 second delay for the voltage drop during the start surge and a half hour delay for the turn on so the battery gets well charged.  

So, until I find a useful low voltage disconnect what I do is turn it on when the battery is charged enough.  That depends on how much I used the night before and how much charge it got the day before.  If there is going to be no sun I just leave it off.  When the remains of hurricane Nate came through my fridge used ice from the day before.  I have on occasion run the compressor some but shut it off before all the water was frozen in order to have a good recharge in the afternoon.  Normally when the top of the water turns to ice and the temperature is 30 degrees the thermostat turns off the inverter.  I have a 12 volt power supply plugged into the inverter.  That powers the thermostat and shuts off the thermostat when it has no power.  When the thermostat, inverter, and power supply all go off, it all stays off until I turn on the inverter by hand.  Essentially I turn it on in the morning and the thermostat turns it off when the water is all frozen.  

Before using it with water / ice I tried it just using it like a regular fridge.  I let it cycle normally.  That took 19 amp hours per day with the extra insulation and it took those amp hours even when the sun was not shining.
 
I would think any condensation would indicate not nearly enough insulation.

Therefore poor efficiency
 
ChetLazar said:
Do you have condensation like the converted chest freezers?

Yes.  I keep a piece of terry cloth towel towel in it.  When it gets wet I switch it with another.  Since the freezer compartment doesn't stay frozen it's like defrosting the fridge all the time.  It's like a couple of teaspoons of water every day.  It isn't a problem like melting ice.  

There's no condensation on the outside.  The area where the magnetic gasket seals isn't even cool.
 
If you were to increase the solar panel to about 200W, the fridge would cause very little discharge on the battery. The battery's role would be to provide start up current during the surge.

In this refrigeration system, instead of storing electricity in a battery to run the compressor when the sun goes down, solar energy is stored in chilled ice. Chilled ice is a lot cheaper than a battery.
 
Trebor English said:
A little bit of fridge makes eating much better than shelf stable only.  It doesn't take much.  

Mine is about 20 quarts minus 2.5 quarts of water/ice so 17.5.  What size is yours?

a bit smaller in size, about $300 larger in price, LoL

BUT, I've also got all the silly things to run it- disco golf cart batteries, nifty expenso inverter thingy, flashy, big panels the Po-Po look at every time they pass

Funny, the economy really took off just as I was buying everything for it...

I'm thinking you need the board's McGyver award!
 
I have decided that I will not use Atlantic Ocean water.  

I was cleaning the fridge.  I took the zip lock bags of water out of the freezer compartment and cleaned and dried the inside of the fridge.  When I went to put the bags back I saw that I had torn one.  I managed to clean up some of the water but some got into the carpet.  It was regular fresh water so there's no problem.  If it had been salt water there would be a problem.  So even though Atlantic Ocean water would keep things colder, I'm going to stick with fresh water.
 
Google polar tubes, or use freezer packs, permanent glued containers rather than relying on flimsy ziplocs.
 
John61CT said:
Google polar tubes, or use freezer packs, permanent glued containers rather than relying on flimsy ziplocs.

How do you freeze polar tubes?  Do you just put them in a freezer?  How long does it take for them to freeze?
 
Depends on the diameter of the tubing, 4" takes much longer than 1", and 2" in between.

Also seawater much longer than plain, in fact you can select eutectic solutions with different freezing temperature points to suit your freezer.

A freezer set to 28° may never freeze seawater.

The maximum energy absorption comes from the phase change, so you definitely want the solution to be solid before transferring to the cooler.

But with a high-salt solution and a very good super-cooler, the tubes will continue to work well for 7-10 days even after the solution has melted, still well below 30° internally.

TL;DR "it's complicated"
 

Latest posts

Top