Freezer to Fridge Conversion

The Back Story

As of summer 2010 the 1.5 year old thermostat on our Woods all-fridge fridge started failing and replacement parts were no longer available. So I revisited my old idea of building a thermostat - and ran across the Mt. Best site about a chest freezer to fridge conversion (linked from WikiPedia). An upright freezer was available for free - and so the dream of converting a freezer to a fridge became a project.

Timeline - Links

Here are links to the important sections below:

Quick Summary

1. The 12 cu-ft Woods chest freezer is drawing 305Wh/day as a fridge (!!) and 1.1kWh/day as a freezer. Both the converted upright freezer and chest freezer use about 1/3 of the energy of our 1993 17 cu-ft Woods all-fridge. Our all-fridge is 856Wh/day while the converted upright freezer, of the same volume, is 341Wh/day a 60% reduction.
2. Chest freezers are more efficient for space utilization. Trying to cram everything from our 12 cu-ft chest freezer into a 17 cu-ft upright was impossible!
3. An all-fridge fridge, or non-defrost freezer is the cheapest to buy, most energy efficient and simplest to repair. A combined fridge / freezer is convient but typically the worst of both worlds.
4. Opening a fridge door is roughly 5% (20C ambient) to 30% (30C) of the total energy draw of the fridge (for my family).
5. A chest freezer converted to a fridge (esp. the super-insulated models) would make the most efficient fridge - but many balk at possible issues accessing the contents. Freezers are available in many sizes and generally power draw is proportional to the interior volume. However, for fridges it's not unusual to see a bar fridge with the same energy rating as a full sized fridge - so it's often not possible to save energy by buying a smaller fridge.

6. Freezers, converted to a fridge, are at least 2.5x more efficient than all-fridge fridges - based upon the lower temperature difference. Ie to estimate the energy use of a freezer converted to a fridge take the Energy Guide kWh/year rating and divide by 2.5
7. Condensation issues bite - freezers for which the shelves double as the refigeration coils (cheap efficient ones) will drip onto the fridge contents. Freezers have a drain and this must be opened to deal with the condensation. We use a drip pan under the converted freezer and empty it weekly.
8. If one is converting a chest freezer then one must use an electronic thermostat.

Energy Efficiency - Conversion Goals

Fridge Opening Effect

Much has been said about how much energy is lost by opening a fridge door. Air has low mass and so the energy loss is low. At issue, I believe, is that cooling works by cooling the coils, which then cool the air, which then cools the fridge/freezer contents. As modern fridges and freezers have a 50% duty cycle (running 1/2 of the time) opening the door at this critial time maximizes energy loss because the cooled air has not cooled the fridge contents. In the case of our upright freezer - there is a signif. mass in the shelves / coils and the compressor only runs for <10 minutes (not 20 to 40 min/hour) as our fridge does.

An attempt to quantify the energy lost by opening an upright fridge was done by comparing night time energy use to daytime energy use. Both intervals were 12 hours. There are caveats - ambient temperatures are higher in the daytime and during the daytime warm items are put into the fridge. We minimized putting warm items into the fridge during the 5 days of data collection. Ambient temperature is very important. Ambient can be 20C to 29C in the summer and during humid weather condensation will leach energy out of the fridge. Fridge temperature is about 5C - so comparing a fridge in the garage (ambient 16C - temperature difference 11C) to a fridge in the house (ambient 25C - temperature difference 20C) without correction for the temperature difference will introduce huge errors.

The ballpark estimate for energy lost to opening the fridge door is 1/3 of the total energy use - for an EnerStar 1993 all-fridge fridge in a household where the door is not opened needlessly or held open for long periods of time.

Conversion Issues

1. no light in converted "fridge" if it's not on - unless one does a bit of rewiring (not too hard) or if one simply replaces the freezer thermostat with a fridge (such as Ranco VC1) thermostat.
2. condensation issues - dripping shelves where shelves are the cooling coils. This is noteable even in October when we've left behind the humidity of summer. The top two coils represent the majority of the condensation in our upright freezer.
3. Thumping. With most upright freezers the shelves are the cooling coils and when the cooling cycle is running those shelves go from 5C to -20C and shift a bit. This makes a thumping noise with ours. The compressor on the 6 year old upright freezer is much quieter than the 19 year old fridge compressor though.

Energy Savings Claims

Claims of 1/15 the energy draw (Mt Best) I can not substantiate in any way!! My measurements indicate a 70% (chest fridge) to 60% (converted upright freezer) reduction in energy use, for a similar volume, can be easily achieved. In the Mt Best freezer conversion (PDF document) energy use, including power-up, during 24 hours was 103Wh with ambient temp between 15C and 21C, fridge temp 4C to 7C, compressor running 90 seconds per hour.

Note: They use a superinsulated Vestfrost SE255 freezer (237 kWh/yr, 239L, ~8 cu-ft.) Our appliances include an upright 2004 Woods freezer (17 cu-ft, 420 kWh/yr), Woods 2001 chest freezer (12 cu-ft, 384 kWh/yr) and our Woods 1993 fridge (17 cu-ft, 430 kWh/yr).

They claim 103 Wh/day running 90 sec or 2 min/hour. That is 36.5 kWh/yr or 0.15 of the rated energy use (as a freezer). However, one would expect 1/3 of the energy use based upon the temperature difference (delta-T 15C vs 40C); so I'm unsure how how a magical doubling of the efficiency happened. The reported runtime of 90 seconds seems low (empty fridge?). Initial current draw is high until the compressor builds up pressure and starts cooling - so very short run times are inefficient. My run times are around 5 minutes with a 4C temperature swing on the 17 cu-ft Woods upright and 1.5 bushles of apples inside. Given the Vestfrost is 1/2 the energy use that would imply 2.5 min. runtimes.
Note: runtimes would be determined by the mass inside the fridge. But more mass means longer run times, spaced out further in time.

Whirlpool EV181NZTQ auto df, 635 kWh/yr, 17.7 cf
Fridge twin EL88TRRWQ 17.7 cf thermo W10424991

Fan only 3.9W with good power meter or 6W with EM100
startup 230W
165W at 3 min 26.9C
185W 10 min 20.4C
180W 36 min 5.7C  0.1 kWh or 167W average
setup to turn on at 4.3C on at 5.9C -

temp controller 2.2W DC wall wart

• BEST site: EnergyConservationInfo.org They measure 0.3 to 0.5 kWh/day use, mention condensation issues
• GreenChange.com
• ASelfSufficientLife.com
• Mt Best Chest Fridge

Hints and Traps

Positioning the Gas Thermostat Bulb

Caveat! The position of the thermostat bulb is critial in a freezer converted to a fridge!! In the conversions I've done the thermostat bulb MUST be touching the coldest coils. As those coils will drop to <-10C the thermal mass determines the run-time of the compressor (how long it stays on). This MUST be less than 5 minutes. The coldest shelf in a freezer will generate most of the condensation (it requires a drip tray or only hold items under it that can tolerate being dripped on) and it drops to <-20C. Food sitting on that tray will be frozen. We use strips from plastic election signs to space food above the shelf.
The temperature control will be finicky. On ours if it's all the way to the warm end the fridge will maintain 6C and cycle on for 5 minutes every 30 minutes to an hour.
Note that very short runtimes (compressor only on for 1 minute or less) are inefficient and long run times (>5 minutes) will start freezing contents in the fridge.

Condensation

1. Open the drain on the freezer, attach a pipe (if one isn't there) and make sure that it has a U in it so that an airlock is created - preventing cold air from slowly leaking out. We used a drip tray which has to be emptied weekly. One could also plumb the drain straight into a sink drain.
2. Drip Trays: I used some plastic election signs to create drip tray (direct the condensation to the back of the fridge so it runs down the back wall) under the top two sets of coils where 99% of the condensation happens.

Thermostats and My Electronic Control Circuit

There are several options to controlling the temperature:

Electronic Thermostats

Electronic thermostats can allow precise control of the temperature and an adjustable temperature swing and often working for heating and cooling. Note that positioning a gas thermostate accurately can also result in low temperature swings.

1. NOTE: for a chest freezer conversion it is critical that you use an electronic thermostat instead of a gas one.
2. Home brew electronic fridge thermostat: used a 9V 200mA DC blob, draws 3Wac
3. Mashmaster Fridgemate MK II Kigital 110V $48+$14 s/h
4. Johnson Controls Digital A419ABC $72 PexSupply.com -34 to 100C $64 (watch out for 24V version!!) Amazon.com

My electronic control was based upon the Mt. Best design with modifications to use parts I had kicking around. As I wasn't about to trust my design in the long run I investigating using gas thermostats. The main issue with my thermostat was that while the fridge used an average of 15W of power, my thermostat required 5W extra!! Proper selection of the transformer for the thermostat is critical to reducing the power draw of the thermostat to <2W and possibly <1W.

Gas Thermostats

Note: Oct 2011 Thanks to a reader who spent the money to do a chest freezer conversion I'm 100% convinced that to convert a chest freezer one must use an electronic thermostat. It isn't possible to locate the gas bulb near the coldest coils.

Note: Dec 2010 I've purchased a Ranco VC1 (fridge) thermostat for the upright freezer and it works well if the gas tube is correctly placed. I had to locate it partially on the 2nd coil/shelf from the top and there it runs the compressor 4 to 5 min. max. The Ranco VF3 thermostat is designed for freezers and both were $33 each.

Note: Nov 2010 I moved the (failing) gas thermostat from our fridge to our upright freezer. It almost fit physically and the temp knob did not match but it worked well AFTER I re-located the end of the tube. It was originally the 3rd cooling coils from the top of the freezer and the top 2 get the coldest and have nearly all of the condensation. The thermostat would seeminly never shut off on the 3rd tray (and things were being frozen) so I moved it to the 2nd cooling coil and then it worked well keeping the freezer at 7C (that's about all that thermostat was doing in our fridge).

Gas thermostats have a larger temperature swing and a crude temperature control.
Note that pretty well any gas thermostat will work to convert a freezer to a fridge; but it must be mountable and have the necessary electrical connections.

1. Johnson gas A19AAT-2C 6' capilary tube 2C differential
2. ETC Supply Ranco thermostats with specs
3. AP Wagner Supply thermostats with specs
4. Mt Best thermostat (PDF) I used this for my basic circuit, however I rejigged U1C as an over-temperature alarm as it's not needed in the circuit (as a voltage isolator/follower).

Freezer and Fridge Specs (2010)

It can be hard trying to find energy efficient appliances. Avoid auto-defrost models and upright freezers and always look at the EnerGuide ratings. Below are some ratings for fridges I have and others for comparison.

• SunDanzer DC Fridges and freezers Energy use (PDF) (DCR165 Fridge 77Wh/day at 21C, 168Wh/day at 32C, 348Wh/day at 43C) (DCF165 Freezer 272Wh/day at 21C, 441Wh/day at 32C, 766Wh/day at 43C) DC?165 is 165L or 5.8 cu-ft
• Mt. Best Vestfrost SE255 chest freezer rated at 237 kWh/yr, 239L or ~8 cu-ft
• 2010 Woods 16 cu-ft chest freezer 360 kWh/yr, aprox $600
• Our 1993 Woods all-fridge fridge 17 cu-ft 420 kWh/yr, 64"H x 30D x 27W (room for +3W +4"H)
• Our 2001 Woods chest freezer 12 cu-ft 384 kWh/yr - NOT Energy Star certified! measured 1.1kWh/day fall 1.25 kWh/day summer
• Our 2004 Woods upright freezer 17 cu-ft 420 kWh/yr - Energy Star, 1.3kWh/day (20C ambient, -18C inside)
• 2010 SunDanzer 5.8 cu-ft fridge 77Wh/day - 28 kWh/yr, $1150 US
• 2010 SunDanzer 5.8 cu-ft freezer 272Wh/day - 99 kWh/yr, $1150 US
• 2010 Freezer Danby 4.4 cu-ft 329 kWh/yr
• 2010 Freezer Danby/Woods 5.4 cu-ft 340 kWh/yr
• 2010 Fridge Sub-Zero 700BR 5.4 cu-ft 204 kWh/yr 5.4cf
• 2010 Fridge Sub-Zero UC24RO 5.7 cu-ft 223 kWh/yr
• 2010 Freezer Danby 5.9 cu-ft 190 kWh/yr
• 2010 Fridge Dimchae 6.4 cu-ft 240 kWh/yr
• 2010 Freezer Danby 8.2 cu-ft 262 kWh/yr
• 2010 Freezer Woods 9 cu-ft 251 kWh/yr
• 2010 Freezer Danby 11 cu-ft 306 kWh/yr
• 2010 Freezer various 13 cu-ft 278 kWh/yr
• 2010 Fridge best 13 cu-ft 278 kWh/yr
• 2010 Fridge Danby (best) 17 cu-ft 320 kWh/yr
• 2010 Freezer Woods 16 cu-ft upright 409 kWh/yr aprox $700 (582 kWh/yr auto defrost)
• 2010 Freezer Danby 17 cu-ft 358 kWh/yr (some as low as 320 kWh/yr)
• 2010 Freezer Woods 18 cu-ft upright 403 kWh/yr

What the @#$#@$!

Why does a converted freezer use 40% of the power draw of a fridge?

1. Compressor and coil efficiency!? Freezer compressors are more powerful and have a larger cooling coil area. The heat rejection coils seem to be the same area (embedded in 2 walls). I am certain that the efficiency of a fridge is lower because of the compressor cycle limiting the cooling coil temperature to around freezing while a freezer will happily go to -22C. I have no clue as to why this would give an increase in efficiency.
2. Temperature difference? Freezers keep their contents at -20C while a fridge is about 6C. With an ambient temp of 21C that gives delt-T of 15C for a fridge and 41C for a freezer. 15 / 41 = 37% or 1 / 2.7 So one would expect a converted freezer to use 37% of it's power rating. This has been measured to be the case.
3. Cold air falling out? A chest fridge/freezer is more efficient than upright due to not having to chill warm air after the door is opened. This was measured to be about 7% at 20C and 30% at 30C.
4. Better and more insulation? Freezers have a 30% greater wall thickness than fridges. I can't imagine that they have better insulation so because of the insulation a freezer to fridge conversion should reduce power by 1 / 1.3.
5. Compressor runtime? The compressor for a freezer runs for shorter time intervals and this should decrease the efficiency. Our 2003 Woods fridge compressor draws 180W shortly after starting, uses about 130W one minute later and settles around 106W after 2 minutes.
6. Penetrations A fridge has a condensation drain which is a small uninsulated hole and slowly leaks cold air. This will reduce summer (high humidity) efficiency.

Energy Use Data - Upright Freezer Becomes a Fridge (Sept 2010)

Here is data from the '04 Woods upright 17 cu-ft freezer converted to a fridge and the '93 Woods 17 cu-ft all Fridge.

Energy Use Data - Chest Freezer Becomes a Fridge (Oct. 2010)

The freezer is rated 1.1kWh/day so I anticipate 1.1/2.5 = 440Wh/day but it used 305Wh/day over 4 days at 19C ambient!!

Note that the 1" foam sheet on the freezer door was removed for this test. When the chest freezer is a freezer the temperature drops like this: 19.5C ambient, 11.5C under 1" R5 foam, -19.2C inside freezer (-17.6 based on thermometer). That means that the freezer door is: (11.5 + 19.2) * 5 / (19.5 - 11.5) = R19

Wood apparently made a convertable fridge-freezer back in the 1980's (?). What
likely killed it was the sensitivity of placing the thermostat.  When converting
a freezer the position of the gas thermostat bulb is very critical and one then
has very little control over the actual temperature.

old Woods thermostat Ranco K59 P489190-22-5P24

Amre Cambridge 519-650-2673 - Can match fridge thermostat with another
- stock of Johnson: A419 $148  A19 $130

Bills Appliance Repair <$100 to replace thermostat 578-9652 351 
Lancaster W (can't find it!)

Reliable Parts 935 Frederick 519-570-0559 - No woods parts

Waterloo Appliance  519-884-2507

Home Brew Thermostat
Tyco relay R10-E1y2 V185 2A coil 185 ohm, 12V aprox 70mA
Need relay G6RN-1A DC12V 18mA, 650 ohm rated 8A 250V
MOV across relay load?  diode on coil side?
Solid state relay such as IDEC RSSDN  10A has 20mA max. off-state leakage according to spec sheet - In reality the leakage is ??? They are $24 vs $4 for Omron G6RN relay

Date            Time    kWh     Wh/h            Night   Daytime Ambient
                                                Wh/h      Wh/h  Temp (C)
Aug 28, 2010    19pm      2.67
Aug 29, 2010    7am       3.11    37      night   0.44
Aug 29, 2010    19pm      3.93    68      day             0.82    28      set thermostat slightly cooler
Aug 30, 2010    6am       4.36    39      night   0.43            30
Aug 30, 2010    21pm      5.25    73      day             0.89    30
Aug 31, 2010    6am       5.63    42      night   0.38            30
Aug 31, 2010    18pm      6.45    68      day             0.82    30      door left open after lunch
Aug 31, 2010    22pm      6.9     113     day     0.45    1.27    30
Sep  1, 2010    6am       7.17    34      night   0.27            25
Sep  1, 2010    19pm      8.1     120                     0.93    31
Sep  2, 2010    6am       8.59    45      night   0.49            25
Sep  2, 2010    18pm      9.18    49                      0.59    25
Sep  3, 2010    6am       9.7     43      night   0.52            25
Sep  3, 2010    20pm      10.41   51                      0.71    25
Sep  4, 2010    6am       10.8    39      night   0.39            22

                                        Average 0.43    0.79    Wh/h per time period
Fridge Energy Star rated 432kWh/yr = 1.2kWh/day, past measurements: 0.77kWh/day in winter, 1.3kWh summer


Fridge vs converted Freezer:
		Freezer				Fridge				Energy Use Per Period				Ambient	Comments			
Date	Time			kWh	Temp(C)	kWh/day	kWh	kWh/day	Freezer	Fridge	Ratio	Period	Temp (C)				
Sep 21, 2010	16:00:00	0															
Sep 21, 2010	19:30:00	0.05	4.4	0.343	0.13	0.891	0.343	0.891	38%		24.0	Freezer not used yet, full of apples			
Sep 22, 2010	06:00:00	0.23	6.9	0.411	0.54	0.937	0.411	0.937	44%	Night	22.4				
Sep 22, 2010	19:00:00	0.42	6.8	0.378	1.09	0.980	0.351	1.015	35%	Day	23.1	added MOV, 111W running, ~600W startup			
Sep 23, 2010	05:50:00	0.57	6.8	0.363	1.42	0.902	0.332	0.731	45%	Night	18.6				
Sep 23, 2010	18:30:00	0.74	4.6	0.352	1.87	0.889	0.322	0.853	38%	Day	23				
Sep 24, 2010	06:00:00	0.9	4.8	0.349	2.28	0.882	0.334	0.856	39%	Night	22.1				
Sep 24, 2010	20:30:00	1.14	6.4	0.358	2.82	0.884	0.397	0.894	44%	Day	23.8	Warm day			
Sep 25, 2010	06:30:00	1.29	7.2	0.359	3.17	0.879	0.360	0.840	43%	Night		3.458	Total days test running		
Sep 25, 2010	12:45:00	1.4	4.7	0.363	3.43	0.887	0.422	0.998	42%		21.9	Fridge power meter hung . no data missed??			
Sep 25, 2010	21:00:00	1.51	6.8	0.359	3.79	0.901	0.364	1.047	35%		21.4	4.063		Fridge little used	
Sep 26, 2010	06:45:00	1.64	5.0	0.356	0.31	0.888	0.320	0.763	42%	Night	20.4				
Sep 26, 2010	19:00:00	1.79	4.7	0.349	0.7	0.876	0.294	0.764	38%		19.6	Freezer just ran			
Sep 27, 2010	05:30:00	1.92	5.3	0.345	1.01	0.862	0.297	0.709	42%	Night	19.5				

Sep 30, 2010 The converted freezer has now become our fridge.


Note: the day and night use at about 21C (reported as kWh/day equivalent) is:
Freezer: 346 and 342 (day and night use the same)
Fridge:  915 and 806 (day use 6.5% higher)

Temperature Swing over two days

Dec 2013

Dec 18, 2013 I now do weekly defrosting - but it's clear from this data that it should be done more often.
For this data the temperature sensor is on the back metal plate by the cooling coils and it's set to turn on at 1C and off at -3.5C which gives a run time of around 8 minutes (less as it ices up). Data samples were captured every 5 minutes. I had to write a Perl script to convert the Thermochron data from AM/PM to 24-hour clock for a GNUPlot script. The raw data is here: bottom (raw csv) and top (raw csv) which were processed to give bottom (text) and top (text). The resulting plot is Temperature Swing over a week.
At the point when the red line peaks I did a defrost cycle (fan on, no compressor) for around 5 hours. This plot really makes it clear why a high resolution Thermochron is very important. The +/- 0.5C resolution of the cheap model gives very coarse data.

Note: Sept 16, 2013. We've had a case where the coil in the fridge froze up and ice started to form on the metal plate covering it. To deal with this I moved the STC-1000 temperature sensor from the side wall to the top right of the metal plate at the back. Some research indicates that the twin, refigerator, model has a defroster. Trying to get the lower part of the fridge < 6C requires a defroster! In order to defrost ours I've broken out the power for the fan so that I can turn the fan on. Then by turning off the fridge and leaving the fan on for 6 hours the fridge is defrosted.

Note: Sept 12, 2013. Temperature measurements with Thermochrons have revealed that while the lower part of the fridge maintains 6C - the middle and up is around 10C to 12C. Also when the compressor turns off it can get quite cold at the bottom as the cold from the coil sinks. A solution would be to make sure that the coil fan stays on for an extra minute after the compressor has run. This was not a problem when the coils were at the top of the fridge. But trying to maintain an constant temperature thruout a volume which is 2x larger than most fridges is a challenge. We now keep foods that require cold lowest.

Energy use, with the STC-1000 thermostat, set at 5C +/- 3C, without the door being opened at 24C ambient from Aug 18 to Aug 23 over 5 1/3 days was 3.14 kWh or 0.589 kWh/day or 215 kWh/yr at 24C or an estimated 180 kWh/yr at 21C

Energy use, with my own electronic thermostat, and nobody opening the door at 22C ambient, from 9:15am Wed Aug 7, 2013 to 2pm Mon Aug 12, was 2.15 kWh, 13:40:26 time running over 5 days 5 hours which gives: 413 Wh/day or 150 kWh/yr turning on and running for aprox 18 min. at a time.

Energy use from Sun July 21 to Monday July 28th was 0.534 kWh/day or 195 kWh/yr at aprox. 22C ambient or 185 kWh/yr at 21C ambient.

My [analog] electronic temperature controller uses 2.2W and runs at 12V using a DC wall blob.
The freezer defrost timer was 2W according to the EM100 power meter and 0.5W according to the more accurate Belkin.

Here is the energy use data for our Whirlpool freezer converted to a fridge:

Power use is measured with an EM100 Power Meter.

Temperature Vs Location

How does the fridge temperature change in different parts of the fridge? This is for the STC-1000 thermostat set to 5C +/- 3C so it turns on a 8C and runs until 5C.

The first 2 columns were done at one point in time and the 2nd two (middle and top temperature) were done later.

Summary: The middle temperature swings from 5 to 8C while the bottom swings from 2.8 to 5.0C and the top swings from 4.2 to 8.1C. So the top is around 3C warmer than the top.

W10474749 FOSHAN FSTB 62" capillary tube - stock freezer thermostat for the Whirlpool

FOSHAN KDF8A 4-30733-012 FSTB  31" capillary - fridge thermostat, way too short to get near the cold coils

VC1 K50-P1110  09-51-27S  Ranco 47" 120cm capillary VaRifix (specs turns on at -2.5C, off at -8C to on -16C off at -25C) - fridge thermostat used on previous upright freezer and is about 15" too short

For the simple op-amp temperature controller the feedback resistors have a voltage swing
which equates to a temperature swing.  This needs to be adjusted based upon the freezer.
Initially the resistor values were the ones used for a chest freezer and manual defrost upright
freezer conversion - but they resulted in runtimes that were too long and temperature swings
that were too large.

Pin 2: 6.47V 3.0C,   3.60V 4.4C   TRIP 3.65V  for 200k over 220k feedback resistors

Pin 2 is the amplified, and offset, thermister reading.

Pin 3 is the feed loop with an offset to prevent quick on-off cycling

Fridge  Pin        Fridge on?   Feedback     Vcc
Temp    2      3      off?      Resistors
6.0C    4.04V  5.90V  on        200k / 470k   10.43V, 1.5V swing, > 40 min runtime
5.2     4.62   -     running
3.5     6.28   4.38   off
3.0     7.12   4.77   off       200k / 1M   850mV swing, 30 min runtime
4.3     6.02   4.77   off
5.9     4.52   5.61    on
4.1     6.03   5.07   off       200k / 2M7  320mV swing, 15 min runtime
~6.1    4.79   5.40    on

Fridge Initial turn-on
startup 230W, running time, power draw and temperature 

165W   3 min   26.9C
185W  10 min   20.4C
180W  36 min   5.7C   0.1kWh energy used

Trying to compare the STC-1000 temperature controller with my home-brew one gives the following data. The STC-1000 thermistor is in the middle back of the fridge, just above the metal plate housing the coil and fan. This area has a larger temperature swing and putting the thermister inside the coil housing resulted in large temperature swings as that area gets very cold while running. My temperature controller thermostat is on the bottom, back side wall.

5:05pm turn on at 8.7C and runs till 3.6C
5:45pm 7.0C  (drop of 3.4C in 40 min or 5.1c /hour
7:01pm 2.0C (compressor just turned off)
7:34pm  6.4C
7:50pm  7.1C
8:05pm  7.7C
8:07pm  compressor turned on
8:08pm  3.9C
8:42 pm 6.1C

This data reveals that the temperature probe for the STC-1000 sees a relatively quick temperature rise after the compressor turns off
and then a constant rise in temperature over time.

4:48am 7.8  (open door and it rises to 8.5C) compressor turns on
4:50am 8.2C
4:52am 8.0C
4:57am 6.4C
5:01am 5.0C
5:01am 4.8C compressor turns off
5:06am 5.3C
5:10am 5.6C
5:15am 5.9C
5:24am 6.5C
5:34am 7.0C