The fridge: AB Coldmachine and the fans.

When I stepped onto the dock after our 26 day day trip to Ottawa, a club member congratulated me on my trip and asked if everything went OK, “sure” I said, “it was great!” “Nothing broke?” he asked, “Nope” I replied. It wasn’t a lie, it was just that I didn’t remember that the fridge compressor was making strangled-guinea-pig-like noises as opposed to the usual Darth Vader sounds, besides, it was still keeping things cold.

IMG_0001A couple of days later the cabin was very quiet, on the trip, with a kids, stereos, motoring and lots of activity on the boat I hadn’t realized that the fridge was ominously silent. I looked in on the compressor under the port dinette seat and realized that there was no air moving, the compressor was very hot. Not having paid too much attention to the compressor in the past I figured it was supposed to be hot – but not hand-burning hot!
Clearly the cooling fans had stopped. The fans are computer fans, nothing special, easy fix I say to myself. I pulled them out and tested them. One was definitely dead, it wouldn’t spin up and when I poked it with my finger it raced then died.  The second made grinding noises and was sluggish, clearly the source of the strangled-guinea-pig noises.

I popped out to a local computer surplus store and got two fans, slightly higher CFM, 91 vs 80 and slightly higher amps, 0.75 vs 0.45A. I figure faster fans would be more efficient and as electricity is not an issue (with my solar panels produce an abundance of juice). In an ideal world I would look for fans with lower amps and higher CFM as I’d found on Digikey, but these were in hand and cheap.

IMG_0004I cleaned the housings, installed the fans and turned the fridge on expecting to hear the Vader like whirr of the air intake. Nothing. I double checked the wiring. Just the compressor running. Sitting in a pool of deflated smugness I realized I should have done some research, but the fans were such an obvious target!

After some research it appears the EMC/brain/control unit (part # ) is very heat sensitive and well known for getting cooked.  It is cooled by the spill over air that flows over the compressor pump after passing over the coils. The unit is surprisingly expensive, same price as replacing the actual compressor, but of course cheaper then replacing the entire unit. Some people  have mounted a third smaller computer fan that is dedicated to cooling the control unit, assuming that this fan would last longer than the other fans…

There is a fridge guru by the name of Richard Kollmann who can test control units. His site is a great repository of tips and trouble shooting.

When I first realized the fans had stopped the compressor was hot enough to cook a hamburger on,  I presume that the unit had overheated. Sure enough a new control unit brought the fridge back to life.  When ordering the part we discovered that because most fridge manufacturers use the same Danfoss components, the part number 101N0210 is the same regardless if your fridge is a Frigoboat, Dometic, Alder Barbour, Nova Kool etc.

Like all systems on a boat, maintenance and constant monitoring are required. The fridge, while non-essential served as a reminder that just because something “works”,  doesn’t mean it should be ignored. My mistake was twofold; I should have at least looked at the fans and cleaned them prior to the trip, I would have discovered their sorry state. Second mistake was dismissing the fan noises because the evaporating unit was still cold. Either course of action would have saved the control unit from overheating.


June 2015 – UPDATE

The fridge runs but never gets very cold. Am looking at the fans as a possible problem.
I might have put the two fans onto the fan circuit, but I don’t think I did, or if I did, there was some reason for it That escapes me now? (see error #1 above).
According to  Richard Kollmann’s trouble shooting tips

“If fan circuit on these variable speed compressors exceeds 1amp compressor start up will be aborted. This condition can be confirmed by disconnecting Black fan wire at module, if compressor runs replace fan.”

If the fan or thermostat don’t solve the problem. It seems I do have a an LED installed, never bothered to look at it because I didn’t realize it was there!

Further update

I  looked at the thermostat innards, all seemed well. I replaced a splitting spade connector, shorted the leads and still it ran as “normal” that being, cool on the inboard side of the evaporator, nothing happening on the port side.

“Place jumper wire across thermostat terminals on electronic module, if the compressor still does not run go to next step.”

If not, then the refrigerant level.

The key to successful troubleshooting is to begin with non destructive testing first. It must be determined if a unit that fails to cool properly is a control problem or a refrigerant system problem. It is very difficult to tell if these small Danfoss compressors are actually running because they are so quiet. Many times the fan noise is believed to be the compressor running. If it is confirmed that the compressor is running, then and only then, should destructive servicing equipment be connected to a capillary tube system.

These small units are very sensitive when it comes to servicing them with refrigerant. If the compressor is a model BD2.5 it must be serviced with R12 refrigerant. If the compressor is one of the following ,BD3, BD35 or BD50, 134a refrigerant is used.

Connecting a servicing gauge set to these units, if not done properly, can be destructive to the system. Here are some simple guide lines to follow when you question the refrigerant level and refrigerant flow conditions:

  1. A system low on refrigerant may be running in a vacuum so do not connect the gauge set when it is running.
  2.  Purge gauge set and hose with refrigerant before connecting it. These small systems only contain 2 to 5 ounces of refrigerant and can easily be contaminated.
  3. The amount of refrigerant pressure in the system when the gauge is connected is only an indication of whether all refrigerant was lost or there is still some left in the system. Pressure in a capillary tube system when it is not running is relative to the ambient temperature of all the components and not refrigeration volume.
  4. If there is no pressure when gauge is connected, pressurize the system and find the leak and repair it. After the leak is repaired Vacuum system to remove air and moisture then service it with the correct refrigerant.
  5. These systems are normally serviced by adding the correct volume of refrigerant by weight, but in the field when a measuring cylinder is not available they can be serviced by pressure in the suction side of the system. This pressure will very based on evaporator size, type of refrigerant and temperature in the evaporator. One ounce plus or minus from the correct volume will greatly affect the power consumption and performance.
  6.  If you have the small Abler Barbour with their small chamber (bin) evaporator the correct suction pressure with a 70˚ F. box and warm evaporator would be 8 to 10 psi for R12 refrigerant and 6 to 8 psi for 134a refrigerant. These readings must be reached in a time window of ten to twenty minutes after start up with a warm box. After twenty minutes suction pressure will drop as the plate temperature drops.


Adding refrigerant did the trick! I brought the level up to 10psi with some R12 from the auto store. I was amazed to see how quickly frost appeared on the evaporator, minutes really. I am not sure how the coolant got low but I will monitor it but at least it works!!! The blue cap, when turned slightly, leaks coolant, but not when fully done up or fully loose?! perhaps it was bumped loose?

Follow up:
The refrigerant was low again after coming back two days later with the fridge off. I left the gage & bottle connected. I filled it again and it worked beautifully.
I re-read some more on-line regarding leaks and several people mentioned checking the seal in the blue cap.  I had assumed the blue cap was just a dust cap, like a car tire, I didn’t think it served any mechanical function in keeping pressure.

Richard Kollman again:

Inspect valve cap you removed to see that rubber seal inside cap is still good. Slow leaks are sometimes traced to a bad cap seal.

Read more:

Here is a good post on emergency refridgerant-line puncture repair.

Leave a Reply