Jump to content
Click here to check out our structured, online appliance repair training courses for rookies and experienced techs.

FAQs | Repair Videos | Academy | Newsletter | Contact

Stay connected with us...

Subscribe to our YouTube channel for lots of appliance repair tips and help! Subscribe to our MST Radio podcast to learn secrets of the trade. Sign up for our free newsletter and keep up with all things Appliantology.

Samurai Appliance Repair Man's Blog

  • entries
  • comments
  • views

3 ways to half-split warm refrigerator problems

Samurai Appliance Repair Man


Whenever you're dealing with a warm refrigerator problem, your first question to answer is, "Am I dealing with a failed sealed system or a control problem?" In other words, you need to half-split the problem between the sealed system and the controls (air movement, defrost system, temperature control, compressor start device, etc.). I'm going to talk about three ways to half-split warm refrigerator problems to either rule in or rule out the sealed system and the controls. 

1. Frost pattern

This is the venerable old skool method where you tear down the freezer to expose the evaporator so you can see the frost pattern on the evaporator coil. A properly frosted evaporator coil will have a light coating of frost on most of the coils except the last rung or two will be frost free. The frosting parts of the coil is where your refrigeration effect is taking place: the refrigerant is boiling and changing phase from liquid to vapor. By the time it reaches the last two rungs, all the refrigerant is boiled off to vapor and the vapor starts picking up superheat but there is no longer refrigeration taking place. 

The frost pattern also gives you information about other problems like incomplete defrosting, lack of air flow, air channelling, etc. All of these are control problems, not sealed system problems. For some problems, this is diagnostically useful information to know. But in the context of half-splitting between sealed system and controls, there are other, more clever ways to do this. 

This is a favorite method of old skool techs and techs who do not understand how sealed systems actually move heat around. That's why you'll see many techs reach for this method first. But there are more clever ways to half-split the problem that do not rely on tearing down the freezer or installing valves. 

2. Compressor amps

In this method, you measure the running amps of the compressor. It can be very effective IF you know what the normal running amps should be for that compressor. If you're going to make an electrical measurement, you need a spec to compare it with to tell you what it means. You can't use the Locked Rotor (LR) amps on the compressor label because this is the starting amps which is much higher than the normal running amps. Sometimes the tech sheet will give a wattage spec for the compressor and you can convert this to an expected amp reading.

If amps are below low normal (however you determine the reference for "normal"), then you know the compressor is not doing as much work (in watts) as it should (watts=amps X volts). The reasons for this are usually either 1) weak pump (eg., failed suction valve a la the LG linear compressors) not moving as much vapor around the system or 2) refrigerant leak (meaning there's less vapor mass to move around the system). Either way, this is usually a pretty good indicator of a sealed system problem IF there's normal heat transfer occurring at the evaporator-- in other words, no evidence of the evaporator clogged with rime ice and the evaporator fan is running normally). As mentioned, the problem for the technician with this method is often knowing what the normal running amps should be. Running amps will also vary by a few tenths with the heat load on the evaporator coil.

This method gets muddier when you're dealing with variable speed drive systems. You need the spec for compressor amps at full speed for this method to be useful with these systems. 

3. Condenser temperature split

This is by far the easiest and fastest method requiring zero tear down or fishing for compressor wires. It is also the least used by technicians because they are suspicious of it. The reason they don't trust it is because they don't understand the thermodynamics of sealed systems. Here's a dirty secret: many techs who do sealed system work have a low level of understanding of how sealed systems actually work; in other words, sealed system thermodynamics. It's too bad because this is definitely an area where you can work smarter not harder. 

Condensers, like evaporators, are designed by the engineers to operate at a certain temperature split. The temperature split is defined as the actual coil temperature minus the ambient temperature. In condensers, the maximum condenser temperature split is 30F. This means that with the system fully loaded with heat (as in a pull down situation) and normal heat transfer at the evaporator coil, you will measure a condenser temperature around 30F above the room temperature. But this is a design maximum, which you may or may not actually measure. In normal operation where compartment operating temperatures are being maintained, this split is typically less, maybe around 20F or less. But the point is that in a normally functioning sealed system, the condenser will always be warmer than the ambient. You can measure this with your IR gun in some condenser configurations or by using a thermocouple probe with putty to stick on the midpoint of the clean condenser coil. Or, if you have a calibrated palm like I do, you can simply feel it with your hand.

The condenser split screening test works with any and all refrigeration systems (within the conditions described above) be they vapor-compression or absorption (like RV refrigerators) because heat (thermodynamics and heat transfer) works the same way. Thermodynamics describes how heat moves within the sealed system; heat transfer describes how heat moves into and out of the sealed system.

It's like with electricity and circuits-- electrons work the same way despite the technological innovations we've seen over the past 20 years (computer control, triac-gated Neutrals, etc.). Electrons require a voltage difference between two points in a complete circuit to power loads. Does not matter if the load is a heating element or a compressor. A load is a load is a load: they convert electrical power (watts) to mechanical or thermal power. The power supply to that load is controlled by switches. Does not matter whether the switch is a mechanical cherry switch, a relay, pressure switch, or triac. A switch is a switch is a switch: they control the power supply to loads. So it does not matter what newfangled technology they come out with, all circuits work according the well-established Ohm's Law, Kirchoff's Law, etc. 

Similarly with refrigeration systems. If you're going to remove heat from one compartment and reject it somewhere else, you have to have a difference in temperature to make heat (BTUs) move. In the evaporator, the coil must be colder than the compartment to force the heat to move from the compartment into the coil and boil the refrigerant. (That is your refrigeration effect.) The heat thus picked up by the refrigerant (plus the heat of compression in a vapor-compression cycle) must be rejected to the environment in the condenser to force that refrigerant vapor to release its heat and condense back into a liquid. This means that the condenser must be warmer than the environment for heat to flow from the condenser to the surrounding air. There are design temperature splits that engineers use on both the evaporator and condenser coils. And, as mentioned, these splits do not happen "just cuz, golly shazzam." They happen because it was a design criteria imposed by the engineers to make it happen.

The one wrinkle to be aware is that in a dual compressor system, the design temperature splits on the FF compartment are lower than for the FZ compartment. And since the FZ and FF systems share the same condenser, you cannot use condenser temperature splits in dual compressor systems to half-split warm FF compartments. This is because you cannot distinguish the FZ split from FF split so the results are diagnostically meaningless. OTOH, if each system has its own condenser, then this method still applies, accounting for the lower splits in the FF system. 

Other than that, it does not matter what new compressor type-- split phase vs. inverter-driven; linear vs. reciprocating-- or any other other whiz-bang they come up with for refrigerators, they will always function according the well-established laws of thermodynamics and heat transfer. 

When I use this half-splitting method on a warm refrigerator (ie., freezer temp above zero), I like to make sure that the evaporator coil is loaded with heat. This will increase the temperature split at the condenser. Ensure the defrost system is working if you suspect that's a problem (force defrost, check amps; eyeball check for external rime ice on the evaporator cover; peeking into air louvers, etc.). Then open both doors and force the evap fans to run. If you're dealing with a variable speed compressor system, make sure the compressor is running at full speed. On most computer-controlled refrigerators, this is just a key dance on the controls. Give it a few minutes and then do your condenser split check. What you're really looking for is something above room temperature. If you only get 10F above, that's marginal but you can't use this test alone to make conclusions about the sealed system. You'll need to combine it with one of the other two methods above. If the condenser coil never gets above room temperature, Houston we have a situation. This condition is positive proof that there's a problem with the sealed system. 

Bonus method: Short-cycle the compressor

This method applies only to split phase compressors because it relies on quickly cycling power to the compressor motor. Inverter-driven compressors often have a boot-up process which prevents the ability to short-cycle compressor power. 

This method is simple: while the compressor is running, unplug the fridge (or turn the cold control off) and then plug it back in. If the vapor pump in the compressor has failed (for whatever mechanical reason) or the sealed system has leaked all or almost all of it's refrigerant, then the compressor motor will start back up and running without doing the hum-click cha-cha. 

Why does this work? Because in either case-- weak vapor pump or leaked refrigerant-- vapor is not being compressed. Which also means there is no (or little) pressure differential between the high side and low side that needs to equalize before the compressor motor has sufficient torque to start spinning the shaft. So there's no work requirement from the vapor pump. All the motor has to do is start spinning the shaft. 

Again, this is a half-splitting screening test which means: 1) the presence of the hum-click cha-cha after short-cycling power does not automatically mean the sealed system is "good" however 2) if the compressor is able to start running after short-cycling power, then this is a positive indication that you have a sealed system problem: either a weak pump or a leak. Don't know which at this point without your gauges but you are now focused on the sealed system, not the defrost system, fan, dampers, thermistors, or anything else. 


Keep in mind that these are methods to half-split a warm refrigerator problem between the sealed system and the controls. Half-splitting does not tell you how a system failed, but points you to the correct system for your next troubleshooting steps. That's the point: the half-split between systems. You are not diagnosing a specific problem with the sealed system, just that it has a problem which is yet to be determined and you are now focused on the correct system for further analysis. 

Every test has specific parameters and limitations. You need to understand what those are to use them effectively. 

It's good to have all three (four) in your troubleshooting arsenal and ready to deploy one or more when you're half-splitting a warm refrigerator problem.



  • Like 10


Recommended Comments

Rhubarb Tau


+1 for Condenser Temp for troubleshooting! BTW a thermal camera makes looking at the condenser a lok more akin to inspecting a frost pattern (accounting for emissivity of the sometimes-shiny condenser fins)



...in a dual compressor system, the design temperature splits on the FF compartment are lower than for the FZ compartment. And since the FZ and FF systems share the same condenser, you cannot use condenser temperature splits in dual compressor systems to half-split warm FF compartments.

I've used condenser temp as a diagnostic indicator on Sub Zero units where FF and FZ share a condenser, but each system has its own dedicated turns, where the FF has the top half of the condenser and the Fz has the bottom half (or maybe vice versa?), I typically clip my thermocouple on one of the turns in the middle of the FF section, and seem to get a useful reading. 

Do you think there's too much heat-load sharing through the shared fins for this to be a useful measurement?



Link to comment
  • Team Samurai
Samurai Appliance Repair Man


1 minute ago, Rhubarb Tau said:

Do you think there's too much heat-load sharing through the shared fins for this to be a useful measurement?

Depends on which compartment you're diagnosing. Its probably fine for the FZ with the larger splits. But the lower splits with the FF can make it ambiguous-- how much heat is radiating off the hotter FZ coils and making the FF coils look okay even if they may not be-- especially with the FF turns on the top half. I don't know the answer to this question but with some field experience using your thermal camera, I'll bet you could develop some good rules of thumb. 

Link to comment



A properly frosted evaporator coil will have a light coating of frost on most of the coils except the last rung or two will be frost free.

With R600 systems the frost pattern can often decrease during disassembly on a hot day, leading to misdiagnosis when using frost pattern analysis.

  • Like 1
Link to comment
Vince Neibert


On 6/22/2022 at 3:24 PM, THRAMICH said:

With R600 systems the frost pattern can often decrease during disassembly on a hot day, leading to misdiagnosis when using frost pattern analysis.

I’ve run into that problem with other refrigerants as well. If possible, I will leave the unit running until I can remove the evap cover. You also need to try and determine if unit just finished a defrost—that can really throw you off. 

  • Like 1
Link to comment

Agreed with @THRAMICH- every r600a frost pattern I’ve checked has disappeared a few minutes after removing the cover. 

  • Like 1
Link to comment
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.