Temperature drops as gas expands, and increases as it is compressed.
In a cooler, the compressor takes the gas and compresses it, making it release heat. This heat is exchanged with the outside atmosphere VIA the "Condenser" coils in back or in the sides. Then it flows to the "Evaporator" coils inside where it expands again, pulling heat into the coolant. And the cycle continues.
The coolant, weather it's R134a, R12, or even ammonia (In big commercial rigs) is in a gaseous state when it passes through the cold side evap coils.
(snippage)
That's not quite right. This is a bit long, but I hope informative.
The compressor takes
refrigerant from the evaporator, which if everything is working right is in the form of a
vapor. The compressor takes it from a low pressure vapor to a high pressure vapor,
adding heat in the process. Then the high temperature, higs pressure vapor flows through the condenser. The additional heat from the compression cycle is given up to the air outside the unit, cooling it enough to condense into a high pressure liquid. The last few coils in the condenser further cool the liquid (this is called subcooling). The liquid refrigerant, still under high pressure, is routed through the
liquid lineto the
refrigerant control, which meters the flow of refrigerant into the evaporator. The refrigerant control can be as simple as a length of capillary tubing (this is the case in most domestic refrigerators), or a compicated as a Thermostatic Expansion Valve (TXV or TEV). It is the refrigerant control that maintains the pressure differential between the liquid line and the evaporator, and this differential is critical. As the refrigerant is metered into the evaporator, it is still in liquid form. As the liquid flows through lower pressure of the evap coils, it absorbs heat, and as the pressure has been lowered, the boiling point of the liquid refrigerant is also lowered, and the refrigerant boils, changing state from a liquid to a vapor. As the refrigerant vapor passes through the last coils of the evap, it gains even more heat (this is called
superheat. The vapor is then drawn back to the compressor via the
suction line.
Depending on the type of refrigerant contol, either the superheat or subcooling can be measured to determine if the system contains the correct amount of refrigerant, and also if there are problems with the refrigerant control, evaporator or condenser. A low charge condition can contribute to icing, but is usually not the main cause, and you already has that checked.
Icing the evaporator is usually caused by insufficient airflow across the evap coil. Granted, a tight door seal will minimize the amount of moisture in the air inside the cabinet, but if there is sufficient airflow over the coil the moisture should simply condense and drain. Commercial units such as this are usually set-up with the evap fan running continuously, and the cycling compressor is controlled by a pressure switch connected to the suction line. As the sensed temp drops, so does the suction line pressure, and when the pressure reaches the corresponding desired temp, the switch opens and de-energizes the compressor, but the evap fan keeps running. As the temp inside the cabinet rises, so does the suction line pressure, until the switch closes and energizes the compressor, repeating the cycle.
OK, so back to your problem. I think there are 2 likely suspects here:
1. The evap fan motor may be going bad. They can overheat, causing intermittent operation. When cool, they start and run, but then they heat up and shut off, which of course will lead to big-time icing.
2. Do you have this thing running on an external (override) t-stat like a JC or Ranco? If so, you probably need to connect the evap fan to non-switched power. With an external stat, the power is cut to everything at the same time. The problem with that is that liquid refrigerant will continue to be metered into the evap for quite a few seconds
after the fan has stopped, and the result is that a small amount of ice will form at the end of each cycle. Another option would be to run the fan on a delay-on-break timer (again requiring unswitched power), so that it runs 30-60 seconds after the compressor stops.