Mitsubishi PUZ-ZM Series freezing up
I’m experiencing outdoor coil freezing issues in heating mode with two different Mitsubishi PUZ-ZM outdoor units, and I wanted to see if anyone else has come across this problem.
First Issue – PUZ-ZM71 Freezing Up
7 months ago, I had a Mitsubishi PUZ-ZM71 installed with a PEAD-M71 indoor unit. From the start, the system kept freezing up in heating mode, even when outdoor temperatures were well above freezing (10-15°C).
One thing I noticed early on was that the outdoor fan speed was very low—only 200 RPM unless the outdoor temperature dropped below 10°C. This meant that in mild conditions, the fan wasn’t moving enough air over the coil, allowing it to drop below freezing and ice up.
I also found that pouring cold water over the outdoor temperature probe would trick the system into thinking it was colder outside, which made the fan ramp up and quickly remove any ice forming on the coil. However, once the probe warmed up again, the fan slowed back down, and ice would start building up on the coil again. This confirmed that the system was controlling fan speed purely based on outdoor air temperature rather than actual heating demand or coil temperature.
After extensive troubleshooting, the entire system was replaced, including:
• Outdoor unit (PUZ-ZM71)
• Indoor unit (PEAD-M71)
• Copper lineset
• Control panel and wiring
• Refrigerant charge (checked twice)
Despite all this, the exact same issue persisted—the system would freeze up in heating mode and enter defrost even in mild outdoor temperatures.
Mitsubishi tested the unit in their own workshop and confirmed that there was an issue, but they never specified the exact cause. They eventually replaced the PUZ-ZM71 with a PUZ-M100, and that completely resolved the issue. One major difference was that the M100’s outdoor fan runs at 450 RPM when needed, preventing freezing and allowing proper operation.
New Issue – PUZ-ZM140 Freezing Up
At the same time Mitsubishi replaced the ZM71 with an M100, I also installed a PUZ-ZM140 (twin-fan unit) with a PEAD-M140 for my living room.
Out of curiosity, I ran some tests on this unit today at 15°C outdoor temperature, and it’s freezing up just like the ZM71 did.
Here’s what I observed:
• At 9°C outdoor temp, the fans run at a higher speed.
• At 13-15°C outdoor temp, the fans slow to 240 RPM, and the coil eventually drops below freezing, leading to ice buildup and defrost activation.
• The PUZ-M100 does not have this issue—it ramps its fan properly and does not freeze.
It seems like the PUZ-ZM series may have a fan control logic issue, where fan speed is based purely on outdoor air temperature rather than system demand or coil temperature. This means that when the outdoor temperature is mild, the fan does not ramp up even when the coil is dropping below 0°C, causing unnecessary freezing.
In contrast, the PUZ-M100 properly adjusts fan speed based on demand, preventing freezing.
Has anyone else had freezing issues with a PUZ-ZM series heat pump in heating mode?
Its so frustrating as ive wasted so much time on this already.
Here is a couple of videos of what’s happening.
Cheers.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Unfortunately those are not sold in north America. Does looks like somebody messed up on the FW for the fan control, if nothing else, running the fan at such low speed is not great for COP as the coil temp is too low. Sometimes the units have jumper switches on the outdoor units that let it adjust certain parameters, I would check the installation manual to see if yours has it.
If all else fails, you can do a non manufacturer approved warranty void modification by putting a resistor in series with the outdoor temp sensor. This will fool the unit into thinking it is colder outside than it really is. You can look a the repair manual for your unit as it will have the resistance curve of the sensor which you can use to size the series resistor.
Cheers for the reply.
Are they sold under another model number in North America?
Yes it does look that way, but how his this possible over multiple different size outdoor units? to note, I have had two PUZ-ZM71s and one PUZ-ZM140, all presenting with the same issue, I could understand if it was solely one size unit, but multiple sizes with the same fault? I had spent a lot of time studying the service manual installer manual etc etc etc, nothing about any switches for different temp zones. Mitsubishi had one of the units back in their workshop and they were stumped with how it was operating.
The resistor is an idea, but then it would have the be removed for summer time as it would cause issues with the fan going to slow in the summer. Another option would be to run the outside thermistor against the coil, that would fool it into thinking its colder outside. But I shouldn't have to do these work arounds on brand new units.
Cheers
R32 is pretty new here, so there isn't a comparable unit.
I would guess the unit was designed for cooling climate so they never did much testing in heat mode. It works, just not well.
Outdoor unit should not start frosting up till near freezing temperatures, at least that is what happens with the air source heat pumps here.
If they have "cold climate" units in your area, you can try swapping to one of those. Those are definitely meant for heat and should work better.
I believe these ZM outdoor units are the cold climate versions, they produce heat at a colder outdoor temp than the M version that is working fine.
Oh and to note, the temps on the video are all in C, so in that video its about 60 f outside where the unit has frozen up.
Have you checked your thermistors resistance to make sure they are all functioning in the correct ohm range. I wonder if it is basically a limp mode that is making it run solely on outdoor air temp. It seems like the first check would be sensors before logic or firmware. But I don’t really have a ton of Mitsubishi experience. My checks would be pressure, superheat/ sub cool (if possible) then sensors. Also in the video you pour water on the sensor, are you basically triggering its humidity algorithm to reduce frost build up?
From the control panel you can read all of the thermistor temps (through request codes) they are all reading correctly, so the units know what the temperatures are. Limp mode does sound logical but why are multiple units doing this exact same fault? Pressures are fine. Im pouring water over the ambient temp thermistor to trick the unit to think its colder outside. As soon as the ambient air drops below 10c the fan or fans speed up and the issue goes away.
So the system is seeing that the coils are at or below 32 and the the outdoor is at 60?
I wouldn’t be particularly surprised if it was a firmware issue as Mitsubishi has certainly had those, but it seems like the compressor speed and rev should still be keeping it above freezing even with a moderate to low fan speed.
Are you able to measure superheat/subcool, or does the unit tell you?
Low refrigerant would be at the top of my list of possible problems. Or some sort of error in the eev. But it could certainly be some software glitch (though you’d think they would have that pretty well worked out at this point…..)
The unit does tell you the subcool and super heat, but im not sure how accurate that would be. It was telling me subcool was around 1c and discharge superheat from 20-40c. They come pre charged for a maximum lineset of 30m, we are within that length. The charge was pulled and replaced twice, the charge has always been correct. Remember exact same issue was on 2x zm71 units and 1x zm140
Unit, with Mitsi testing one of the ZM71 units in their workshop with the same freezing results. They also ran a third brand new ZM71 in their workshop with “inconclusive” results, they then decided to send me an M100 and that works flawlessly.
Those numbers seem very off and would indicate low refrigerant or restriction in your refrigerant line. I dont know about you other unit, but maybe it’s a common component causing the issue. Since both numbers are off it seems like it’s not the fan
Here is a picture of gauges from 6 months ago. If I remember correctly, the pressures struggled to climb above 300 until the fan sped up. Once the fan was up to speed the temps and pressures were within spec if I remember correctly.
The copper linset was replaced once, (the one that was replaced was fine) and the bigger unit is on another linset + Mitsi used their own lineset in their workshop, so thats ruled out. The charges are to what the units require, so thats ruled out. I guess it could be a faulty component, I find that strange though across multiple units.
Yes I’m not saying it’s for sure the charge, it’s just almost certainly not the fan. When you increase your fan speed you may be melting the frost but you are exacerbating the superheating. You want that number in the range of 15, not 30-40. It could well be a software issue, causing the eev to close to much or maybe the filter is clogged, or something along those lines. A good technician should be able to diagnose these one at a time to figure out what the issue is.
The unit is probably throttling the fan because it is showing such a high superheat
Here is another picture of the gauges after it was running for a bit. If its the charge then Mitsi have pre charged them all wrong and the charge weight on the side of the units are wrong. We ran it without the filter and in the ceiling with the plenums removed. Currently the bigger ZM140 unit just has the ceiling unit hanging in the ceiling without the plenums or ducting, that is how I tested it in the above video.
Another thing we did notice, is the compressor was slugging when the fan was spinning slow.
Deleted
Freyr, just thinking about your comment regarding the unit potentially keeping the fan slow to reduce the superheat. Isn't the superheat in heat pumps controlled by the metering valves rather than the fan speed of the outside unit? to note, all the numbers and pressures were taking in the winter when the heat pump was working in heating mode. The unit works fine in the summer in cooling mode.
Yes to my knowledge that is now they are controlled. I guess it doesn’t exactly make sense to me as if the fan speed was the only thing increasing, the super heat would increase. But if the fan speed increased, and the refrigerant increased to keep it at saturation throughout more of the coil, then the superheat would decrease. I feel like there is a disconnect between the fan speed and the eev at your high outdoor temp. But honestly, the control algorithm is above me, and there well may be other factors at play. If I were you, and the Mitsubishi techs/company didn’t have any other thoughts, I would plug all your data into a reasoning model like o1 or o3 and see what spits out, and maybe you will get more components to test (ie, what signal is the fan getting), what sensors should you test. I think you could narrow the problem down pretty effectively.
Maybe others will chime in, I don’t think I have much else to add except test everything till you find the root….which maybe isn’t reasonable and Mitsubishi should just replace it.
Cheers for the help.
I have done just that multiple times with chat gpt, it keeps falling back on a potential issue with the fan control logic.
One point, through the request codes you can see the fan output steps and fan rpm, each step represents a fan speed, the lowest step on all 3 models of heat pumps seems to be step 5, step 5 on the original ZM71 was 200rpm, on the other problematic ZM140 step 5 is 240rpm, on the M100 unit that has no issues, step 5 is 470rpm. So when the outside temp is above 10c, all 3 units are being told to run the fan/fans at step 5, but because the M100 fan is running so much faster, it doesn’t freeze up.