GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Issue with attic-located Mitsubishi ducted cooling

kylel | Posted in General Questions on

We have been struggling with a poorly performing Mitsubishi ducted mini-split setup since 2016 that we purchased shortly after moving into our 3 story, 1896 townhouse.  The ducted unit is in the attic is for cooling only the 3rd floor.  Whenever it would get hot, like over 90F outside, we would see that the 3rd floor temp would go way over the 73F temp setpoint, going up to over 78F.  The installer, who was a Mitsubishi diamond dealer, tried adding another duct and even added a second outdoor unit to move over two other ductless units I have in the house off of the main outdoor unit but it never solved the problem.  Since then, the Mitsubishi dealer was acquired by a company that does not sell Mitsubishi so the level of support is minimal.  I have not had any luck discussing this problem with Mitsubishi.  

At this point, the setup of interest is (all fans are set to auto speed):
– MXZ-5C42NA2 outdoor unit
– MSZ-FH06NA ductless on 1st floor – MHK1 Setpoint 74F 
– MSZ-FH09NA ductless on 2nd floor (in bedroom) – MHK1 Setpoint 72F
– MVZ-A18AA4 ducted unit in the attic (it isn’t a conditioned attic so it gets hot) MHK1 Setpoint 73F 

Deciding to take more of this into my own hands to understand the problem, I set up temperature loggers in various positions related to the attic ducted unit:
– Inside the ducted unit (in the filter area)
– At the outlet of the ducted unit in the plenum 
– At the MHK1 thermostat that controls the unit
– In the attic near the ducted unit to measure the temp around the ducted unit
– I also put an inexpensive anemometer on a duct outlet to get a sense on if there was a drop in airflow that could indicate icing in the coils 

Also, to see how the MSZ-FH09NA ductless on 2nd floor (in bedroom) was performing while the ducted unit was running I also set up temperature loggers:
– Inlet of the unit
– Outlet of the unit
– At the MHK1 thermostat for the ductless unit

Attached are the graphs of what I found (including some other details).  Some key findings:

       The attic temp peak temp was 101F

       The temperature at the thermostat for the 3rd floor ducted unit went up to 77.4F (73F setpoint) 

       The 3rd floor ducted unit was only providing 11F of cooling on the warmest day (6/23/20)

       Looking at a longer period of time (from 6/23/20 08:00 until 6/27/20 0:00) I can see that the typical AC In-Out varies from 9.5-14F when the unit is working to cool the 3rd floor.

       The airflow (the cheap Bluetooth anemometer I used is not very good so doesn’t have much resolution and is a bit noisy) showed that the airflow was increasing when the temperature at the thermostat was 76.5F so I don’t think that there is icing plus the outlet air isn’t very cold. 

       On the hot day, the 3rd floor ducted unit was essentially full on until 4:56 the next morning at which time there was a noticeable temperature response at the 2nd floor ducted unit.

       Also included is the house power during peak temp at the 3rd floor thermostat.  No obvious power spikes, I think, to indicate a serious issue with the unit.    


Regarding refrigerant charge, the total line length is approx. 176 ft.  The company that took over the original Mitsubishi dealer, in order to try to figure out the issue, emptied and refilled the unit last year with 9 lbs 9 oz of refrigerant (a number that they calculated).  The unit’s factory charge is 8 lbs 13 oz.  Using the calculation in the service manual, it looks like the adjustment for line length would be 1.08*(176-98)/5 = 16.9 oz which would indicate that the charge probably should be closer to 9 lbs 14 oz (so 5 oz higher than what they filled).  Also, keep in mind that this was filled last year so there is always a chance for a leak that might have dropped it below the fill level.  

My questions are:

       Any idea why I am only getting an 11F temp drop across the inlet and outlet of the ducted unit on the hot day (and up to 14F in the best typical case)?

       Would being 5 oz low in refrigerant explain this?

       Is there more data I need to collect?  Anything I might be doing wrong with my data collection so far?  

       Is there a known issue with having the MVZ-A18AA4 in an unconditioned attic?

       Does the interaction (sudden temp drop at 4:56) for the 2nd floor ductless important?  Looks like too much is being diverted to the ducted unit but at the same time the ducted unit wasn’t able to provide enough cooling.

       Any suggestions on getting help from Mitsubishi?  The original Mitsubishi dealer that sold me the unit that it would control to a setpoint of 72F and that has never been the case on hot days. 


I’m in climate zone 5A.  Much appreciate any suggestions anyone can provide to hopefully finally resolve this drawn out situation.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. joshdurston | | #1

    Wow, I have to say this is one of the most impressive posts I've seen on the GBA Q+A. Good work collecting some data.
    I think you've proved that you have a refrigeration problem (not an airflow problem), the attic unit doesn't seem like it's performing with that low of temperature drop.
    My FH09 (1:1) will have a 20C temperature drop across the coil (I've seen outlet temps as low as the 5C range).

    I'm no expert, but have some questions:
    Does the attic run better if you disable one or both of the 1st and 2nd floor units or at least set the fan speed manually to low?
    It would be invasive but it would be interesting to swap the attic unit to a different set of ports on the outdoor unit. That might show if the expansion valve is the problem.
    Are the liquid line temperatures (entering the indoor coil) about the same on all the units, what about the suction line temps leaving the various units?

    Is your condensate draining properly? High condensate levels can trip some faults and cause problems.

    1. kylel | | #2

      Hi Josh,
      Thanks for the reply and the compliment! Yeah, sure is nice when you have the data and these new Bluetooth sensors make it easy and affordable. Good to know that getting larger temp drops on this type of unit is expected.

      I will make it an experiment to shut off the other two units and see what happens to just the ducted unit.

      I will also have to see about measuring the coil inlet and outlet line temps. Seems like there would be some more useful info there.

      The condensate drain seems to be ok. I flushed it a few weeks ago and it was clear.

      I’ll save swapping the line set locations until later. Hopefully that won’t be needed but could be something to try if I don’t make much progress.


  2. Expert Member
    AKOS TOTH | | #3

    There has been a couple of threads about similar issues, seems to be a parameter setting issue and/or insufficient charge:

    1. kylel | | #4

      Thanks for the link. Sure sounds pretty similar. Seems to at least be pointing to lack of charge being an issue.

  3. Yupster | | #5

    Amazing job with the data collection. High SEER units like these are VERY sensitive to refrigerant charge. So it definitely could be that. 5 oz. of refrigerant is the amount that would fit in a 6' charging hose, so it's not a lot!
    I would recommend before you adjust the refrigerant levels to calculate the actual operating capacity of the unit. You can do that with the formula BTU = (H2 - H1) x 4.5 x CFM. You have to know how to measure enthalpy and the cfm you can get from a static pressure measurement and the blower charts. H2 = return enthalpy, H1 = supply enthalpy. This will tell you if the system is producing anywhere near the right amount of btus. If it is producing somewhere in the right range, your problem could be elsewhere. If not, now when you adjust the refrigerant, you can recalculate this and know exactly how much you improved the system. Any good hvac tech should be able to do this for you. Enthalpy is super easy to measure with the right tools.

    1. kylel | | #6

      Ok, yeah 5 oz doesn't sound like much and probably smart to do more measurements of the system before making changes. I may use seeing if they can perform enthalpy measurements as a way to find an HVAC tech who knows what they are doing. Any other measurements that the better HVAC techs usually do?

      1. Expert Member
        AKOS TOTH | | #7

        Before going down the road of HVAC tech, check if the parameter in the previous thread is set correctly. It is something you can change through the existing thermostat.

        1. kylel | | #8

          Are you referring to the Target ETm setting? I don't think that is something I can change through the thermostat. The other post seems to indicate that it is a dip switch setting on their MXZ-8C48NAHZ unit. On my MXZ-5C42NA2 I don't see a reference to it in the service manual so perhaps it might not have this option. I could increase the static pressure through the remote but I'm pretty sure when I've tried that in the past it didn't solve the problem. Let me know if you think there is something else I might be missing, though.

          1. Expert Member
            AKOS TOTH | | #9

            Maybe give Mitsubishi a call to see if there is something similar on the non-hyper-heat units.

            There are a lot more parameters that can be adjusted and finding documentation on it is unnecessarly hard to find.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |