Mitsubishi mulit-zone short-cycling in moderate weather
![](https://secure.gravatar.com/avatar/ab87ec3c6f74d621ea49e0deceab8976?s=52&d=https%3A%2F%2Fimages.greenbuildingadvisor.com%2Fapp%2Favatars%2F200x200%2Fm.png&r=g)
We have 1650 square foot house in Albuquerque (zone 4B), built in 1947. I listened to the most recent Green Architect podcast where they talk about Mitsubishi multizone heat-pumps short-cycling in moderate weather. In the podcast, I was hoping for a solution or advice for owners in this situation, but it was more about getting the sizing correct from the install.
We are in shoulder-season here now, with temps in the 50s and 60s during the day, and mine is doing the same thing. Here’s what I have:
Outdoor:
MZX-8C48NA
Indoor:
MSZ-FH06NA (x4, two bedrooms, dining, and study)
MSZ-FH15NA (living area and kitchen)
MSZ-EF09NAB (master suite with bath)
The house is comfortable, despite the short cycling I hear(when in the backyard) the outdoor unit doing during the day. I have to do the usual thing of adjusting the setpoint based on the outdoor temperature. Right now, keeping it at 66 results in the house staying at around 68.
The house is old construction (mix of CMU, and stick framing) with virtually no insulation, but the air leakage isn’t bad for an old one, probably because it never had ductwork installed. A friend recently did a blower door test and it was 1690 cfm50, and I’ve sealed a few big holes since then. Part of the house (about 1000 square feet) is over a vented crawlspace, and the rest is slab-on-grade additions.
Recently I had an energy audit done (for free by the local electric utility) and he thought my usage looked really normal, roughly 13000 kwh last year. He estimated (I guess based on weather data and seasonal fluctuations) that 27% of that was heating and 22% of that was cooling.
On the podcast they described tight, well-insulated houses in Maine short-cycling in shoulder season, and I thought that my older, leakier house in Albuquerque might be analogous.
Seems like one of the solutions could be to turn the system off once the outdoor temp has reached a certain temp, but what is that temp? Open to hearing other ideas.
Thanks!
Mark
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
![](https://www.greenbuildingadvisor.com/app/themes/greenbuildingadvisor/dist/img/modules/detail-library/thumb1.png)
![](https://www.greenbuildingadvisor.com/app/themes/greenbuildingadvisor/dist/img/modules/detail-library/thumb2.png)
Replies
>"Seems like one of the solutions could be to turn the system off once the outdoor temp has reached a certain temp, but what is that temp? "
Before turning it off completely, start by turning off the zones that seem to run least often as the shoulder seasons approach. By shifting some load of those zones to the zone that are still active the compressor cycles will be longer.
The minimum compressor output of the 8C48 is the MAXIMUM output of an FH06, so the fewer FH06 zones that are active the fewer spin-ups it will do just to serve a miniscule amount of load for just a few minutes would be occurring. I suspect the FH06 in the bedrooms would be the first to try turning off, since an FH06 is usually oversized for even the design heat load of a typical bedroom. If it's getting too cool in the bedroom at night, of course turn it on, but only when the room is in use.
The Mitsubishi system sends bypass refrigerant through the "off" zone heads even when the head itself isn't calling for it, so there will be some amount of heating going on when any other head is actively heating. In gross oversizing cases a nominally idling head can even overheat/overcool a tiny-load room just on the bypass refrigerant. So turning an FH06 "off" with the remote is really just "sort of off".
Without knowing the layout of the house or the design zone loads it's hard to make intelligent guesses as to the order in which to turn off the zones, but if the dining or study have open doorways to the living/kitchen area those become prime candidates. You may be able to ride through most of the shoulder seasons with just the FH15 running, turning it's temperature setting up a bit if the adjacent "off" zones are running too cool, as long as it doesn't make the the living & dining area too warm. With just a bigger head running it can even modulate the head a bit, where as a bunch of FH06 with low loads will just cycle on/off, and if the cycles are too short to overlap with other heads, cycle the compressor on/off with it.
Here's a floor plan showing the head locations. The red ones are the FH06 units, the blue is the FH15, and the green is the EF09. The mud/laundry room is unconditioned. The middle bedroom really does have a door to the hallway! You can see that the rooms are all open to each other except for the bedroom doors, which we keep open until our daughter goes to bed at night before us. But they are not big openings, just the size of a doorway.
Today I tried turning off all of the FH06's. I left the EF09 on because that room is mostly disconnected from the rest of the house and is on the north side. What you said about the 'off' heads getting refrigerant is totally happening, and possibly making our house even warmer than with the heads on! We had a high around 60 today with rain, so no solar gain today. The temp has dropped to 40 and I turned on the one FH06 in the southern-most bedroom since our daughter went to bed, and the temperature in the house has remained very comfortable.
At first it seems counter-intuitive to turn off heads if the compressor is short cycling, because I thought we would want as much of those little loads as possible on the outdoor unit so it would actually use it's minimum capacity, but I guess I was thinking about it backwards. Since those little loads from the FH06's were not presenting themselves simultaneously, it's better to get one bigger load request from the FH15?
Thanks so much for the help!
>"Since those little loads from the FH06's were not presenting themselves simultaneously, it's better to get one bigger load request from the FH15?"
That's right. When the shoulder season loads are so low that the rooms get up to temp quickly the heads cycle the compressor on/off as quickly and often as the heads do, with little overlap. With most of the heat being supplied by fewer heads the number of spin-ups of the compressor go down, and the lengths of the cycles go up.
It's too bad the FH15 isn't blowing toward the kitchen door, where it would also be blowing at least some of it through to the den. As it is it looks like with the door open the FH06 in the den would also be heating/cooling the kitchen. With no exterior walls the kitchen's heating load is miniscule, possibly even negative (the refrigerator alone might be enough to keep it at temp most of the winter.)
The first bedroom FHo6 to turn off and leave off would be the middle bedroom, which has half the amount of exterior wall of the corner bedroom, and probably half the heating load or less. If the occupant of that room prefers a cooler sleeping temperature that one might only need to be turned on during cold snaps.
Similarly the den has conditioned space on 3 out of 4 sides, and could probably be simply left off most of the time if the door to the kitchen is kept open when the den is unoccupied. At it's rated 8700 BTU/hr output it doesn't take much time for an FH06 to bring a 175 square foot room with only one exterior wall up to temp. If it's part of a sorta-insulated stick framed part of the house, even at design temp heat load of that room is probably less than 2000 BTU/hr, maybe even less than 1500 BTU/hr.
If the dining room isn't used except during meal times, turning that FH06 off except when needed shovels a bit more load onto the FH15 too. The recovery time on that room will likely be longer than with the den, since there is more exterior wall, more load, and it may be starting out at a cooler temperature, unless it's getting a lot of solar gain due to it's southerly and westerly exposures. During the cooling season it will probably run longer duty cycles than the master bedroom's FH09 since it gets a slug of heat from solar gain, some of it much later in the day.
I'm sure with experimentation you'll figure out what combination works best during different weather cycles & seasons. Don't be afraid to turn a few of them completely off when there is little to no load. Turning them all off except the FH15 probably work just fine during the lighter load days.
Thinking further down the road, as little as 2" of continuous EPS or 1.5" of polyiso on the exterior of the CMU walls would bring it up to code, and would provide steadier room temperatures. An R10-ish sweater over the house would even-up the indoor room to room temperatures by quite a bit too. With an EIFS stucco type finish EPS is a bit easier to deal with than polyiso, which would need a different, more expensive exterior finish/siding option. Products such as InSoFast EXe 2.5 over both the CMU and framed sections would bring it up to better than code, and is a reasonably well designed retrofit that goes in pretty easily (at a higher material cost, of course), as the name implies. (Not that you were really looking for another expensive home project... :-) )
I'm experiencing short cycling in the summer that leads to a musty smell in the bedrooms. I have a msz- fh15na paired to an muz-fh15na which never has the issue as it does the living room, kitchen ,hallway etc.. But the 3 bedrooms each have their own wall unit ( 2 msz-fh09na's and a msz- fh06na) connected to a mxz-3c24nahz2. I believe this outside unit only " throttles" down to 6000 btu so in the well insulated bedrooms - closed cell spray foam , plenty of attic insulation , zip system sheathing.... the temperature is reached before the humidity is removed . The only solution seems to be to put the units in dry mode but then it gets like a meat locker.
The problem is your heads are probably 3x to 4x oversized for the load in the room. This is why a head in every room is terrible idea even though most major manufacturers promote this in their brochures.
The way I've gotten around the issue is by blocking the intake on the unit to reduce the airflow. This reduces the effective capacity of the head and also lowers the outlet temperature increasing dehumidification. All modern units monitor the coil temperature and won't freeze up but you'll end up short cycling the outdoor unit even more. I don't know how this effects the long term durability of the system, but much better than sleeping in a meat locker.
For a bit of extra cooling, I like to use this to cover the inlet:
https://www.nationalgeographic.com/magazine/2019/09/
I was wondering about the effects on system durability with what you tried. I've tried to accomplish a similar thing by running the fan on whisper speed as to cook it down more slowly with mild success. The installer wanted a unit in each bedroom for the winter, and possibly for a better payday for him...., But I wanted a branch box and ducts for the bedrooms but he resisted. I wonder if that would have made a difference. Ironically in the winter I usually just run the main living room/ kitchen unit and leave the bedroom doors open. I'd also like to see if kumo cloud has a setting where dry mode is engaged at a certain humidity then automatically switches back to a/c at another humidity level. I don't have it myself but I've been thinking about it. Thanks.
The only way to know if this is a problem is with a power meter you can use to monitor the outdoor unit.
For my multi split, as long as I have at least one other head running, the unit will not cycle.
If I do reduce the flow too much, you can see it slowly ramp up on power then ramp down to base load. I would consider it more of modulation than cycling. If this was happening without a base load, I could see it being hard on the outdoor unit.
P.S. The link was meant as a joke.
Cldlhd It sounds like your system is too large for your cooling load. This is a risk when you sizes the system to heat a house when it is -20 outdoors .
The first question is what part of the world are you in?
What is you indoor temp and humidity?
Have you tried turning off 2 of the heads so the last one is doing all the work?
Walta
What's with the link?
I'm in the suburbs of Philadelphia. Well when it's hot and humid I usually set the air conditioning to 70 the humidity in the main living area is usually in the 50's to lie 60's but the bedrooms will vary more, from low 50's to 70. I assume this is depending on if the outside unit is running. I was aware of short cycling humidity issues with Central air systems but was hoping Mitsubishis wouldn't have this issue due to the inverter technology. I wasn't aware of the fact that the multi-head outside unit had a minimum BTU output that much higher than the individual wall units. You're living you learn I guess. Right now with the bedroom doors open and only my main unit in the living room running the humidity in the bedrooms is 51%.
I thought it was a joke but ya never know. I have a Sense energy monitor but I don't have a ton of confidence with it accurately reporting all the usage from the whole system ( head, compressor, fans etc . )
Post a month's worth of daily heat pump energy use and I'll calculate some relative COP vs load data.
Short cycling is probably less of an issue that you think - sometimes it even improves things (eg efficiency, latent removal).
Sense should work well enough. The heads consume next to no power (~20W to 40W). You should see roughly 500W for a 6k head running and around 800W for a 9k.
Cycling on these is terrible for humidity removal as the fan continues to run. All the moisture on the coil gets re-evaporated when the outdoor unit cycles off. This is the musty smell you mention.
You might need to reduce the number of heads you run so at least one will not be cycling to get better humidity control. One on the second floor in a sunny room with the door open is your best bet.
Don't know why I can't reply to Akos but have to reply to myself but.... It's a single story house, if I only run the living room head and keep the doors all open I don't get the musty issue as I assume the load is big enough to keep it running longer. I figure that's the concept you're getting at with just running one of the 3 bedroom units but at night with the doors closed that won't work. I thought maybe running all 3 with the doors closed might work. I was thinking maybe there would be enough demand to keep the outside condenser above it's 6000 BTU minimum? The bedroom units are designed to go down to 2000 BTUs if they have their own outside unit in a one to one setup.
If your multi is like mine (and most) it will run the indoor units at their rated capcity, it will not module them. So a 6k head will only run at 6k output or zero. No in-between.
The only modulation on the outdoor unit is to match setpoint to the combined capacity of all the indoor heads on. So if you have a 9K and 6k head on, the outdoor unit will run at 0, 6k, 9k or 15k depending on which heads/combination are calling for cooling.
So if you run an over sized head in a bedroom, it will cycle and because of re-evaporation do almost nothing for dehumidification.
There is no easy way around this except for blocking the airflow on the unit a bit to reduce its capacity.
The better solution is to replace a couple of the heads with a single properly sized ducted unit feeding multiple rooms. Pretty expensive change though.
That's what I originally wanted - a ducted unit in the attic feeding the bedrooms and a ceiling cassette for the main living area. The contractor, supposedly a Mitsubishi diamond elite contractor, resisted saying they rarely installed anything but wall units, the wall units are more efficient blah blah blah. In hindsight if I knew more I'd have pushed for it the way I wanted or found someone else. Funny enough a work buddy of mine I met after my install hasxa friend who commercially installs Mitsubishis and he's currently getting the exact setup I wanted.
Dana, thanks for the thorough and helpful response! The last two nights (and days) I've been running just the FH15 and EF09, with all of the FH06's in 'fan' mode blowing on medium. The warm air that the FH06's produce when receiving bypass refrigerant is enough to heat the house such that the rooms with the 06's are are a few degrees warmer than the rooms with the heads that are actually on. Should I be concerned about this?
Last night the low temp was around 30, and the house stayed totally comfortable and an even temp with those two heads running and the 'off' heads blowing the bypass refrigerant air. It seems like most of the work is being done by the 09 running in the Master (north side of building, so it is the coolest room), and the 15 doesn't seem to come on much at all.
I wonder how low the temps need to be to actually need to turn on any of the 06's. Seems like my system might be in the category of 'gross' oversizing. Quite the learning experience here!
I'm wondering if adding adding a mhk1 thermostat for the FH15 would be of any help. Thoughts?
Btw, I am always looking out for expensive home projects :), and giving our house a foam sweater would be great. But the definite next project though is the crawlspace: insulating and making it conditioned. The original part of the house is over the crawl. If you look at the floor plan it's the kitchen, living, and two southern bedrooms and bath. Possible project after that is getting a new roof with insulation and PV, but we'll see. The stucco is in pretty good shape so insulating the walls is not something that will happen for a while.
Thanks again for all your expertise!
Mark
Hi Mark -
Just wondering a bit about your airtightness. You stated 1650 sf for house. If you average height is 8 ft, that means your house is about 5200 cubic feet; is that right?
If it is, that means that 1650 cfm50 translates to 19.5 ACH50--VERY leaky. But you had an energy audit done; did they do a blower door test confirming the 1650 cfm50 your friend got?
In any event, if these numbers are correct, really surprised that the energy auditor did not emphasize how much benefit you would get from air sealing.
Peter
Mark did mention fixing a few big holes since then. The part of the house over the vented crawl space would be a prime location for sniffing out more.
Maybe I'm doing the math wrong, but here's what I was thinking:
1650 square feet, 8 foot average ceiling: 13,200 cubic feet.
1690 cfm is 101,400 cubic feet per hour.
Air exchanges per hour would be cubic feet per hour
of the fan divided by cubic feet of the house, right?
That leads to 7.7 air exchanges per hour. How did you get 19.5?
Thanks for the help!
Mits's 15- and 18-class units are not nearly as good for small spaces as the 6-, 9-, and 12-class units are. The minimum fan speeds of the 15/18 units limit their ability to modulate as low as the 6/9/12s. I have a two-head mini-split system with a 9 and 15 and the 9 never cycles (always cooling - never just the fan) but the 15 does. This tells me that it's not an issue with the compressor's minimum capacity, but instead that the minimum fan speed of the 15-class unit does not allow that head to modulate as low as one with a lower fan speed would.