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

Understanding Minisplit Short-Cycling and Turndown Ratio

benrush | Posted in Mechanicals on

I understand that a “somewhat” oversized mini split will generally run at a more efficient rate, as opposed to an undersized mini split, which will run full-out, and therefore be less efficient. It sounds as if the problems with oversized mini splits would be 1) wasted money on oversize equipment, and 2) decreased dehumidification, due to the equipment not running enough. Am I about right, so far?
Is there more to “short cycling” than what the two words sound like? (i.e. “cycles” that are “short”?) Is this a matter of perception? -like, I think the cycles are kinda short, and Dana REALLY thinks they’re short, but Anon3 thinks the cycle lengths are just fine? -or is short cycling a more specific, inarguable phenomenon?
On a related note- I’ve read that programmable thermostats don’t save money with mini splits; if we turn a mini split off when we leave- it will run full-out to catch up when we return, which will be so inefficient that we’ll waste all the energy we saved by turning them off in the first place. Wow. Really? How long would a mini split have to be off to actually save energy? I bet a precise answer would depend on half a dozen factors- but is there a rule of thumb so I can easily achieve some confidence that I’ve made the right decision most of the time?
How about turndown? What does that mean? Under what circumstances should we care about the turndown of some particular equipment? (Am I even using the term correctly?) I thought mini splits were selected on the basis of size, SEER, manufacturer reputation, cost, and (if you also want heating like I do- near Chicago- in zone 5 but close to 6) how well it will heat at low temperatures. Do I also need to be thinking about turndown when I select equipment?
As always, thank you for your wisdom. I look forward to your feedback.

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. Kenny78 | | #1

    Dana will no doubt write an essy( which I am grateful for all the education) but the lay mans explanation and knowledge(me) is thus: the turndown ratio of equipment is a percentage of max output, giving you a minimum modulated btu/hr. The mitsubishis in 9 and 12,000 varieties have a minimum of around 1700 btu/hrs. Most competitors minimum is ~3000 btu/hrs. If the load for your conditioned space is less than the minimum, it cycles. This is a problem for many multisplit systems because you can only modulate to the compressor minimum unless multiple zones are calling.

    In regards to programmable thermostats, I don't know about a "total" waste although I would think them uneccesary unless your equipment was grossly oversized.

    There are further nuances to indoor units which include a larger coil inside(for more dehumidification?) and cfm rates( occupant comfort and/or reaching very long rooms?) that are beyond me.

    As to short cycling, it is undesirable for dehumidification and for almost all equipment is bad for longevity.

    We purchased a home that had been built with window units. It only had a 24,000 btu in the main room(half the house) which kept the house tolerable in comparison to no air conditioning. Over the winter of 2012-13 after 1 year of the window unit I found a used Goodman 12,000 btu heat pump for $100 dollars(add $300 for the bad compressor they didn't remember and another load of gas) and I'm off to the races. Now I tell this to emphasize that I regret not buying a top tier unit that had an inverter compressor and operations below 40 degrees. But before monkeying with this thing I had not yet become convinced of mini split awesomeness. I secretly pine for a significant component failure that I can't bubba back together so I can put a real good unit with low modulation in its place. The Panasonic cassette that I installed at my mothers was Italian and Japanese and blows the socks off this rebranded midea unit. If I had one of the "good" we could keep the remote at a higher setting and still keep the house dehumidified to keep the crawl space odor and stuffiness at bay, which until all the finished floor and subfloor come out is unfixible due to the manner ofconstruction. For cooling this odd 1300 sqft house this unit is oversized due to tree shading and ~9" thick limestone veneer.

    Almost all the mini splits and even window units have a place, but careful and deliberate consideration is the watch phrase with HVAC equipment. Least here at the top of climate 3a anyways.

  2. GBA Editor
    Martin Holladay | | #2

    Q. "Is this [short cycling] a matter of perception? -like, I think the cycles are kinda short, and Dana REALLY thinks they're short, but Anon3 thinks the cycle lengths are just fine? -or is short cycling a more specific, inarguable phenomenon?"

    A. If you are hung up on the phrase "short cycling," you can call it "cycling" instead. A ductless minisplit is most efficient when it operates for 24 hours a day. If the load is less than the appliance's minimum output, then it has to turn itself off. That is cycling. You don't want that to happen every day.

    Q. "How long would a mini split have to be off to actually save energy? I bet a precise answer would depend on half a dozen factors- but is there a rule of thumb so I can easily achieve some confidence that I've made the right decision most of the time?"

    A. I'm just guessing here, but let's say 24 or 48 hours. If you are going away for the weekend, it might make sense to turn off the minisplit. (Or it might not -- your choice.)

  3. benrush | | #3

    Ken, Martin,
    Thank you both.
    My mind is still adjusting to the concept of invertor compressors.
    Will sizing a mini split correctly- so that it's running all the time- also allow it to achieve maximum dehumidification?
    So, if I want a mini split to operate 24 hours a day, I need to add "minimum output" to my previous list of selection criteria? (i.e. size, SEER, manufacturer reputation, cost, how well it will heat at low temperatures (zone 5, in my case), and also minimum output?)
    Is "turndown" a related concept? Is it called turndown in the product literature?
    Thanks again,

  4. Kenny78 | | #4

    There might be a turn down ratio in the literature or it might state a minimum modulated output. You will want to size the equipment for the heat and cooling needs of the conditioned space. Read some of the experts post here about aggressive manual j calculations. Dana and others can point you to Internet resources to find accurate weather data for your locale. A modulating unit(inverter) will ramp down while still holding temperature until it hits the minimum output. Hopefully as Martin points out, it will never-more realistically minimize cycling. While it is modulating in cooling it will be dehumidifying.

  5. Expert Member
    Dana Dorsett | | #5

    The minimum modulated output at your average load conditions is the relevant factor. Since HSPF testing tests the minimum & maximum output at +47F as well as the COP @ 47F at the "nominal" or "rated" modulation level, it's usually specified in the submittal sheets. With Manual-J or other load tools it's useful to compare the load @ 47F with the minimum output @ 47F when sizing mini-splits for heating.

    For cooling load test SEER at the nominal rated output at 80F indoors, 95F outdoors, at a specified indoor & outdoor dew point. They usually test the min & max capacity under the same conditions as well, and it's usually listed in the submittal sheets. If your 1% outside design temp is only 83F and you like to keep it 75F indoors, the minimum modulated and maximum level will be higher than in the spec, but it requires a better mini-split model to infer just how low it can go.

    Lower minimum-modulation levels are better as a rule. Running continuously at low speed uses less power than duty-cycling, spinning up larger heavier compressors & blowers for several minutes several times per hour to achieve the same average output. This is true in both heating and cooling mode.

    Mini-splits will DEHUMIDIFY better when the compressor is running at a higher speed than at it's lowest-speed / highest-SEER mode, but it'll still be dehumidifying some. But it can only dehumidify when it's actually running, so when duty-cycling the average dehumidification will be pretty low, and some of the condensed moisture on the coil re-evaporates into the room during the idle/off cycle.

    Most mini-splits have a "dehumidify" or "dry" mode that optimizes for dehumidification even at low speed. In humid climates that's probably going to be the most optimal for comfort. It will have a lower than advertised as-used SEER in dry-mode than in normal cooling mode, but will be more comfortable at a higher set point than it would be in normal cooling mode. eg: 50% RH @ 80F is quite a bit more comfortable (and healthier) than 70% RH @ 75F.

  6. benrush | | #6

    Ken, Dana, thank you both.
    Dana, I'm having a little trouble following your first two paragraphs.
    In the first paragraph, are you pointing out that the published minimum modulated output of a certain mini split merely tells us how it performs at 47F, and therefore I should be more concerned with how it will perform in my specific climate?
    In your comment about Manual J, are you saying it's useful to compare a heating/cooling zone's load at 47F to a mini split's minimum output at 47F? If we're trying to size the equipment so that it runs all the time, I would have thought it would be much more important to consider the zone's load at, say... 60F and 80F, when heating and cooling (respectively), are barely needed.
    In your second paragraph, are you pointing out that SEER is tested in a lab under certain specific parameters, and that real world performance will be different?
    Again, I really appreciate your efforts to help me understand these points. Sorry for asking you to go over it again.

  7. davorradman | | #7

    Are there in general any minisplt brands that have better minimum modulated output than others, comparing models in the same power and/or price brackets?

    I am from Europe, so we don't have the same models on the market, but it's similar.

    Though, I don't see Toshiba mentioned here anywhere, and they seem to have the best rated min/max capacitiy.
    For example, the cheapest Toshiba I can find here says this for loads:
    TOSHIBA RAS-10BAV-E / RAS-10BKV-E (700$ installed cost)
    Cooling: 2,5(0,67-3,1)kW
    Heating: 3,2(0,7-3,9)kW

    These are the numbers at ~47f outdoor. Can these numbers be trusted?

    Problem is that in the datasheet, it is not stated at which indoor temps are those modulating ranges rated. I am guessing 70f.

    Can they be even lower, if temps are different? For example, if i were to set it as 65f indoor, will it go lower than 0.7 in heating, or if I set it to 80f in cooling, will it go lower than 0.67kw?

    Also, since modulating is so important, I find it strange that this cheapest entry level series (Mirai) has a seemingly better modulating range (it goes lower at 47f outdoor) than the more expensive tier (Suzumi Plus). This is why the numbers seem of too me.

    Not sure if you have any experience with Toshiba minisplits in the USA, but if anyone can have a look at the datasheet and help me make more sense of it.

    My heat/cooling loads are being calculated by a professional right now, but I have also done it myself and for my purposes (cooling, and heating in the shoulder season in the evening) sizing is not the issue here (almost certainly the weakest available model will be chosen, 9000btu).
    Just unclear on the modulating range.

  8. cldlhd | | #8

    I have the issue with short cycling causing dehumidification issues in my bedrooms. I have Mitsubishi mini splits, four units total. Three are in the bedrooms and one is in the main living area. The three bedroom units are connected to one outside unit. Even though the Mitsubishis are designed to go down to 2000 BTUs if connected to their own outside unit when they are connected to a multi head unit they are governed by what the outside unit can throttle down to. In my case that's 6,000 BTUs before it cuts out so with well insulated bedrooms the room usually cools down before the humidity is adequately removed

  9. Rykin | | #9

    It's actually worse than that. The 6000btu is only The lower limit for the compressor if only one indoor unit is running. If a second unit is running, the compressor kicks up even higher. I recently had this problem when I had a dual unit Mitsubishi installed. I thought the 6K would be divided amongst the two units, but my energy monitor showed that was not the case. I contacted Mitsubishi and got a hold of an engineer and they explain the technical reason as to why it functioned this way. In short, I have a 12K and a 9K unit on one outdoor condenser. The engineer told me that his best guess is that each unit would run at half speed minimally. That means the lowest output for the two units combined is about 10,000 BTUs.
    So, my units short cycle. On mild days, I can only run one unit to get adequate dehumidification.

    1. ianrking | | #10

      Yes, that's right - that the minimum compressor capacity is different depending on how many indoor units are running at once. Mits's recent data sheets are not at all consistent in how they report this. Data sheets for older models show a bit more information. A bit frustrating to say the least....

      1. cldlhd | | #11

        Also, I was wondering about this. If you have three units running inside at once, wouldn't it make a difference? If all three were calling for cooling versus one? I guess the answer to that is obvious but I started running my mini splits ( I have one outside compressor that powers the three bedroom units and one that just runs the mini split for the general living area) in All three bedrooms instead of just the two that are occupied. Hoping this would make a difference. Personally, even though I know it's probably less efficient and not great for the equipment, I just used the Cielo breeze controller in each room and programmed it to completely shut the unit off in the room. When set temp is reached. That prevents the unit from blowing all that humidity off of the coil back into the room.

        1. ianrking | | #12

          As an example, look at this service manual for Mits's MXZ-A generation of multi-splits. The minimum capacity of the outdoor unit is a function of how many indoor units are operating at any one time.


          1. PBP1 | | #13

            Thanks for the info, I looked through quickly but didn't see the portion on minimum capacity versus number of indoor units operating at any one time, might you have a page number? I do see/have known about the units/duct/no-duct charts.

            I have a "rightsized" Mitsubishi MXZ-3C30NAHZ2 with three ducted heads 9K, 12K and 15K, comfort has been good even at - 10 F. A cold month (ave 20 F) will be around 50 kWh/day; whereas, a shoulder heating month (ave 50 F) will be about 10 kWh/day and a low heating/cooling month (ave 60 F) about 5 kWh/day.

            Also, the OP from 2017 states: I understand that a “somewhat” oversized mini split will generally run at a more efficient rate, as opposed to an undersized mini split, which will run full-out, and therefore be less efficient.

            Not sure if that was unpacked as to comparing oversized and undersized variable speed compressor ASHP units.

            Oversized is likely to have a higher minimum output. Thus, for lower demands, it turns into an on/off system at its minimum output, which is inefficient (as if it wasn't a variable speed compressor unit).

            An undersized unit may have its lowest COP at max output, where, being undersized, it constantly tries to meet demand, however, it's not turning on/off (maybe more frequent defrost though). An oversized unit for that same demand may have a higher COP (as it's not at max output); however, as stated, that oversized unit, for low demands (more common than highest demand), will be an on/off unit - bad for efficiency and ASHP longevity. Also, probably bad for comfort too.

            Here's a link with a plot of a fixed speed compressor on/off unit versus a variable speed compressor unit (I believe this assumes a fixed outdoor temperature). Both curves meet (pretty close) at the 100% demand. Peak COP for the variable speed compressor unit is around 40% of max load (demand).

            Given that max demand is usually a few days a year and that lighter loads dominate the heating season (say 8 months from Oct-May), undersizing is better than oversizing. An undersized unit having a lower minimum output will be less of an on/off unit compared to an oversized unit having a higher minimum output.

            But that again turns to the question for Mitsubishi and minimum being related to how many heads are on at one time. If you can turn one head (or more) off during light demand, is that beneficial?


  10. ianrking | | #14

    @PBP1. Look at starting on page 7. For different combinations of indoor units the specs show a different minimum capacity. The example I've attached 2C20. For a single operating* unit the min capacity is 5,400btu/hr. For two operating units, 7,800btu/hr.

    *The 2C20 cannot have fewer than two units attached to by inference, the chart must mean to communicate that capacities when a # of units is operating - not attached.

    1. PBP1 | | #15

      Thanks, yes, I knew about that. But I'm still unsure if turning one of three units off changes the minimum. For example, consider three units with a minimum of 9000 and switching one thermostat/controller to "off", would that result in a lower minimum (< 9000)? Or, is the outdoor unit programmed to know it has three units regardless if one or more are turned off such that the minimum remains at 9000? Also, in my set-up, the units do "run" individually at times. For example, one unit has satisfied its zone and is not "running" while another unit is "running" to satisfy its zone. Seems like the minimum in the chart (9+9) is for the case when all (both) units are "running". If all units are running at the same time, then whole house demand is generally substantial and the elevated minimum not too much of an issue (which may be compared to the sum of the minimums of three individual units in my case).

      1. ianrking | | #16

        I think it's the former (i.e. "consider three units with a minimum of 9000 and switching one thermostat/controller to "off", would that result in a lower minimum (< 9000)") because based on the MXZ-2C20 situation, you can never have fewer than 2 units so the lower minimum attributed to a single unit can only be for the case when the other unit is turned off.

        I also think the minimum changes when one of the units is cycling i.e. fan is on but not calling for cooling. I have two units attached to an MXZ-2C20 and I notice that when both units are running and one stops calling for cooling, the other one goes a bit wonky for a bit until compressor can ramp up/down appropriately.

        1. PBP1 | | #17

          Thanks, I'll try to pay more attention to behavior, unfortunately my Efergy power meter is on my non-ASHP circuits, maybe I'll get another for the ASHP and do some experiments.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |