Mitsubishi heat pump defrost cycles
Ducted Mitsubishi Smart Multi (MXZ) heat pump and both the power usage and the duct temperature in supply next to blower drops for 15-20 minutes with each defrost cycle. The behavior is the same at any temperature from -10F to high 30s F outdoor temperature. Fan on high or low I also see the exact same behavior and same also when defrosting every 40 minutes or every 150 min.
“Officially” the defrost cycle is a couple minutes but reality is 20+ minutes of no to low heat. Everything I have read from Mitsubishi and other companies with respect to defrost is that it should bounce back to pre-defrost output quickly.
My local DSG reps and Mitsubishi tier 2 support who have maintenance tool data tell me everything looks normal with the unit.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
That doesn't quite look like defrost. During defrost you would see the power first drop to 0, back to near full and back down to 0.
That looks more like cycling.
I agree that it looks like cycling but it is 100% a defrost cycle followed by 10 minutes at 48Hz. Others with Mitsubishi Smart Multi units have shared similar power cycling behavior. Here is the compressor frequency change chart during two defrost cycles from mtool data.
Ok. It looks like defrost than ramp up to low speed to reheat coil and see what the room load is. After the 10m at 42hz, it starts its modulating bit.
The defrost cycle is only the bit where the inverter shuts down/restart/shut down. It is not that long.
I guess they could have done a bit better controls and go back to the operating point before defrost instead of this.
No clue about their design strategy but the end result is a nearly 40-50% heat output capacity loss on high humidity days with frequent defrost compared to low humidity days.
I don't understand how this is normal as their design book indicates a 5% defrost capacity loss at temps under 25 or 20F.
These do not seem like they are designed for New England with lots of cold and high humidity days.
For those 10min the unit is still running but at about 25% output. Most high humidity days around me happen near freezing, so the unit will already be running at low modulation.
There are also a bunch of jumper settings for the outdoor unit. You can look through the install manual here and see if there are adjustments there you can make to improve.
https://mylinkdrive.com/USA/M_Series/R410A_Systems-3/Outdoor_Equipment/R410A_Outdoor/MXZ_SM36NAMHZ2_U1?product&categoryName=R410A_Outdoor
Thank you very much for the link and the comments/ideas. I will take a look although believe the run/defrost related jumpers are all set at the default or shipped settings.
Just very perplexed as to why it would run at 25% for 10 minutes after every defrost even when super cold. That behavior doesn't match their application notes on mxz defrost behavior, other power usage defrost cycles I have seen from similar zuba or pva unit owners, or one of the advertised benefits of hyper heat being fast defrost recovery.
Hard to imagine that this would be by design or if so I am having a hard time understanding why it does this.
Maybe this is a programing issue or sign of another problem?
Where are you getting those graphs? How are you getting all of this data? I have two new Mitsubishi installs that I'd love to see how they are performing, even though they seem to be providing extremely good comfort.
I really believe these units should be providing real time info on energy usage and performance but they don't.
I installed an emporia vue energy meter to get per min resolution (or higher for past few hours). I also installed a high resolution thermometer in supply duct near blower to get high resolution temp and airflow and can calculate delta T and cop. It is data logging but not web enabled so takes a little work. Also have outdoor temperature/humidity weather station. Mitsubishi logs or at least measures almost all of this information but doesn't make it accessible.
I don’t like the shape of those graphs. What I see is a compressor running pretty much at the 2/3 or more of capacity with ever increasing pressure and with very little to no modulation. Something. Is off methinks.
On defrost cycles- in humid weather around freezing a defrost every 45 mins is about right. At single digits outdoor your defrost should be at least roughly every hour and a half of running.
But as I said- the overall graph is off.. you should see a ramp up, followed by a step down to a flat portion and then downslope once the set point is near until it shuts off. Do you run your indoor fan on "High"? Maybe try "Auto" just to see if the compressor will modulate better. It is not necessary more efficient on "Auto", but at least you will know system can modulate.
The pressure spikes before defrost are typical and correct though. Once the outdoor exchanger plugs up with ice pressure spikes and defrost is initiated. So defrosts look alright. I am surprised of the slow ramp up after the defrost though.
Overall, defrost looks ok, frequency at lower temps should be less.
Thanks, this is a Smart multi 4 ton MXZ-SM48NAMHZ so it is a large City system. rated at 54,000 btu. I am in the middle of CT and design temp around 10F, give or take. My plots are running full out at 5, -8F and the others are showing some modulation.
With this ducted unit the minimum target supply air temperature seems to be 100F, giving close to 36,000 btu minimum output at any temperature. (32F delta T * 1050 CFM min * 1.08) ( The issue is that if defrosting every 45 minutes to an hour the unit can no longer make 36,000 btu/hr as the defrost or defrost recovery cycles are 20 minutes on off or low). This leads to high peak power usage cycling at any temperature when defrosting.
Longer run times in low humidity conditions I do so the unit modulate some (assuming I am in the narrow range of allowable modulation which is give or take only when above 36,000 btu and below rated output of 54K ).
Do you have a power graph at say 40f? This is when direct vapor injection is not active so you should see very low (relatively speaking) consumption and good modulation...
Here are a few hours with temps in the low 40s showing power usage. Have some actuall run data from mtool on these days and the compressor ramps up around 50 Hz.
As far as I can tell the reason it doesn't modulate lower is that the supply air target is 100F with a minimum delta T over 30. Heat output at minimum is significantly over my load.
Those graphs at warmer outdoor temps look alright... There is some modulation taking place and power consumption is in line with my expectations.
This must be a larger home as this behavior is typical for a slightly undersized systems- they will work alright at higher outdoor temps, but run without much modulation (and higher than normal energy consumption) when the direct vapor injection cycle operates below 30-40ish outdoor temp.
Beyond the speculation above - Ask the tech to check if there is a firmware upgrade for the outdoor and indoor units. Reduce indoor flow by running fan on "Low" or "Auto" and see if the system behaves better.
I believe all is good actually with the unit if it is slightly undersized and given a large heat load... if you dont count the slow defrost recovery which might be an older firmware on the outdoor unit.
Thanks and appreciate the speculation as I think you are mostly on point. Overall yes slightly undersized heating a larger space and runs without much modulation and higher than normal energy use. With fast defrost cycles I would expect much more modulation under 35F but haven't seen that due to lots of time recovering from defrost. Low humidity days seem decent.
The slow defrost recovery is crushing because it generally runs with shorter heat output cycles at any temperature. Above 35-40F it modulates on and off and at lower temperatures it is trying to recover from long defrosts running on low followed by trying to recover on high - then going into defrost again.
The other day with frequent defrost at high single digit temperatures I calculated a heat output of about 28,000 btu/hr averaged over 8 hours. That is below heating load (and a good 50% reduction (or more) in capacity from I can get when in the low 40s with output on high) and the house gets cold.
I agree that this looks to be something where a software or firmware fix could make a huge difference, ideally a couple years ago.
During the week with Mitsubishi monitoring data last year it defrosted some 140 times.
I Have to ask is this unit connected to the kumo cloud?
I have yet to see a power graph from anything kumo cloud that looks like it is modulating to my eyes.
Walta
No Kumo cloud. Just a standard mhk2 thermostat. I wish the thermostats output power usage and cop and all the data that enabled the ability to assess performance. Should be required.
I am also in CT, on the water, but I have so many questions. My new Mitsubishi installs seem to be working great, at least from a comfort level. We've never had such even heating and cooling. But how do I know if they're working properly, defrosting properly, cycling properly? All I have is seat of the pants comfort and bills from Eversource.
I do have the KUMO cloud, but that app seems very basic. Is there a way to drill down on that thing to see performance characteristics? Your analysis makes my eyes glaze over....I'm just a homeowner, not an HVAC specialist.
Or should I just forget about it? Seems like Mitsubishi should have some sort of way for us to see if their equipment is working as designed.
Thanks
Nick in Old Saybrook
Mitsubishi does not monitor your installation. If you dont contact the installers they dont know if your setup works or is dead. They are not proactive so to speak.
Buy yourself an Emporia Vue or similar and install (or pay somebody to install) it in your load center. Observe the amount and shape of the energy consumptions. Educate yourself. If you have questions- post here- there are a couple folks that will "decipher" your graphs to a point where at least you will know if you might have a potential problem or all is good.
Not to mention that knowing your energy consumption under different scenarios (weather, fan speed, setpoint, etc.) will help you with optimizing your comfort/ cost of operation.
I agree on the emporia Vue and a weather station as the places to start. Wish it was easier.
Here is what performance/energy usage looks like in the mid 20s with light snow or high humidity. 7 defrosts in 4 hours and output low enough each hour that it cannot recover a degree if it tried.
Just 20 minutes with power draw over 2kW then 13 minutes of defrost/defrost recovery. Only 14minutes with draw over 6kW.
Low humidity steady state would be about 3.5kW according to past data.
Ask tech to check firmware on outdoor unit…
Deleted
Look familiar? I am seeing this chart in my sleep now. I thought it was cycling but then I caught it in the act of defrosting. SUZ-KA18NAHZ. MHK2 thermostat. Western PA.
Hasn't anyone been able to jailbreak Mitsubishi's communication? The only data I have are electrical power (IoTaWatt), outside and inside temps, outside humidity.
That is interesting and similar, what is it doing after defrost for 5ish minutes on low? Does it do that after every defrost cycle?
There was a post here the other day about someone using a esphome microcontroller to try and monitor their system, maybe something there. https://www.greenbuildingadvisor.com/question/too-frequent-defrost-cycle-whats-the-normal-ratio-of-heatdefrost-when-its-this-cold#comment-285349?