Hello David. I recconize your user name as being on houzz for some time so I'm optomistic you will be letting us know what works and what doesn't as you attempt to remedy this. Some might ask what difference that make's. Only serious gearheads know how benificial it is and keyboard warriors don't understand or care.

Reduced air flow can cause discharge to be a few degrees cooler than required/ desired. At start up,extra cold coil condenses alot of humidity which shortly ice's over coil. Metering device modulation can open letting air across coil to thaw ice then cycle repeats.

I hear you regarding "The evaporator air filters and the condenser coils are clean. Nothing is obstructing either" but I'd like you to look again at evap fan blade and BOTH SIDES of evap coil plus look through coil while holding a light on oppisite side. M-split filteration is on the sorry side and alow debris getting to coil. If that were not bad enough,partially due to desighn and partly due to size, m-split evap coils are difficult to access,clean and service. M-splits are super sensitive to balance so take you time and make sure you get as much flow as possible. It go's without saying but stay away from acid cleaners or anything else that must be completly washed off because you will have to rely on condensate rinsing some residual cleaner when put back in operation.

Austin, thanks. Couldn't a leak be just a connection opening up in the lineset?

I think (hope) Mitsubishi is a reputable company. If it does need something like a reversing valve or a coil, I expect (hope) they will stand behind it. Cause if not, it's a pretty expensive disposable plastic toy.

Klem, thanks for the suggestion. I'll try that and report back. Cross your fingers.

You are correct, I've been on and off Gardenweb since around 2003, on when I have some free time, off when my life gets busy. It's kind of nice when long time friends recognize you. :)

Austin, thanks. Couldn't a leak be just a connection opening up in the lineset?

a flare nut leak. sure it's possible, but I doubt that. Usually if it's gonna be a flare nut leak it will happen much sooner than 8 years in.

This isn't necessarily about reputable company as the warranty that came with the unit is all they are on the hook for. In that time frame warranty was probably 5 yr parts, maybe 7 yrs on compressor.

Now the warranty is similar unless installed by Diamond contractor then you can get 12 year parts. I don't think there was this difference in warranty coverage in 2013 when this unit was installed.

I suppose it's also possible what Klem suggested, just highly unlikely. Usually if problems are associated with cleanliness you don't have to dig for it. Given that you said it's clean, I take that at face value in a forum board setting and you identified that without being asked.

With all that said Mitsubishi is a larger mfg. so if you're still going to find parts for this particular model your chances are marginally better than with an opposing mini split manufacturer. There's a minor more strength in "numbers". However, distributors aren't known for keeping parts they may never get rid of... so there's much more to it than meets the eye.

I cleaned the slightly dusty evaporator coil with an old paintbrush and vacuum cleaner. When I restarted the unit it went down to an outlet temp of about 46 degrees and stayed there for over an hour. A few hours later the room temperature was down to 71 degrees and the humidity was 41%. Success, or so it seemed.

However, in the heat of the day today (82 degrees) it's back to very short cycles, even shorter than before. Outlet temp falls to 41 degrees, then instead of holding for 1-4 minutes as before, it immediately shuts down the compressor until the outlet temperature gets up to around 65 degrees. Room temperature on these short cycles has risen to 78 degrees.

I wonder if the outlet temp falling to 41 degrees is significant, because the unit ran for a long time when it wasn't going much below 46. I don't know what would cause it to try to run colder and give up, though.

Maybe it's worth a more thorough cleaning of the evaporator coil with some kind of mild liquid cleaner - suggestions?

If that's a no go, then I guess it'll be time to call in an HVAC tech and hope it can be fixed. I don't have the training or the tools to mess with refrigerant on my own.

I hadn't give thought that you might think I was infering dishonest about cleanleness until A Air commented on taking people at their word. Truth is I've lost count of how many times contractors and employees told me a coil was clean when it wasn't. Their warn't lying,they simply didn't see dirt until I showed them how to make certain. I did the same until wiser men showed me different. I still haven't mastered how to perdict what third parties ( mfgr,dealer and distributor in this case) will do. Hopfully you don't need to go that far so let's do what we can on our own before finding out.

I just believe this is well worth persuing since it cost's nothing and will do no harm. I can't say the same about tapping into sealed systems. Manys a perfectly charged system has been contaminated,caused to begin leaking or worse when needlessly opened up. As a rule of thumb,as compressors get smaller, sensitivity increases across the board and skill to deal with it become's increasingly neccessary.

If not useful on this unit,I have some tips for confined coils. Adapt as the need arises. I carried mirrors including one similar to those dentists and hygeniests use. It's sometimes amazing to see what's around corners. Much quicker than dissassembly only to find everything ok. A garden sprayer will handle most coil wash jobs but in addition I carried pressurized containers with different coil washes inside. I often prefer pushing dirt back out in direction it came from. A 1/8" soft copper pipe with spray head can be snaked into back side of coil and using high pressure liquid or plain air ( depending on whether pressure tank is upright or inverted) can drive dirt out to side I can see. Eventhough you can't see wand and spray head,the spray can be directed by watching dirt come out side which is visiable. I figure this can save tech time and do better work when dealing with M-splits. If a wet vac hose with 90* floor tool can be manipulated to catch dirt and liquid as it's expelled,you can do 1st class on central system coils. Bigger spray head,more water and more pressure.

I would use no rinse evap coil cleaner on this ms but it usually only come's in gallons. I have no objection to diswashing or laundry detergent if that's all one has.

If you have a digital thermometer that reacts within secounds of temp change, and an amp meter,do this. Watch and record running temp and amps. Do amps increase above rated run amps before compressor cut's off, do temps progressivly change as amps change and other readings. Since I've lead us this deep,I should at least say why I think this could be air flow related. A freeze up caused by low refrigerant can be expected to start iceing at point where it enter's coil then progress until air flow begin's to drop then continue progressivly covering coil in turn progressivly restricting air flow. That's a clean coil with inadiquate refrigerant. A properly charged coil drop's temp rather even across whole coil. A properly (and to a degree, inadiquate) charged coil with inadiquate air flow from dirt restriction or too low fan can be expected to frost then shortly ice across intire coil surface. Anywho,this will be one less thing tech need's to do if one must be called. Your amp-temp chart might also prove helpful.

On a mini split I doubt it would run long enough to generate ice maybe frost, because there's low pressure controls the OP has already confirmed in saying the compressor cycles off after so long.

Mini split is also considered a critically charged appliance, so essentially it would only take a few ounces of refrigerant to cause problems one way or the other. (over or under charged)

8 years old... easily within the time frame of developing freon leaks. If I've seen it once, seen it a million times. ( playing the odds based on info presented. A guess as it were.)

True that I've been told condensers were clean when they weren't... but this is a forum board until proven otherwise.

• Dirty filter. This is one that you can handle on your own. If the filter in a forced air HVAC system is too dirty, it can create so much airflow resistance that the system overheats and shuts down.

• Faulty capacitor. Air conditioners and heat pumps use capacitors — start and run capacitors, to be more specific. The run capacitor is there to maintain a charge, and if it is faulty then the system won’t be able to run continually as needed.

• Refrigerant leaks. Again specific to heat pumps and air conditioners is the matter of refrigerant leaks. If there is not a sufficient amount of refrigerant in a system due to leaks, then serious damages could develop. Short cycling may be an early warning sign of such problems.

Klem, give yourself 3 pats on the back. Maybe more, because you nailed it.

After what you wrote I dug deep into the indoor unit here. I found the squirrel cage (or chipmunk chorus line cage, since it's small in diameter and runs almost the full width of the unit) caked with dirt. I should have taken a picture, but put it this way, when I pulled it out big chunks of black grime fell out all over me and the floor.

I thought that the filters were keeping that gunk out! The evaporator coil was only a bit dusty so I figured the filters were doing their job. I can't figure out how the blower could get so filthy when the coil wasn't.

One possible contributing factor might be that the HVAC tech who I had do PM on this unit in (I think) 2017 may not have bothered to clean the chipmunk cage.

The blower took lots of scrubbing. On your advice I also cleaned the coil, using just plain water in a spray bottle, a brush, and a shopvac to pull the dirty water out. Rinsed out the condensation pan, put it back together, and now, guess what.

Now it's too cold in here! But I'm not really complaining. :)

As a bonus, the unit is much quieter now.

Thanks for staying on top of this. Thanks also to Austin for the warnings about parts availability. That's pretty scary. You have me wondering where I might look for a similar used or junked unit that I could strip down for spare parts.

Aaw shucks David,tweren't nothing any keyboard warrior wouldn't have done for you.

As a bonus, the unit is much quieter now.

Thanks for staying on top of this. Thanks also to Austin for the warnings about parts availability. That's pretty scary. You have me wondering where I might look for a similar used or junked unit that I could strip down for spare parts.

Interesting resolution for sure. Hard for me to imagine how that wheel got so clogged up. I guess mold growth is likely how that occurred. Odd really, and quite rare for a unit that old just to be cleanliness issue.

The spare parts using a junked unit would take a lot of luck to find and 9 times out of 10 the junked unit you find won't have a working part that you will need.

Let us know that you make it thru the summer with this unit without further trouble. Curious to see where this goes...

Hi Austin. Am I the only one who thinks that 8 years is not very old?

I expect a gas furnace to last at least 30 years, and an air conditioner or heat pump to last at least 20.

In 2 of my rentals I have Trane gas furnaces and air conditioners from the mid 1990s. Another one has a Comfortmaker (no air conditioner) from about the same era. They are still working well despite the minimal maintenance that rental units get. Just the usual occasional repairs, blower motors, draft inducers, one control board, an igniter, a condenser fan.

Another rental, I don't remember the furnace brand, real old school, standing pilot, natural induction. Again it had had the usual repairs -- blower motor (3 times in 30 years), gas valve, limit switch, thermocouple. I would have kept it longer, but the tenants kept complaining that it was too noisy, and wanted air conditioning. Good tenants, pay on time, so they got what they asked for. Also a rent increase to pay for it.

I even have a cheap Frigidaire window air conditioner from 1998. Still works as well as new, which isn't saying all that much.

Maybe I'm just unusually lucky.

And I suspect that you are right, that was probably a combination of dust and mold on the split's chipmunk cage.

I looked on the big auction site expecting to see lots of listings for used coils and such for mini-splits. Did not see any. There are thousands of listings of used parts for old furnaces, often at asking prices barely less than new parts. If many mini-splits die in 7-10 years, where are they all going when HVAC dealers haul them away? Right to the scrap yards? Crazy.

Hi Austin. Am I the only one who thinks that 8 years is not very old?

Equipment of previous years was made really well. It wasn't uncommon for it to last 20 years with minimal fuss.

So what changed? Plenty. Freon became the culprit of ozone depletion. So government phased it out. As more and more people bought air conditioners, the load on the grid in some areas posed problems of brown outs and rolling blackouts. The equipment of old was modified to become more energy efficient.

The cost of metals used to make the machines and various parts became thinner and smaller to cut costs, but due to efficiency reasons the physical size of the machines grew larger compared to older less efficient machines.

Compressors became more advanced, analog switches were replaced in favor of solid state or digital varieties (communicating) thus furthering the advance of energy efficiency.

New refrigerant was developed to combat the ozone depletion problem. Years later it's been decided that this new refrigerant is contributing widely to global warming and is now being planned to be phased out again in a few short years. (Jan. 2023 -- California finds out first).

The new, new refrigerants aren't much better than the one it is replacing in terms of global warming, plus they have a new problem: "mildly flammable".

8 years isn't that long: Once we could buy a phone that would last 20 years cost $20. It hung on the wall in the kitchen. Now we walk around with computers in our pockets costs hundreds if not a thousand or more. We replace these gadgets regularly as new models come out at an ever faster clip. On average every 2 years or so.

Autos: about 6 years and the value of a car from when it was new is a smidgen of it's former value.

8 years doesn't seem that long but in comparison to other things we spend our money on and how frequently we do it without second thought. In the mini split HVAC realm this is more complicated than regular forced air ducted.

Why? because that cable from the out door unit goes to the indoor heads. That cable is proprietary. The manufacturer controls the life of that system by simply moving on to a new model. You can't buy something that doesn't exist.

You can't mix and match different brands of mini splits like you can with regular forced air ducted. (Of course to do this, you have to know what you are doing and there are tradeoffs for doing so...)

For further explantion here's some visual content for you:

This one has to do with differences in materials. Thinner copper walls and progression to all aluminum evaporator coils.

There is an old Lennox in this one. Which is quite rare now as to age of equipment being replaced. Typically between 10-15 years old is the norm, some less than that.

Most problems you can name with daily life is rooted in over population or lawyers. When both forces converge,society is split on who to blame so they just blame one another. There are as many people who would prefer you go without air conditioning as there are that would have you own air conditioning that last's a life time. You can't please them all.