minisplit: big name or DIY-friendly ?

  • Active since 1995, Hearth.com is THE place on the internet for free information and advice about wood stoves, pellet stoves and other energy saving equipment.

    We strive to provide opinions, articles, discussions and history related to Hearth Products and in a more general sense, energy issues.

    We promote the EFFICIENT, RESPONSIBLE, CLEAN and SAFE use of all fuels, whether renewable or fossil.
I don't think a Nest or an Ecobee (or any other 3rd-party thermostat) can be integrated with a minisplit.
You may be right on that. I think many use proprietary communication between the wall unit and whatever control unit is used.
 
So I've hit the jackpot. A friend who does minisplit installs locally (not up in the mountains, where I'm interested in installing a minisplit to replace the electric-resistance heat) offered to purchase a big-name brand unit for me from a local distributor. So, full warranty coverage. Less than $100 above the online mail-order price.

In exchange for having an un-paid helper, he's allowing me to tag along on some of his installs, and hopefully gain some knowledge and experience. He's crazy about some push-to-connect connectors for refrigeration (like plumbing Sharkbite connectors), not because he's afraid of flaring, but because of the rotational freedom they allow at the connection. So no worries about doing flares (although I've done a few on gas lines, with no issues). And there's a small-time HVAC guy up in the mountains, who'll come do the freon work for labor plus freon.

This is the unit I'm installing: https://ashp.neep.org/#!/product/51107/7/25000///0
The thing kicks tail down to subzero outside temps, and turns down to less than a hair dryer at 47 degrees. Maximum indoor/outdoor sound levels are 45/50 dB respectively. Needs a 20amp circuit.

I'm psyched.
 
Great, looks hard to beat. Look forward to your install reports.
 
So, it's been awhile, but I finally got my Fujitsu minisplit running. One reason it went so slow is that it's at a vacation house, 3+hrs from home, so I only get a chance to work on it a day or two once or twice a month. The other reason is that it took me awhile to convince myself that there were no refrigerant leaks. Using the equipment that my friend lent me, I pressurized this system to 450psi using nitrogen. It seemed to hold fine for several hours, but had then dropped 25psi or so by the next morning; temperature drop could not explain this much loss. But then I came back a month or so later, and it still had 350psi; we reasoned that if there were a leak, it would not have any pressure at all after a month. I re-pressurized it and the exact same thing happened; no discernible loss after 4hrs, but a good bit by the next morning. I finally decided it was the pressure gauge assembly that was the culprit. Called in a local guy to vacuum-pump and commission the system.

First real test was disappointing. It didn't seem to be blowing much hot air. If I hit the "POWERFUL" button on the remote, it blew hot for a few minutes (re-assuring that the thing was at least capable of heating) but then started blowing cool. Consultation with friend revealed that the unit senses room temperature at the air intake (the top of the indoor unit); for some reason I thought it sensed it at the remote, like some other brands do. Friend suggested two things: run the room's ceiling fan (blowing upwards, of course) and adjust the output louvers to make sure none of the hot air was being blown down onto the big piece of furniture just below the indoor unit. The notion was that I was getting "short cycling" due to hotter (than the room as a whole) air getting to the sensor at the intake. It seemed to work. The next night it worked great - plenty of heat, and not so much variation in the temperature and quantity of air output.

So I'm very happy. (Though there's a nagging fear that there's still a VERY slow freon leak - slow enough that it only went down from 450 to 350psi in a month back when I was testing it). Trying to decide if it's worth putting a second one in, into the MBR (I put in a small subpanel to feed the first one). If the house were continuously occupied, it'd be a no-brainer; but as a vacation and occasional rental, maybe not so much.
 
By spring you should know if you have refrigerant left!
 
  • Haha
Reactions: RustyShackleford
So, it's been awhile, but I finally got my Fujitsu minisplit running. One reason it went so slow is that it's at a vacation house, 3+hrs from home, so I only get a chance to work on it a day or two once or twice a month. The other reason is that it took me awhile to convince myself that there were no refrigerant leaks. Using the equipment that my friend lent me, I pressurized this system to 450psi using nitrogen. It seemed to hold fine for several hours, but had then dropped 25psi or so by the next morning; temperature drop could not explain this much loss. But then I came back a month or so later, and it still had 350psi; we reasoned that if there were a leak, it would not have any pressure at all after a month. I re-pressurized it and the exact same thing happened; no discernible loss after 4hrs, but a good bit by the next morning. I finally decided it was the pressure gauge assembly that was the culprit. Called in a local guy to vacuum-pump and commission the system.

First real test was disappointing. It didn't seem to be blowing much hot air. If I hit the "POWERFUL" button on the remote, it blew hot for a few minutes (re-assuring that the thing was at least capable of heating) but then started blowing cool. Consultation with friend revealed that the unit senses room temperature at the air intake (the top of the indoor unit); for some reason I thought it sensed it at the remote, like some other brands do. Friend suggested two things: run the room's ceiling fan (blowing upwards, of course) and adjust the output louvers to make sure none of the hot air was being blown down onto the big piece of furniture just below the indoor unit. The notion was that I was getting "short cycling" due to hotter (than the room as a whole) air getting to the sensor at the intake. It seemed to work. The next night it worked great - plenty of heat, and not so much variation in the temperature and quantity of air output.

So I'm very happy. (Though there's a nagging fear that there's still a VERY slow freon leak - slow enough that it only went down from 450 to 350psi in a month back when I was testing it). Trying to decide if it's worth putting a second one in, into the MBR (I put in a small subpanel to feed the first one). If the house were continuously occupied, it'd be a no-brainer; but as a vacation and occasional rental, maybe not so much.
Any updates on the minisplit?
 
Any updates on the minisplit?
About a year in, it's working great.

Not sure if I mentioned this above, but you know how wood heat is so comfortable because of radiant heat transfer ? Well, it also works in reverse. So if the mountain house has been sitting with the inside temps in the 40s and 50s, the minisplit can heat up the air to what should be a comfortable temperature. But it's not. Because all that cold mass, especially the stone wall around the fireplace, is radiating cold to the people (or we're radiating heat to it). The thing that clued me in is that after being up there for a day or two, it's quite comfortable.

So I installed a wifi adapter, and I turn it on a day before we come up. Even two days is good.
 
  • Like
Reactions: Zack R
Mini-splits are not great at correcting big temp swings such as going from 50 to 70*. But then again, I don't imagine many types of heat besides wood or pellet stoves can pump out the type of heat needed to warm all the inanimate objects at the same time. Maybe a FHW system could at least make it seem more comfortable fairly quickly since the heat is usually along the length of exterior walls (as opposed to a small vents).

If I lower my Fujitsu system down by several degrees, I'll let my pellet stove get it back up near where I want it, then shut it down and start the MS.
 
Definitely a "rule" for minisplits is set them and forget them at one temp unless its a seasonal place with some advance notice for occupancy. Quick response can be had for a price but few people are willing to put in extra units that only may run 5 or 10 percent of the time. Radiant heating and cooling factors in, when the surrounding surfaces are either too warm or too cold they will cause the occupants to feel too hot or too cold even though the air temps are set normally. Few are willing to wait, so the "solution" is crank up the heat to try to get the surfaces up to proper temp. This works as long as there is extra capacity.
 
Yep, sounds right. Seems wasteful, but from now on,I'm gonna start pre-heating (via the wifi adapter) 48 hours in advance.
 
Ideally you would have temperature sensors directly coupled to walls and floors and an outdoor temp sensor. An outdoor reset function would adjust the time to turn on the heat pump on the past response speed of the building based on outside temperature.
 
  • Like
Reactions: RustyShackleford
Ideally you would have temperature sensors directly coupled to walls and floors and an outdoor temp sensor. An outdoor reset function would adjust the time to turn on the heat pump on the past response speed of the building based on outside temperature.
The wifi adapter does tell me the outdoor and indoor temperature (I guess it gets the former off the internets) so I can kinda do that myself.
 
The only trick would be to insert a temperature sensor in the face of a "cold" wall which will lag the interior air temp. During a heating cycle, the walls rate of temperature change would lag the interior air temp due to the thermal lag. There will still be steady state conduction loss through the wall but the rate of the difference between interior air temp and the wall temp would at some point decline and flat line which would show that the space is warmed up. I used to be the engineer who told the controls engineers to figure out the details on how to pull this off ;)
 
  • Like
Reactions: RustyShackleford
At the end of the day I don’t think it’s that waistfull. You are heating the same mass to be same temperature. You would have increased thermal losses so the higher temp for 48 hours but that’s probably small compared to the heat needed to rais the temp of the house and all its contents.

I do think the single head unit heatpumps probably suffer more from “cold walls” than a ducted system where every room has warm air circulating. I also think our perceptions are a learned experience. I know in the mornings no to sit in a cold leather chair without a blanket on it first. I know where are the “hot” spots are and avoid all the cold ones.
 
  • Like
Reactions: RustyShackleford