DIY minisplit in 2023
DCcontrarian
| Posted in Green Products and Materials on
Someone just bumped this thread from 2018 on DIY minisplits:
https://www.greenbuildingadvisor.com/question/diy-mini-split-advice
so I thought it might be good to start a new thread to see what’s out there now and what’s changed.
So what is the state of the art in DIY right now? Is it still Mr. Cool?
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Search and download construction details
Replies
I'm not necessarily knowledgeable about the "state of the art," but I can tell you that I've considered some DIY options recently, and this is what I've come up with:
1) Mitsubishi/Fujitsu. Still the best, and by a fairly wide margin, at least in terms of reported efficiency and Mitsubishi especially for it's turn down ratios. The 6k and 9k units can modulate way down to around 1500 btu/hr. But if you DIY these you need a vacuum pump, a gauge set and various hoses, a regulator and nitrogen tank rental, a flare tool, etc etc. That's going to add around $1000 to the cost.
2) For pre-charged quick-connect linesets, I concluded that MrCool is still the best. But significant downsides include a much smaller turn-down ratio that Mitsubishi and significantly worse efficiency. Another downside with the quick-connect linesets is that you have fixed lengths, i.e., 15 ft, 25 ft, etc., and if your lineset run is not that exact length, you have to coil and store your excess lineset. Advantages include that the only tool you really need to install (besides a drill and a hole saw) is a torque wrench.
3) Another option is something like the Ephoca AIO. Much much easier install -- there are no linesets at all. I would also think that risk of refrigerant leak is less since this is a sealed system from the factory, with no lineset, so it never gets unsealed during installation.
It does require two 6" exterior wall holes. That's a big downside compared to the 3" or so hole required for a mini split lineset. Supposedly the Ephoca unit is pretty well sound and air insulated, so you shouldn't get a ton of noise infiltration or outdoor air infiltration through the holes... but needless to say you will presumably get more of those things with two 6" holes, no matter how well sealed the unit is, than you would get without the two 6" holes.
Another big downside is price. $4,000 for their most basic 120v unit. Much more expensive if you get the unit with the built in ERV. Also, pretty inefficient compared to most mini splits, and although it works at low outdoor temps, it loses a lot of capacity compared to a low-temp mini split.
4) Window unit heat pumps like the Gradient Comfort and the upcoming Midea. Honestly, for a DIY option, this is probably what I would go with IF I had standard up-down double hung windows in my house. Alas, I have horizontal sliders, so this is not even an option for me.
pnw_guy,
Here is a real Heath-Robinson style idea. What would stop you cutting a slot in an exterior wall the size of the piece that joins the two parts of the Gradiant Comfort?
Hah, thanks for the suggestion. It hadn't crossed my mind but I suppose it is a possibility...
For the $5k price of the Gradient I will pass when it requires me to cut up my wall (as opposed to merely sliding open a window). But if the Midea installs the same way and is significantly cheaper, I might actually consider it. At the moment I simply can't stomach the idea of paying $5k for the relatively inefficient, poor cold weather performance of the Gradient when I could get a top-end Mitsubishi profesionally installed for $5k including labor.
Why is it so hard for the DIY systems to have comparably turn down capability?
For well insulated, air tight houses - it seems like this would be an essential feature…
I don't necessarily think it's "hard" for them to have good turn-down. I think that for budget and optimization reasons, the manufacturerers of those units just don't do what Mitsubishi, for example, does to get such a low minimum capacity.
My opinion of the DIY units with zero loss lineset couplings hasn't changed. Still a lot of money for mediocre equipment, and the challenge of fixed length linesets. Better to buy the equipment you need/want and tool up to install it yourself or hire out connecting the lineset and evac IMO.
One new development I did find was the contents of this DIY kit. It's a way to purge, and pressure test, the lineset without intentionally releasing refrigerant. Basically same method as I've seen in some factory manuals but without using refrigerant from the condenser which is verboten in the US.
https://www.pioneerminisplit.com/products/wys009a-20-diy-bundle
The Pioneer lineset flushing stuff looks interesting, it seems like a very DIY-friendly way of getting the moisture out of the lineset during installation. There's lots of stuff on YouTube about using a vacuum pump to get the moisture out, but that requires about $200 worth of tools that most people won't have.
It seems like any unit that is pre-charged with refrigerant becomes DIY if you can get the lineset clean.
It's amazing how cheap these systems have become. Just looking on Amazon I saw this one, 12,000 BTU for under $700:
https://www.amazon.com/Conditioner-Ductless-Inverter-208-230V-Efficient/dp/B08WCTRSW4?th=1
"It's amazing how cheap these systems have become. Just looking on Amazon I saw this one, 12,000 BTU for under $700:"
Really makes the fact that Gradient Comfort is charging $5000, and Ephoca is charging $4000, all the more mind blowing. Gradient, especially, was pitched as a cheap alternative to a ductless mini split.
$5000 vs. $700? Who's the cheap alternative now? lol
And the 120v version is $30 less.
"It's amazing how cheap these systems have become. "
I recall seeing all the major brands for sale in the big box type of stores in Japan. The prices for Panasonic and Mitsubishi were under $1,000 so a fraction of what those same units sell for here. I think it's just amazing how expensive they are here in the US.
Back in 2019 the only quick-connect DIY system was Mr. Cool. There is now a new entrant to the market, Perfect Aire, which seems to be consistently cheaper than Mr. Cool.
My understanding is that the Perfect Aire system is the same underlying units as the 4th Generation Mr Cool; often cheaper too. I can't guaranty this but if you look at the performance data they look quite similar so I suspect that both companies sourced the product from the same manufacturer in China.
Also Blue Ridge has a DIY mini split. Not sure it is the same as the Mr Cool line
I was going to go the DIY route with Mr Cool or Perfect Aire. Instead I bought a Fujitsu and installed it myself, then hired a neighbor who is an HVAC guy to vacuum the lines and do the final commissioning of the system. This saved me $2000 and took less than a day.
>>>but that requires about $200 worth of tools that most people won't have.>>
but that can easily change with....200 bucks
My suggestion is find a HVAC tech looking to make a little money on the side. Friend of a friend etc.
It is carpentry and a little wiring for you, an hours work for the AC guy
Hang the units and have everything ready to go, literally ready to turn the gas valves
If I was in a spot, I think I would call a larger place and ask for a service call. Tell the truth to the tech who shows up, either he drives away or vacs the system.
Tip him well and you have a new friend.
"If I was in a spot, I think I would call a larger place and ask for a service call. Tell the truth to the tech who shows up, either he drives away or vacs the system.
Tip him well and you have a new friend."
hah hah... that's actually a great idea if you are having trouble finding somebody. My area has at least a dozen "big" HVAC companies. I'd be willing to bet you wouldn't have to have too many service techs out to your house before one would agree to do it.
After all, flaring, pressure testing, and vaccuuming falls pretty neatly into the "service" category!
So here's my take on the decision tree:
Either:
I. Go with one of the quick-connect DIY systems: Mr Cool, Perfect Air or Blue Ridge; or
II. Go with a standard minisplit and either:
How does that sound?
That sounds about right, but personally I'd keep the Ephoca and Gradient units in the decision tree, as they are definitely worth considering.
But I'm not so sure about the chemical lineset flush (option B.2). Does that really accomplish the same thing as a deep (sub-250 micron) vacuum? Does it really accomplish the same thing as pressure testing the linset to 300 psi with dry nitrogen? I doubt it. I feel like the risk of a tiny leak not being caught would be pretty high with those kwik evac kits.