Install ductless mini-split

Money goes in, heat leaks out
kevm14
Posts: 15238
Joined: Wed Oct 23, 2013 10:28 pm

Re: Install ductless mini-split

Post by kevm14 »

FWIW we had a last hurrah summer heat wave and I got to observe how everything functioned.

As a reminder I currently have an 18k in the great room, 9k in a bedroom, another 9k in another bedroom, and an 8k window unit in the master. So that's nominally 3 tons of mini-split plus another 8k of window unit for a total of 44k BTU or 3.67 tons total cooling. The result? Excellent! Great room never got above the 72F set point. The kids' rooms units both contributed to the house. I keep the downstairs set to 72 and the upstairs set to 71. My theory is because the main area of the house (living room, dining room, kitchen) have no dedicated cooling, this area is relying on the overflow cooling from all of my other zones. And to encourage the upstairs to help the downstairs, I set them a degree cooler. This seems to work well because the downstairs main house thermostat (for the heat) never went above 72 which is pretty awesome. Also, I have noticed the master bed window unit actually will cycle even to maintain 71.

It's hard to imagine what I'd do with another 10k of cooling (18k two zone mini-split minus the 8k window unit that I would retire). I guess I could just not use it if it wasn't needed. Or let it regulate itself. Or since the downstairs would have a zone now, I could set the upstairs higher than 71, like 72 or even more during the day (unless someone is up there).
kevm14
Posts: 15238
Joined: Wed Oct 23, 2013 10:28 pm

Re: Install ductless mini-split

Post by kevm14 »

Another data point is using the square footage sizing calculation. The downstairs falls around 856 sq ft and the 18k inside unit specifies 750-900 sq ft. Sounds like 18k right? Not so fast. Let's start with a 12k living room unit, adding up all the theoretical units and see what happens.

Code: Select all

Unit location	Unit		Min sq ft	Max sq ft
Great room	18k single	750		900
Ian's room	9k dual		375		450
Owen's room	9k dual		375		450
Master bed	9k dual		375		450
Living room	12k dual	500		600
				2375		2850
My house is about 2,316 sq ft.

The reason for this is because all 3 bedroom units are oversized at 9k, but that's the smallest unit they make. The bedrooms are between 200 and 300 sq ft which are all below even the minimums. So the oversized bedroom units balance out the "undersized" 12k living room unit.

If I somehow wanted to just shut off all the bedroom units and rely on the living room unit for first floor conditioning (plus great room I guess) then yes, it could make sense. But I don't think I would operate the system like that. In most situations the bedroom units will be assisting (heating or cooling). This tells me that the 18k unit is right.
kevm14
Posts: 15238
Joined: Wed Oct 23, 2013 10:28 pm

Re: Install ductless mini-split

Post by kevm14 »

I do think we'd want the 27k system for optimal heating. But I think the 18k would be more than sufficient for cooling. So is it worth spending the extra money on the bigger system? Well that kind of depends on if there are any downsides to "over-sizing." And I still do not know the answer. My main concern would be how far the 27k system can throttle itself down. As I have discussed before, it has higher minimum outputs (in heating and cooling) than the 18k condenser, as you'd expect. And in the summer, I want maximum dehumidification. Having too high of a minimum would lead to increased system cycling OR overcooling, neither of which I really want. In the winter, the concern would be perhaps I am running the 18k harder and into less efficient capacity ranges. Also depends on if I am really trying to not use the furnace, or if the heat pumps just augment it and keep things more comfortable across the various parts of the house (probably the latter).

I don't know if I mentioned this before, but I have already observed that 9k in the bedrooms is more than necessary, especially for cooling. But so far in heating season I haven't seen any of my 9k units get too stressed. So my theory is the 9k will rarely if ever be at max capacity. So if I had a 12k in the living room, and an 18k condenser, it would likely be able to max out that 12k inside unit because the 9k would rarely be running at max. If I chose a 27k condenser, that is theoretically 9k more BTU. BUT, if I am comparing 12k to 18k inside units, that is only 6k difference. In real world situations, my suspicion is that I am looking at more like a 6k increase in output if I go with the 27k system. Suddenly it seems like diminishing returns. The "cost" of choosing the 18k would theoretically be increased furnace usage (hard to say how much though - 6k or even 9k BTU isn't that much considering the furnace outputs something on the 113k range with the 1 GPH nozzle). There could also be a cost if increased electricity usage because the 27k would be operating at a higher COP a lot of the time.

It also goes back to should I optimize for cooling or heating. If heating, the 27k makes some sense. The square footage the living room unit would be servicing really does specify the 18k inside unit, even if my total house capacity doesn't require it. If designing for cooling, the 18k for sure. And aside from the specs I obtained on minimum output, I'd maybe like to consult Mr Cool on this to see if there is other information that could help.

I think my questions/concerns would go like this:
With a 27k system, say in an overnight cooling situation, and the living room unit is just turned off, but the bedroom 9k is on, am I ever going to run into minimum capacity issues with a 27k condenser running a 9k inside unit? Or will it be able to throttle itself down just as much as if it were connected to an 18k condenser? I do not want overcooling and want max dehumidification.

I guess also for cooling, am I going to be annoyed with the 18k living room unit? Is it going to do the same dumb stuff my 18k single zone unit does with these too-high minimum fan speeds and thus poor dehumidification under moderate conditions? That said, if the answer to my 9k question was positive, I could always just....turn off the 18k living room unit for cooling unless it was actually needed (which would be rarely based on how this summer went). And I'd end up using the 18k living room unit more as a heating system. I do think it would do a much better job warming the downstairs as an 18k. I know it's only 6k BTU but that's 50% greater and it probably does throw the air better. Fundamentally I'm asking if the condenser drives the concerns or does the inside unit drive the concerns. So these are my questions for Mr Cool I guess.

From what I observed on my two zone 18k, it throttles down to crazy low output, both in heating and cooling. What I don't know is if this is more a function of the 9k inside units or the 18k condenser. Based on my experience with my single zone 18k, it actually seems like the inside unit is what controls minimum output. The minimum fan speed of my 18k is VASTLY higher than the minimum fan speed of my 9k units. Like it's not even close. The 9k can throttle down below the noise threshold of the room, and moves a tiny bit of air. It's impressive. Meanwhile my 18k moves significant air even at what I assume is the minimum.

So what does the 12k do? I believe the 12k is packaged into a 9k size. But the coil may have more density and I don't know if that also means the fan controller (or fan motor) is different. Hell the fan blades themselves could be different. I just don't know. But between the 12k and 18k (which is a size larger), I would assume the 12k would have more "acceptable" minimum fan speeds. If this is my main concern, then that also seals the deal for the 12k.

But it turns out I COULD use a 27k outside unit with a 9k bedroom + 12k living room! That would give me the full 21k BTU output. That's only 3k extra. If 6k was marginal, 3k is not worth anything. But the upside could be more efficient operation in all conditions, as long as there is no minimum output problem with the condenser itself. Plus if the 12k and 9k were somehow at max output, the 27k condenser would still not be maxed, and would be operating more efficiently. Something to think about now that I say all that.

That leaves me still at the 18k dual zone with 9k bedroom and 12k in living room, unless it turns out the 27k condenser has no downsides, and I decide to pair it with a 9k upstairs + 12k downstairs. If the 18k gen 4 behaves like my 9k gen 4s, maybe I'd just do that. But I have a feeling the 18k fan strategy is different because it's still fundamentally a point source of heating/cooling, and the 18k would be expected to reach twice the area of the 9k.

Refined questions for Mr Cool:
- For the gen 4 18k and 27k multi-zone condensers, are there any minimum capacity differences? Can the 18k throttle down further before having to shut off? By AHRI data, the answer is yes but I want to ask Mr Cool. My concern would be a situation where I have just one 9k inside unit running - is the 27k going to have a higher min output or will it let the 9k inside unit throttle down as much as it it were connected to an 18k condenser?
- For the gen 4 12k and 18k inside units, what are the fan profile differences? Does the 18k fan not throttle itself down as much as the 12k (or 9k)? This is behavior I am observing in my gen 3 18k inside unit and want to know if it translates to the gen 4. Does the 12k perform more like the 9k in this regard? I wouldn't put anything smaller than a 12k in the living room but want to know if there is a downside to choosing an 18k (aside from cost and wall size) in the living room.

Again, in theory, if it turned out that the 27k condenser can actually throttle down as far as a 9k inside unit wants (just like my 18k dual zone system with two 9k inside units), then choosing the 27k makes sense. IF YES, then the second question is, figure out if there is a reason to decide between the 12k or 18k inside unit. I think this is the decision tree I am looking at.
kevm14
Posts: 15238
Joined: Wed Oct 23, 2013 10:28 pm

Re: Install ductless mini-split

Post by kevm14 »

It seems like I did not document my over-charge debacle with the great room unit. I have it in e-mail. But I rectified the excessive pressure around Nov 2023. Fast forward to a week ago and I noticed the heating performance was less than it should have been. Put a gauge on it and, yeah, 300 psi is all she wrote. Supposed to be 400-450 in heat mode.

Today I added 1 lb which got me up to about 390 psi. Added another 0.5 lb and unfortunately that got me to 490 psi. Had to vent down to 450 or so, at which point the inside unit fan ramped up to full speed (86F set point) and that dropped pressure ultimately to a bit under 425 psi, or right in the sweet spot. Due to this, I'd estimate it took something between 1 and 1.5 lbs and probably closer to 1 lb given that the last 0.5 lb went from 390 to 490 and I ended up at 425. Linear math says 1.175 lbs which is probably in the ball park.

It's back to good working order.

Also, I leak checked a bunch of stuff at the condenser. Basically all the fittings and valves at the condenser were checked both with the spray bubble product and the electronic sensor. Nothing. But, I did replace the valve core, and also used Nylog on all 3 brass caps (2 cover the shut off valves and 1 covers the service port). So maybe that will help. If not, then it'll be low in another couple months, and I guess I'll have to check the rest of the line set which is going to be pretty involved. I guess I could also try sniffing the condenser coil and inside unit coil.

The other good news is, Mr Cool is standing by with my leak check results. It is still under warranty, after all, though I am not sure if all types of leak are covered. I think condenser leaks stand the best chance of being covered, followed by the inside unit and then again followed by the line set. Just my theory but maybe it's all covered to 5 years.
kevm14
Posts: 15238
Joined: Wed Oct 23, 2013 10:28 pm

Re: Install ductless mini-split

Post by kevm14 »

Just to better document, this is from last weekend when I diagnosed the low charge issue.

Not running, static pressure. Would be nice to learn how I might use the temp scale of the gauge. Here it's pointing to about 45F. Outside temp was around 30F so I'm not sure what this is saying. Inside temp would have been 65F though so if this is based on the inside unit, then I guess this means the charge is low? Or this means nothing.
20240206_170347.jpg
Running the system at 86F set point, ramping up to check pressure. 300 is about as high as it got.
20240206_170513.jpg
20240206_170656.jpg
20240206_171043.jpg
20240206_171046.jpg
kevm14
Posts: 15238
Joined: Wed Oct 23, 2013 10:28 pm

Re: Install ductless mini-split

Post by kevm14 »

Now here are the pics from this weekend, corresponding to my post yesterday.

Starting to check for leaks.
20240210_104050.jpg
20240210_105611.jpg
Original weight of the R410a cylinder. It was pretty low. 12 lbs.
20240210_114658.jpg
This is the low side gauge. The scale is different but the temp conversion is the same. I think this is in A/C mode, 62F setpoint. The room wasn't a ton warmer but it was definitely in cooling mode. As you can see, this wouldn't qualify as "too low." Well, I guess it shows 115 psi which is 5 psi under spec. I have learned that you cannot use the low side pressure to judge anything about the state of charge. It's too low but it seems normal-ish by this reading. You certainly wouldn't judge to add 1 lb from this! I would remind that when the system had too much charge, the low side was still like 180 psi which is the top of the range. That said, 180 seems too high for what I would consider normal A/C operation.
20240210_115708.jpg
After adding 1 lb.
20240210_115738.jpg
Of course low side pressure didn't change, so I had to switch to heating mode to see the result. Again, 86F set point to make it run wide open. Initially it was a touch over 400 psi but that ended up settling to like 390. The bottom of the spec is 400 psi in heating mode.
20240210_120951.jpg
Waiting...
20240210_121127.jpg
As I said, it dropped to a little under 390 psi which told me it needed more refrigerant.
20240210_121222.jpg
Close-up shot of the valves and adapters for some reason.
20240210_121834.jpg
I let it run at a modest set point during a lunch break and when I came back, I re-tared the starting weight. Still around 11 lbs.
20240210_140435.jpg
Back in cooling mode to charge. This seems to indicate slightly higher low side pressure but this really isn't telling me anything about the charge. Technically it looks like it reads between 120 and 125 so it did come up a bit but man for adding a pound, that isn't much change. If you decided to shoot for, say, 150, that would probably be too much even though that's the middle of the spec. 150 on a much higher demand day, though, is quite possible.
20240210_140441.jpg
After another 0.5 lbs charge added.
20240210_140641.jpg
No noticeable change to low side pressures....looks like about 125 psi despite now being over-charged.
20240210_140646.jpg
Back to heating mode to get the real data. Oops, 470 psi and climbing. Spec is 400-450 psi.
20240210_142138.jpg
It ended up peaking at 490 psi or so, which I guess I didn't get a picture of. So I vented to reduce pressure down to 450. I let it run for a bit at 450 just to make sure it was in the new steady state. I went inside and noticed that the fan had ramped up to high, whereas it was NOT on high when pressures were over spec. After the fan was able to ramp up, that started dropping the pressure further. 430 psi here.
20240210_144337.jpg
Then a touch under 425 psi. It seemed pretty happy here and this is also the middle of the spec.
20240210_144956.jpg
I left it here. So I added 1 lb to get pressures from 300 to 390 let's say. Then another 0.5 lb got me to 490. Whoops. Vented down and ended up at 425 so I guess I probably added 1.2 lbs or slightly under in the end.

And here is the final static pressure. This time it's pointing to a little under 75F. I'd say the room was in this range so if this is the inside coil temp, then I guess that's right. Outside was 55F. Again, I have no idea if this reading means anything at all, though it's clearly higher than when I started.
20240210_150129.jpg
Now we play the waiting game. I may try the sniffer on the inside and outside coils.
20240210_150129.jpg
Attachments
20240210_115738.jpg
kevm14
Posts: 15238
Joined: Wed Oct 23, 2013 10:28 pm

Re: Install ductless mini-split

Post by kevm14 »

Performance seemed bad so I checked pressure. A hair under 400 though when cranked it was doing heat with the louvers pointed down, just not as good as it should have been. So I added 0.25 lb refrigerant assuming that would get it right back to 425. Nope. Ended up doing that two more times for a total of 0.75 lb and finally pressure was at 430. BUT, performance actually seemed to decline. This is a clue.
20240222_123313.jpg
20240222_124622.jpg
20240222_125049.jpg
20240222_125626.jpg
20240222_134816.jpg
20240222_140803.jpg
20240222_142805.jpg
20240222_153022.jpg

3 days later, performance actually seems worse. Last night I left the unit on 65F and closed the great room doors. With the usual furnace set back of 58, I expected the room to be ....about 65F this morning. Nope. It was 59F. Something is clearly not working right.

Threw the gauges back on it and now pressure won't go over 300 psi. I vented "some" refrigerant and I THINK performance picked up a bit but at this point I believe the compressor is damaged. I submitted an e-mail to Mr Cool this morning. At this point I think I am going to angle for a whole new system. I'd like to start over with a new gen 4 system. I should have never touched the refrigerant levels last summer when the fan was clogged with gunk. It ran with a significant over charge for a few months of A/C and then a month or two of heating season before I realized what was going on. Been problems ever since. So yeah, it is possible that none of this would have happened had I just cleaned the inside unit like I was supposed to.
kevm14
Posts: 15238
Joined: Wed Oct 23, 2013 10:28 pm

Re: Install ductless mini-split

Post by kevm14 »

After a ton of back and forth with Mr Cool support (as opposed to Ingrams), they agreed that I need another condenser. Sheesh.

That said, and given the fact that I think I have leaks, I am going to see if I can get Ingrams to do more for me. My claim is that I was skeptical of the new condenser as a solution the first time and it lasted under a year before problems began, so I am going to angle for a brand new gen 4 system. It's been a few days and they haven't responded to my e-mail so this may take some time, and right now I have Mr Cool wanting me to confirm my info so they can send me a new condenser. What I didn't ask is if I have to pay for shipping. I guess I can do that. Ingrams made me pay for shipping the first time.
kevm14
Posts: 15238
Joined: Wed Oct 23, 2013 10:28 pm

Re: Install ductless mini-split

Post by kevm14 »

Received the freight delivery yesterday while I was on my walk.
20240329_122938.jpg

I finished up work and decided to just start the replacement because, why not?
20240329_151859.jpg
First I took a static pressure reading with the old condenser which I was going to use to ask Mr Cool about the refrigerant charge level. Looks like about 165 psi, 65F in room, 52F outside. I think this is still considered overcharged.
20240329_153546.jpg
20240329_153634.jpg

Noting how it was wired.
20240329_154344.jpg

Disconnected the electrical and refrigerant lines. Removed mounting bolts. I was able to lift this thing off myself and had it off an hour after starting the project.
20240329_161652.jpg
Digging the upper connections out of the tar-based sound absorption material. Kinda messy.
20240329_163432.jpg
Had the new line set connected and basically hanging down.
20240329_172528.jpg
Snapped the line set covers back on (after smudging them up nicely with the tar stuff that was all over my gloves). A little over 2 hours has elapsed.
20240329_172534.jpg

Next it was time to heave-ho the new condenser up on the mount and then connect all the electrical and refrigerant lines. I was able to do it but it was kind of a lot. 3 hr mark.
20240329_181436.jpg
Another static pressure test before running. This time showing 160 psi with 66F room temp and 51F outside (pretty much the same).
20240329_181947.jpg
20240329_182046.jpg
Turned it on full heat and it took a while but eventually crested 450 psi. Ended up having to purge a bunch of refrigerant to keep it from going over 450. Then when I turned the temp down and the compressor slowed down, pressure actually spiked again so I purged even more. Wish I didn't have to do that but I do not want this one to run in an overcharged state. By the way it was making a ton of heat inside, on high fan. And the compressor sounded healthy again. All good stuff.
20240329_185152.jpg
After all the purging I wanted to take another static pressure reading. Maybe a shade under 160 so oddly not much lower? 50F outside but coil temp was 77F since it had been heating for so long.
20240329_190532.jpg
Energy efficiency ratings. Apparently there is a SEER2 now. And HSPF2. It scores better on SEER2 but worse on HSPF2. Whatever.
20240329_191018.jpg
End result: back to working properly. Wish the charge wasn't sort of an unknown. I will probably see if I can get some guidance from Mr Cool on the static pressure reading.
Post Reply