He could be right - almost never happened before because almost never is a customer watching or checking what they are doing while having professional experience with similar designs, processes and issues. Customer does not know it all but is picking up on processes and results that are at the very least questionable.

I've asked for a plan for Tuesday (the actual trained tech. coming). They have a chance to recover some of the faith I originally had. If the trained tech. (he is also the General Manager) stays and sees that things are done methodically and properly that could be a good sign. I'm cynical, so what might happen is he comes talks to his tech. and lays out what to do and leaves to help the company owner with sales calls.

exactly. He should have said "We've been probing the depths of crappy installs for some time and making more and more money every month. I'm disappointed that someone finally figured us out" but that would require accountability for ripping people off.

My question: why are they using a different line size than specified? Is there an issue with distance and flow resistance? Or is this just what he had on the truck?

I think all the ports on the outdoor unit have the same connection sizes. However, one of the indoor units calls for a larger diameter line set, maybe because of distance. So at the outdoor unit there needs to be a reduction.

Does anyone here know someone at US Mitsubishi air conditioning group? I'm thinking someone at mid-management level or higher.

Maybe someone here has a first degree connection with an upper manager at Mitsubishi via LinkedIn. See if you can start a conversation and then forward to me if he/she will agree to talk with me.
Thanks.

In case anyone is still interested.....
Apparently there was a meeting on Monday at the AC company and then people showed up here with some new equipment. Things were better but not great.. could not easily pull below 700 microns. In the afternoon they came back with a new hose that was the vacuum type. Big improvement - pulled to 450 microns.
I pointed out that the new hose brought this morning wasn't the vacuum type. If you want to pull a good vacuum things have to be correct and in good condition. The last? problem is when turning off the ball valve to hold the vacuum - it adds 100 microns.
Anyway things are getting better.

Well.. here is another update. They tried a bunch of strange approaches including pulling with two different pumps with hoses going to the same place.
Finally went back to the setup we had yesterday but had the same issue of going from 460 to 540 when turning the valve in an attempt isolate the pump and start the decay test.
We called Mitsubishi tech support and they confirmed the pump to 500 and then hold it for one hour specification. It is commonly done in the field. The spec. would not be as it is if it was very difficult to pass.
I'm not sure what is going to happen next. I suggested a few days ago that we try a process of elimination approach but they didn't want to do that.
I guess at some point they give up and I hire someone to come find the leak and take that money out of what I owe them.

For what it's worth, I installed a two ton Mitsubishi about 2.5 years ago and was able to pull a vacuum to the specified micron levels with a middle of the line pump from Robinair and the automotive a/c hoses that came with my Robinair manifold set (I added a micron gauge to the manifold and appropriate fitting adapters to the lines). I also pressure and leak checked using nitrogen (swept the system as the manual described as well.).

It sounds like the guys working on your system are not sure of what they are doing, and/or not following best practices steps.

Good work there. You probably had no brazed joints and good quality flares. Possibly used Nylog and all the things usually recommended.
They ran the pump all night was 413 microns at 2:50AM and 408 microns at 5:30 AM.
So it is dropping very slowly. The Navac pump is good we tested it by putting the micron gauge right on the pump and got 3 microns. I called Navac this morning to ask what they think is going on. We might be pulling against a small leak - that's my concern. The decay test will be the key thing to watch.

Yes to the flares and Nylog. I had never flared copper tubing (lot of experience flaring aluminum tubing on aircraft, however), so I practiced repeatedly on scrap pieces until I was confident I was doing it correctly. This was a first project of its kind for me, but I do have many decades of tools and experience with automotive a/c, and am willing to read and follow directions, so was reasonably confident I could do the install myself, and thankfully it was successful (took me forever, though, particularly the step of routing and properly bending the copper lines through two walls, as that was definitely a first for me :)).

Hopefully you/your guys will be able to find the slow leak if that is the case.

I asked them to bring a new core tool and try again. Yesterday morning he showed up with a new core tool. So we now have new: core tool, micron gauge, pump, pump oil, and vac. hose. Pulled 413 in 15 minutes. Looking good - so we went for the final decay test (purple line).
See graph. There is no more I'm going to get from these guys so we charged the system. There is a reason it is going up. Hopefully still something not sealing in the instrumentation. I might buy a leak detector and go around seeing what I can find.



http://forums.pelicanparts.com/uploa...1696076846.jpg

Write down what the charge pressure was and check again next year. A good leak detector should find any problems, but I do not know how good the cheaper ones are. Did they braze the line set to the evaporator core onsite or was that pre attached?

They charged by weight. Done. No pressure measurement or readings of any type that I could see. They used the gauge set to get the 410A into the outdoor unit (compressor) but I didn't see any adjustment of pressures/etc. like balancing the sub cooling and super heat and all that business.

Roger that - re. a leak detector - I was thinking of getting one. Been reading about the different types.

The line sets are attached to the indoor units and outdoor unit via flares. During the pressure test of 600psi they found a bad flare and then redid the pressure test. The one thing I didn't like was brazing lines together up in the attic when I specifically asked to not have that.

I would bet that you are leak free with that deep of a micron reading.

Not many companies install them but I still like having a liquid line sight glass for the moisture indicator and the indication of a solid column of liquid to the expansion valve. Nice to have the visual.

For those interested in mini split AC. I'll continue with where I am at this point.
The system is in and "working". The performance is not quite what I was expecting but maybe I need to adjust my expectations.
We had a hot day and here's what happened. I recorded info. on the one wall unit cooling the large family room.
Here's what happened. You see the coil exit temp. going up when I think it should have held. All the heat sources (sun, attic temp. etc.) were going down.

In any case you might like my graph.http://forums.pelicanparts.com/uploa...1696898095.png

^^^

Was the compressor running, even at a very slow speed (assuming it has a variable speed compressor like mine does), at the time the evaporator temp was rising? If so, I doubt that is normal.

Currently my mini split is set on 69; room temp is 69 and the compressor is not on, so the evaporator coil is also 69 (I know that temp because I use an aftermarket controller that has a fins/evaporator coil sensor).

When I have the aftermarket controller set to a very cold temp (47-50F, which is the temp range I sleep in, year round), then the evaporator reaches temps of 45-to-near freezing while supercooling my room and then rises/levels off once those cold temps are reached.

Oh yes, the compressor pump was running I think on the high speed and the outdoor unit fan was also running. The unit is a brand new Mitsubishi 42,000 Btu 5 port Mini Split.

For those still interested - The AC contractor wanted me to do another test while recording the temp. going in (coil inlet) the wall unit. So I did that, see graphic.
Also, I called Mitsubishi again - got transferred to customer relations. The lady that answers that line knows me now. :-) Basically the only option is to continue trying to work with the contractor that installed the system. If they need help they can contact Mitsubishi or their distributor that administers the Diamond Contractor program. The distributor is supposed to have experts that advise or can come on site. The other thing she mentioned a few times.... if I don't have confidence in the original contractor get another licensed AC tech to come and take a look. Certainly not a bad idea but not happy to pay for install and fix labor coverage and then have to go to another company to get a diagnosis.
Anyway here is the last test.

http://forums.pelicanparts.com/uploa...1697036241.jpg

It makes you wonder how many installs are working but not working well. Owners don't know they just think it's a bit more expensive and less efficient than they had expected....

Bingo! At this point I'm somewhat happy to say the AC contractor now agrees something odd is happening. Now we can cut the bravo sierra and start diagnosing the issue.

^^^

There is a good chance that they either did not get all of the air and moisture out of the system after all (possibly due to an unfound leak, as indicated by one of your vacuum charts), or did not add the proper amount of refrigerant (too much or too little), given the strange rise in evaporator temps even when the compressor is running.

Mini-splits come with the compressor pre-loaded with refrigerant that is sufficient for whatever length of piping the install manual specifies (more must be added if that length is exceeded); with your complex system (multiple evaporators/head units), I'm betting the installers did not correctly calculate how much additional refrigerant to add.

When set up correctly, those Mitsubishi's are capable of bone chilling cooling, way beyond what a normal person (me being abnormal! :D) would ever want.

Good thinking.
1 - They/we did many evacuation runs with the last couple under 500 (~420) microns. From all my reading under 500 is good. However 420 is certainly nothing to write home about. I think the system is dry enough to operate to spec.

2 - They were aware that a calculated amount of R410A needed to be added - two people checked the calculation. Of course they could have still put in not enough.

3- Your last point is the best - When properly installed and setup I should get "beer" cold air.

He now admits something is off.

I think I'm finally getting somewhere on this. What do we know now:
1 - The one GL wall unit appears to be somewhat okay running by itself. The controls work and it tries to heat and cool.
2 - The GL unit doesn't, at this point, have the capacity to cool the big room enough.
3 - When two or more of the smaller bedroom units are turned on the "big" GL unit all but stops cooling.

This brings to mind a question. When the GL (24,000 Btu) unit is running alone is it cooling near its capacity? How would we test this? Fortunately the CFM for the 5 different fan speeds are in the service manual. I should have thought of this before... it is easy to calculate the cooling provided (Btu/Hr.) using the very common equation Q=1.08*dtF*CFM. So I tested the unit while on full cooling (min. set point of 61F) at the 5 different fan speeds and this is what I got.
What do you think now?

http://forums.pelicanparts.com/uploa...1697387791.jpg

Just for a data point my man cave garage is 24x26x12 foot high ceiling . Has 2x6 walls with R19 fiberglass insulation . The ceiling is also R19 . Two garage doors are insulated . So overall a fairly tight building .

The Mitsubishi 24k mini split can get the interior to meat locker quickly . I have never timed it but my guess is less than an hour after turn on . Likewise in heat mode it heats quickly . Just a single head unit .

Past experience with a 3 bay garage in my previous house again with a 24k Mitsubishi was similar . I guess in both cases my installer did a great job . When installed properly the performance is excellent in my opinion .

Good data point. Thanks. Your MC is very similar size to our family room. We have a vaulted ceiling with good insulation. However, we have two large windows that get sun until about 2PM or so and part of the room is open to the kitchen.
The contractor is being a bit quiet right now. At some point I might need an independent evaluation and see if the system is capable or has an issue.
If you have a way to measure air temperature and don't mind checking - maybe you can measure the coil input and exhaust temps for me to compare. I've seen a few measurements in the high 40sF but those don't last long. Probably will require a ladder to get to the top of the coil. The controls work by looking at the set point and air temp. going in the coil. That's what "people" are saying.
Thanks in any case.

@rfuerst, If you want to do some measurements I suggest.....
Fan speed 3, vanes down 3 "notches", setpoint 61F. I assume you have a GL24 unit but if different please send out the model #.
Thanks.

hcoles, those head units have three temperature sensors - one on the evaporator's in pipe, one on out pipe, and a room sensor that is beneath the cover, to the side of the LED display/sensor for remote control (there are a couple of small slits on the side of the cover and the room sensor is there - it basically looks like a short, thin black wire . . . can't see it without removing the cover/case however). I know this because I had to make mods to those when installing my aftermarket controller.

With my fan speed on the second from highest setting (not the "powerful" setting), and the remote control temp set at 61 (my aftermarket controller not operating at the time), the air exiting the evaporator is consistently in the 45-48 degree range (with aftermarket controller running the exit air temp gets much lower).

My model is GL24 (a/c version only - no heat pump function).

Based on what you've described above, it sure seems like there has been an error made in the amount of refrigerant added. Unfortunately, your best option at this point is probably going to be getting someone else to look at/test the system, as long as you can find a tech that you are confident he/she knows what they are doing (based on your remark about installer business going quiet).

Good info. Thanks. At this point I think you might be right re. getting someone to look at the system. My contractor gave me a sheet from Diamond System Builder. I have seen 48-50F discharge temps. from my GL24 but not for very long. Maybe as you say the system is low on 410A. But there are no codes flashing from the LEV sensor indicating low charge.

Okay people, the saga continues.
During hot days the large family room unit goes to a high (+65F) discharge temp. when it should keep supplying cold air. The contractor agrees something is not right.
I've been watching them closely e.g. during the pressure and vacuum processes because I found the brazing "issue" and ground wires placed in the panel on the wrong bus. Then the vacuum process had poor performance due to worn out equipment/etc. Anyway - I'm watching them like a hawk.
Now the GM and lead tech. wants to come and work on the system - AND - doesn't want me to be near where they are working. I suspect many reasons but of course not sure why he would all of a sudden change how we are working. My main concern is he will again not do a good job during the vacuum process and doesn't want me to see him do that.

Am I being unreasonable to want to know exactly what they are doing?

Put the shoe on the other foot. When you are working on something and it isn’t going well, do you want someone right there watching, asking questions and making comments? I say, let them have a go at it while you are away and see if that makes a difference.

A couple simple things to check with the contractor. Is the vacuum pump a minimum 3-5 CFM capacity? Is pump oil clean, should be changed after each use as contaminates, moisture are captured in the oil. Does pump pull below 500 microns when “blocked” off? Does the tech know triple sweep, double evacuation procedure? Did they purge the line set with nitrogen before brazing to prevent oxidation? Looks like they used foscopper brazing solder on the lines, not silver solder or silfos or stabrite. You should talk to the tech intelligently before he starts to work as rwest suggests, so he knows you’re pretty well versed as to what’s happening with the system. Good luck, hopefully the problem will be resolved.

Bob, thanks for the interest and comments.
FYI - When the system was first installed they pressure tested with nitrogen and vacuumed the system. This is where we hit a snag. I read the Mitsubishi manual that describes the vacuum process and specs. Their equipment initially was not capable because it was worn out and needed maintenance. I figured all this out and they finally showed up with much better equipment and pulled an "okay" vacuum.
So, I feel justified in being skeptical and wanting to watch closely to what they might do. I understand it is not comfortable to have someone watching over you when working. The bottom line is at this point - when someone makes a big change (very open to don't be near us) without explanation - it makes me feel they are untrustworthy.

I've always maintained that if you want to watch (not under all circumstances, depends on what tools are being used) go ahead because I know what I'm doing and it doesn't make any difference.

But stay quiet.

Like Zeke says, stay quiet but keep your eyes open. You’ve been armed with plenty of info from the board to have a pretty good idea of what’s going on. I wonder if the techs are NATE certified? Years ago, I had my techs attend Carrier and Trane tech schools to learn how their equipment should be serviced including understanding a psychometric chart to properly charge a system. At least the hvac owner is standing by his installation and hopefully the issues will be resolved. Best of luck….

The contractor is part of the Mitsubishi Diamond Contractor program. This is supposed to mean at least one employee has been to Mitsubishi for training. I think there are also some on-going training requirements.
Re. psychometric chart - as I understand mini-split systems don't "use" the chart in the same way as standard AC. E.g. there appears to be only one tap (on the pressure side).

My contractor has one person and that person lives ~2000 miles away, TX vs CA.

Minor update. The head guy from the contractor called and we discussed why sometimes people don't like to be watched/etc. So I guess all is going okay - I can still watch what they are doing bit I guess be careful what I say. I don't know - positive progress I guess.

Taking over 2 months to correct their usual and customary installation strongly suggests nobody has taken them to the mat on their crappy work before.

I've been reading about mini-splits and watching YouTube, there is a lot of info. out there. What I'm getting is... Mini-splits are a different animal compared to what installers worked on 20-30 years ago. They are less forgiving of poor installation practices that worked okay in the past. One guy on the HVACtalk forum, said residential work doesn't pay enough for us to do good work. My feeling is that someone that knows what they are doing can do correct work without too much more time/materials being used. For me, I'm willing to pay the additional amount but many probably go for the lowest bid. The issue is you don't know all this until you have selected your contractor and watched what they do.

I think some of you here are HVAC technicians or very interested so here is an update.
Their best technician came Friday Nov 17 to find out what is going on. Most if not all of the LED flashing codes indicated a low refrigerant level. So they decided to pull the charge and weigh it. It weighed 3.5lbs instead of 9,8lbs. So they (tech. phoned manager) concluded there is a leak. Pressurized with nitrogen to 600psig in attempt to find the leak(s). Hooked up a digital pressure gauge set Testo 550s. The pressure was dropping. Found a leaking flare fitting right at the outside unit. Checked all the coils and flare fittings at the inside units, using a sonic sniffer, and didn't find any leaks. Closed the isolation valves and hooked one tap on the Testo on the "machine" side and the other on the "house" side. This was a way to isolate where a leak might be if there is one remaining. Initially the machine side was dropping but we were in the afternoon and things were getting colder so not sure what is happening.

I'll attach graphs for comment.

Thanks.

http://forums.pelicanparts.com/uploa...1700663473.jpg


http://forums.pelicanparts.com/uploa...1700663498.jpg