aunsafe2015
Well-known member
Last summer I posted several threads about mini splits and mini-split installation. I received very helpful replies from many of you. The mini split has been up and running for several months now, and I thought I would return and share some observations about the unit itself and the installation process. Next paragraph is the short version. Longer version is after that.
If you are doing a mini split install as a partial DIY, make sure you have a decent HVAC guy lined up in advance, or else be prepared for delays and headaches. If you have other things you'd rather be spending your time doing, consider just hiring a highly regarded company to do the whole project start to finish for you, to save the time and to have more certainty about the warranty. My mini split works great but temperature and humidity fluctuations are a little more than I would have expected given the supposedly great modulating capabilities of the unit, but that might simply be due to the challenging environment my unit is in (garage with a garage door and one of the exterior walls uninsulated). The effect on my electric bill has not been noticeable even though during warm months I keep it at 76 degrees 24/7 and in the cold months I’ve been keeping it at 68 degrees 24/7.
--Long version begins here--
First, if you are not doing the entire job yourself start to finish, make sure you have a reliable contractor lined up in advance (and ask how they plan to do the job. Some of them claim that a micron gauge is not necessary, for example, which I think most folks on this forum would disagree with). At least where I live, it was hard to find somebody to do just the pressure test, vacuum, and turn on. It seems like the contractors with the best reviews online aren't really willing to do the job unless you also buy the unit from them and have them do the entire install. I knew this in advance, but it still turned out to be more of a headache than I thought it would be. If I had just hired a highly regarded contractor to do the entire job start to finish, my HVAC probably would have been up and running 2 or even 3 weeks sooner than it was.
Second, I bought a lot of tools and equipment (lineset, condenser pad, linehide, electrical whip, various tools). I probably went overboard. But after paying the HVAC contractor for the turn-on, the electrican who ran the 240V line from my main electrical panel, etc., I probably only saved about $1,000 versus contracting the entire job out. Yeah, I have some new tools and a bunch of new knowledge, which is certainly worth something. But for the amount of time and effort I spent researching and doing my part of the project, I'm not really sure the savings was worth it. Plus, if I ever need warranty work, there's at least something of a question mark as to whether Mitsubishi would honor a warranty on a unit purchased online. So for the $1,000 or so extra I would have paid to just have somebody do the entire job, it would have saved me a bunch of time, the unit would have been up and running a few weeks sooner, and I would have a 10-12 year warranty. So I'm on the fence about whether it was ultimately worth it to do this as a partial DIY. If you already know what you are doing (I did not) and/or already have most of the tools (I did not), then doing it as a partial (or complete) DIY would probably make a lot of sense and could easily save $2,000 or more.
As to the unit itself, I have it installed in a ~300 sq ft garage. One wall is shared with finished living space of my house. Of the 3 exterior walls, 2 are insulated with R-13 open cell spray foam, so should be pretty tight. The other wall has a 9x7 garage door (R-10-ish) and the wall portion of that wall is not insulated. So perhaps a challenging application for an HVAC unit.
It does a good job cooling and heating. Maintains cooling setpoint of 76 even when it is 95 or 96 outside (I have the 9k BTU unit). Maintains heating setpoint of 68 even down to 25 degrees outside (hasn't gotten colder than that yet). One thing that surprised me though is that the temperature swings are a little greater than I would have guessed, given the modulating capabilities of the unit (supposedly able to go from 1700 btu all the way up to 12,000 btu). When setpoint is 76 in cooling mode, it will generally range from 75-ish to 77-ish. When setpoint is 68 in heating mode, it will generally range from about 67 to about 71. So, perfectly comfortable, but again, given the inverter and modulating capabilities, I would have thought it might stay within a tighter temperature range. Even the single stage ducted system in my house keeps a tighter temperature range. Perhaps the issue for the mini split is that the garage, with one uninsulated wall and a 9x7 garage door, is just a challenging application too difficult to maintain as precise a temperature. In any event, it does a good job and I'm perfectly satisfied. I just though the temp fluctuation would be even lower than it is, given the modulating capabilities of the unit.
(For the record, I have the MHK1 thermostat which senses temperature at the thermostat rather than at the head unit. Also, my lineset is only about 10 ft long. Mitsubishi has no minimum lineset length as far as I know, but I cannot help but wonder if the lineset is so short that the system is a little overcharged with the factory default refrigerant charge, though the installation manual says nothing about removing refrigerant for short lineset lengths, so I believe my installation job was fully compliant with the manual.)
As for relative humidity during warm months, it gets up to about 58% in the garage at night-early morning hours when the mini-split is not working very hard. But by mid-late morning when it is up into the mid-80s and beyond, the RH starts to drop and will get all the way down into the low 40s by mid-late afternoon. The spray-foamed walls should be pretty tight, but the garage door probably is not, perhaps explaining why RH rises pretty quickly at night. Again, given the modulating capabilities, I would have hoped it would keep the RH down a little bit further at night and early morning. But having done a little more research, my current understanding is that when the unit is operating at the lower end of its capacity, like 1700 btu, the coils probably aren't getting cold enough to be removing much water. So the dehumidification capabilities are probably pretty weak in nights and early morning. To me, this at least partially defeats the purpose of having such a wide modulation range. If I had known I wouldn't get much dehumidification at the lower capacities of the operating range, I would not have been so concerned about the minimum operating capacity of the unit (I specifically chose this Mitsubishi unit b/c it modulates lower than other comparable units, except for the LG Art Cool Premier which can actually go all the way down to 1,023 BTU). Yes, the unit does have a "dry" mode, but when it is in dry mode you can't set the temperature, and it overcools. The one time I tried dry mode it did indeed lower humidity, but it also cooled it down to about 72 degrees which was significantly lower than the 75-76 I wanted.
I’ve actually used a data logger to collect some data about the temperature fluctuations and I may create a separate thread about that when I find the time to do so.
Thanks again to all the folks who gave feedback along the way. Hopefully this thread will be helpful to others in the future.
If you are doing a mini split install as a partial DIY, make sure you have a decent HVAC guy lined up in advance, or else be prepared for delays and headaches. If you have other things you'd rather be spending your time doing, consider just hiring a highly regarded company to do the whole project start to finish for you, to save the time and to have more certainty about the warranty. My mini split works great but temperature and humidity fluctuations are a little more than I would have expected given the supposedly great modulating capabilities of the unit, but that might simply be due to the challenging environment my unit is in (garage with a garage door and one of the exterior walls uninsulated). The effect on my electric bill has not been noticeable even though during warm months I keep it at 76 degrees 24/7 and in the cold months I’ve been keeping it at 68 degrees 24/7.
--Long version begins here--
First, if you are not doing the entire job yourself start to finish, make sure you have a reliable contractor lined up in advance (and ask how they plan to do the job. Some of them claim that a micron gauge is not necessary, for example, which I think most folks on this forum would disagree with). At least where I live, it was hard to find somebody to do just the pressure test, vacuum, and turn on. It seems like the contractors with the best reviews online aren't really willing to do the job unless you also buy the unit from them and have them do the entire install. I knew this in advance, but it still turned out to be more of a headache than I thought it would be. If I had just hired a highly regarded contractor to do the entire job start to finish, my HVAC probably would have been up and running 2 or even 3 weeks sooner than it was.
Second, I bought a lot of tools and equipment (lineset, condenser pad, linehide, electrical whip, various tools). I probably went overboard. But after paying the HVAC contractor for the turn-on, the electrican who ran the 240V line from my main electrical panel, etc., I probably only saved about $1,000 versus contracting the entire job out. Yeah, I have some new tools and a bunch of new knowledge, which is certainly worth something. But for the amount of time and effort I spent researching and doing my part of the project, I'm not really sure the savings was worth it. Plus, if I ever need warranty work, there's at least something of a question mark as to whether Mitsubishi would honor a warranty on a unit purchased online. So for the $1,000 or so extra I would have paid to just have somebody do the entire job, it would have saved me a bunch of time, the unit would have been up and running a few weeks sooner, and I would have a 10-12 year warranty. So I'm on the fence about whether it was ultimately worth it to do this as a partial DIY. If you already know what you are doing (I did not) and/or already have most of the tools (I did not), then doing it as a partial (or complete) DIY would probably make a lot of sense and could easily save $2,000 or more.
As to the unit itself, I have it installed in a ~300 sq ft garage. One wall is shared with finished living space of my house. Of the 3 exterior walls, 2 are insulated with R-13 open cell spray foam, so should be pretty tight. The other wall has a 9x7 garage door (R-10-ish) and the wall portion of that wall is not insulated. So perhaps a challenging application for an HVAC unit.
It does a good job cooling and heating. Maintains cooling setpoint of 76 even when it is 95 or 96 outside (I have the 9k BTU unit). Maintains heating setpoint of 68 even down to 25 degrees outside (hasn't gotten colder than that yet). One thing that surprised me though is that the temperature swings are a little greater than I would have guessed, given the modulating capabilities of the unit (supposedly able to go from 1700 btu all the way up to 12,000 btu). When setpoint is 76 in cooling mode, it will generally range from 75-ish to 77-ish. When setpoint is 68 in heating mode, it will generally range from about 67 to about 71. So, perfectly comfortable, but again, given the inverter and modulating capabilities, I would have thought it might stay within a tighter temperature range. Even the single stage ducted system in my house keeps a tighter temperature range. Perhaps the issue for the mini split is that the garage, with one uninsulated wall and a 9x7 garage door, is just a challenging application too difficult to maintain as precise a temperature. In any event, it does a good job and I'm perfectly satisfied. I just though the temp fluctuation would be even lower than it is, given the modulating capabilities of the unit.
(For the record, I have the MHK1 thermostat which senses temperature at the thermostat rather than at the head unit. Also, my lineset is only about 10 ft long. Mitsubishi has no minimum lineset length as far as I know, but I cannot help but wonder if the lineset is so short that the system is a little overcharged with the factory default refrigerant charge, though the installation manual says nothing about removing refrigerant for short lineset lengths, so I believe my installation job was fully compliant with the manual.)
As for relative humidity during warm months, it gets up to about 58% in the garage at night-early morning hours when the mini-split is not working very hard. But by mid-late morning when it is up into the mid-80s and beyond, the RH starts to drop and will get all the way down into the low 40s by mid-late afternoon. The spray-foamed walls should be pretty tight, but the garage door probably is not, perhaps explaining why RH rises pretty quickly at night. Again, given the modulating capabilities, I would have hoped it would keep the RH down a little bit further at night and early morning. But having done a little more research, my current understanding is that when the unit is operating at the lower end of its capacity, like 1700 btu, the coils probably aren't getting cold enough to be removing much water. So the dehumidification capabilities are probably pretty weak in nights and early morning. To me, this at least partially defeats the purpose of having such a wide modulation range. If I had known I wouldn't get much dehumidification at the lower capacities of the operating range, I would not have been so concerned about the minimum operating capacity of the unit (I specifically chose this Mitsubishi unit b/c it modulates lower than other comparable units, except for the LG Art Cool Premier which can actually go all the way down to 1,023 BTU). Yes, the unit does have a "dry" mode, but when it is in dry mode you can't set the temperature, and it overcools. The one time I tried dry mode it did indeed lower humidity, but it also cooled it down to about 72 degrees which was significantly lower than the 75-76 I wanted.
I’ve actually used a data logger to collect some data about the temperature fluctuations and I may create a separate thread about that when I find the time to do so.
Thanks again to all the folks who gave feedback along the way. Hopefully this thread will be helpful to others in the future.
Last edited:
