GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Feedback on DIY HVAC Plan

kurtgranroth | Posted in General Questions on

I’m building an 800 sq ft single story guest house in Phoenix AZ almost entirely DIY. It is the first house I’ve ever built, but my goals are to reach a high performance threshold like Matt Risinger, only with 1/10 the budget (if that) and zero experience. What could go wrong?

I worked up a tentative HVAC plan for the house and want feedback on it, if possible. One overriding goal is to have maximum performance for the absolute smallest amount of money and do it 100% DIY.

I’ve attached the floor plan of the house along with one possible layout of the HVAC components.

Some background: I did an online Manual J for the house and it claims about 11937 total BTU cooling. 5293 of that is the glass doors, since there is a lot of glass on all walls except the East — almost 40% of the North wall is glass! 2400 is in “appliances” which may or may not apply — this guest house is forbidden from having a stove, so the fridge, hot plate, and toaster oven are going to be it. Will they add 2400 BTU like a typical 220v or gas stove would? Don’t know.

I am almost surely going with a Mr Cool DIY option since I will be installing it myself, and they dominate in that space. The Manual J suggests 12k BTU but maybe oversizing to 18k makes sense? What if the Manual J isn’t good enough; the temp tops 120F; and the 12k can’t handle it? I flip-flop on that.

I’m proposing mounting the wall unit in the foyer, just above the bathroom door. This is mostly because it’s the high point in the house — the foyer has 9ft ceilings. The bedroom/bathroom/kitchen have 8ft ceilings and the great room has an 8ft soffit surrounding it, with only the center at 9ft. So it “fits” best above the bathroom door.

I have notable concerns about that. First, the instructions explicitly say to keep it away from doors and it’s literally by three of them. Second, is that one unit going to evenly cool the entire space? Am I going to need to put a transfer duct between the bedroom and great room? I was willing to pay an expert to design this part just to ensure that it works evenly, but couldn’t find one that would work so small… but yeah, very concerned that this is sub-optimal.

For ventilation, I’m shooting for around 20cfm and one pair of Lunos e2 HRVs should handle that. I’m thinking of putting one in the bedroom and one in the great room so that it pushes and pulls air back and forth through the house. Maybe that would even be enough to mix the conditioned air and negate the need for the transfer duct? Those locations are picked because that’s where I want the fresh air the most and, coincidentally, they are very shaded so hopefully they will have the coolest possible air to start with for their cycle.

Then there’s the bath fan in the shower. I’m planning on a bog standard Panasonic Whisper style fan that is either timer based or humidity based. No ERV/HRV, though – an ERV/HRV with enough cfms would easily double or triple my entire HVAC budget!

So hear me out: My thinking with the bath fan is that it will definitely be exhausting air faster than the house “normally” can provide it, so it will take the easiest input route… which is whichever of the two Lunos e2 tunnels is actively pulling air. That only does 20cfm and the bath fan does over 50cfm, so wouldn’t it just suck in air faster than the Lunos e2 can cool it, but it would still be fresh and filtered air? Does that make sense or is this just nonsense?

I did debate putting another exhaust fan through the wall behind the toilet, for odors. But I’m gambling that the one fan in the shower should be enough for that, too.

There will not be an exhaust fan in the kitchen. Like mentioned before, this cannot have a stove. There might be fumes and such from the hot plate and toaster/convection oven… but if it’s an issue, then I can either retrofit a recirculating hood or (more likely) construct a filter of some sort to handle the problem.

I’m open to literally any thoughts y’all might have on these ideas.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. aaronbeckworth | | #1


    Where are you planning on placing the outdoor unit for the Mr Cool? Are you sure that the line sets will reach your proposed indoor location?

    1. kurtgranroth | | #3

      Aaron, the outdoor unit is going to be on the East wall to the right of the indoor unit on the layout picture. That's about 10 ft horizontally and give-or-take 8 ft down, so I should have plenty of pre-charged line set.

  2. kyle_r | | #2

    I really doubt you will get adequate cooling in the bedroom from one ductless unit in Phoenix. I would consider putting a ducted unit in the ceiling of the hallway with a short duct run to each room. You could diy the ducting. The lines won’t be precharged but you can hire an hvac tech to pressure test and vacuum them out for a few hundred dollars. Or buy the equipment to do it yourself.

    1. kurtgranroth | | #4

      Yeah, the even cooling is a huge concern. I think I'd be more likely to go the ceiling cassette route over a ducted anything in any multi-zone case, though, since I have an unconditioned attic and handling ducts up there when I didn't plan on them in advance might be more of a pain that it's worth.

      My issue with any non-official DIY solution is that, from what I've seen, the warranties are only available when installed by a pro. So I could get a sweet Mitsubishi; buy all the equipment to charge the lines; install it all... and then if something breaks later, I'm out of luck since it would have zero warranty at that point.

      My thinking is that maybe I can mix the air well enough with a through-wall transfer duct, if the air isn't mixed well enough by default?

  3. Expert Member
    CARL SEVILLE | | #5

    If your building is very tight and well-insulated I expect that the single ductless mini split will do a reasonably good job of conditioning the building provided you leave the doors open most of the time. I have a single ductless head on one side of my first floor which is over 1300 SF and it conditions the space fine, similar with the head in my 2nd floor hallway - as long as the doors are open, it conditions the bedrooms fine. My only issues have been 2 out of 3 Mitsubishi coils have failed, and even under warranty, the labor costs are high to replace them.

    1. kurtgranroth | | #7

      Thank you! Yeah, there is only the two doors in the entire house and only the bedroom one matters, from my perspective. It'll almost surely be open most of the time and will have a 3/4" gap under for airflow regardless. I'm still wondering if an in-wall transfer duct would be necessary or just help or do anything at all...

      1. Expert Member
        CARL SEVILLE | | #9

        A transfer grille won’t do much. A bath fan near the head ducted into the bedroom would help when the door is closed but probably not worth the cost if you’re keeping them open most of the time.

        1. kurtgranroth | | #10

          Ah, referring to it as a 'duct' was probably the wrong word. It would be more of a transfer 'fan' -- like this 75cfm model: The thinking is that I would have it installed in the wall between the great room and the bedroom - it would pull the air from near the floor in the great room and pump it just above the wardrobe in the bedroom.

          I may be wildly overthinking this, but I picture the mini-split mostly dumping the air down the foyer into the great room. The transfer fan would then be pushing a bunch of that chilled air into the bedroom, which might even create an air cycle of sorts.

          I have zero experience with this sort of thing, though, so I don't know if it would actually work like I think or if it would be pointless.

          Your experience of having just the one head unit per floor and it conditioning the bedrooms just fine does give weight to the argument that it might be mostly pointless. Hmm..

  4. Trevor_Lambert | | #6

    20cfm of ventilation is sufficient for one person, barely. If you expect more than that, you'll need a lot more cfm.

    1. kurtgranroth | | #8

      Okay, here's my thinking on that, in three parts.
      First, the IRC dictates ventilation at 1cfm per 100 sq ft + 15cfm per bedroom. In this house, that would be (1*(800/100)) + 15 = 23cfm. The Lunos e2 is rated for up to 22cfm, which is almost bang-on.

      But second, is that even necessary? The exterior doors will be opening and closing a decent bit, since it's a retired couple that like the outdoors. During most of the year, the window and at least one door will be open a substantial amount since the weather is gorgeous. And even without all that, there's the incidental use of the bath fan - that'll be pulling many times the steady amount for a minimum of 30 min per day and more. Heck, even 30 minutes can be as much as 50% of all of the air in the house exchanged.

      And third, the IRC recommendation equates to 0.22ACH and ASHRAE comes out to 0.47ACH, by my calculations. Even a passive house will rarely hit those numbers and that's with extreme levels of air sealing. I am certainly aiming to have the highest level of air sealing I can get, but I do recognize that this is the first first house I've ever built and so practically the chance of me getting even under 1ACH (and realistically, maybe even under 3ACH) is incredibly low. Thus, even if I didn't do any purposeful ventilation at all, it seems like just accidental ventilation via the house leaks would be many times the amount of air changes required.

      So... yeah, that's my thinking why shooting for the IRC minimum might work for me.

      1. Trevor_Lambert | | #12

        I don't know how the IRC arrived at their number. My real world experience tells me that 20cfm per (adult) person in a well sealed home will maintain maybe 800ppm CO2, at best.

        Opening of doors has minimal impact on whole house CO2 levels, unless the frequency is over the top, like dozens of times per hour, 24 hours a day. I can't see that being the case. Open windows works great; surely there must be at least some periods of time when the weather is not conducive to that. I would guess it's pretty significant; Phoenix is not what I'd call a temperate climate. If it was really that conducive to having windows open most of the time, then A/C also shouldn't be much of a concern. I can't see why you'd run the A/C while having a bunch of windows open. The bath fan is even less helpful. 30 minutes at 100cfm works out to about 2cfm averaged over the day. It only takes a couple of hours for the CO2 levels to start running away, and it takes a lot longer to bring them back down to where they were.

        The ACH numbers you're quoting are at 50Pa. In a house with a 1ACH50, in still air that's going to be nearly zero. There are plenty of studies showing that even leaky homes, like over 10ACH50 or even 15ACH50 still have excessively high levels of CO2 at least some of the time. Surely you've been in a house that felt stuffy, despite being an old, leaky home. That perceived stuffiness is elevated CO2.

        1. kurtgranroth | | #13

          Thanks for your thoughts on this Trevor! Yes, you are right that my reasoning as I wrote it is absolute bunk!

          I tracked down the original reasoning that ultimately convinced me to go with one pair of Lunos e2 and the starting point was the Lunos e2 docs themselves -- they recommend just one 22cfm pair in my case:

          From my understanding of this, both the IRC/BSC and ASHRAE agreed on the same ventilation rates until maybe 2013 when ASHRAE doubled their rate. It seems that many knowledgeable people disagree with that new rate. Joe Lstiburek - in my mind, possibly the most vaunted building scientist alive - actually recommends ignoring the new standard. I know that Matt Risinger has also poo-pooed the new standard as being too high.

          But one of the reasons for posting my plan for feedback was to get real-world experience and so your experience carries a lot of weight for me. I need to think about this a lot more... and maybe get a good quality CO2 monitor just to get an idea of what I'm dealing with in my old-school unventilated current house, as a baseline.

  5. exeric | | #11

    I'll be interested in how your project turns out. In the past I've complained about the looks of the indoor units on mini splits. But between my concerns about the carbon footprint of my propane fireplace and the need for better cooling as summers continue to heat up in California I've altered my opinion. For my 1100sq house I am installing a DIY 18000 BTU Mr. Cool mini split just like you.

    I've heard good things about them and can only hope they are true. One nice thing about Mr. Cool mini splits that even the bIg boys in that space don't have (Mitsubishi, Fujitsu) is that they have a "follow me" type of remote control. That means it has a temperature sensor in the remote and wherever you move it the temperature in that room get compared to what you have set for a reference. That means if you move to a room that is farther away from the mini split it will order the unit to try to boost heating or cooling to reach the temp you set in that more distant room. It comes at the expense of efficiency because it will over compensate the rooms you no longer inhabit. But that seems like a good compromise to me in a small home and reduces complexity by not needing a multizone mini split. It seems like a pretty nifty implementation.

    1. kurtgranroth | | #14

      Yeah, I've floated the idea of replacing our existing central AC units in our primary house with mini-splits to my wife and she placed a hard veto on anything with a wall mounted indoor unit. So if we eventually go mini-split in our main residence, then it'd have to be ceiling cassette or ducted! Heh.

      The "follow me" feature in the Mr Cool is, indeed, a very compelling feature. I also see that you can point the louvers in pretty fine grained directions. That MAY come in handy? I do wish you could move some louvers in one direction (into the bedroom perhaps) with the rest in another direction (straight to the great room) but that doesn't appear to be possible.

      Mind you, when I first heard "follow me", my initial thought was that the remote was tracked and so the louvers would follow to point to wherever the remote was physically. I envisioned just carrying the remote from the great room into the bedroom at night and the head unit would automatically switch where it was pointing. I was mildly disappointed when I read up on what it really was. Still handy, just not as "magically" handy as I envisioned.

  6. exeric | | #15

    I thought I'd update my own experience installing a Mr. Cool DIY 18000 BTU mini split. It's not an easy job and you have to enjoy doing DIY to really make it work for you. That said, I'm very impressed with its performance in my 1100 sf two-bedroom home. I did a few things wrong, or in the incorrect order, while installing it but it still works great with no kinked, discharged, or cross threaded line sets. I ended up manhandling it pretty violently because of my nonconventional line set routing but it came through unperturbed by it. That's a testament to a well-designed unit that works correctly even when installation is DIY.

    The biggest roadblock for most people doing their own installation will probably be the wiring. Because I have a background related to that (aircraft avionics installation) it wasn't a problem for me. However, if you need assistance in that area don't do it yourself. If you pay someone to do that work it will easily eat up as much money as the mini-split itself. Something to consider if you want to go the DIY route.

    With my Mr. Cool running on automatic mode, it heats up the living room to the set temperature in about an hour. And the bedrooms come within 1 degree F of the set temp within 2-3 hours. It's definitely throwing some air directly into those room with their doors open. I have no problem with uneven temps. Overall I have no complaints in its operation and as yet have no complaints. It pretty quiet and seems very accurate in its temperature without engaging in any wild temperature hysteresis pattern. Still early days and haven't test its cooling features so take this with a grain of salt. I'll try to update this if things change for the worse, which I don't expect to happen.

    1. kurtgranroth | | #16

      Thanks for the report! I am, indeed, curious how difficult it will be to install it. I'm assuming that it'll be very straightforward based on the videos I've watched since I am building the house almost entirely on my own, from scratch, so how hard could a DIY-focused product be? Last words, right?

    2. user-7327173 | | #22

      Hi Eric,
      I'd love to hear more about your mini-split setup. I live in San Luis Obispo in a 2-bedroom 1100 sq foot house built in 1905. My load calculations are 15k cooling and 10k heating, but they drop to about 10k cooling and 8k heating when I apply double-pane windows, which I intend to approximate by insulating my single pane windows. I'm planning to bring the attic to R-50 and the crawlspace to R-19 but the walls will stay uninsulated. How many heads have you installed?

      1. exeric | | #23

        Hi, my non ducted single head mini split has met all my expectations and more. The part of California where I’m at is in a small valley that’s isolated from marine influences. It’s cl 4. It gets colder in winter here than all the surrounding counties. The two bedrooms on my small house are together on one side of the house. My natural tendency is to push against conventional wisdom if it makes logical sense. In this case the conventional wisdom is that homes always need more than one mini split or a MS with multiple heads.

        My house doesn’t seem to need that. There’s a straight shot from one exterior living room wall to to both entrances to the farthest rooms in the house, the BRs. It’s about 22 ft. So I just instlalled it on that living room wall. It works perfect. One bedroom is always about 1.5 F different from the living room and the other is about 2.5F different.

        The heating and cooling loads using Coolcalc are slightly higher than yours but not much. I erred on the side of more capacity than I may have needed because I’m only using one single head mini split. The extra capacity was a little insurance because of the less flexible nature of my system. I’m very happy with my system.

  7. Deleted | | #17


  8. big__o | | #18

    The good news is you can afford to oversize in Phoenix since your air is dry. So an 18k unit shouldn't raise humidity to stupid levels like Florida.

    I'm building an 750sf house and I plan to use a ducted system. Even if I have to build it below the ceiling.

    1. kurtgranroth | | #19

      Have you decided on what ducted system you'll be using? Will you be installing it yourself?

      1. big__o | | #21

        I have three midea units in my ongoing build. If they work out well I'll stick to midea. Daikin has the best reputation but is 2x the price of midea or gree.

        I have installed myself. Nothing to it. Very straightforward. I don't have a warrantee but I'm saving 20k not going with a pro so I guess I'm self warranteeing.

        I really like the ducted because you can deliver air where you need it, can install a better air filter, the return is far from the supply so you are not creating a loop, and the thermostat is remote instead of being inside the unit.

        You can make your ducts or have a sheet metal company make the supply and return ducts for you. I've done two so far and they were less than $80 each

  9. kurtgranroth | | #20

    FWIW, I reached out to the Lunos folks to ask if using a non-communicating high-volume exhaust fan like a bath fan or range hood could use the existing Lunos e2 "tunnels" for makeup air or if it would be detrimental to the system. In particular, I was concerned that a few hundred cfm of air blowing "backwards" through the active unit might damage the fan.

    I was told that the Lunos e2 system monitoring the air flow and if it exceeds the rated volume, then it will simply turn off the fan. At that point, the two tunnels will essentially be pure passive makeup air sources. The system should auto-start back up when the air flow returns back to normal.

    Sweet. That's just about perfect.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |