Noisy Minisplit in Low-Heating Mode
I have a 42k Mitsubishi system (MXZ-5C42NA2 – non hyperheat), located in Seattle (zip code 98107), installed on the north side of my 2000 sq ft, 1946 house. The system has 4 heads, all the “FH” model, 6,9,12 and 15k.
I noticed that in heat mode, when 1 or 2 units are calling for 1 degree of heating, the system is pulling about 900-1100W (as seen in the Sense app) and as it modulates up and down, it sounds like a diesel truck/apache helicopter is idling outside. When you check the outdoor unit, the ground is definitely vibrating, but somehow the inside of the house magnifies it. The sound is noticeable even at the furthest points from the unit, like upstairs, on the other side of the house.
The installer has already come out, confirmed the unit is operating properly, all lines are secured and not vibrating, and added these cork insulators under each foot, but this has done zero to fix the issue. They are willing to come back out again and take drastic measures like depressuring the unit to repack the ground and/or adding gravel, foam pads, etc. I can tell I should probably suggest the course of action, as they’ve never encountered this issue. I also noticed that the pad bolts are loose and can be easily pulled up.
The ground the unit sits on is compact dirt, with nothing underneath. You can jump in the area and the vibrations easily resonate, however.
Attached are photos and video of the install, and the house. The layout pictures have old models, as this is from a previous quote.
What should the best course of action be to mitigate these bad bad vibrations?
https://youtu.be/kpe0E7FIWUs – Outdoor Unit
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Hi Michael,
The issue of noisy minisplits has come up a few times in the Q&A. The sounds you are hearing could be related to the defrost mode, which kicks on when the outside coil accumulates ice that needs to be melted off. Although, you didn’t mention your climate zone, so this may not apply. Regardless, you might want to read through these Q&A threads:
Mitsubishi minisplit popping noise during heating.
New minisplit is making strange noises.
This is not an issue that is related to those questions (popping and defrost) and does not relate specifically to defrost mode. I believe there is an installation issue with the pad, as there should not be that much vibration transferred into the ground and then the house.
I have two ground-mounted mini-splits (Fujitsu) right outside my living room, and there isn't even a hint of vibration in the house other than the odd time during defrost mode. And even then it's barely noticeable and only if you're standing right at the window/wall they're in front of.
The main differences between my install and yours is that mine are on concrete pavers (not a plastic base), they're a bit further from my house (but only a few inches), and they aren't bolted down. The rest is pretty much the same, with the pad being on compacted soil and the units sitting on those rubber vibration isolators. You can definitely feel the unit vibrating if you put your hand on it, but nothing seems to get transferred to the ground, so I think it must be the base or maybe they're bolted down too securely? I know with one of my indoor ducted units that I had installed vertically, when I mounted it to the wall in a closet it gave off a lot of vibration through the framing. But once I loosened the bolts a bit (it's sitting on a return plenum, so the bolts aren't holding it up or anything) that disappeared.
If you really can't solve the issue with it as-is, the next thing to try would be spring isolation mounts. They're not cheap, but they would likely solve your problem.
It sounds to me like an unusual amount of vibration and also an unusually conductive soil. If you are positive there is nothing wrong with the equipment, I would suggest a mass-damping concrete pad--something big enough to absorb vibration, not just the plastic base I think you have or lightweight concrete blocks. The isolation blocks your technician installed I've used for washing machines and they do a little but not much.
Is snow unlikely event in your part of the world?
The insulation instructions say it should be above the depth of any snow you are likely to get.
I like the idea of a 16-18 inch thick slab under the unit then we know it will not wiggle.
Let’s see a photo of the lines going into the house maybe we will see something.
Walta
Thanks for all the inputs! Snow is an unlikely event. I am in the situation of vibrations on very conductive soil that is magnified inside the house.
From the above comments, it sounds like a concrete/thicker slab would help out the most.
You can see the lines int he latter part of this video https://youtu.be/kpe0E7FIWUs.
I've found the fan can cause some low frequency harmonics. It's a little bizarre since the unit feels essentially silent and almost vibration free when I stand next to it and touch it, but because it's mounted on my house wall some low frequency pulsing vibrations are magnified in the house. Surprisingly just at lower loads, once the fan and compressor speed up I don't notice it any more. I don't think the fan is actually out of balance, I think it's the pulses of air coming off the blades in my case. You might be able to adjust the fan in or out of the "cone" to tune this but I haven't tried.
My point is you might want to try and figure out if it's the fan or compressor causing the vibrations.
Commercially I often see equipment mounted on spring suspended vibration isolation mounts to decouple pumps, chillers, and boilers from the floor of the mech room.
I suspect as others have suggested a high mass poured concrete pad, with properly selected (for the weight) rubber vibration isolation mounts would likely eliminate most of the problem.
"I've found the fan can cause some low frequency harmonics. It's a little bizarre since the unit feels essentially silent and almost vibration free when I stand next to it and touch it, but because it's mounted on my house wall some low frequency pulsing vibrations are magnified in the house. Surprisingly just at lower loads, once the fan and compressor speed up I don't notice it any more. I don't think the fan is actually out of balance, I think it's the pulses of air coming off the blades in my case. You might be able to adjust the fan in or out of the "cone" to tune this but I haven't tried."
I think you've described exactly what is going on with my unit. It's right next to my bedroom window and I get a low frequency warble/hum that I only notice in the morning shortly before I wake up. If I go outside it's very quiet and it's not until I put my head nearly on top of the unit that I can hear the low frequency noise. I honestly feel like a bit of a crazy person trying to explain it to my HVAC company since it's so quiet outside but inside it's the kind of noise that's just variable enough to keep me awake. Did you ever try and fix it?
Does anyone have a link to a recommended concrete pad? Something like this https://www.ecomfort.com/Diversitech-UC2448-3/p68192.html?new_bottom=1 or heavier? The 42k condenser weighs about 188 lbs I believe.
Michael, if you're going to do it, I'd go all the way--form and pour at least an 8-12" thick pad, or even 16-18" as Walter suggested. A 2' x 4' x 1' pad will weigh about 1200 lbs. The one you linked to weighs 32 lbs.
The point is to attach the unit to something so big and heavy that the unit stands no chance of moving it in the smallest way.
If you have a pro pour this slab they will have to pay for the minimum delivery where I live it is 3 cubic yards. So it costs you no more money to make it be massive.
If this is a DIY job you will be mixing yourself several bags of premix it was 3-4 dollars a bag again the bigger is better.
Walt
Not to belabour the point but I have never seen a ginormous concrete pad under a mini split. I am not opposed but I thought something like these rubber risers would sufficiently mitigate the issue. https://www.ecomfort.com/Rectorseal-RMB16L/p31117.html
I am fairly confident I can’t get the installer to pour a concrete pad but having them buy something off the shelf and install it they could do. Either way, they’ll have to remove the refrigerant to reset the unit.
Most rubber feet are not soft enough for any attenuation at the frequencies you talk about.
Based on your description the sound you are getting in the house is in the neighbourhood of 40Hz. To get some transmission loss there you need isolators with a natural frequency of around 4Hz, so that means around 0.5" static deflection.
Most of the rubber feet type isolators you see sold would have 1/10 that deflection at best, they do squat for low frequency isolation. This includes the rubber block you linked to.
As other have said, the best solution is a very heavy concrete pad.
You can try getting some spring isolators for your unit as a first test if you don't want to pour the pad. Just make sure they are sized for at least 0.5" (more is better) of static deflection under the weight of your outdoor unit. These add only a couple of inches of height, so you would not have to remove the refrigerant to do this.
Hey M_Sea, did you ever find a solution to this? I'm in Seattle too with almost the exact same equipment, house, and noise issue. I've recently tried 7/8" isolation pads under the mount on the plastic/foam pad that the installer provided and I think it helped a little. I think the idea of mounting on a high mass concrete pads sounds promising. Wondering if you did that or found any other solution.
M_Sea
I have a Mitsubishi unit. I can hear a low frequency hum but only if I stand inside immediately opposite where the unit is located outside. Last winter I did have an incredibly high pitch noise that was something like those high pitch fire alarms. It was so high pitch it was making members of my family feel physically ill and we could not stay in the house. The house was not completely finished until April so we able to mostly avoid it last year.
My installer's technician claimed it was the defrost cycle, but I did not accept that the defrost would produce a noise that would make us feel ill. It was not logical.
I had a different company to do other unrelated work. This "independent" company confirmed that the noise was abnormal. The other thing that this independent contractor confirmed is that the noise was being transferred into the house via the line set (copper pipes). The line set is attached to the heat pump. If the heat pump is vibrating, the line set is going to be vibrating. That vibration continues into the house and there is nothing really to attenuate that vibration along the line set. In my case, the noise was travelling about 25 feet right to the end of the line set (to the only indoor unit I have).
The indi. contractor suggested that if the installer refused to fix the issue to call the manufacturer (Mistubishi) to register a complaint - based on his own experience with these types of issues. He said the goal of the complaint was to get the manufacturer involved in resolving the problem. This would be the last step if the installer gives up on you.
This issue may be less common than other, but you see that J_Sea has the same issue. None of these issues are isolated. I know that if the noise returns this winter at my place and I can't get resolution, I am going to measure the noise level - see link below. If Mitsubishi does not want to get involved, I will seriously consider filing a warranty claim.
measuring noise info
https://www.ccohs.ca/oshanswers/phys_agents/noise_measurement.html
Just some ideas if ever you can't find a solution.
I got the installer to remount the unit on the plastic pad with 1" spring mounts, this removed virtually vibration to the ground and reduced the humming a fair amount. Unfortunately, in heating mode when the unit is either at low rpm or high rpm, I can hear the droning. I surmise that the lack of insulation (!) in my 1946 house is the culprit.
But I also have a slightly vibrating lineset (>150ft of lineset) in normal heating mode, leading to some resonation with the indoor units. No idea what the resolution could be for this, either.
I also have defrost events in the 40-45 degree range (F), periodically 60-90 min. I see no ice on the coil before the unit calls for defrost, but I do end up seeing water dripping from the unit and then about 10 mL of water on the pad. Seems odd that it is calling for defrost in these conditions, but I had the installer check the thermistor and there were no issues or strange codes/blinking.
The best way the system seems to work is once the house is at equilibrium, do not touch it and avoid >3 degree setbacks. I found that at 35-45 degrees it will be operating in heating fine but then I call for a larger temperature increase and it immediately goes into defrost (facepalm)...