Mitsubishi Mini-split fried twice due to a power quality problem of some sort.

[Gasquet]

I have a Mitsubishi Mini-split that has had all the low voltage boards fried twice during a short power outage. No fuses blown on the LV boards, whole house surge protector didn't trip/protect, no breakers tripped. The third heat pump is off line until I can solve this challenge. HVAC tech and commercial electrician have both tapped out with the issue not yet resolved.

The first time this happened the only thing of note was that the single phase voltage was a bit high at 249V. Spec is <253V, which is close/concerning. Utility says it's within spec so not their problem.

After repair a second surge protector (why not) was placed on the circuit to the heat pump. Another power outage happened frying the low voltage boards a second time in less than three weeks. Now and only now after the second event, there is a voltage imbalance between the two hots and ground of about 9V (varies). The imbalance goes away when all loads are shut off and scales roughly with the number of loads turned on and the expected size of the load.

Grounding was vastly improved with the help of a commercial electrician with no impact to the voltage imbalance.

Mitsubishi tech support says the voltage imbalance max spec is only < 3%. This seems to be a weakness that potentially explains the vulnerability of these systems to damage. It's hard to source replacement boards in the U.S., likely because the replacement rate is abnormally high. I was reading the following thread https://forums.mikeholt.com/threads...re-the-same-conductor-with-ac-voltage.102415/ which talks about how the logic signals (low voltage) is sent through the same conductors as the power to the air handler (yes, some models can be ducted).

I don't want to go through this again. The heat pump is currently off line. It's also getting hot out and window units are struggling.

Voltage imbalance is likely an issue with the neutral. But if after a visit from the utility next week results in no willingness to take any corrective action, what do I try next to protect this system? I feel like installing a true isolation transformer, and referencing the secondary to ground to remove the voltage imbalance at the heat pump. Love the heat pump technology overall when it is working. I think the design is not robust enough in my opinion and at $600/board they are making a profit on this design flaw susceptibility to slight over voltage (253V ain't great) and voltage imbalance of <7.2V. That's going to be violated pretty often during a power outage.

It doesn't help that I am experiencing a short power outage roughly once per week since the end of March with only a few exceptions.

What would you do to baby these systems. Buck/Boost transformer or full on isolation transformer?

 

[ptonsparky]

Why are you experiencing short power outages? Utility wide, or just you? Just you? get someone in to find the problem. Underground service or overhead?

Your unit is looking for the 240 volt +-, Line to Line. It does not care what the Line to neutral voltages are.

 

[petersonra]

First off, no amount of grounding will improve your situation. Whatever money you spent on that was a total waste.

Surge protectors are not helpful if there really is a design defect in the boards themselves. They don't clamp down on a few extra volts. You need to see a few hundred volts extra before they start to do anything. The clamping voltage depends on the type of spd but in some cases it may be well over 1000 volts p-p.

It's entirely possible that the voltage difference from line to ground is just normal voltage drop on the neutral conductors. What you need to be measuring is the voltage from l to n. If it is not the same you might have a loose neutral upstream somewhere. But if it is a 2 wire system it just won't matter much.

 

[ptonsparky]

Coordinating the POCO and a knowledgeable EC is suggested.

 

[ptonsparky]

Nor does it care what the Line to Equipment Ground or Line to Earth voltages are.

 

[don_resqcapt19]

...

Voltage imbalance is likely an issue with the neutral. But if after a visit from the utility next week results in no willingness to take any corrective action, what do I try next to protect this system? I feel like installing a true isolation transformer, and referencing the secondary to ground to remove the voltage imbalance at the heat pump. Love the heat pump technology overall when it is working. I think the design is not robust enough in my opinion and at $600/board they are making a profit on this design flaw susceptibility to slight over voltage (253V ain't great) and voltage imbalance of <7.2V. That's going to be violated pretty often during a power outage.
...

The line to neutral or line to ground voltages are not seen by the heat pump, so that is not the issue. With a line to line nominal voltage, you can expect to see up to ±10%. So your utility saying 253 is their maximum is well below + 10%.

In my opinion, a piece of equipment designed for a nominal voltage system of 240 volts that can't successfully function on 260 volts it poorly designed junk.

As far as the utility looking at the line to neutral voltage, they need to do that with a device such as the "beast of burden" of some other device that provides that type of loading. That type of testing will show if there is a neutral issue on the line side of the meter. A load side neutral issue is on your electrical contractor to find and correct.

 

[ActionDave]

Grounding was vastly improved with the help of a commercial electrician with no impact to the voltage imbalance.

When I get ready to retire and lose the will to do anything worthwhile or productive I think I get some business cards and a website and sell myself as a grounding expert. I bet I end up making more money.

 

[don_resqcapt19]

First off, no amount of grounding will improve your situation. Whatever money you spent on that was a total waste.

Surge protectors are not helpful if there really is a design defect in the boards themselves. They don't clamp down on a few extra volts. You need to see a few hundred volts extra before they start to do anything. The clamping voltage depends on the type of spd but in some cases it may be well over 1000 volts p-p.

It's entirely possible that the voltage difference from line to ground is just normal voltage drop on the neutral conductors. What you need to be measuring is the voltage from l to n. If it is not the same you might have a loose neutral upstream somewhere. But if it is a 2 wire system it just won't matter much.

Exactly...improving the connection to earth never solves any electrical problem. There are cases where it can "mask" the problem, but it never solves it.

 

[ptonsparky]

When I get ready to retire and lose the will to do anything worthwhile or productive I think I get some business cards and a website and sell myself as a grounding expert. I bet I end up making more money.

Don't forget to give yourself time for those naps that seem to become more necessary after retirement.

 

[winnie]

What is connected to the low voltage boards? Are they completely internal to the machine, or do they have external low voltage wiring?

How is the external unit mounted? Brackets on the wall or on a plastic pad, or sitting on concrete?

Jonathan

 

[Fred B]

Some of what is said in op sounds like a bad neutral but not an open at service. An outage or shut down will temporarily improve the situation until the impedance increases on the neutral due to loads applied. The imbalances seen could actually be worse than measured and creating the situation burning out the control boards.

 

[ptonsparky]

Some of what is said in op sounds like a bad neutral but not an open at service. An outage or shut down will temporarily improve the situation until the impedance increases on the neutral due to loads applied. The imbalances seen could actually be worse than measured and creating the situation burning out the control boards.

IF the control boards are referencing a line to neutral or ground some where.

One conductor failing or faulting to earth frequently means the others are in the process of failure. Sometimes it shows up via L-L voltage, others L-N.

It would help to have simultaneous current and voltage measurements to determine if it is an undersized neutral conductor.

but

A competent electrician with a set of test lights and a couple 120 volt 1500 watt heaters should be able to find a neutral problem.

 

[David Castor]

I'd be curious to know the reason for all of the outages. Switching transients and or reclosing can create voltage spikes, although these are usually dampened quite a bit by the time they get through the system to your house. I'd probably send the power company the invoice for your repairs and see what they say. Since you're in Austin, I suspect you get a fair amount of lightning.

Surge suppression is most beneficial the closer it is to the equipment being protected. I'm not clear where this second suppressor was installed, but I'd suggest putting one as close to the outdoor unit as possible - like on the outdoor fused disconnect. I'm not convinced this will solve THIS problem, but that is best practice. Every installation I see around here has a suppressor on the disconnect box. And we don't even get lightning.

Good luck.

 

[Gasquet]

Nor does it care what the Line to Equipment Ground or Line to Earth voltages are.

If that were true, then why does it have a <3% specification on voltage imbalance? The compressor may not care, but the low voltage boards do. There is some digital communication on the power lines implemented, so this is where I think the weakness comes in. Go back to the link above and read the other thread. It's pretty relevant.

 

[Gasquet]

I'd be curious to know the reason for all of the outages.

I know there are power quality issues from the utility that everyone on this group is focused on. But it is a utility. They just do open circuit tests and say everything is fine and move on. I am in conversations with the utility. No moment on their end yet. ASSUMING these power quality issues never get fixed, what can I do to protect the heat pump?

 

[Gasquet]

Surge suppression is most beneficial the closer it is to the equipment being protected. I'm not clear where this second suppressor was installed, but I'd suggest putting one as close to the outdoor unit as possible - like on the outdoor fused disconnect.

Done. It failed to save the unit the second time.

 

[Gasquet]

What is connected to the low voltage boards? Are they completely internal to the machine, or do they have external low voltage wiring?

How is the external unit mounted? Brackets on the wall or on a plastic pad, or sitting on concrete?

Jonathan

One is inside the air handler. Two are inside the compressor housing. I don't have the benefit of circuit diagrams unfortunately. HVAC tech calls Mitsubishi tech support and they spend about an hour going through specific measurements and then conclude a board is fried. It gets replaced and the process repeats until all three boards are eventually replaced. The air handler gets power and communication from the compressor unit (both on the power lines). There is no low voltage wiring and unit came with a wifi transmitter to the thermostat.

 

[Gasquet]

First off, no amount of grounding will improve your situation. Whatever money you spent on that was a total waste.

When the ground rod is sitting horizontally no more than 3" in the soil on top of solid limestone, one doesn't really have a ground. I did need to improve ground. The secondary structure didn't have a ground.

All of this is independent of figuring out how to adapt to a power quality problem that likely will not be fixed.

 

[Gasquet]

The line to neutral or line to ground voltages are not seen by the heat pump, so that is not the issue. With a line to line nominal voltage, you can expect to see up to ±10%. So your utility saying 253 is their maximum is well below + 10%.

A bit of miscommunication. I wasn't clear. The 253V max spec was for the heat pump, not a utility spec. So using a buck boost transformer to step things down a bit is one way to make and adaptive improvement at the heat pump.

As far as the utility looking at the line to neutral voltage, they need to do that with a device such as the "beast of burden" of some other device that provides that type of loading. That type of testing will show if there is a neutral issue on the line side of the meter. A load side neutral issue is on your electrical contractor to find and correct.

Something like this is planned for this week. I have sent in enough line monitoring data that they are at least willing to come out a second time and re-test things to show me that the issue is not on their side of the meter. Assuming that test passes, I still have a problem without some kind of corrective action.

 

[Gasquet]

Focusing the discussion. Would you use an isolation transformer, or just a buck boost transformer (assuming that this heat pump does indeed have a reference to neutral in the low voltage circuitry)? Also, assume that for now, we will not find the problem with the neutral. Normally there is heat associated with an issue on the neutral. Can't find heat.