Diagnosing and Repairing a Budget Inverter Air Conditioner That Failed After a Teardown
イチケン / ICHIKEN
In 2025, the creator behind this video took apart a low-cost Comfee air conditioner for a teardown review. They concluded that, despite so-so build quality, it was usable, and they reassembled it with the plan of installing it at home. That installation never happened. By summer 2026, when they finally set it up to test, the unit no longer cooled or heated. The video follows their attempt to find what broke, which they suspect was damage from their own disassembly, and to fix it at the component level.
The Symptoms
The indoor unit ran normally. The outdoor unit behaved oddly. Right after power-on, it made a sound as if something inside was moving. After about 20 to 30 seconds, the outdoor fan began to spin, then lost momentum and stopped within a few seconds.
The outdoor circuit board has an indicator LED. When the fan stopped, the LED's flashing stopped too. The creator read this as the board's own power supply collapsing during startup. They also watched the refrigerant line and saw no change in temperature, so they concluded the compressor was not running at all.
They admitted the cause was likely their own handling. The previous video involved removing the outdoor board and desoldering the IPM (intelligent power module), the power semiconductor at the heart of an inverter air conditioner. The creator suspected that the soldering iron's heat, or static electricity, may have damaged it during removal.
Because the unit had already been disassembled, they did not expect a normal repair service to accept it, so they decided to handle it themselves. They warned viewers not to try this at home. A DIY repair that ends in an air conditioner fire may not be covered by insurance or warranties.
Working Through the Checklist
The creator removed the outdoor board unit for closer inspection. The box conveniently had a circuit diagram printed on it, along with an outdoor-unit inspection checklist for troubleshooting, and they worked from both.
Fuses. The first check was fuse continuity. Besides the obvious fuse, two brown box-shaped parts on the board are also fuses. All of them showed continuity, which ruled out a blown fuse.
Fan circuit. The outdoor fan did spin briefly, so the creator treated its drive circuit as working. By the same reasoning, the power input and the bridge diodes that feed it were also functioning. That narrowed the suspects to:
- the PMV
- the temperature sensor
- the four-way valve
- the compressor's IPM
- the compressor motor itself
PMV. This is the electronically controlled expansion valve, mounted deep inside the outdoor unit. The rattling heard at power-on comes from the PMV operating, so the creator considered it fine.
Temperature sensor. They had not specifically disturbed it in the earlier teardown and assumed it was unlikely to have broken, so they did nothing with it.
Four-way valve. This valve is operated by a coil. The creator heard a click at power-on, which they identified as the internal solenoid actuating, so they considered it working as well.
Isolating the Compressor Side
That left the compressor section. The question was whether the fault lay in the compressor itself, its motor, or the inverter that drives the motor.
To test the motor for an open or shorted winding, the creator measured winding resistance. An open winding would read extremely high, and a short would read 0 Ω. The measurement was 2.2 Ω, so they judged the motor to be intact.
They then noted that a shorted motor could itself damage the IPM or blow fuses. To rule that out, they disconnected the wires between the IPM and the motor and ran the unit again. The symptoms did not change: the outdoor fan started, then everything stopped. From this, they concluded that the motor and compressor were not the problem and the fault was on the circuit board, with the compressor IPM as the prime suspect. Every other part had shown signs of life, but the compressor had never run even once.
The IPM Tests: Inconclusive, Then Decisive
The IPM sits under a heat sink. It contains IGBTs plus the circuitry that drives them. For a quick check, the creator used a tester's diode mode. If the IGBTs and diodes inside are undamaged, the diodes should show a forward voltage of about 0.6 V. The readings were about 0.6 V. By this test, the IPM looked healthy, and the creator said at that point they did not know the cause.
Next, they removed the IPM entirely, put the board back in the air conditioner without it, and ran forced cooling mode. The indoor unit started and the indicator LED flashed for a long time. The creator worried this was another failure. Then the outdoor fan started and kept spinning at high speed. The power supplies no longer dropped out.
The creator took this as the conclusion of the failure analysis. The compressor IPM was almost certainly broken, and it had been dragging the rest of the board down. Since the power-stage diode check had passed, they speculated that the damage might be in the IPM's internal drive circuitry. They repeated that they had probably broken it through a poorly done disassembly in the previous video.
Why Repair at the Part Level, and Where to Get the Part
The creator explained that normal repairs, including those by manufacturers, replace the whole circuit board or assembly rather than individual parts. For this model, they could not find a replacement board online. They also did not expect the manufacturer to sell a board to an unknown individual. That left a component-level repair.
The IPM is a Mitsubishi Electric SLIMDIP-L. According to the creator, it is not something you can simply buy at a Japanese electronics parts shop. Its buyers are appliance and industrial equipment makers, so small quantities are very hard to obtain. They expected that asking a distributor for a single unit would probably be ignored. Even if it worked, they guessed the lead time would be several months.
Somewhat reluctantly, they turned to AliExpress, where SLIMDIP-L parts were for sale. The creator stressed that this is not an official channel. They thought most such parts are probably production-line surplus or components recovered from scrapped boards, meaning refurbished. They also noted that a part pulled from a used board might actually be more trustworthy than one that merely looks new, though either way there is a chance of receiving a defective unit.
The part they received looked very close to genuine: package shape, fine print, and laser markings. The creator said they were not sure, but did not think it was a second-hand part. Their guess was that a genuine SLIMDIP had somehow found its way onto the market.
Installing the Replacement and Testing
The creator mounted the new IPM with thermal grease, reattached the heat sink, soldered it in, and powered up in fast-start mode. This time they could hear the inverter driving the compressor, which they had not heard before. It started with a beep and then settled into running. The air coming out was cold, the refrigerant pipes became ice-cold, and the motor drive noise was clearly audible. They declared the repair a success.
Before reassembly, they applied a moisture-proof coating to the board. The original factory coating had been removed during the work, and they explained that running the board without it could lead to damage from moisture or foreign matter. After full reassembly, a final test showed the inverter starting, the outdoor fan running, and the refrigerant lines going cold.
The Creator's Assessment
In the end, the fault was the failed compressor IPM, which the creator believes they damaged during the earlier teardown. They fixed it with what they called a rough-and-ready method: an IPM from an unofficial source that looked like the real thing.
Weighing the effort involved, their view is that replacing the entire circuit board would be the better option than part-level repair, consistent with how appliance manufacturers handle repairs. Even so, they said they were glad to learn that a repair like this is possible.
This video is brought to you by JLCPCB.
Does everyone remember this air conditioner? This is a Comfee air conditioner. In 2025, I did a teardown review of this air conditioner. It's cheap, and the quality is so-so. I wrote a review that gave the impression that it's okay to use it. We had planned to install this at home and keep it running for a long time, but I missed the chance to install it at home. It's already 2026.
So, about this summer, I put it together to test if it worked, and I noticed that something seemed off about how it was working. Let's try running it.
The indoor unit is running without any problems. The outdoor unit is also making a "beep-beep-beep" sound, and it's trying to move. I'll wait here for a while. The outdoor unit's fan is running, but it just freezes like this.
Here's what it actually looks like. Is there something wrong with the outdoor unit? The air conditioning and heating aren't working anymore. The outdoor unit's fan runs for a few seconds, but after that, for some reason, it stops.
After I did a teardown review last year, it was good that I put it together, but after that, I didn't check to see if it was working. I left it as is, and when I finally went to put it on, it seems like things have taken a turn for the worse. It seems I broke something during the disassembly and reassembly process.
So, today, while analyzing the malfunction, I'll find out what exactly broke. And if it can be repaired, I'd like to fix it. But first, let me introduce today's sponsor. Please.
First, let's take a look back at how far we disassembled it in last year's video. After removing the circuit board inside this from the outdoor unit, I had also taken off the IPMs, which are power semiconductors and serve as critical components for inverter air conditioners. If there are any damaged parts, I think it might be a motor malfunction, a circuit board malfunction, or an IPM malfunction.
The IPM seems the most suspicious, especially since I was applying quite a bit of heat with the soldering iron when removing it from the circuit board. It's possible that static electricity may have caused it to break.
Normally, as soon as a problem occurs, it would be easiest to send it in for repair, but since there's no way they'll fix items that have already been disassembled due to our circumstances, this time, I'll figure it out on my own.
I'll go ahead and fix it this time, but if a fire starts in the air conditioner due to a DIY repair, there's a possibility that various types of insurance or warranties may not apply. Please do not try this at home under any circumstances.
Now let's actually get started. Let's take a look at the circuit board as it is, with the components exposed. As I've been saying for a while now, the indoor unit is operating without any problems. Let's take a look at the outdoor unit. Right after turning it on, it makes a sound as if something is moving inside. If you leave it alone for about 20 to 30 seconds, the outdoor unit's fan will start spinning, but it loses momentum partway through and comes to a stop. I'm not really sure.
Since there is an indicator LED on this circuit board, we'll take a look at that movement as well. Just like before, the outdoor unit's fan will start spinning, but it stops right away. At the same time, the flashing LED on the circuit board also goes out. During startup, it feels like the power supply to the circuit board itself is collapsing. I'm looking at the refrigerant line, but there's no change in temperature, so the compressor doesn't seem to be running.
To analyze this a little further, I removed this circuit board unit itself. Let's take a closer look. I removed the circuit board. Helpfully, the circuit diagram is drawn in this box, so using this circuit as a reference, we'll figure out where the problem lies. There's also an outdoor unit inspection checklist so that you can troubleshoot as well.
First, check the fuse for continuity. OK. This is fine, too. There are also two brown boxes. These are also fuses. I have verified the continuity. Therefore, it does not appear to be a fault caused by a fuse.
Next is the circuit for the outdoor unit's fan. In the circuit diagram, it's this part. In the actual circuit, it's around here. The IPM is located under the heat sink. Regarding the outdoor unit fan, as you saw a moment ago, it was spinning, so since there are no issues with that part of the circuit, we'll move forward. The fact that the outdoor unit's fan is running means that things like the power supply connection and the bridge diodes are working without any problems.
In that case, the PMV, temperature sensor, four-way valve, IPM for the compressor, or compressor motor look suspicious.
First, regarding the PMV, this refers to an electronically controlled expansion valve. When I turn on the power, I hear a rattling noise coming from the outdoor unit. This is because the PMV is running. It's located quite far back inside the outdoor unit. It's working when I turn it on, so since this is also fine, we'll move forward.
Regarding temperature sensors, since I didn't mention them specifically last time, I figured they probably wouldn't break, so I haven't done anything in particular. As for the four-way valve, it operates using this coil. I think I heard a "click" when I turned it on. That's the sound of the solenoid inside operating. Since this is also fine, we'll move forward.
Next, I'm thinking this part is probably broken. Let's take a look at the compressor section as we've envisioned it. Is the compressor unit broken? Or is the compressor motor broken? Is the inverter that drives the motor broken? I'll look into that.
First, let's check to see if the compressor motor has a broken wire or a short circuit. You can do this by measuring the winding resistance. If there's a break in the wire, the resistance will be extremely high, and if there is a short circuit, the resistance will be 0 Ω. Since the winding resistance is 2.2 Ω, the motor itself seems to be working. If there's a short circuit, the IPM could be damaged, or a fuse could blow.
Test the operation with the wire connecting the IPM to the motor disconnected. As expected, the symptoms remain the same as before. Just as the outdoor unit's fan starts spinning, the outdoor unit stops operating. Based on what I can see here, it doesn't feel like the motor and compressor are broken; it's definitely inside the circuit board. The IPM, in particular, seems quite suspicious.
As I suspected, the IPM for the compressor is the most likely culprit. I've been able to verify that most of the other parts are working, but the compressor motor hasn't run even once. The IPM is located under this heat sink. This is the IPM. The IPM consists of an IGBT and its drive circuit. I'll use a tester to perform a quick diagnostic check to see if it's broken.
As long as the IGBT and diode in the IPM aren't damaged, it should display a forward voltage of about 0.6 V for the diode. In fact, since the display shows a value of about 0.6 V, the IPM doesn't seem to be broken. Since the IGBT and diode in the IPM don't seem to be broken, I don't know the cause. Remove the IPM for now.
I removed the IPM. Try assembling it in this state for now. Try installing the circuit board without the IPM in the air conditioner. I'll check to see if it works. This is forced cooling mode. I hope it works.
The indoor unit is running. The outdoor unit's fan isn't running yet. The indicator LED in this area is flashing. It's taking a long time. I wonder if that's no good.
The fan started spinning. It's spinning rapidly. That's amazing! Can something like this really happen? If you try turning it on in this state, believe it or not, the outdoor unit's fan suddenly starts spinning rapidly. The outdoor unit's various power supplies aren't dropping out anymore.
As a result of the failure analysis, it looks like the IPM for the compressor is almost certainly broken, and it's safe to assume that it had stopped working. In addition to IGBTs, IPMs also contain the circuits that drive them. It's possible that the circuitry in that area has malfunctioned. In my last video, I didn't do a very good job of taking it apart, so I wonder if I broke it.
I'm going to start repairing it now. Normally, repairs would be done by replacing the circuit board. In fact, even if you have the manufacturer repair it, they will replace the entire circuit board. Regarding this product, I couldn't find a replacement circuit board online. Even if I ask the manufacturer to sell me a replacement circuit board, they certainly wouldn't sell it to someone whose background they don't know, so this time, we'll be repairing it on a per-part basis.
So, next I'll buy the parts I need for the repair. The IPM uses Mitsubishi Electric's SLIMDIP-L. However, it doesn't seem like you can just buy this SLIMDIP easily at electronics parts stores in Japan. Since only home appliance manufacturers and industrial equipment manufacturers buy them, it is extremely difficult to obtain products in small quantities. Even if you ask a parts distributor to sell you just one, it'll probably be ignored completely. Even if it were possible, I think we're talking about a lead time of several months or something like that.
So, even though I really didn't want to do it, this time, I decided to source the items from AliExpress. For some reason, Mitsubishi Electric's SLIMDIP-L is on sale. Of course, this isn't the official channel. I think most of them are surplus items from production lines or refurbished items recovered from discarded circuit boards. If it looks brand new, it might be a refurbished item, so components recovered from discarded circuit boards might actually be more reliable. However, if you do that, there's a chance it might be a defective unit.
Here's what I actually received. It looks pretty much like the genuine article. It's not just the shape of the package; even the fine print and laser markings are pretty close to the real thing. I'm not sure, but I don't think they're second-hand items or anything like that. A genuine SLIMDIP somehow turned up. I suppose that's how it is.
Next up is this SLIMDIP from an unofficial source. We'll mount it on the circuit board. Apply thermal grease and install the heat sink. I'm going to solder.
Let's turn it on right away. Fast startup. It moved. I hadn't been able to hear the inverter making any noise until now. This is the sound of the inverter running the compressor. After it makes a "beep" sound at first, it works like this. That's nice. The wind is cold, too. The refrigerant pipes are also quite ice-cold.
I'm using an IPM to run a motor-driven compressor, and I can clearly hear the drive noise, too. I think that's fine. So, the repair was a success.
I'll reassemble it, but first I'll apply a moisture-proof coating. Since we've removed the moisture-proof coating that was originally on the circuit board, if you continue to use it as is, it may be damaged by moisture or contact with foreign objects. Once you've finished assembling it, turn it on again, and we will perform a final functional test.
I just heard the inverter start up. The outdoor unit's fan is also running. That's nice. The refrigerant lines have also gotten cold. The repair is complete.
So, we identified what broke in the air conditioner and completed the repair by replacing parts. As for the cause, one of the two IPMs—the one for the compressor—had broken down. I think I might have broken it when I took it apart in my last video. In the end, I fixed it using an IPM that looks like the real thing, obtained through unofficial channels. The repair is complete, using some rough-and-tumble methods.
Considering the effort that went into it, it looks like it would be best to replace the entire circuit board instead of repairing individual parts. Even with the approach we took this time, I'm glad to know that it can be repaired. In fact, the repairs performed by appliance manufacturers are not on a per-part basis. After all, the replacement will be done on a per-circuit-board or per-assembly basis.
If you found this video helpful, please give it a thumbs up. If you have any questions or concerns, please feel free to leave a comment. Thank you for watching the video all the way to the end.
Article published