Anything I could add pales in comparison to the intelligent comments I'm reading. Just to let you know, I give you a TON of credit for giving this a go and hope like hell you get it figured out :-)

With that sort of a blow up the gate driver ic's are probably also toast. There will also be resistors in series with the gates and they may be open. There will be zener diodes protecting the gates and they should be changes regardless. You should change all the IGBTs whether they test good or not. Then you need to see if your short is gone and if not test all remaining semiconductors removing them as necessary until you find the short. Also be sure to remove all the blackening from the board and the heatsink as it will be conductive. You may get lucky but it's a lot of work and you still blow it up the first time you turn it on. 20 years experience repairing switched mode power supplies up to 24kW. Gillian :)

So basically these high-frequency inverters have three main sections 1. High-frequency inverter that turns the 48VDC into ~15KHz+ 220VAC, then rectifies to turn it into 380VDC. The super high frequency lets you get away with a very small transformer (2x the freq= 2x the power for the same transformer, so a 15khz transformer can carry as much power as one 60hz 200x as big. NOTE: The transformers used have very little iron in them as iron saturates at higher frequencies. ). These were the K100E08N1 (80V-100A) mosfets and that sideways transformer you see there, and also 2 diodes somewhere on that side. MOST of the time this is done with a transformer center tap and a half-bridge, the middle of the transformer is connected straight to the + and the mosfets take turns switching - on the left and right sides. This creates a square AC wave which works super efficiently to boost the voltage but at 15KHz it's not useful whatsoever, thats why we rectify it to ~380VDC and store it in some caps. 2. 380VDC to 220VAC. The part that blew in this case. Since we're working with high voltage, the current is low and so we don't need nearly as many or powerful mosfets. Generally these are IGBT's rated for 600V. This will be a FULL H bridge configuration, so requires at LEAST 4 IGBTs, or a multiple of 4. And generally has some sort of inductor+caps to smooth out the SPWM into a full sine. 3. MPPT,separate circuit which generally consists of big inductors and mosfets to drop the solar inputvoltage and increase current, Depending on the setup it may be used to directly charge the 380V caps for much better efficiency. GENERALLY when these cheapo inverters go kaboom, it's not just the mosfets that fail. This is because their circuits are designed so on the edge that the second the mosfets fail their gates short out and the gate drivers+resistors are so low-power that they blow as well (in the split second that the inverter blows up). So my bet is that the gate drivers + resistors are probably blown. Hopefully the rest of the spwm circuit isnt blown, cause those are probably a bit complicated if they aren't off-the-shelf EGS002 modules. ALSO when working with any switching circuit like this always try to use a current limited power supply, 1A is more than enough to get it running. If there's a short, 1A won't do anything to the mosfets.

VERY LIKELY the MOSFETs on the OTHER heat sink are also damaged. Then there are DRIVER transistors feeding the MOSFETs and they might be damaged as well. You also MUST check BOTH ways with the diode setting on your meter. One way there should be '.4 or .6, and the other direction should read OPEN (OL)' You MUST check diodes in FORWARD and REVERSE polarity (.4 and OL)

To check the transistors you could get a LCR-T7 transistor/IC checker. Also, totally support the idea of putting a current limited supply on the battery input to trace it back. But if you don't have a current limiting supply, also just work back from the battery connectors for components that are across the rails and which could be shorted then pop them out.

The guy who this was is a NICE guy. Glad to see you guys worked out a deal. God Bless.

Good call on getting transistors from a reputable store. Very interesting video. I enjoyed it.

I love disassembly videos. I hope you managed to get it up and running.

I believe the measured "dead short" could be the large caps at the battery side in the inverter. You could use some resistance (4 pieces of 12V light bulb connected in series) to charge up those caps or checking whether there is a real short or not.

Appreciate the video - even though you have a dead-short. This is real life and I (for one) appreciate you sharing :) I overloaded a Reliable 3000w by trying to run a 12,000BTU AC - and it also resulted in a dead short on the 48v battery input. Looked at things but, like you, did not see anything I could repair. Will be interested to see if you make any progress on this.

Great video. Looking forward to learning with you on this one

Good effort made! I don't know the kind of multimeter you are using but confirm this: set the meter to buzz on a short circuit as you did, connect the +ve lead to the inverter input battery +ve and the common to the inverter negative input if it buzzes its a real short, open the inverter where you find the 16 transistors connected to the heat sink, some of them if not all have failed, identify them and replace them to solve the problem. sometimes it may not have shorted but it will buzz when you change the meter leads. good luck.

@4:50 is a security screw. You need a cheap component tester. They are like < $10 now and no tech should be without. Diodes usually die shorted. If there is enough power, they will blow through the casing. If you make a detailed closeup video of the boards and traces of both sides, I for one would enjoy troubleshooting it from afar. We can all troubleshoot it from our end. Tracing the input to find the cause of the short, is the first step. You meter can't provide the threshold voltage for the IGBTs.

8:05 Those IGBTs have a turn on voltage of 4 volts which is probably higher than what your meter produces

Hey, I have the same inverter, and also had the same problem. If only I had seen your video before I repaired it. I also had a dead short on the terminals. The culprit was those smaller transistors on the other side of the board. They visually look okay but they’re all shorted out. My inverter is EXACTLY the same but a different brand name.

If you don't have a 48 volts constant current power supply, use a load between the positive of the battery bank and the positive of the inverter. A good load for this test can be heat element, an 120v iron or incandescent lamp 120v 500w or even four 12 volt 55 watts halogen bulbs in serie. Remember that most MOSFET are very sensitive to static energy discharge. Just a little spike in the gate could short permanently. To avoid the you need to be touching ground.

The caps and components related to the 48V system will be rated for around the same voltage, caps should be around 63V rated for example. The HV and LV sides will be separate systems and the 48V won't share the 450V big mains caps with the HV side. The only other thing I would check is isolation of the heatsinks and to ground

9:48: That pad behind the FETs is not only a thermal pad, but also an electrical separator. The metal pad on the back of the FET is also connected to one of the pins (I think the Collector, but not sure). So to keep them from all shorting out, the pad is used. Not all the FETs will be activated at the exact same time, or turn off at the exact same time. They pulse on a different times to help give a sine like wave. Then smoothing capacitors and the inductors (the donut looking thing with thick heavy wire wound around it) take the somewhat jittery wave and make it a nice clean sine wave.

Try to find a schematic and trace the circuit to determine what components were exposed to the mains voltage. I would also check all the caps and resistors that are located near the transistors for short or open conditions. Unfortunately, you soldered all the transistors back in. When you had them off, would've been the best time to check for shorts. Good luck.

When you need to to remove those power devices, I would cut the legs first since there is no need to keep the blown devices intact, and that will also minimize damaging the board especially the multi layers boards. BTW, the Gate driver devices may be also damaged too.

2:27 - one wire (2 ends) is an inductor. 2 wires or more (4 ends or more) is a transfromer

usually igbt have a zener diode and resistor connected to the gate for protection from transient voltages , when they blow they usually take them out too , you have yo follow the gate of the igbt back to the zener and resistor likely tiny little surface mount , as for the short on the battery side sounds like there is still a bad igbt , they can also be damaged if you take too long soldering them in , you need to set your soldering iron to a lower temperature and do the soldering within 3 seconds per pin , give them time between pins to cool down as the heat can damage them. great video well done , keep up the good work :P

If you have a scope and a signal generator you can build a magnetic probe and trace on the 48V battery connection. Test setup by shorting function generator leads together and move magnetic field probe near test wires. Function generator is on full power (~10Vpp ,20khz sine wave). Typically the output is around 50 ohms so the current is roughly 100mA peak. Magnetic probe is 50turns of magnet wire maybe 1/2 inch diameter. A small solenoid coil without center core would work. Coil connects to oscilloscope input set on 10mV/division. Holding the coil near the shorted function generator should show a sine wave the closer the probe is to the signal and coil core lined up with transmit wire. The coupling is magnetic with this setup so you can trace the shorting current to the offending bad part. With power off (0V on 48V battery connections), hook up the signal generator to the 48V connections and start probing the wire/traces. You are measuring coupled current and not efield with this setup. It also possible to wrap the coil with copper foil and connect to the scope ground to reduce any e field pickup. E field shield. Good luck and let us know how it goes. Kevin

At least you got about $300 worth of individual components, at single quantity prices. So still seems like a good investment for someone like you ;) When checking battery input impedance with an ohmmeter, always reverse the leads and try it in opposite polarity, in case there is residual capacitive charge giving u a false resistance reading. Remember an ohmmeter actually injects voltage from its internal battery into the circuit to measure resistance. The odd screw is a security head. There are bits for that. Keep it up :)

2:32 Center bottom there is an 8-pin IC. It looks like there is a white spot on top. Make sure this is not a blow out point on the IC 5:22 The red parts are high voltage capacitors. The one on the far left looks like it "could" be damaged. Make sure there is no damage to the casing As for the FETs, I would have replaced all of them. Don't take any chances with this kind of failure.

Loved Your video, The input short is at the primary side(input) , pls check the input side, and try the voltage injection method to dig down to the problem, To Check MOSFET or IGBT (4066) you need to have a voltage gate supply of 5v at least then u will see approx 0 Ohm by C-E to verify the working of the transistor. Appreciated Your video. The inverter was damaged by a back-feeding grid, this is Not a Gird Feed inverter. keep Output of this inverter isolated from input and never mix up.

The five in a row IGBT's are four output PWM H-bridge, and fifth one I believe is charger buck switcher, along with TO220 diode next to it. The four output PWM IGBT''s were blown as result of applying AC input to inverter output. Other four IGBT at other end of heat sink are synchronous rectifiers on output of transformer. These four were likely okay. When output PWM IGBT's are blown, need to check low value series gate chip resistors and four 8 pin IC's, high side/low side driver devices on back of board that drive the PWM IGBT gate drive. They can blow out as result of IGBT's being shorted.

9:55 This is where a toothbrush comes in handy. Dip it into the alcohol and then brush the board. It removes the crud easily. a few seconds later "i suppose you could use a toothbrush" 😄 yup, exactly!

Transformers have at least 2 windings. Primary and secondary. What you have is a coil (inductor) used to filter out noise, or most likely part of a low pass filter or current inrush inhibitor. When it's called L1 or L followed by a number, it's an inductor. T1 or T any number, is a transformer.

Great video dude, great responses dudes....

I would check the boost state FETs. The battery voltage needs to be boosted high enough to create a 120V sine wave and it looks like it uses that EI core transformer in a push pull configuration next to the H bridge IGBTs. If the input FETs are blown then you would get a short measured across the battery input.

Did u try to charge the start capacitors before power on? And did you check all the fuses?

The igbt need about 6 to 10 volts to turn them on. When you checked the dc input you have the red connected to negative showing a short. The input probably has a diode across it for protection. Test with the red from the meter to positive and you should see the meter change as it charges up the capacitor inside.

yes these are High Frequency inverters I have been repairing quiet alot of those , all Mosfets have to be replaced as well as checking the reservoir capacitors, there are also ICs called Gate driver ICs which drive the mosfets those also have to be checked they are 8 pin DIP ICs which drive the H Bridge of the Mosfets, The other thing to check is the IGBT transistors on the other heatsink. there is also a 2pin rectifier near the IGBT transistors to be checked.

the drivers for the gates usually blow when the fet does....fun job

Here's some info to chew on. First, even though the MOSFETs have been replaced, most likely the driver ICs or driver transistors are toast as well. They should be tested and replaced as needed. To test the MOSFETs, turn you VOM to the diode test (usually on the ohm range) as the diode test has enough voltage on the leads to turn on the Gate of the MOSFETs. Oh, almost forgot, there is usually a small resistor in the gate lead of the MOSFETs. Check those. Usually, but sometimes you're lucky, when the MOSFET let the magic smoke out, they take out those resistors as well.

Scorched rosin flux is hygroscopic (absorbs moisture from the air) and can cause an intermittent high impedance short at a transistor gate. Circuit works on dry days and fails on humid days. (learned from experience) Acetone (used in fingernail polish remover) works better and faster than alcohol to remove flux residue. Use pure acetone (from a paint store) as fingernail polish remover usually has lanolin or other additives. The old Fluke 8050A multimeter could measure very low resistances and that allowed tracing the location of a short. If you find a currently available multimeter with similar capability, please let me know. A short in a PROM chip was found by following the PCB circuit to the point of lowest resistance. The chip enable was internally shorted to ground and the chip still worked when removed and was copied to a good replacement part. If the short resistance increases when you remove a part of the unit, it is likely that the removed part contains at least a part of the short. This makes life easier when looking for multiple shorts as caused by a massive failure. Hope this helps.

I think the IGBT's switch the AC side voltages... if your dead short is on the DC inputs, then the short is likely somewhere on the DC side of the board. Try isolate the input board from other boards and test again. My guess is it's one or more of those small transistors on the other side of the unit

start by taking one board out check for short move to next and so on

I would say the fact that the dead short is there even powered down and it’s easy to see with a meter should make finding it relatively easy. Wouldn’t bother with passives initially a dead short of that magnitude ie 0 ohms is probably either as somebody suggested a physical short on the board or another semiconductor failed. Second samuels suggestion with the thermal check recommend a current limited power supply but would also suggest disconnect the smaller boards and check them all separately it is a bit safer and helps to narrow down quicker. Just go through the semiconductor stuff first I would highly doubt a passive component as they generally don’t handle lots of power and if they do they generally blow open and if this is at mains voltage It’s hard to miss. Basically check all the transistors that look biggish then work your way down oh and from memory that inverter might have a large diode on the 48 volt supply don’t know why that would blow from the other side of the inverter but it’s easy to test.

What is inverter soft start failed and what precautions should be taken to avoid this circumstances?

Fun fact, but the "shorting" on the battery side may be a bad reading due to the big capacitors parellel to the battery charging by the multimeter. Try holding the multimeter leads for many seconds on the battery input and see if you get a non-short circuit reading.

How do I adjust the factory settings for such an inverter?

Great Video Thank You! God Bless!

Thank you sir,,you showed a video...good idea in troubleshooting

Thanks for info & video

I believe the battery soft start which looked health physically OK has a dead short. That 19 mosfet there is a dead short. Also the voltage you are using to charge the gate and source across maybe low. Your meter has 9 volts.

I'm sorry Brother, I think your very smart, I would not have a clue how you could check the continuity, on all the part's, when there is so much solder paths, How do they test the compactors. Thank you for your reviews. Sincerely your freind from Mo Rick and God bless you and your family and are Great Country

AC backfeeds will damage all the DC Capacitors in Line to the Batteries input so I will advise you check the Capacitors. Also check the other MOSFET bank on the other sections of MOSFETs... Any AC misfire is likely to damage all DC amplification transistors especially in an H-Bridge Configuration. David Oniya ( Nigeria)

Hello. Can you tell what is that rather small card you removed last before taking out big board out of the casing? My inverter is not charging and i think that card is acting up so i need to look for a replacement. Thanks.

In my experience, counterfeit transistors can actually work just fine. Same with other chips, a lot of my DC to DC converters use blatant TI rip offs but work very well.

Good job.

short on battery input could be capacitors low ESR or capacitor short on battery input. MPPT stage with short ( protection diode or mosfet) DC / AC inverter with short on input ( diode or mosfet shorted ) Good idea to limit amps on battery input and look for any component heating up with FLIR cam

Hi you can't open the igbt transistor in this way. but you can test the mosft in this way by clicking the gate of the mosft it will operate

well done bro..

Normally in a inverter there will be a two set of mosfets .N-CHANNEL and BJT .. n channel are use to switch low voltage DC and IGBT are there to switch high voltage DC to main AC

Sugiero poner en serie con el positivo de los 24 volts una lampara de filamento de alta potencia. Puede ser una de luz alta/baja de carro. Si al conectar la alimentacion, la lampara se enciende y no el inversor, entonces hay seguro un componente en corto. No te des por vencido tan rapido. Debe haber una fuente switching de baja tension para los integrados de control. Es lo primero que debe arrancar al conectar las baterias, puede estar en corto. Te deseo mucha suerte desde Argentina

That soldering wick is containing flux too and therefore it looked dirtier afterwards ... cause it melts away and you can see it in the first solder wick points ... you see the heat "melting something" before the tin kicks into the copper.

was that a fuse to left of two caps @ 12:38. F3 maybe. Wouldn't cause a short but worth checking. Might try to find schematic. Nice replacement though. I though you had it.

Hello, 3000 watt solartronics full sinus inverter fault light is on. When the inverter is turned on, a relay tripping sound is heard, then the fault light is on, what do you think could be the problem?

Please can you give me a guide lines. I have a similar Inverter.. But all mosfet transistor are ok.. I tested them. But the the inverter did not come on at all. Not display?

@5:56 we love those Ali price but man it is crappy when it goes wrong... Thank you sir, good luck!

Hi All I have one of these came with 8 pannels, the chap said it packed up so he bougt a new inverter, can any one point me in the right direction to start diagnosis, can find faults was a heating engineer long time ago but need some guidence manufacturers litruture would be a good start. cheers.

so this solar inverter start works properly after reflow bad transistors?

I would have checked the tracks while the MOSFETs where desoldered, It looked to me that some are shorted If the bord such black burn marks you should remove all the burned PCB because burned charred PCB material can be conductive. I don't know how many layers that board has but it can be that some inner layers are shorted, in any case, you should probe the board and the traces when the MOSFETs were out.

There should be a protection diode across the positive and negative for reverse polarity protection.

I have a mppt 2400w 24v planning to put a shelly plus 1 using (12v) on the switch(AC) is that possible so i can turn the the switch(AC) as a wifi/bluetooth?

The comments on this video are mind boggling, with lots of great ideas; I'll say here one of my favorites to find shorts: you feed in, a VERY LOW VOLTAGE, and feel for a component heating up; KEEP increasing the voltage and checking; the hot one is the shorted one ; i have ONLY used this in repairing computer motherboards, but i guess that if you scale it up,it should also work on Inverters?

4:55 That screw is a type of security screw. The specially designed tip has a hole in the middle of it where that post goes into the bit. Altho, this is the largest one of those I've seen and especially odd that there'd be one in one of these devices. I would expect to see this on the "outside" of the case to keep someone from going in, not from removing the pcb.

Did you check if the board was still a dead short before replacing the FETs? It could be a transient supressor.

Hi, Thank you for your video, can you help me. I have a 5.5 kw 48 volt hybrid inverter my there is fault 59 after its pass 03..

Thanks for the info. Can you tell me the part no. For the Mosfits for both sides thanks ac and dc

The screw you thought was defective is actually meant to keep it from being removed. The tool to remove it has a hole in the end to accommodate the pin in the center of the screw plate. They are usually a Torx head, Phillips or Allen head ( predominantly Torx) made to keep anyone but a repairman from removing. Normally, they are placed on the outside (microwaves have them) but, I assume it is there due to the capacitors. Did you recheck all of plug-ins you removed?

Very good

how can i get one damage inverter like yours

Hı , what is the big transformer in the middle for? And this inverter 2 MPPT or 1 Mppt

The mosfet 3MNB0120-3 is damaged in my inverter, what should I replace it with?

Did you check inverter drive?

I have a PIP-5048MK, not sure what the difference is between it and the MG model

Here's an item for your electronics bench. I find it very handy. ebay: Multifunction Transistor Tester Diodes LCR - TC1 Full Color Graphics around $20. LCR and transistor checker. runs off of USB power. Not sure it has enough voltage to check fets however.

hi, check if the blue transistors are shorted or not (near capacitos) 12:40/13:40

In high frequency online ups the AC output neutral is same as battery minus. So i guess due to the incident every component between these are affected (DC fuse, mosfet, rectifier diodes, transformer, capacitors, IGBTs and other components which supports these power components). Might be the reason why it is showing a short at the battery terminals. Not sure if this might add some more help.

When I asked if these could be AC coupled they said there would be catastrophic failure. I guess she was right.

You want around 100v rated caps on the dc charge side , the factory 63v is crap they go to 50v when charging so expect reduced life if not replaced .

Chek the 4 optocoupler of the damage mosfets

You are asking for advice, however, in your statement that the unit was damaged by feeding grid current into the unit, that made understanding near impossible, as most of these hybrids are designed to feed grid/gen current in to charge battery during non pv periods. Some clarification on that is advised prior to asking for information from us. Does this unit in fact allow grid/gen input? If not, where did the 'electrician' feed such into the unit? Without this info we would be merely guessing. There are triacs in some of these that require actual voltage from the battery and/or grid to activate the circuit. It would be useful to test each terminal to the case to see if it was a dead short or a mere design condition you've encountered.

please make an update video on these repair attempt

Hello what kind of circuit board is that at minute 2 ? we have 2 inverters, one with this board and one without this board. thanks for an answer.

For me I have a problem.my display went off . what can I do make it display again please help.

Solar inverter 3kw, when we switch it on , it gets off automatically without giving error code, what is the problem?

5:58 the probes were backwards. you actually tested the internal parallel free-wheel diode that you can notice in the datasheet.

Morning Im Lost with an PIP-5048 Inverter There are no Error codes All FET's And Igbt are in order When power from util nothing switches on when powered from battery Inverter switches on When powered from both Inverter indicates utility Voltage 238 V but battery doesnt charge Inverter Indicates Battery Charging But no current or Voltage to Battery When Inverter on and running battery can be switched of and inverter stays on Any ideas on where to look next

Just a trick, to turn on mosfets on that way only works on mosfets or other components that have an gate and not a base. Like someone else said it could also require a higher voltage to turn them on

I use quite a few Aliexpress components. They all seem to have worked fine. Sometimes they are counterfeit or out of spec rejects but mostly they seem to be overruns. But I would NEVER take a chance in an inverter or similar with potential fire risks.

Bonsoir merci Pour un convertisseur hybride mppt qui s'allume et et ne donne pas de courant erroe 01 fault et les ventilateurs sont en forme que faire

i have some electronics back ground so a damage inverter could be a nice project for me

The filled screw probably picked up solder during the wave soldering process. Should pop right out with a scribe tool or other sharp pointed, hardened tool.

check if the gate resistor are ok and the mosfet driver.

My Inverter is reading the volts from my panels but not the amps or watts what was actually wrong with the one in the video before working on it.

Try to call or eMail MPP Solar in TAIWAN and you will get pretty good answers and support ... add all the details you had heard from the owner and make a reference to this video. The pre- and after sales are brilliant and you can not throw this manufacturer in the typical chinese pot cause it is not that what you consider to be China cause Taiwan is the island and the american allied known for quality and sustainable products. No, I am not a Taiwan employee, just a buyer of their products and I have benefitted a lot from their pre-sales service when I was not sure what product family I should or could take for my project and they had helped a friend with a 5 year old inverter where something had been blown up after a thunderstorm and a blitz. He got a spare board within 3 instead of 5 days (from taiwan to germany) and his unit was working again. And the technicians are willing to help to repair just the pieces that are dead if you are familiar with soldering. They try to help and not to sell their boards which is quite unusual. So you might give them a try cause that would be the fastest and best solution you can get and report about in the upcoming video how you were able to fix it. Better than reading and judging and following all those comments of pros and amateurs, cause mpp Solar is the developer and manufacturer of these inverters that are sold all over europe under different brands and logos. They know their stuff and I guess you will understand in a second how valueable their support can be once you get in touch with them. Might have saved you 6 months waiting for spare parts and ideas. Just write a short mail with a link to this video and then you should get support cause my friend did it exactly that way with his 5 year old inverter. Might be that they ask for measurements and then he made another update video and send the link to them and found the cause within 3 days and fixed the whole issue in 1 week. Not bad considering the time delay and the distance from Germany to Taiwan ... the german customs need far more time to do their work but this pcb arrived untouched, therefore lightning fast. The damage of the 5 year old inverter was cause by lightning or Blitz and its power peak / impulse. Repair was a board replacement that had cost 80$ ... but it was not a big board. The reason was that the special spare part needed was not available in europe during first lockdown and would have cost 30$ + european shipping alone, so he took the complete board which was the easiest exchange .