Me thinks you've had a bit of lightning about.

The problem will be figuring out the values for these.

The thing that looks like a diode is a thermal fuse. They operate at a particular temperature, as well as at a particular current. Hopefully there will be marks on there indicating the temperature and/or current. Typical ratings are 15 A or 20 A. This might carry the PV current, so if you can't find markings, size to that. If P7 is of of the PV inputs, and there is connectivity from P7 to one end of the fuse, then this seems likely. 

Movistors usually have a code on them; sometimes you can guess the ratings from that. They also have an energy (number of Joules) rating, but that's proportional to their diameter; if you replace with one of the same size (diameter and approximate thickness), this usually will have the correct rating. Its best not to have to guess the rating for movistors.

[ Edit: Example from above:

The "561" from the part number represents 560 V standoff voltage (like resistor code: 5, 6, and 1 zero). ]

The gas arrestor is obviously connecting to chassis/earth through the large via with the screw icon next to it. You likely can't get a rating off the blown one; if so, any of about the right size will provide some protection.

Edited January 10, 20251 yr by Coulomb

9 hours ago, Coulomb said:

Me thinks you've had a bit of lightning about.

The problem will be figuring out the values for these.

The thing that looks like a diode is a thermal fuse. They operate at a particular temperature, as well as at a particular current. Hopefully there will be marks on there indicating the temperature and/or current. Typical ratings are 15 A or 20 A. This might carry the PV current, so if you can't find markings, size to that. If P7 is of of the PV inputs, and there is connectivity from P7 to one end of the fuse, then this seems likely. 

Movistors usually have a code on them; sometimes you can guess the ratings from that. They also have an energy (number of Joules) rating, but that's proportional to their diameter; if you replace with one of the same size (diameter and approximate thickness), this usually will have the correct rating. Its best not to have to guess the rating for movistors.

[ Edit: Example from above:

The "561" from the part number represents 560 V standoff voltage (like resistor code: 5, 6, and 1 zero). ]

The gas arrestor is obviously connecting to chassis/earth through the large via with the screw icon next to it. You likely can't get a rating off the blown one; if so, any of about the right size will provide some protection.

The gas arrestor just evaporated. It was falling to pieces before I could desolder the pins.This MOV is also so burnt that you can't see the markings Only the thermal fuse is legible and I will replace it. It's interesting that only these elements exploded since the weather was nice and there was definitely no lightning 👀 

Thank you for your help anyway, I'll probably buy another one if I can't fix it. This time maybe from a more decent company with normal software 💀

I'm looking at a VMIV 3kVA board which looks identical at that part.

The thermal fuses are Aupo BF172 (data sheet seem to say it's a 10A)

MOV is 20N561K

Gas arrestor is 2RM600-8

Sorry I'm on my phone and pictures are tricky to rescale and upload with it. I can send pictures later if you need.

 

On 2025/01/08 at 4:41 PM, Coulomb said:

Are these the capacitors in series with the transformer windings? With a diode across them?

From memory yes those were the ones that finally did it.

I did some digging and found 2 shorted transistors. Because error 09 does not occur by itself if the elements that exploded are only PV protections. Any suggestions on what else might need to be replaced besides these two?

Check the resistors on the gate of those IGBTs. Do you have an oscilliscope to check the driver circuits?

The bus softstart circuit might also be damaged, a quick check would be if the MOSFET driving the primary of TX2 is short circuit.

I use a high voltage IGBT/transistor/LED tester on the IGBTs from the DC-DC circuit to the H-bridge. I have had some pass the basic multimeter test but their break down voltage is out of specification when tested. It's saved expensive failures of replaced parts. See below of failure vs pass. (it's also very handy for checking zener diodes). It was USD20 odd from Aliexpress. EEVblog forum has some technical discussion on it here https://www.eevblog.com/forum/reviews/completely-non-sketchy-mini-high-voltage-transistor-tester-(emeco)/

Edited January 10, 20251 yr by Shadders
Grammar and spelling corrections

W trybie bypass, podwójny inwerter max11kw pobiera 1000 W 5 A. Uszkodziłem go, podłączając fazy ln i nl do zasilania odwrotnie. W momencie przełączenia na bypass, został uszkodzony. Ładuję akumulator z paneli, produkuję napięcie dla domu, ale w momencie przełączenia na bypass, pokazuje stałe obciążenie 5 A. Proszę o poradę, gdzie szukać uszkodzenia i przepraszam za mój słaby angielski.----IS THIS ELEMENT RESPONSIBLE FOR MEASURING THE AC NETWORK?

Edited January 13, 20251 yr by ares43

On 2025/01/13 at 4:18 AM, ares43 said:

W trybie bypass, podwójny inwerter max11kw pobiera 1000 W 5 A. Uszkodziłem go, podłączając fazy ln i nl do zasilania odwrotnie. W momencie przełączenia na bypass, został uszkodzony. Ładuję akumulator z paneli, produkuję napięcie dla domu, ale w momencie przełączenia na bypass, pokazuje stałe obciążenie 5 A. Proszę o poradę, gdzie szukać uszkodzenia i przepraszam za mój słaby angielski.

In bypass mode, a dual inverter max11kw draws 1000 W 5 A. I damaged it by connecting the ln and nl phases to the power supply inversely. At the moment of switching to bypass, it was damaged. I charge the battery from the panels, produce voltage for the house, but at the moment of switching to bypass, it shows a constant load of 5 A. Please advise where to look for the damage and sorry for my poor English.

Please post in English, poor or not.

It's possibly not damaged. These inverters draw amps of reactive current when input is connected to output and also the output of the DC->AC converter. So this isn't nearly 1000 W, it's nearly 1000 VA, probably only a few watts of losses in the capacitor of the DC->AC converter's filter.

On 2025/01/13 at 4:18 AM, ares43 said:

----IS THIS ELEMENT RESPONSIBLE FOR MEASURING THE AC NETWORK?

Yes, that measures AC current in some models:

From https://forums.aeva.asn.au/viewtopic.php?p=94438#p94438 .

AC voltage is measured by strings of resistors on the main board, along with an op-amp on the control board.

Edited January 14, 20251 yr by Coulomb

When I bought the inverter, it was initially connected correctly and there were no such indications after switching to bypass. I bought a pre-meter and when connecting the pre-meter, I confused the phase (l) with zero (0), at night, when the batteries ran out, it switched to bypass and by morning it had recorded 10 kWh, the consumption was about 2200W. During disassembly, I noticed the error and connected it correctly and the consumption decreased to about 1200W. The pre-meter and the electricity meter show the same.

I measured all the resistor ladders, diodes and transistors and everything seems fine.

theoretically,

The question is, could the IGBTs have been partially damaged?

Is the element responsible for the measurement (Op07c in the photo) damaged and is sending incorrect data to the processor?

Today I connected the inverter to the battery, the consumption is 1.7A, does it seem correct?

Then I connected the network and gave 220V to the input when I measured the output it turned out that there was no constant phase from L to ground OK 112V and N to ground 112V when battery powered constant phase is present. no ideas.

Thank you very much for your interest in this problem and I am full of admiration for your knowledge and willingness to share it.

 

 

 

 

 

 

 

Today the mosfets arrived, it was worth waiting 7 days. I didn't install the fuses on the PV side, I don't know why they blew up so I'll let it go for now. Tomorrow I will test when the sun shines, 230V works again and the error disappeared, nothing but mosfets was damaged. Resistors and control are all working.

On 2025/01/16 at 7:05 AM, ares43 said:

when the batteries ran out, it switched to bypass and by morning it had recorded 10 kWh, the consumption was about 2200W. During disassembly, I noticed the error and connected it correctly and the consumption decreased to about 1200W.

I suspect that nothing got damaged.

On 2025/01/16 at 7:05 AM, ares43 said:

Is the element responsible for the measurement (Op07c in the photo) damaged and is sending incorrect data to the processor?

I very much doubt it. The op-amp is reading millivolts across a very low value shunt resistor. If anything would  blow up, it would more likely be that shunt resistor. I'm sure it's fine.

On 2025/01/16 at 7:05 AM, ares43 said:

Today I connected the inverter to the battery, the consumption is 1.7A, does it seem correct?

Yes, that's around 85 W, that's about right for an 11 kW model.

On 2025/01/16 at 7:05 AM, ares43 said:

Then I connected the network and gave 220V to the input when I measured the output it turned out that there was no constant phase from L to ground OK 112V and N to ground 112V when battery powered constant phase is present.

When AC-in is connected to AC-out (usually the case with these inverters if AC-in is present), the inverter disconnects its own AC-out neutral to earth connection, and relies on the neutral to earth link in the house wiring. It sounds like you don't have such a link. What voltage do you measure neutral to earth when the inverter is not connected?

 

For god sake i  cant read this ics name ...the schematic has 3 ices 2 of them are lm358

Uc3845 

The third not clearly readable 

Pleas3 try read it mr @Coulomb

On 2025/01/07 at 11:19 PM, Shadders said:

@wael_fatheIs the inverter stable when on and no load?

Is the 08 an bad reading or is the DC bus voltage going too high?

I have previously repaired a 08 error by replacing the small ceramic capacitors on the output circuit of the SG3525 which feeds the DC-DC IGBTs.

Yes stable....what cap u replaced ...is it in output or where exactely @Shadders

Edited February 10, 20251 yr by wael_fathe

4 hours ago, wael_fathe said:

The third is not clearly readable 

Please try to read it Mr @Coulomb

My guess is that it's an 8-bit microcontroller (CPU with flash or mask memory, RAM, PWM, I/O etc). Possibly  a PIC chip.

I read:

B/S C 3 2 F 7/T 2 6/5 0/C/other

None of those combinations pans out with a quick internet search. Sometimes you can get a clue from the manufacturer's logo on the chip.  Or you can read the part number directly, if it's not obscured. I use a cheap USB microscope for this.

But this will be completely proprietary, with no hope of replacing it and programming the firmware. So if it's faulty, all is pretty much lost anyway. So I think it doesn't matter too much what it actually is.

Edited January 29, 20251 yr by Coulomb

On 2025/01/29 at 4:20 AM, Coulomb said:

My guess is that it's an 8-bit microcontroller (CPU with flash or mask memory, RAM, PWM, I/O etc). Possibly  a PIC chip.

I read:

B/S C 3 2 F 7/T 2 6/5 0/C/other

None of those combinations pans out with a quick internet search. Sometimes you can get a clue from the manufacturer's logo on the chip.  Or you can read the part number directly, if it's not obscured. I use a cheap USB microscope for this.

But this will be completely proprietary, with no hope of replacing it and programming the firmware. So if it's faulty, all is pretty much lost anyway. So I think it doesn't matter too much what it actually is.

one day i will discover as this  is common inverter  in my country

however its  pretty weird that  a  micro controller  exist in the feedback loop  of  smps  battery charger ?

On 2024/12/10 at 8:42 PM, Coulomb said:

The output smoothing capacitor (the "C" of the "LC filter") is only about 10-20 μF (for 3-5 kVA models); these are large rectangular film capacitors, non-polarised. Did you mean the bus capacitors (typically 2 x 470μF for 5 kVA models, polarised), or confusing the two, or... ?

I believe that the 2SA1020 is PNP. 

Almost certainly not. My understanding is that most of the time this power supply operates on battery voltage, so that would be twice as high on the 48 V models. These circuits can operate over a wide voltage range, but I don't think that wide.

Have you replaced the SG3525? Also check for bad solder joints on the tiny capacitors. I had a King 5 kW that had the timing capacitor intermittently open circuit. Also check the larger chip capacitors out of circuit; those things are barely able to be called capacitors, and can drift and change capacitance with applied voltage. The larger valued ones (usually only used for bypassing) are the worst, as they need the more extreme permittivity materials, which have the worst parameter spread.

 

On 2024/12/11 at 6:51 PM, Coulomb said:

Zero rings sounds like it's open circuit (or you tested the wrong pairs of pins). If the transformer is open circuit, I can't see how that would shut down the power supply when the 3525 is enabled. [ Edit: Though maybe your tester is testing for shorted turns, and rings in the case of no shorts. ]

Did your tests include the small MOSFETs (UT304G) across the outputs of the NPN/PNP push-pull pairs? If these are shorted or not turning off, that might cause the power supply collapse.

i was able to solve the   smp   problem  it seemed  12 regulator is shorted  though  the short is not appearant in the 

dmm  but  when repalce it   it worked  i remamber i also find  shorted  2sa transistor  ....

how ever i never  be able to solve the  inverter porblem .all sections ok    the  caps  fill to 370  but there is no  

pulses  to the dc -dc  and some times  all the inverter  g oes  off the  pcb  has  blackned  sections  near  by the  smd pnp and npn drivers that  are near the input dc-dc  fets  ....i closed the case and return it to the customer  after  an  

nightmarish  atttempts to repair it  ..to no avail  i  draw conclusion that the burned  fet  area  is the one  responsible  for the  strange shutdown 

On 2023/11/10 at 8:47 AM, Mischel68 said:

As promised, feedback on error 52.

I was able to repair the inverter.
3 battery mosfets were burnt out (short circuit).
This also causes damage to the conductor tracks.
I swapped the mosfets and added wire jumpers to the conductor tracks.
The inverter works again. Let's see how long.

Thanks again for the help and all the work you did.

greetings from Germany

Mischel

pretty strange how it charges normally with bad  dc  side  components 

for charging to work all  dc-dc component have to be in good state  

The inverter worked and in less than minute fets get hot ....when i unsolder and scoped this same signal over all fets  ...to me seems not notmal normall signal have that leaning  from up and from bottom

This one is too straight in top

What say you?

What can cause all 4 signals to get deformed ...seems impossible...unless its common fot all othem which will lead to sg3525?

@Coulomb

 

 

This is what I found when fixing a similar fault:

I have no idea why my tops are so bent in the middle, and yours are so flat.

The above is from this post:

https://forums.aeva.asn.au/viewtopic.php?p=65282#p65282

Edit: Yours doesn't seem to be going negative; perhaps I misunderstood what you were saying. Is your -12 V supply working? 

What volts per division was you oscilloscope on?

Is yours at the SG3525 output or at the MOSFET gates?

Edited February 12, 20251 yr by Coulomb

16 hours ago, Coulomb said:

This is what I found when fixing a similar fault:

I have no idea why my tops are so bent in the middle, and yours are so flat.

The above is from this post:

https://forums.aeva.asn.au/viewtopic.php?p=65282#p65282

Edit: Yours doesn't seem to be going negative; perhaps I misunderstood what you were saying. Is your -12 V supply working? 

What volts per division was you oscilloscope on?

Is yours at the SG3525 output or at the MOSFET gates?

The wave form appareantlly sound....the reason feta get hot is one 100 ohm resistor find open ...replaced it and every thing fine

 

 

Test it under 700watt load all fine

The disaster happen when i applied ac..input.......it charged good then in less than mint dc side fets exploded violently and my house ac triped off

 

 

What a waste of time am extremely  depressed..what cause invetr work fine  as inverter and explode as a charger

@Coulomb

3 hours ago, wael_fathe said:

The wave form apparantlly sound....the reason FETs got hot is one 100 ohm resistor find open ...replaced it and every thing fine

I hope you didn't replace a resistor marked 100 (10Ω) with a 100Ω resistor! The 100 code means 10 followed by 0 more zeroes, as you probably know. I'm not aware of any 100R resistors in the MOSFET area.

3 hours ago, wael_fathe said:

The disaster happen when I applied ac..input.......it charged well then in less than a minute DC side FETs exploded...

And I hope that if you did make the above slip-up, it wasn't the result of the big bang.

3 hours ago, wael_fathe said:

What a waste of time am extremely  depressed.. What can cause an inverter to work fine as an inverter and explode as a charger?

These setbacks can be heart-breaking. It could be other components in the MOSFET gate drive circuitry. The buck IGBT is only used when charging; at other times, its freewheel diode just passes bus current. It's possible that there was something wrong with the buck IGBT, or its gate driver. Gate drivers! 😠

On 2025/02/13 at 3:42 AM, Coulomb said:

I hope you didn't replace a resistor marked 100 (10Ω) with a 100Ω resistor! The 100 code means 10 followed by 0 more zeroes, as you probably know. I'm not aware of any 100R resistors in the MOSFET area.

And I hope that if you did make the above slip-up, it wasn't the result of the big bang.

These setbacks can be heart-breaking. It could be other components in the MOSFET gate drive circuitry. The buck IGBT is only used when charging; at other times, its freewheel diode just passes bus current. It's possible that there was something wrong with the buck IGBT, or its gate driver. Gate drivers! 😠

I will comback again for for it...for now i will try to repair easy faults...need time to forget the disaster....some ferts went into smoke

 

Never thought in an inverter that is 48vdc 5k  draws only 1 amp  when working as inverter all fets and  igbt are super cold....every thing was telling me that it be a legendary repair

 

Tell the big bang

 

 

 

 

 

 

today  i beat 06  error    by  only  resolder  broken track in the ac  igbts ....it was  cut under the  pcb   

 

inverter come in with  badly blown ac igbts 

i replaced  all igbts and  bad  zeners and  one bad  coupler ....checked the waveform with all  igbt out it was fine

 

installed  all  igbts  in again  error  06

 

replaced all  coupler and l44 diodes  and all zeners   .....no use  still   error 06

replaced  all  power supply caps   and  7912  and  even   tl074 compartor  ....still  error  06

 

only when i left heat sink to find fine  cut  due to igbt explosion ....repair the  pcb  and installed the igbt back

 

the bad connection happen in the collector of  one  of the ac igbts   once  fixed 

06  disappear 

Edited February 24, 20251 yr by wael_fathe

On 2025/02/12 at 5:42 PM, Coulomb said:

I hope you didn't replace a resistor marked 100 (10Ω) with a 100Ω resistor! The 100 code means 10 followed by 0 more zeroes, as you probably know. I'm not aware of any 100R resistors in the MOSFET area.

On 2025/02/12 at 2:29 PM, wael_fathe said:

there is  actually 100 ohm in clone models ...it is 101  it tested 100 still yet it was bit blackned  so  i replaced it  

today i cured random dc-dc igbt burn despite the fact that they have reasonable waveform

by replacing all caps in the smps with 1600mf 25vdc....never put1000mf

the 220 appear and no more burnt igbts

i also repllace lm7912 am not aware which one cured the random igbt blow

Edited March 17, 20251 yr by wael_fathe