PIP inverter repairs and hardware modifications

[coulomb]

Came home to all my power off this evening and found I had a dead short in one of my inverters

Commiserations.

Taking it apart I found one igbt and one MOSFET in pieces

Usually the MOSFETs and IGBTs fail in pairs. So there are probably less obvious failures elsewhere.

I'm probably going to replace all 16 MOSFETs and the 4 igbts. Would anything else be recommended to change?

Yes, the 4 electrolytic capacitors that protect the MOSFETs. There are also another 4 IGBTs, but the other set is likely OK. But verify.

Also check all the gate drive circuitry. Often resistors and diodes, sometimes transistors, blow up in the gate drivers. There is nothing quite as disheartening as replacing 16 MOSFETs and finding that they blow immediately because the gate drivers were also faulty.

Could it have been something else that caused this or just unlucky?

Most likely it was the medium quality capacitors that dried out. Replacing with long life types earlier might have prevented this fault. See the index for recommended capacitor and MOSFET replacement part numbers.

[jdavies1994]

Also check all the gate drive circuitry. Often resistors and diodes, sometimes transistors, blow up in the gate drivers. There is nothing quite as disheartening as replacing 16 MOSFETs and finding that they blow immediately because the gate drivers were also faulty.

Thank you for the response, are they the four pin SMD components directly under the MOSFETS? Just asking because I only usually deal with DIL gate drivers. Q50-Q57

[daniel1979]

Hello to all,

first, thanks for all who try to repair and successfully repair their Pip Inverters.
Its an exiting thing to learn more about electronics. The Pip Repair Theme gives me
a big chance to learn more and more, because the pips electronic is mostly possible to
learn and to understand...

So here is me new in the Forum, found on Google , searching for pip repairs...

And here is my Pip Problem:

First, its a little Different Model then i saw on the many Pics here.
Its a 500Vdc Voc Model. I bought 2 of them.

Unfortunately one Unit quit its Job after 2 Months only. It just went off, i was in the Room at the time, nothing to hear and no smoke.
So its not so difficult, buts a lot of work to find the errors.

It was a broken R215, D53, Q36, U10
I changed everything to new, the Diode was not easy to find

But not the inverter not come up either. I always test it with the Battery Connection only.
Maybe it comes up with PV Input, i think i will test this tomorrow.

The Problem is now, the U10 still has no Vcc, should be Arround 15V normaly. For me i have 0V
Also i have nothing like 150V on R211 or D57, so the Problem is the DC/DC Converter "Area" ?

I also took some Measurements arround the U9 Area and found no shorts or something...


And now some Pictures and a good time to everyone


..oh i tried to add photos by draging it to the edit plane...didnt work today...i try tomorrow

[daniel1979]

[jdavies1994]

Well that would cause it, ESR of 0.37Ω

[jdavies1994]

Does anyone know what part number D4 D7 D8 D44 and D45 are? They're all the same and are marked L437 but online searches aren't coming up with anything also nothing of use in the service manual

Cheers

[coulomb]

Does anyone know what part number D4 D7 D8 D44 and D45 are? They're all the same and are marked L437 ...

I have them marked as L4; the 37 might be a date or lot code. L4 is the marking code for a BAT54, a common SOT-23 dual Schottky diode.

[jdavies1994]

I have them marked as L4; the 37 might be a date or lot code. L4 is the marking code for a BAT54, a common SOT-23 dual Schottky diode.

Thank you very much Coulomb means I can get the parts ordered now!

[jdavies1994]

Well the issues persist, it's not blowing fuses anymore and at least it now turns on but I have error 06 as soon as it goes to start.

I believe this is voltage to high on output but as my meter is auto ranging I have no way to prove this until my new meter arrives

Any suggestions? I have ruled out voltage dividers and the relays that I have seen suggested elsewhere as these are all ok.

[coulomb]

If the inverter output voltage is normal, and as you say the 8 resistors at the inverter output (R147 etc and R135 etc) then the rest of the voltage sensing is on the control board, connected to pins 9 and 10 (with pin 1 at the left edge) of the control card (CN9). If the output voltage really is too high, then you'd have to suspect the full bridge and its gate drivers. I note that D4/D7/D8 etc are part of those gate drivers. If those diodes failed, other components nearby are suspect.

[ Edit: added context for pins 9,10. ]

[jdavies1994]

I'm only getting 20VAC when it throws the 06 error so it's definitely going to be something to do with the voltage detection circuit. But what I'm struggling to work out, all resistors up to and on the daughter board following ntye traces are ok I've completely ran out of ideas here

[coulomb]

I'm only getting 20VAC when it throws the 06 error so it's definitely going to be something to do with the voltage detection circuit.

I wondered for a while whether you might have a really old PF0.8 model (4 kW/5 kVA) with firmware that perhaps used fault code 06 [ edit: was 07 ] for low inverter voltage as well as high inverter voltage. But a check of PF0.8 firmware 52.30 indicates that a low inverter voltage would be reported as fault code 58.

Of course, if as seems likely from this post that you have clones, they may well be using different firmware. In fact, it appears that your post triggered the original Do I Own a Clone post.

Edit: So I think you should be open to the possibility that your inverter uses fault code 06 for low inverter voltage, as other Voltronic Power firmware does, meaning that there may not be anything wrong with your inverter voltage sensing circuit.

[jdavies1994]

I wondered for a while whether you might have a really old PF0.8 model (4 kW/5 kVA) with firmware that perhaps used fault code 07 for low inverter voltage as well as high inverter voltage. But a check of PF0.8 firmware 52.30 indicates that a low inverter voltage would be reported as fault code 58.

Of course, if as seems likely from this post that you have clones, they may well be using different firmware. In fact, it appears that your post triggered the original Do I Own a Clone post.

Mine is the new style withe the scc on top of the heatsinks but thats not to say it isn't old firmware. It does however state 72.something not that means a whole lot with clones.

Although mine is a clone the part numbers contained withing the service manual all seem to marry up with what I have so I suspect it's one of the better clones but who knows.

Back to the fault at hand I've been a bit of an idiot, I've measured the voltage at the inverter output. Which has a relay upstream causing the low voltage duh! So after measuring at the voltage dividers I have seen 300vac all the way up to 500vac but as it's powering for a flash I don't get an accurate reading the meter just spikes and it's gone. So it's definitely not low voltage, you mentioned high voltage would relate to the pwm circuit? I dont suppose you mean the t350 optos and the qa1/qb2/qc1/qd2 igbts and surrounding resistors? Or do you mean q27-30 it's MOSFETs and surrounding resistors?

I only ask as this all tested ok with the multimeter after replacing the t350 optos and the qa1/qb2/qc1/qd2 igbts and surrounding resistors.

[coulomb]

I wondered for a while whether you might have a really old PF0.8 model (4 kW/5 kVA) with firmware that perhaps used fault code 07 for low inverter voltage as well as high

Oops, I meant fault code 06, not 07. Edited now.

Mine is the new style withe the scc on top of the heatsinks but thats not to say it isn't old firmware.

The style with the heatsink on top is older, ending in about August 2015 (for genuine models).

Top heatsink.jpg (15.41 KiB) Viewed 713 times

It does however state 72.something not that means a whole lot with clones.

I just checked 72.40, which came with my October 2015 manufactured inverter, and it's the same (inverter low voltage is fault code 58).

you mentioned high voltage would relate to the pwm circuit? I dont suppose you mean the t350 optos and the qa1/qb2/qc1/qd2 igbts and surrounding resistors?

Yes. There are more parts surrounding it than resistors, there are five diodes including two zeners, and two capacitors. This is the inverter-proper, the main DC-AC converter, where the AC voltage in question is generated.

Or do you mean q27-30 it's MOSFETs and surrounding resistors?

Those are the bus end of the main DC-DC converter. Faults there might possibly push up the bus voltage, but I think the inverter proper is more likely where the problem lies.

[jdavies1994]

Every component in this photo has now been replaced, tested and checks out ok.

Is there more to the pwm circuit that I'm not seeing apart from the signal to the optocouplers?

I really appreciate your help columb, if you want to PM me a PayPal address I'll be happy to send over some beer tokens

[coulomb]

Is there more to the pwm circuit that I'm not seeing apart from the signal to the optocouplers?

Yes, below the bottom of the photo, all the way to TX7 and also D1, D40, D19. These are the power supplies for the drivers. With the inverter running, I believe you should be able to see the +18 V and -5 V with respect to the emitters of IGBTs QA1 etc.

[jdavies1994]

Finally! It's all working again, the zeners had failed so replaced all 6 in that area (massive pain to order got lost in post twice!)

This is all thanks to you Coulomb.

Please send over you PayPal address as i'de really like to buy you a beer but flying to Australia isn't feasible at the moment

[ohdearibrokeit]

so, I'm a long time lurker, been quietly reading the forums on the pip units for a while now. unfortunately had one of my units let the magic smoke out, and i'm now doing some diagnostics. several questions:

is it common for the dc/dc mosfets to all fail and represent the same failure mode? reading no resistance g-s and diodes all read 0.63v fv.

what on earth is the deal with q47-48-49 etc? i'm getting resistance readings in the megaohm range across b-e and b-c, but all diode forward voltages are showing 0.62v. what's the most reliable way to test these?

and how on earth are you supposed to know which pins are 12 and 13 etc on the driver ic?

i've checked the 10 ohm ad the 200 ohm resistors and verified the driver transistors q50-57 are bad, is there anything else i should check?

i think it goes without saying im following the service manual here...

incidentally if anyone was wondering, yes it exploded (sparks out the side of the unit), only 2 resistors show physical damage, and yes it was a code brown...

many thanks guys, any advice and pointers are greatly appreciated

[Revlac]

Been some time since I repaired mine, in fact its run longer since the repair, than from new.
I can't remember exactly how I tested Q46 to Q49, but they where ok.

For U9 it might be better to look up the data sheet for that SG3524 its a common chip used in Many things, if the driver transistors are ok there is a good chance that ic is ok too, best to check and be sure.

3.2 Check the bus side components
3.2.1. Power devices
DC/DC IGBT: Q27/Q28/Q29/Q30

Best to check these, almost always a pair of these are blown but may not be visible, and there drive components.

There is some diodes not mentioned on the repair manual that I have, they where also blown open circuit, they where part of the gate drive on the IGBT's that were blown, Diodes are ZD31 and others in that area.

There are other things to check, follow the manual mostly and others might have some input as you progress, Its very important to find and replace all busted components or it blow them up again, a successful repair is very satisfying if all goes well.

[MasterCATZ]

PIP-5048MG

what would have caused the IRGP4066D to blow , Inverter switched off sometime in the night and when I rebooted it (no solar or grid connected just batteries ) the IRGP4066D blew

also what upgrades should be done ?

[MasterCATZ]

and what is the trick to release the board ?

[coulomb]

and what is the trick to release the board ?

Did you find the screw under the small internal manufacturing date sticker between the battery terminals?

[coulomb]

PIP-5048MG
what would have caused the IRGP4066D to blow ,

A possible failure mode is as Weber found after one of his inverters blew up unexpectedly. See the last paragraph of this post.

also what upgrades should be done ?

The capacitor upgrade, and possibly the MOSFET upgrades, if they are not already 100 V parts. See the index in the first post for details.

[ Edit: added "the last paragraph of" ]

[Revlac]

@MasterCATZ Have you found a service manual?
Some of the later models are using STGW80H65DFB, instead of IRGP4066D.

[jdavies1994]

Finally! It's all working again, the zeners had failed so replaced all 6 in that area (massive pain to order got lost in post twice!)

This is all thanks to you Coulomb.

Please send over you PayPal address as i'de really like to buy you a beer but flying to Australia isn't feasible at the moment

It appears I have spoken too soon now all hooked up again I have an error 60. When seperate they both work fine and when hooked up in parallel the one that blew up gets an error 60 after around 2 minutes. You can see the VA gradually creep up whereas the load stays at zero on that unit. And on the other unit they both creep up together. I'm suspecting a parallel board, I don't suppose you have any other suggestions?