MPP Solar / Infinisolar MPI 10k hybrid inverter died - fuse goes off when synchronizing with grid

[Stefanseiner]

Hello,


after one year of usage my MPI 10k seems to be defective.
You can see the whole project here -> 100KWh DIY 18650 Powerwall in Germany




Without grid connection it runs fine, but when I connect to the grid then the external fuse (Siemens 3x B16A) goes off.

at minute 0.50

- I double and triple checked all the wire connections including grounding and external circuit breaker
- I changed the fuse to a slower "C" rated model
- grid voltage is stable round about 230 volts


I changed the three phase wires in all possible combinations to avoid wrong connection
B / G / S fuse off S lights up in Siemens fuse
G / B / S Error 09 line phase loss
G / S / B fuse off B lights up in Siemens fuse
B / S / G Error 09 line phase loss
S / B / G fuse off G lights up in Siemens fuse
S / G / B Error 09 line phase loss

B = brown phase wire = phase 1
G = grey phase wire = phase 2
S = black phase wire = phase 3


Do you have some idea what's wrong or what could be defective?

Thanks and regards

Stefan

 

[houseofancients]

hi stefn, i've not had this problem when i had them , but diy tech repairs did and had to replace a board

 

[Hedges]

It may have connected to grid while not synchronized. Or phases swapped, except it worked before and you've tried all now.

Try putting a light bulb in series with each phase. Perhaps it will connect, and since no power delivered it might keep running and remain connected.
Then you could check voltage across each bulb. Frequency too, if your meter measures that.

Doesn't mean you can do much to fix it, unless something obviously broken inside or your replace a PCB (if you know which one is bad.)

What I did, when my inverter didn't connect, was put meter between inverter and grid. One inverter could be synchronized and another out of phase (mine are separate stacked inverters) so voltage incorrect. Not synchronized, voltage goes up and down. Floating, can show voltage then pull together by meter impedance.

If something became grounded, either on grid side or your side, that might cause the problem. Your 3-phase service - do you know how it is supposed to be grounded? Check utility phase to ground voltages, and yours. If one side was "Y" with center grounded and the other side "delta" with corner grounded, there would be a conflict. But I don't know why it would connect without qualifying all phases as synchronized.

 

[Stefanseiner]

Or phases swapped, except it worked before and you've tried all now.

I installed a 3-phased circuit breaker and swapped two phases, thats why the "phase loss" error sowed the first time. I changed it back but there are some new warning messages

 

[Stefanseiner]

I had contact via mail with MPP-Solar.
They told me to make some tests and I made a video of this and send it to MPP-Solar.

After analyzing the video they suggest that the mainboard is defective.
It is sad for me that the inverter dies after 11 month of usage only, but much more sad that they now say I have to pay for the spare parts although the failure is within the warranty.


At this point I am very disappointed of MPP-Solar and I don't know if I should recommend inverters from Voltronic any more when other people are asking me :-(

But I still hope we can find a solution - and I will keep you up to date

 

[houseofancients]

I had contact via mail with MPP-Solar.
They told me to make some tests and I made a video of this and send it to MPP-Solar.

After analyzing the video they suggest that the mainboard is defective.
It is sad for me that the inverter dies after 11 month of usage only, but much more sad that they now say I have to pay for the spare parts although the failure is within the warranty.
View attachment 88095

At this point I am very disappointed of MPP-Solar and I don't know if I should recommend inverters from Voltronic any more when other people are asking me :-(

But I still hope we can find a solution - and I will keep you up to date

mpp solar is a voltronics rebadger and reseller.
warrantee issues were for me the main reason to use a european reseller.

that being said, they should provide you with a new mainboard if the inverter is still within the 1 year warrantee period

 

[Stefanseiner]

after several discussions MPP solar is willing to provide a new mainboard for free but with shipping costs of USD 110.
They do not accept it is within warranty time but a one-time goodwill.
I think it is because I write the process of warranty transaction of MPP Solar public at this forum (and others too)

With this clarified I took the service / disassembly manual and searched for the mentioned mainboard of the MPI 10k
-> Download of the service manual as pdf
it is the left pcb with the big coils and heatsinks on it


I took pictures of all cables I had to remove so I can assemble the inverter later on the right way


here it is


at the back side I found an area with a damaged / non-existant resistor and some white dust around. The service manual says this is the "INV/REC module" for the "S-Phase". I guess the IGBT above is damaged and produces the short at the 230V AC side which is causing my problems with the falling circuit protection



I wrote MPP-Solar back with this new knowledge and pictures, let's see what's going next.


btw. speaking of warranty:
I wonder, why MPP Solar only gives 1 Year of warranty to the end-user because Voltronic, the original manufacturer gives 2 years warranty to their reseller. This means that with a failure between 1 and 2 years Voltronic replaces the defective part for free to MPP Solar but they cash up the customer.

So I've done a little research which reseller gives how much warranty time:

• Voltronic 2 years (I asked them via mail)
• FSP 5 / 3 years
• MPP Solar 1 year
• EASun 1 year (down at the footer of their website)
• Infinisolar 2 years (see Italian website)

 

[Hedges]

It looks like solder fillets of that resistor broke, rather than melting.
Dot in the middle suggests it was glued on for wave soldering (which soldered the thru-hole pins).
Not sure what made the white "crazing" lines around it. Are those cracks? Seems to be a crater at the epicenter.
Maybe mechanical stress caused that damage and caused part to pop off.

Looks like those 1k resistors are stacked 2s2p, which would still be 1k but 4x the power handling. With one missing, if the midpoint was also connected (2p2s) it would be 1500 ohms with one missing. But I think it is 2s2p, so 2k.

 

[Stefanseiner]

no this aren't cracks, it's a kind of white dust.

I can't find the missing resistor. Either it slided deep inside the inverter while carying it the way through my garden to the shed, or it exploded completely

 

[Stefanseiner]

The new mainboard for my MPI 10k arrived
positive: shipping war very quick and took only 1 week from Taiwan to Germany
negative: 110 USD shipping costs have to paid from me

Here it is. Seems that it's a new revision with a slightly different pcb layout




I 'm going to insert the new mainbord these days and hope, that my MPI 10k will run again after this repair

 

[Stefanseiner]

Last weekend I changed the mainboard and yesterday I startet the inverter.

But as I turned the solar-input switch at the bottom of the inverter to "on" I heard a sizzle at the same second and a little smoke came out of the right upper side of the inverter.

Now the inverter goes in an endless reboot loop and shows "error 03" which means "dc bus voltage soft-start is time-out"

here is the error log from solarpower software


Now I sent the video to MPP-Solar support with the question, at which area I should search for the damage.
At my first view to the (new) mainboard I couldn't find any damages, but I guess the failure comes somewhere from the solar-input side because the smoke was there after I turned the solar-switch.

Do you have any ideas?

Now I am curious what the MPP-Solar support may answer.