MKS-II-5KWB Repair Please help

[dangraham]

Hello Forum,
A little background, I am originally from the UK, I have used the PIP style inverters in the UK on a few off-grid installations, some single units and some single phase parallel installations. This was about 5 years ago and I never had any significant issues with these inverters, they were extremely cost effective and once commissioned proved to be quite reliable.

Fast forward to now, I currently reside in Phuket Thailand, we have had numerous issues with the grid supply so I decided to install an off-grid solar setup. We imported 25x 360w panels, 32x 280Ah LiFePo4 cells for 16S2P 48V 560Ah, 3x MKS-II-5KWB for 3 Phase operation, and we have a 3 Phase 30KVA generator. We also imported a Growatt 5KW grid-tie inverter to export if we are away from the property for extended periods.

The installation went reasonably well, I got the batteries installed and tested, I got the inverters mounted and configured, I started to commission the system and found numerous issues with the domestic wiring that took a while to resolve, electrical installations in Thailand are generally held together with spit and bubblegum. I had the system operating well, I have 8* panels to two inverters and 9* panels to the third, we were feeling very happy with the installation and then the problems started.

After approximately 3 days of reliable operation and no changes to the system one of the inverters exploded, the component on the SCC marked F1 (I presume a fuse) exploded, and also the majority of the battery side mosfets. At this point we presumed it was just a defective unit, so we had to re-wire the system, and the house for single phase and put the remaining two units into single phase - parallel operation. This was a bit of a pain in the bum, a lot of additional work, a loss of 1/3 of the solar harvest, but was sufficient while we negotiate with the supplier regarding replacement / ordering additional inverters (the alibaba suppliers are really not very helpful when it comes to customer service)

So after approximately 2-3 weeks of reliable single phase - parallel operation, again with no significant changes to the system, another inverter exploded. I was messaged by one of the Thai staff that the inverter was beeping, I looked and one inverter was displaying fault, I looked up the fault code and it referenced overload, the inverters are configured to auto recover from overload but this obviously did not work. I asked the Thai staff to observe me doing a system reset so they would be able to do this if I am not available. I proceeded to show him, Isolate the AC output, the AC Input, turn the inverters off, Isolate the solar, Isolate the battery, everything is off. Now to start everything up, turn on the battery, turn on the inverters, wait for inverters to start-up and Bang! the thing shot molten mosfet legs out the vents in furious anger! Scared the S@!? out of us!

So now I have removed this inverter and inspected inside, the failure is almost identical to the first unit, a small capacitor leg has burned out on the SCC board and almost all of the battery side mosfets have blown. I am now down to one inverter and I am feeling very reluctant to order replacements of the same, I am astounded at the way these units have failed, I know they are cheap but can things really be this unreliable! I also now do not have any grid supply as a fall back option, if this last inverter blows I will be stuck running the generator 24/7 for power and this is not acceptable. I can probably source another inverter locally as an emergency option if the worst comes to the worst.

Ideally I would like to repair / upgrade the two broken units I have, I will also be in communication with the supplier to see what they can offer but I do not have much faith in the Chinese to do this, the cost of returning the units and sending new ones makes this difficult. I have been reading some of the posts regarding the repair / upgrade of these units and various components, I am a little concerned that my units already have the CSD19505KCS mosfets and 80V capacitors. I will also need to try and test the mosfet drivers before just replacing them, and I still don't know why the units failed in such a catastrophic way in the first place. Is it possible to repair them without knowing what the original cause of the fault condition was?

This forum appears to have a wealth of information regarding the different variants of this inverter, I would welcome any advise on the next practical steps to take with these units before they end-up being salvaged for parts and build my own SCC from the components and have to buy alternative inverter. I am hoping I can make one good inverter from the two blown units, although I will have to replace the mosfets, any advice on testing the driver circuits would be helpful and any upgrade options other than the CSD19505KCS?

I will try and attach some pictures of my problems.


Original Installation


3Phase operation charging from solar and generator at 9KW






First SCC component failure


Original SCC F1 component for comparison


First FET failure


Another view of first FET failure


Single Phase fault warning


Single Phase fault warning


2nd Failure SCC cap failed


2nd failure FET


2nd failure FET another view

Thanks in advance for any advice you can offer.

[coulomb]

Hello Forum,

Welcome to the forum.

We imported 25x 360w panels, 32x 280Ah LiFePo4 cells for 16S2P 48V 560Ah, 3x MKS-II-5KWB for 3 Phase operation, and we have a 3 Phase 30KVA generator.

First warning sign: I've never heard of an Axpert MKS-II-5KWB. That certainly doesn't mean that they don't exist, it's just the first red flag I see. More below.

So after approximately 2-3 weeks of reliable single phase - parallel operation, again with no significant changes to the system, another inverter exploded.

That's certainly really bad luck.

I don't know why your images aren't showing in the Chrome browser; I was able to use my moderator powers and see the images referenced by the img tags, and they seem to be perfectly fine. Maybe there are just too many in one post, or the total pixel load is too much (they are all rather large images). [ Edit: "Open image in new tab" also works, and the images load in Firefox. ]

I don't know the high voltage Solar Charge Controllers (SCCs) very well, but this struck me as odd:

MKS II SCC fault.jpg (81.04 KiB) Viewed 477 times

The font that the PCB part number is drawn in is different to what I'm used to seeing from genuine Axperts. It's a tiny thing, and they may have just changed PCB software or a different engineer likes a different font. But there is also this:

MKS II MOSFET failure and label.jpg (118.37 KiB) Viewed 477 times

In the blank area on the barcode label, the barcode is missing. On the photo below,I see a second small sticker under the large label, with what looks like the barcode and serial number. This is highly unusual for a genuine Axpert. Again, they may have recently changed their practices, but it's been consistent since 2013 so far. So I'm afraid it seems to me that you actually have three clones, not three genuine Axpert/PIPs manufactured by Voltronic Power. If this is the case, they may not have copied the circuit properly, or used inferior components, and who knows what else. They might be exploding because the cloner just didn't do a good job and/or doesn't understand the original design well enough to copy it properly. Finally, the the manufacturing date sticker which should cover a nearby screw is missing:

Faulty MKS II 5KWBs.jpg (117.75 KiB) Viewed 477 times

I hope I'm wrong. If they're clones, you may have extreme trouble getting the faulty machines replaced (or not, depending on who the manufacturer and/or resellers are).

[ Edit: I originally indicated that the black screws are bad; in fact black screws are good and silver is more indicative of clones. Sigh. ]

[coulomb]

I mean to add that unlike the 145 V max SCC models which connect only to the battery, the 450/500 V max SCC models connect to the internal 400-500 V bus. The fact that the MOSFETs have failed is a bad sign; it means that the fault travelled through the isolating high frequency transformer to the battery-side full bridge, so a large part of the entire inverter-charger could be affected. So repair could be involved.

[dangraham]

Hi Coulomb, I appreciate the quick reply. I must admit I do not know who the original / genuine manufacturers of these inverters are? I know there are many of the same style inverters and it is very difficult to know what is genuine and what is knock-off with the Alibaba suppliers. In hindsight it was naive of me to assume that they would be genuine or good quality, I spent a long time speaking to many different suppliers to ensure they would support 3 phase and parallel operation, this was challenging enough, during these discussions you assume that the item will preform as advertised, add on to this the fact I was also dealing with multiple suppliers and the whole process was rather draining.

Anyway it is what it is now and I need to try and salvage from this situation if possible, worst case scenario is I have the components to build a charge controller and look for an alternative inverter. I will post some more pictures for you and hopefully they may shed some more light on the situation, I will be preparing a response to the supplier tomorrow to explain the current situation and how defective these units are, any background information on the original manufactures would help with these discussions, i will try and get some replacement units from them for what it's worth and possibly run them with more conservative settings if they are not of original quality, as I expect they are not. Thanks again for your help!

Not sure why the images are not showing, they do show for me when I view the post, I will resize the images to the smallest setting on my phone before posting, hope this helps.


















Slight damage to the track on the underside


Not very uncommon but not ideal to see jumper cables, not sure if this is common in other units?

It still baffles me why the 2nd unit would fail the way it did? there was no solar input, AC input or AC load, it was just booting up from a power off situation. Battery was connected and both inverters were switched on, it is very strange that it would fail in such a catastrophic way in this scenario, i must admit I still have a lot of posts to read on this forum to completely familiarise myself with the overall architecture of the unit.

[dangraham]

So another thing I wanted to ask, if I was to repeat this installation with the same requirement of 3x 5KW inverters operating in 3 phase, what would be your recommendation on the model / brand of inverters to use? I have considered the idea of 3x Victron Quattro's or similar but this is a serious expense (20K Aud) and I would never recover the capital cost, better to be raped but the local grid mafia who charge double the national rate for an appalling supply.

I would be interested to know who the original manufacturer "PIP" and where I can obtain genuine units from, ideally direct from them, and support the installation requirements etc.

Thanks,

[paulvk]

The capacitor that burnt is not a capacitor it is a NTC thermistor so over current burn out.
What is the arrangement of panels and what are their full specs I suspect an over voltage problem.
Can you also read the part numbers of the fets.

"I would be interested to know who the original manufacturer "PIP""

MPP Solar https://www.mppsolar.com

[weber]

Search on eBay for "PIP-5048MG" (not MGX) from maximum_solar (eBay reseller for MPP Solar).

[weber]

Like paulvk, I worry that 9 of your panels in series may be sailing a little close to the wind, with regard to the maximum open circuit voltage of 450 V. How many cells are there in your 360 W panels? I see most with 72 cells, but some with 96 cells.

Assuming 72 cells, I see Voc's of around 48 V at 25 °C. And I see temperature coefficients as high as -0.29 V/°C. In that case, the temperature of the panels would only have to drop to 18 °C for their Voc to go up to 50 V, and therefore 9 of them being 450 V.

Did the failures happen on cold mornings? With the second one, did the failure happen after you reconnected the PV array?

[coulomb]

Hi Coulomb, I appreciate the quick reply. I must admit I do not know who the original / genuine manufacturers of these inverters are?

All the genuine Axperts, PIPs, etc are designed by Voltronic Power in Taiwan, and manufactured in mainland China by Voltronic Power. But they don't supply any of their "own" models; they are all supplied by resellers (the customers of Voltronic Power), and VP brand them according to the requests of the various resellers. VP don't want to talk to owners at all, so you always have to go through your reseller for support (parts, firmware update files, warranty returns). The resellers will resort to asking VP only if absolutely necessary. This is a apparently the "Original Design Manufacturer" model.

I know there are many of the same style inverters and it is very difficult to know what is genuine and what is knock-off with the Alibaba suppliers.

Yes. That's why I prefer to go with Ebay.

Anyway it is what it is now and I need to try and salvage from this situation if possible, worst case scenario is I have the components to build a charge controller and look for an alternative inverter.

These high voltage solar charge controllers charge the battery via the DC bus and the DC-DC converter in reverse (pushing power towards the battery). They are not like the 145 V max SCCs where you could take them from the inverter-charger and put them in a separate box and run them stand-alone. Though that would also require a firmware change, since as they come, they are reliant on undocumented commands coming from the inverter-charger's main DSP (CPU).

Not sure why the images are not showing, they do show for me when I view the post, I will resize the images to the smallest setting on my phone before posting, hope this helps.

They are still not showing for me in Chrome. They work just fine in Firefox, so I'm switching to Firefox for these posts.


The above is a non-genuine date sticker. Genuine manufacturing date stickers have the year printed, and the appropriate month blacked out. Another clone indication.


The above metal bars look non-standard as well. The bent-up tabs are unusual. It might be the connection to the primary of the transformer somehow. Again, it might be a recent change by Voltronic Power, but I doubt it.

It still baffles me why the 2nd unit would fail the way it did? there was no solar input, AC input or AC load, it was just booting up from a power off situation. Battery was connected and both inverters were switched on, it is very strange that it would fail in such a catastrophic way in this scenario...

I'm wondering if you don't have pre-charge. You have a very stout battery, with short circuit current probably over 20 kA. With three inverters in parallel, even though the wiring is thick and direct, it would have considerable inductance. With 3 inverters in parallel, there would be a frightening start-up current, though you might have been starting with only one inverter connected. Contact bounce on the battery switch in conjunction with the very high surge current and inductance could cause high voltage spikes at the battery terminals. These could have caused punch-through of the MOSFETs. My own system with a smaller (320 Ah nominal) LFP battery, has a contactor with a high pulse power resistor across it for pre-charge. That limits the start-up current to some 18 A, 3 orders of magnitude less than without pre-charge. Pre-charge is over in less than a second. It also prevents wear of the main battery switch or contactor.

[ Edit: fixed quoting. ]

[paulvk]

The circuit with the blown capacitor and NTC thermistor appears to be a small power supply not related to the solar charger there is a bridge rectifier after the input from the common mode filter and thermistor.

[coulomb]

The circuit with the blown capacitor and NTC thermistor appears to be a small power supply not related to the solar charger

UIU believe it's a small DC-DC to power the electronics of the SCC.

there is a bridge rectifier after the input from the common mode filter and thermistor.

I'm mystified by what the bridge rectifier is for. In case the PV is connected backwards perhaps? But it doesn't seem to be large enough to take maximum panel current (18 A).

Edit: Initially I though this was another sign of a clone, but the originals seem to have them as well.

Edit 2: There doesn't seem to be any point to powering the electronics if the PV is connected backwards. Ah, wait; perhaps there is. Perhaps they detect the wrong connection, and turn the MOSFETs on hard to short circuit the PV input. Though the free-wheel diode in the MOSFETs (if present) would be doing that anyway, just with slightly more voltage drop. Or maybe there is something else that they can do to protect against reverse connection.

Edit 3: Looks like it's the AC SPS; see 2 posts down.

[paulvk]

No it appears to be a typical mains power circuit configuration the two pin connector is its supply and the 4 pin is its output.
From what I can see it does not connect to the PV.
It was good that he posted the back of the board so I could follow it.
Now that to me says a surge on that input so now where does that two pin cable connect.

[coulomb]

No it appears to be a typical mains power circuit configuration the two pin connector is its supply and the 4 pin is its output.
From what I can see it does not connect to the PV.

Ah! Brilliant! It could well be what's called the AC SPS. It seems to push a small amount of power into the DC bus, perhaps enough to supply the idle power, but also powers other circuits through another secondary winding. Below is from the Axpert VM II service manual. The part numbers don't all line up, but I think the general circuit does:

VM II AC SPS.png (157.55 KiB) Viewed 425 times

Typical of many Voltronic service manuals, the schematic isn't quite high enough resolution to be readable.

Now that to me says a surge on that input so now where does that two pin cable connect.

I can't read much from the schematic, except that the two inputs end in "L" and "N", strongly suggesting Line and Neutral. [ Edit: probably AC-in. ]

That does indeed put a different complexion on things, thanks Paul.

Edit: In the VM II service manual, it shows the AC SPS to be on its own small (~70 mm square) PCB, not part of the SCC PCB. It also seems to seems to have one output, so the circuit that the low voltage secondary powers is probably just its own control circuit.

[ Edit: Added my guesses as to how the labels connect together. ]

[Revlac]

Those small Blue Capacitors on the SCC (with the legs burnt off) look like they go to that screw hole in the corner, I assume that would be earth? But can't see where or if that was connected.


Also I can say that 2 or more inverters running off the same battery bank (everything else isolated), If one, for some reason has a blowup it can take out another that is running on the same battery bank meters away, been there done that.

Repaired and going well since.

Really need to find the cause of this before doing the same again with another inverter, I haven't until now, seen all 16 mosfets blown, usually just eight and 2 igbt's on the rectifier stage.
Please Correct me if I got something wrong.

It looks like a nice tidy setup.

Cheers Aaron.

[paulvk]

So if its a bus charge supply running off the AC mains in then high voltage spikes on AC in, maybe the generator or grid , maybe lightning so get voltage suppressors for AC input ask more questions of what happened the first time.

[dangraham]

So I managed to contact with MPP solar & Voltronic and it looks like the new versions of the inverters are quite different from this old style. At this stage I think my best option is to write-off the bad experience with the Alibaba supplier and chalk this one up to an expensive lesson learned. They have confirmed that the units are clones and this likely explains why they do not function correctly and randomly fail. There are a number of components I can salvage from these units and will be useful in other future projects, but this has come at a high cost to say the least. At this stage I don't really feel comfortable attempting to repair the units because the source of the original failure is still unknown and likely to fail again.

For anyone interested in the Alibaba supplier this is the link to their company profile https://ouyad.en.alibaba.com also www.ouyad.com

As you can see their profile page makes it look like they are a reputable company, but please I hope my bad experience with this supplier and the cloned inverters may hopefully prevent someone from having the same issues I have suffered. I will attempt to get a refund or some acknowledgement from them but I do not hold much hope in this regard.

This is a copy of the email from MPP Solar confirming the inverters are clones

I would like to thank you all for the valuable input and I will keep you updated on the developments, hopefully I will have a new set of inverters soon and a fully functional system. Right now I am praying this last inverter doesn't crap it's pants and I am left with just the generator I have dialled down the settings to keep things as conservative as possible.

Regarding some of the previous questions, the solar panels are 360W OCV is 47V and VMP is 37V so 9*47=423V I am in Phuket Thailand and they are installed at a very horizontal angle so I don't think over voltage would be the cause, early morning voltages are in the 200-250 range and temperatures never below 25deg C, by noon voltages are are pegged at VMP and we have about 2-3 hours of peak performance, also the failures have happened without the solar connected. The remaining inverter is operating with 9* panels still connected.

I had considered a pre-charge on the battery isolator but discounted this due to Thai staff needing to reset the system "has to be fairly basic" and the capacitors should absorb any intermittent spike from the isolator bounce and no mention of this being required in the documentation. I may look to install this as an additional protection feature for the new inverters. My e-Bike runs at 110V fresh of the charger and I use a 1000w halogen floodlight bulb as a pre-charge resistor for this, the power caps in the motor controller are quite large

[weber]

So I managed to contact with MPP solar & Voltronic and it looks like the new versions of the inverters are quite different from this old style.

They didn't send you a brochure for the inverter model that yours are clones of! It is the PIP-5048MG, as I mentioned earlier. See
https://www.ebay.com/itm/MG-Solar-power ... SwNxZfTfM0
And don't forget the parallel kits are sold separately:
https://www.ebay.com/itm/Parallel-kit-f ... Swm41bOwAb

I recommend you get the the MG because they have been around long enough to get a lot of the bugs out, and they will be a drop-in replacement for what you have now. The new removable-display models (MGX and MAX) are likely to be as buggy as hell. AFAIK people are still having problems with the removable-display models introduced two years ago (MK and GK). Steer clear of the removable display models I say. If others have more up to date info on this, please chime in.

I updated my "Voltronic Zoo" post yesterday in response to your need. Check it out:
viewtopic.php?p=82538#p82538

[coulomb]

This is a copy of the email from MPP Solar confirming the inverters are clones

Interesting. But MPPSolar aren't the manufacturers; they presumably don't have access to the list of other resellers to pronounce whether they are clones or not. They seemed more interested in selling you new inverters.

For months, Weber and I believed that MPPSolar were the manufacturers, and they often act as though they are. Maybe they were the first resellers, and perhaps still are (one of) the largest resellers.

[paulvk]

They have confirmed that the units are clones and this likely explains why they do not function correctly and randomly fail.

I am not sure that is true , I think even if you install new inverters the cause of this failure may still be there,
I feel it could be a voltage spike on the incoming grid/generator to the inverters and with that blue capacitor
(but is it a capacitor maybe its a MOV so can we have a picture of them with the writing on them)
with a burnt leg it means its a spike to ground this is typical of lighting or maybe a problem with the neutral to earth connection.

[weber]

... I think even if you install new inverters the cause of this failure may still be there,
I feel it could be a voltage spike on the incoming grid/generator to the inverters and with that blue capacitor
(but is it a capacitor maybe its a MOV so can we have a picture of them with the writing on them)
with a burnt leg it means its a spike to ground this is typical of lighting or maybe a problem with the neutral to earth connection.

I agree the cause may still be there. And I agree they are likely MOVs.

[francisco.shi]

I just had a look at the picture. I would be 99% certain they are MOVs.
When MOVs take a hit they always destroy themselves in a spectacular way. Capacitors on the other hand rarely explode like that.
That means there must have been a (substantially) high voltage spike for the MOV to have blown up this way.

[coulomb]

I just had a look at the picture. I would be 99% certain they are MOVs.

Though the designators strongly suggest capacitors:

MKS II blown caps or MOVs.jpg (54.77 KiB) Viewed 349 times

There are capacitor of about that size in iMiEV on-board chargers that explode. But capacitor or MOV, it suggests a large voltage surge, and I now believe that this is part of the AC SPS. So lightning or some other surge on AC-in.

[dangraham]

So the plot thickens, I was messaging two of the sales reps from Foshan Ouyad Electronic Co. Ltd. through the Alibaba messaging system, I asked that she sent the relevant information by email to both myself and business partner who need to make a decision on how to proceed and what they can offer.

Then started to receive emails from a “Nancy” using a completely different email address not the standard sales rep addresses. overseas01@ouyad.com overseas02@ouyad.com

Nancy [Zoeie@protonmail.com] Asking about the failed parts and to send address for delivery etc. It immediately looked like a scam, I messaged back through the Alibaba system complaining about it and asking them to clarify why the communication was not coming from the company email address?

Then a third person started to reply back through the messaging system and via the company email Alex [roberthu@ouyad.com] asking who was Nancy and what was she asking us for? Obviously the access and security controls to these Alibaba sales reps accounts are not very good so be careful about what information you give and always be aware that any information given on Alibaba can be used to try and scam you.

Alex has stated that he will send replacement boards……. I have now been emailing him for nearly 3 weeks to get conformation that they will send both the main PCB and the SCC, he has said “I will try..” he has also said they will not cover the shipping cost but offered for $125USD. Still trying to get confirmation they will send both boards and when they can send.

In the mean time I have been discussing with MPP about the PIP8048MAX inverters, also mentioned the potential issues as warned here, they say this has all been resolved with recent firmware updates. I have also confirmed their warranty situation and policies, so looks like I might give these a go and see how they fair.

Interestingly the last 5KW inverter is still running fine, no changes have been made to the system since the removal of the 2 failed inverters, it still has 9* panels connected and we still switch the generator on and off as required for 3 phase in the workshop and topping up the battery. All 3 were installed and operated exactly the same apart from one failed inverter only having 8* panels.

I have opened another thread on my inverter-welder-charger project if anyone is interested to offer some advice.

http://forums.aeva.asn.au/viewtopic.php ... 6b8776b0ef

Thanks,