diy solar

diy solar

Mystery used Sol-Ark 12K Full Failure (Fixed itself)

Usually the AC PWM IGBT H-bridge devices. They often take out the gate drivers also. If they short out, they can also blow out the battery side MOSFET's.

On SolArk's, these devices are buried below several layers of PCB boards on back heat sink requiring removing a couple hundred little screws.

This is picture of a MPP HF inverter that had AC input source hooked up to inverter AC output.

View attachment 185407
I don't think the battery input is shorted, the battery seemed to handle being hooked up okay, granted I didn't check for dead short and I should have but it read like 53V on the screen so I doubt it's shorted
 
Ah ok. Well if you wired the AC onto the wrong end I’m not sure the inverter can be blamed for not detecting the AC and refusing to start.
The inverter has to be active, independently putting AC output with battery power that is out of sync with applied AC on output terminals to blow the H-bridge IGBT's.
 
Also worth noting, in that video the Inverter shown is not backfeed capable, though I'm not sure if that means solark can handle a reverse connection like that better or not.
The inverter has to be active, independently putting AC output with battery power that is out of sync with applied AC on output terminals to blow the H-bridge IGBT's.
 
Capacitors usually blow correct? Blown capacitors in my experience emit alot of smoke and residue and will smell bad in the box forever basically lol. Neither was detected inside this unit from basic opening of panel. Is there some other way the DC bus could fail less noticeable from over voltage? Also not so DIY, working with capacitors is fairly common for me.
My point is that you may have large energized capacitors that need to be discharged before working on the board. Not that they have blown.
 
My point is that you may have large energized capacitors that need to be discharged before working on the board. Not that they have blown.
Gotcha, yeah that's something I can handle assuming I could find the issue and it seemed fixable
 
The inverter has to be active, independently putting AC output with battery power that is out of sync with applied AC on output terminals to blow the H-bridge IGBT's.
Wait so that's quite different than just reversing the wires and turning on?
 
What do you mean by reverse? Wire to AC out instead of AC in?
I'm not sure I'm just trying to get to the bottom of that the other person is talking about as far as this potential failure to judge how likely it is to be related to my issue based on the difference in Inverter architecture and likelihood of someone doing it on accident. For example if he means feeding the load side with grid once everything is already on and working correctly then that seems unlikely to be done on accident but if he means just simply reversing grid and load during installation and then turning on can cause this then its more likely which then tells me to dig deeper for damaged fets or igbts
 
I'm not sure I'm just trying to get to the bottom of that the other person is talking about as far as this potential failure to judge how likely it is to be related to my issue based on the difference in Inverter architecture and likelihood of someone doing it on accident. For example if he means feeding the load side with grid once everything is already on and working correctly then that seems unlikely to be done on accident but if he means just simply reversing grid and load during installation and then turning on can cause this then its more likely which then tells me to dig deeper for damaged fets or igbts

I was reading this thread as reversing grid and load initial installation. Seems quite plausible.
 
If it were me, I would step back and check the basics. A lot of assumptions tossed around already for just starting to troubleshoot.

Please make sure you check some videos to make sure you don't do some damage (I get nervous giving people ideas...), but there are some screws by the MPPT inputs that you can check voltage on with a known good meter. That way you can see if there is actually some solar input getting to the unit, and not some type of other wiring fault. Check all switches and breakers. Possibly a complete power off, and just start with one source.

If you don't take it one step at a time, its just a frantic game of hunting for problems all over the place.

Once you are done with all the connections, carefully check settings one at a time. Things like TOU settings can make a unit not use the battery. In that case you would only see grid pass-through, like you are seeing now.

Slow and steady.
 
If it were me, I would step back and check the basics. A lot of assumptions tossed around already for just starting to troubleshoot.

Please make sure you check some videos to make sure you don't do some damage (I get nervous giving people ideas...), but there are some screws by the MPPT inputs that you can check voltage on with a known good meter. That way you can see if there is actually some solar input getting to the unit, and not some type of other wiring fault. Check all switches and breakers. Possibly a complete power off, and just start with one source.

If you don't take it one step at a time, its just a frantic game of hunting for problems all over the place.

Once you are done with all the connections, carefully check settings one at a time. Things like TOU settings can make a unit not use the battery. In that case you would only see grid pass-through, like you are seeing now.

Slow and steady.
Alot of that I did already do, and as far as TOU stuff is why I wanted to do a factory reset of all the settings but it requires a password, I'll check into that manually.

I appreciate the advice on checking MPPT but even with no solar at all it should have turned on with battery alone.

Having turned on the screen with battery alone suggest the battery is connecting to something inside so reverse polarity diode and battery side mosfets seemsm okay
 
If it were me, I would step back and check the basics. A lot of assumptions tossed around already for just starting to troubleshoot.

Please make sure you check some videos to make sure you don't do some damage (I get nervous giving people ideas...), but there are some screws by the MPPT inputs that you can check voltage on with a known good meter. That way you can see if there is actually some solar input getting to the unit, and not some type of other wiring fault. Check all switches and breakers. Possibly a complete power off, and just start with one source.

If you don't take it one step at a time, its just a frantic game of hunting for problems all over the place.

Once you are done with all the connections, carefully check settings one at a time. Things like TOU settings can make a unit not use the battery. In that case you would only see grid pass-through, like you are seeing now.

Slow and steady.
This thread did remind me of some basics to go back and check like the power button wiring and TOU settings that you mentioned, will check into it when I'm messing with it again. Although I don't see why in TOU the unit would be off? But I've never actually had to mess with that as I've done offgrid and grid tie with no backfeed systems with these solarks so far
 
This thread did remind me of some basics to go back and check like the power button wiring and TOU settings that you mentioned, will check into it when I'm messing with it again. Although I don't see why in TOU the unit would be off? But I've never actually had to mess with that as I've done offgrid and grid tie with no backfeed systems with these solarks so far
That is a good point. I don't think it ever says off due to TOU, but it could not charge or discharge due to those settings. If I recall correctly, on mine the power button was lit when it said off on the screen.

One other thing I thought of, is to check basics like grid settings. There are modes for different voltages, and frequencies.

Good luck on it. I still have hope for you that it is a fully functional unit.

-------

One last edit. At least one version of the sol-ark manual says if the Normal light is not on, to call them. I'm guessing they will call back soon. They have even been working Saturday when I contacted them.
 
Last edited:
After a good sleep, grid settings come to mind too. If it was set to the wrong grid setting, you could get the AC too high alarm I would think.

Let us know if you have a breakthrough!
 
So.... bizarelly after messing with all kinds of settings and trying various things, it decided to turn on at one point and now works as expected, so big yay!!!

Not sure what exactly did it, TOU was off, power button looked okay, I remember I turned off grid selling, turned on BMS Error Stop, activate battery and some other adjustments but no idea what actually made the difference since I made several changes in setting at the same time.

Tried hooking up grid, load, solar and battery and it all seemed to work okay, after some reboots it still worked okay.

??? Go figure...
 
Back
Top