15 kW Output Board Connections

Should work.  If you're directly running the transformer wires to an output panel (like you said), then you only need the first and last connections here (L1 -> AC O/P L1, and N -> AC O/P N).  The control board regulates from 120v, not 240v, so the L2 connection is not necessary.

Filter capacitors across the output of the transformer are also a really good idea to get a clean sine wave (these are on the L1 - N side on the board here...though I do recommend removing all but one of the three stacked up in a row, to help reduce no-load current).  Also would be a good idea to put one of the (removed) filter caps from L2 - N to clean that up as well.

We just finished testing this rig.... it runs all test loads just fine, but we found two weird problems:

1)  Doesn't re-start.  Runs fine when you start it (PWR SAV OFF), but once you switch it to "OFF" it will not re-start in either mode (no loads) unless you unplug the battery and cycle it.

2)  The voltage is low.  Seeing 98/95/193 volts instead of 120/120/240.  I adjusted VR2 to correct level, but when we re-started it, it did the same thing.  VR2 is now maxed out, and the voltage readings are still the same (198-201v max).

This is a 48V rig with an AS9 transformer.  It ran our 240V central A/C just fine, and a few 120-volt loads, albeit at a lower voltage.

Are you sure it has an AS9 transformer? 🤪  15kw PJ inverters have historically always used the AS6.5.

Unbalanced output phases on the transformer is a PJ characteristic, consider it some added character thrown in for free...

Bit stymied on this one.  When you say it doesn't re-start, does it throw an error (ALARM light, beeper)?  Does it try to start the transformer and then error?

Is this a "stock" inverter as originally shipped by PJ, or is it a hodgepodge of parts put together in the same chassis?  (PJ has a habit of changing resistors on boards, and this could have quite an affect on the end result, particularly if it's not how it was tested at the factory.  Resistor voltage boards--the removable ones on the "output board"--have changed from older to newer inverters, and switching these out can easily cause a 20-30v shift.)

What battery voltage are you providing, and what voltage is the transformer "O/P" (low side)?

I'll have to check the transformer's low side... I can't remember the actual values off the top of my head.

I ordered the components for this inverter a while back from Sean, and specifically asked for the AS9 (this was going to be a big boy inverter).... so they are more than likely not completely matched... but it still runs great.  I could not stall it, no matter how many loads I put on it.  

I'm still lost as to why it won't re-start.  We ran it for two hours -- both load and no load -- without any issues, then turned it off.  When you turn it on again, the second LED lights up (INV ON), but nothing comes out.  No fault codes, nothing... almost as if it were on standby... except it's not.   This is by far the strangest thing I have ever seen.

If you cycle the power by removing the battery, then it works just fine when powered back up.  It's almost as if it doesn't see the switch again.  This is an older main board, so there are no DIP switched on it, as far as I can see.

EXTREMELY GOOD INFO TO KNOW!  I am currently assuming that you did not get the WiFi upgrade kit with it?

That bit of info puts an entirely different spin on the system and the issues...

So I have a very good idea as to what the problem might be...but I need a bit more info...

Big question: Is it running the stock PJ firmware, or is it running the redesigned GS firmware?  An easy way to find out is to trip an alarm (low voltage, or heat a temperature sensor up with a soldering iron), and see what happens. 

• If it rapidly beeps (also flashing the ALARM) LED, it's GS firmware.  PJ might beep, but I don't think the LED flashes with it.  (Been so long since I've used stock PJ firmware, I can't stand it!) 
• GS firmware also may briefly beep when you turn the power switch off.

If it's running GS firmware...there's a stupid "battery charge protection" circuit on the PJ board that I might have "broken" the reset code for--and without it, the CPU will try to drive the FETs, but this "protection" circuit will block all the output signals, so the FETs won't do anything.  This would explain the first issue you mentioned.

For the second issue, if it's got GS firmware, that would REALLY explain why it's running low output voltage--and this is almost a given for an out-of-the-box fresh programmed inverter.  Gotta love PJ...I adjusted the CPU's ADC (analog to digital converter) for a longer sample time...and suddenly they started regulating low voltage, even with the knob all the way up.  Worse, this is not an across-the-board change: different inverters regulate quite different output voltages with this change.

But if it's got GS firmware, then I am very confused as to why Sean would sell it without a WiFi board...?

This one definitely has PJ firmware... there is one long, continuous beep if the batt voltage is out of range.  I think we purchased this just before the WiFi board came out... we had it running continuously for a long time..... and never even realized there was a problem (because we never shut it off).  I didn't even notice the voltage issue, because our building is 208v anyways.

Gee, who knew that the PJ code is glitchy.

Is the code actually loaded onto the processor, or is on the signal board?

The less said about the PJ code the better 🙄.  I'm amazed that the inverters function as well as they do, and not surprised at all that they tend to blow up in the transition to and from battery charge to inverter mode.  Once I get enough GS stuff finally done, I'll go back and finish revamping the PJ firmware, and provide them with an upgraded firmware.  Which will be...smaller than their original firmware.  With a less efficient compiler.  And with a lot more features.

The firmware is actually written to NVM (non volatile memory) inside the processor.  It's actually an MCU (microcontroller unit), not a CPU...but PJ calls it a CPU, so I let the lingo slide.  A CPU would require external memory to store the program; an MCU has the program memory internal.

So with it having stock PJ firmware, I'm at a bit of a loss as to why it's not resetting the stupid "shutdown circuit."  Low voltage could potentially be due to the hodgepodge parts, and the resistor boards being different than originally used.  These will easily change the output by 20-30v.

Nice of you.  And probably smart since my prediction is that within a year from now, powerjack/upower inverters will ship with a "new" control board which will be a knock off of yours anyway. 😄

I haven't given them any code for it ("we have no code, we cannot test inverter before ship it!"), so it's completely useless for them.  As far as Jack probably sees it, we're starting out looking similar to what they started with...and just haven't yet figured out that their way is better.

I even offered to make them a new control board that's bent towards their ideas anyway, was told it wasn't necessary.  Still plan to do it, though.

All of the suggestions that turned into Genetry Solar inverters were ideas I first suggested to Jack anyway.

So while getting ripped off is a possibility, I don't think it'll happen that way.  They even (started) to pay me for my time on the new CPU code...but I still haven't gotten them a fully useable firmware version, so they've held off for a bit.  While I don't know the Chinese culture very well, I feel that we're on pretty good footing at least so far.