For troubleshooting the power board, I think you can unplug the ribbon cable coming from the spwm board and apply power thru the pre-charge resistor to check if there is any change in current draw. In my case (aliexpress power board) current draw is about .05A or less.

I check installed mosfets for shorts by ohming out between the gate/drain gate/source drain/source. In my case usually the ~23KHz leg of the H-bridge short out and the 50/60 Hz leg are still good. So, I have good installed mosfets that I can compare readings with.

Morning Guy's,
              Well first thing I took TinyT's suggestion and took the ribbon cable off the power board and the short was gone   Now with the ribbon cable off I connected up the resistor bank and pretty well straight away the voltage got to Vbat  

So it does look like the power board is OK

Turned off the resistor bank and the voltage fell away pretty quick so my thought off mounting the resistor bank and using that 450 amp on/off switch. So when the inverter is off it's simple case of connecting the resistor bank as it also connected to the non contact on the switch so the caps can charge then the on/off switch can be turned on and the feed wire can be removed to take the resistor bank offline.

So the job now is to get the brain board off and find where that short is.

Regards Bryan

  Quote  

So the job now is to get the brain board off and find where that short is.

If no short is evident on the brain board when you get it on the bench it might be that that it was driving both high-side and low-side FETs on at the same time.
Checking the gate drive signals should show if that is the case.

Well took out the brain board and to be sure replaced the 7805 as it had copped a beating over the years with the heatsink attached.

Now when I power it up from the 10 amp PSU




As soon as I put in the ribbon cable on the board everything is shut down and all that gobblegoot on the screen does suggest to me when I programmed the nano's the files could of been corrupted, but thats only a guess.

Now both the leds come on too where the green power board led is on without the ribbon cable and when cycling the on/off switch the brainboard led does come on and off.
Edited 2026-02-09 11:06 by Bryan1

Well went and got my good laptop only to find it won't program the nano's so setup a 5 metre USB lead and used MX25 to program the nano's. I did see clearly the nano reset on the brainboard and I took the lcd off to program it.

So went and tried the test again where the LCD still has that gobblegoot and when the I turn off the inverter and add the ribbon cable on starting no current draw on the PSU but the volt meter stayed in the millvolt range.

I am wondering as I have moved the nano folders numerous times the files may have come corrupt so I had to go back to Sept last year in my emails to find Poida's email and now got fresh files to try.
Edited 2026-02-09 13:37 by Bryan1

Ok reprogrammed both nano's with those files from sept last year and no change at all the lcd is showing the wrong character set yet the defines in the serial file do correspond with the pinouts on Mikes LCD which I used. One would of thought once programmed the screen would come up with the initial screen.

Now as soon as I put the ribbon cable on after turning off the brainboard a dead short is shown so clearly as I can't find any shorts and the way the LCD is acting I have the wrong files for this board I do think.

Now the files I have

nano_serial_to_parallel_LCD_2.ino
pico_2_heavy_filter_bv_better_stop_prefect_symmetry.ino

Regards Bryan

Hi Bryan,
Well done getting to this point in the build, I suspect the baud speed might be a bit out,
nano_serial_to_parallel_LCD_2.ino   is 9600
sSerial.begin(9600);


I don't have  this one...pico_2_heavy_filter_bv_better_stop_prefect_symmetry.ino

See here https://www.thebackshed.com/forum/ViewTopic.php?TID=9409&P=47
looks like no LCD for this one, correct me if I'm wrong.
Very busy have to get back to work.

Thanks for that Aaron I did check the lcd file and yes that sSerial.begin(9600) is there and I did get another copy of the inverter file but still no change.

Now as soon as I connect that ribbon cable the short happens and I did check each pin on the 10 pin connector with no faults found and I even put my scope on and couldn't make any sense as every pin I checked was just noise.

Just checked now and although those 2 heatsinks are separate they do show a dead short so next job is finding some isolation tabs to isolate the fets before I try again.

Regards Bryan

Edit: Ok got all 10 of the front fets isolated and the front heat sinks are now isolated and thats not the problem as still got the same problem. Now in the morning I will isolated the fets on the rear of the board.

Now just to be clear the pcb's for this project were bought off that group buy a few years ago.
Edited 2026-02-09 16:51 by Bryan1

Just thought I could power up this control board with the Mad power board as a comparison...not the same as yours but might be a little help.
this is without the LCD or Toroid and no capacitors connected, 1 Fet per Leg (for testing) and was originally set for 48v, powered to 25v ans used 0.029, on or off was the same as under voltage was enabled.


Tried powering up quickly to 31.5v and the green power light come on, obviously used a little more power (0.072a) before under voltage lockout and then the green power light went out.


Your power consumption may be similar or higher.
I will try to dig a little deeper to see were the problem short might be.
Fault finding is a better terminology  
Edited 2026-02-09 20:36 by Revlac