For me if one Mosfet or driver transistor blew or measured to be bad, I replace them all (24 Mosfets plus 4 transistor pairs) - learned from my earlier bad experience.

And if a plated thru hole barrel plating gets pulled from unsoldering, I make sure top and bottom copper layers for the thru hole are soldered to the pin (works only for 2-layer pcbs).
Edited 2026-03-10 08:06 by tinyt

Ok with step 2. VS1 and VS2 tied to ground ribbon cable removed 24 volts and 320mA input turned on voltage on psu dropped to 20.34 volts and current draw 340mA.

TIP31 1.5 1.9 900mA

Measured the Vbat pins on the 10 pin connector 1.7 volts.




So there is something blown me thinks

Edit:

Tried again lowering the current to 250mA and the pictures shows the result
Edited 2026-03-10 09:07 by Bryan1

Ok removed the VS1 and VS2 to gnd wires and the current dropped to 51mA connect Gnd to VS2 and the short happens. Remove short the led comes on and....

TIP31 17.93  VBat  17.5

D25 17.41
D26 17.44  17.72

18V zener 17.93

Edit: that was interesting powered it up and decided to measure the voltage on pin 1 of the TIP 41 and TIP 42 chips.

On VS1 and VS high side 17 volts
On VS1 and VS2 low side no voltage.
Edited 2026-03-10 09:39 by Bryan1

Ok the following scope pic's are WITHOUT VS1 and VS2 tied to ground

6. CH1 to R1
  CH2 to D1




7. CH1 to R26
  CH2 to D7




8. and 9.

CH1 R27  
CH2 D18





so no action from the high side

Looks like picoverter H01 and H02  signals are missing. So in the picoverter board, replace the two IR2184 chips and do steps 8 and 9 again.

Edit: I think the TIPs high side are shorted. are they newly replaced? If not replace them first before replacing the IR2184 as the shorted TIPs can blow the IR2184s.
Edited 2026-03-10 10:39 by tinyt

I did put new IR2184 chips in and when going to 8. the scope has seen the connection and just that flat signal on both channels now on 9. the scope doesn't even pick up a signal.

Anyway got an email this morning from Aussie post so those new chips should be here soon.

So finally off this project for a day or so and somehow waiting for Aussie post is a good feeling  
Edited 2026-03-10 10:45 by Bryan1

Ok checked the tracking this morning and just got the confirmation that package of TIP chips has arrived  

So let the fun start again and see it I can get that energy meter to come to life.

Ok got the new TIP chips in hand so set up the tests again with the H01 and H02 TIP chips removed.

The short fault when VS1 and VS2 are connected to ground is still there.

So 24.1 volts 250mA current thru the resistor bank.

VS1 and VS 2 tied to Gnd

14.84 volts 250mA current

Remove VS1 and VS2 from Gnd

24.1 volts 46mA

Ok removed the low side TIP chips and still got that short with VS1 and VS2 to gnd.

So with Vbat on 24 volts 46mA current draw decided to measure the mosfet solder joints and found something interesting.

VS1 low side Gate 0 volts Drain 13.58 Source 0 volts

VS2 Low side Gate 0 volts Drain 17.44 Source 0 volts

H01 High side Gate 13 volts Drain VBat Source 13 volts

H02 High Side Gate 17 volts Drain VBat Source 17.44 volts

Edit:

Tied VS1 to Gnd turned it on and the TIP31 was at correct voltages now connect VS2 and turn it on VBat 2 volts 250mA current draw
Edited 2026-03-11 10:17 by Bryan1

check C8 for short or orientation.
Also C21 for short.
Edit2: Also C7
Edited 2026-03-11 11:14 by tinyt

Removed from the board

Checked C8 93uf

Checked C21 94uf

Checked C7 104nf

turned it on let the caps charge then connected VS1 and VS2 to ground and short circuit as Vbat on 1.2 volts 250mA current draw psu showing 15 volts

Now it is VS2 where the fault lies as connecting VS1 does work.
Edited 2026-03-11 11:32 by Bryan1

Earlier you said VS1 to GND and TIP31 at correct voltage. except when VS2 is grounded Current draw is 250 mA. Is this still true?

In the schematic the top row section, I think there is some short. We have to look for it.

With no power, compare resistance reading at the pads of C8 and C5. Follow polarity mark on the silkscreen for the capacitors.

Ok got the linux box going now here are the pic's of VS2 on the board.


Top Side



Bottom Side

Board is almost bare, and still we have a problem.
Still with no power compare resistance readings across the pads of C8 and C5.

C5 measured no measurement where C8 is a dead short, measured the 10K resistor which did prove to be 10K so left it in.

No probing using VBat - with one probe does NOT make contact around the C8 area with the other probe.

Checked VS with the VBat connections underneath and no contact so no shorts there.
Edited 2026-03-11 12:14 by Bryan1

Use a magnifier and check for copper trace short between the pads of C8.

See red Arrow


Edited 2026-03-11 12:09 by tinyt

Ok got a sharp knife and ran between the track and pad and the exposed copper does show the dead short. I also had a look with my 45X magnifier which does confirm the short.

So I am thinking cut the track from the TIP 41C to the cap and make the connection under the board using the cap legs to make the new circuit path.
Edited 2026-03-11 12:25 by Bryan1

  Bryan1 said  

Ok got a sharp knife and ran between the track and pad and the exposed copper does show the dead short. I also had a look with my 45X magnifier which does confirm the short.

So I am thinking cut the track from the TIP 41C to the cap and make the connection under the board using the cap legs to make the new circuit path.

I would just cut a little bit of the shorting copper.

See yellow below, You can paint the exposed copper with a little bit of nail polish or something similar.


Edited 2026-03-11 12:36 by tinyt