I am to the exciting bit of my Wiseguy inverter testing. Tests 14/15 where the control board and the power board are still on current limited supplies, two 560uF caps are in place, the choke and toroid are connected and a 1uF capacitor is across the a/c side.
At about 12-13V applied to the power board the a/c output starts and the output wave is shown below. Increasing supply voltage >13v causes current draw to increase above 500mA (tried up the 1A so not just on the margin). The wave form looks like its missing one quarter (would look OK if it was not missing). Any suggestions where to look. Checking the Power board for something around the wrong way?
Inverter a/c Output

---

F4 H7FotSF4xGT

Current at 14V should be around 300ma. I don't have wiseguys info on the PC I'm on at the moment. So assume you are measuring AC output?

I ask because it would be more like half the waveform missing if start vert ref is mid screen.

What was the controller outputs like, I think WG had them in his tests, but I'm not sure?

Yes measuring AC output. The vertical 0 is where the 1> pointer is on the screen grab above. AC or DC couple shows the same thing. There is no earth connection on the AC output except when the DSO probe is connected.

I have not looked at the input waveforms, will have to research what they should look like.

What will happen it I pull the Opto's out one at a time to narrow it down. Will the FETs just not turn ON or will they always be ON.

That is not the best way, the current is low at 13v, and you have small caps, you could look at the opto in and out wave forms.

So full waveform but with distortion. Have you checked that 1uf 300vac rated cap across the output, usually 3uf is a better value, just making sure it's not something simple like a dodgy cap. You can take the Toroid off and use a small RC network across the board output, but need to look up the values, been a long time since I touched mine.

The output of the transformer can never have a DC offset so it will always be fairly symmetrical and AC or DC coupling will make no difference. The primary side will give a better indication.
If you don't have a differential probe 2 standard ones will do (assuming the scope is 2 channel) with one trace inverted to make comparison easier. If it has math functions subtract ch2 from ch1. If single channel save the waveforms for each side and compare them on the computer.

What looks like a negative quarter cycle could be an inductive spike from the transformer, rounded off by the output capacitor, if one or two MOSFETs aren't switching on.
That would also explain the higher current as there would be a DC component to the waveform. The DC resistance of the primary is almost zero.
At full voltage perfect symmetry is essential to avoid saturation.
Edited 2026-02-28 17:45 by phil99

FYI: I was alluring to swapping the toroid for a RC load to not only remove the toroid from the equation, but at the same time, allowing measurements on the power board without having a nail as a load for the FETS if a slip happens. Not interested with DC offset.

Hi Gerry - on holidays again so away from pc. I would be looking at the opto couplers/drivers. One reversed maybe or cheap sockets gave me that result (optos were oriented ok). Look at the gate drives - no supply needed on FET main supply only the 12v from controller.

Check orientation of the smd drive transistors.

There was also an issue on some early PCBs that needed a short on one trace to ground be removed. Not sure if I can access the info here but it was in the thread when boards were being ordered.

Thanks for the ideas. The short to GND on R6 I am aware of and was fixed during construction.
Just trying to get to grips using the DSO and what I should be seeing at various points.

Gerry if needed, I have some close up pix of the power board on my thread, but I have more if you want to post a pix of the board, I can compare it, otherwise I'll look for pix of that version board and post if needed, I made a few.

Mike whats a holiday? I don't remember them  

If all of that reads funny, I'm in a rush  

What is a holiday ? = Stop having fun in the shed and take me on a trip somewhere where there are no soldering irons, CRO's or meters to distract you! The holidays will continue until morale improves    

Well that looks better. Thanks for the pointers everyone. It was good to track it down having a look at the signals with the DSO. I learnt a lot more than if I had just visually noticed the bent pin on that FOD3182 Gate Drive Optocoupler.

I used two channels on the DSO with MATH A-B to do the probing of the signal. I could not see anything like the expected pulse type waves anywhere.
I had the LCD connected and could see the PWM% on that display was 0.1%  With Mike's (@KeepIS) later software the PWM won't start until VCAP is close to VBAT. I disconnected the VCAP sense wire between the Power Board and the Controller and connected it to the power on the Controller board. Mike's note below applies also if power board is connected but not powered.

PWM% was now 100% and the waveforms looked a bit like I expected. I probed the outputs of the Optocouplers (pins 6&7) and one did not appear to be giving an output.
I pulled it out to change it and noticed a bent pin that was not getting into the socket. I adjusted the pin a put it back and it had output.
I then powered up the Power Board and all was well, 250mA current when 48V applied.
So the original post shows the output with one Opto not driving, which also results in greater than expected current draw due to the imbalance.

The note from Mike's Latest Changes.txt

If you want "Test mode" without Power Boards, make a plug
for the Controllers SPWM output connector and connect the
Vcap pin (both Vcap pins in Dual Mode code) via a 1K
resistor to the controller DC input line.  


Gerry

Congrats! Seeing that first clean sinewave it’s just like love at first sight hey lol