Distortions in SPWM inverter?

Images:

FvM said:

As a "blind" guess, seeing the filter (reactive load) bringing up the problems suggest that they are related to body diode reverse recovery. IRF740 has a slow body diode, you better avoid to make it conduct a considerable share of the load current.

Click to expand...

So, the problem is that the filter tries to maintain the current which have nowhere to go during dead time, and body diodes can't handle it or turn on in time?
Then, it should be as simple as adding external diodes to the FETs.
I'll try that.

Meanwhile, i've uploaded the images to the board storage - should work now.

FvM said:

None of the links presently works for me. Why don't you post your pictures at edaboard as others do?

Click to expand...

Strange. Is it that the site not accessible, or the images not loading?
They are generated by the scope itself, can that be a problem?
Would a re-coded one open - http://orbides.1gb.ru/img/spikes-all_c.png ?

Why - people tend to follow the path of least resistance.

- - - Updated - - -

Adding diodes (SF28) antiparallel to the FETs had no noticeable effect.

I guess the links don't work due to a server problem.

Adding diodes (SF28) antiparallel to the FETs had no noticeable effect.

Click to expand...

Most of the current will still flow through the body diodes.

FvM said:

I guess the links don't work due to a server problem.

Click to expand...

Um, including the two on the top, that are attachments on this server?
Or you do see them?

The server problems are with orbides.1gb.ru

I don't see much on the waveforms, except
- although the voltage overshoots are probably not critical as such, they shouldn't be there. Either the layout and bus capacitor inductance is too high, there's an excessive high current peak, or it's a probing problem.
- consecutive switching periods are considerably different. Either the control signal isn't perodical or other conditions, e.g. the bus voltage aren't stable

For a complete diagnosis, other signals (bus voltage, output current, drain currents) must be recorded.