Mosfet rise curve in deep.

Supply Voltage about 25 v.
Yellow is HS gate.
Blue is Drain.
No snubbers at the H bridge, only TVS diodes for protection.

you can see the current pick up here - in the red circle : ~ 700 nS - so faster GD may well be in order !!!


Or is it 70nS ( zoomed in ) ?? OP should state clearly.

Easy peasy said:

you can see the current pick up here - in the red circle : ~ 700 nS - so faster GD may well be in order !!!
View attachment 192400

--- Updated Jul 16, 2024 ---

Or is it 70nS ( zoomed in ) ?? OP should state clearly.

Click to expand...

I understand that you suggest to boost the current in driving.I will agree with you!
But, I didn't understand what this faint voltage drop for 700 ns in the red circle means and even before the gate voltage rise..? The scopes take place with short piece of copper for reducing the extra noise.

....Or is it 70nS ( zoomed in ) ?? OP should state clearly....

..You mean to make further zoom at this 700 ns?

It is not clear if the x10 zoom is on - which would mean approx 70nS current pick up time - or whether the x10 zoom is off - meaning 700nS

70nS is better, and 50nS better for losses.

the only thing that can make the Vds dip in that small way is a pick up of current - there must be some time skew on your scope, as:

if you probe Vgs and Vds at the same time - you should see Vgs near the threshold voltage when this dip occurs !

That "faint voltage drop" might be the FET driver switching current kicking the ground ('scope attached to rail?) up and difference down. That's one thing that's going on during the "warmup".

Thank you all..!

Dick freebird

The gnd terminal of my oscilloscope has no wire, I have put a small piece of copper to reduce the distance.

Here we see pulses while the inverter is running.
Supply voltage 27 volts
Dead time 500nano
Driving voltage 14 volts

In yellow the pulse of the gate,
in blue the drain pulse at the same time.
Oscilloscope Pin Zoom to x1.
Pause time for sampling pulses at
5 μs

Pictures 1,2,3: High Side Gate and Drain.
Pictures 4,5,6: Low side G and D.
I add an deferential mode choke,and the Details is in the Picture 7.

I see that have a very clean pulse than previous, ringing is Low for no load condition.
Also i see that we have changes at the timing relationship between Vg rise,and Vd falls and opposite.
In practice the system have a very stable working, stable sweet noise at the choke i add.No load with a 1000va toroidal transformer is 10.9 watts.

Easy Peasy also i cannot see any more this small drop for 700 ns in the Voltage. And yes probably the zoom was at x10 , this seems and at voltages in the oscilloscope.Were is this important time you notice??

What's your opinion for all that changes ?

dick_freebird said:

That "faint voltage drop" might be the FET driver switching current kicking the ground ('scope attached to rail?) up and difference down. That's one thing that's going on during the "warmup".

Click to expand...

But it isn't - if anything large gnd currents from gate drive would increase the Vds signal.

you can see the other fet turning on here:



The other waveforms show classic ZVS switching due to the inductive nature of the load on the H bridge - switching off fast gives near zero current in fet channel before voltage, Vds rises.

The wave forms will be some what different at full power

Good evening to all !

At the point we have reached, and based on the measurements and data we have, would it be safe to agree to load the output with some load ??

I imagine it would be very useful to publish the results for the rest of the forum users, especially new technicians.

Today I also include some pictures of the choke coil and of the voltages and currents in the power supply, without load.

Picture 1: Yellow voltage AC transformer input,after choke.
Black, output voltage.
In the power supply we see the Supply voltage, and current.

Picture 2:Choke .
Transformer 1000va.

Picture 3: One 10000 μf cap close to the mosfets, and others around the PCB.

If you look closely to the pictures,in the white pcb the MOSFET pins are directly soldered to the thick 3x 10 mm copper traces for fast heat transfering.Of curse thick cables and a big heatsink will add for the load test.

I wish you the best of luck and outcomes for your testing !