IGBT driver undertanding and repair help

[Gazello]

Hello,

i am trying to fx a welder with a broken h bridge IGBT modul. The driver is partly broken , a zener diode and the npn . I tried to understand the driver, but it is out of my understanding of regular drivers.

I think for fast turn off the diode /capacitor arrangement is a negative voltage pump, but not sure.

The zener ( seems to be a 2,4 or 2,7V ) makes no sense for me at all.

And confusingly nothing to reduse current flow to gate in the upper path , directly connected to gate ??

It is the same circuit in HIgh side as LW side only flipped the transformer connections .


if someone has the patience to explain the current flow and what makes what , i would appreciate  it.

I draw it maybe in a weird way but it should be right , to check i attached the original pictures and added the red lines for botton traces .  Fell free to check if i made a mistake.

thanks a lot
MArkus 

[T3sl4co1l]

That seems reasonable enough.  The zener and capacitor get charged to some peak voltage during the positive swing (through gate capacitance, I think?), while during negative swing, the transistor is turned on, shunting gate to -V quite aggressively (PBSSxxxx are quite powerful little transistors).

Turn-on will be limited by the 56Ω and transformer, so, pretty slow.  This is reasonable for a ZVS resonant type inverter, and probably still isn't terrible for a square pulse type forward converter.

The transformer should output about 20V peak then, I think, and the transistor sees +15/-5 Vge (if it's more like a 5V zener as I would think to choose, but even a mere 2.4-2.7V will still help out a bit).

IGBTs tend not to turn off very well at Vge = 0, and some extra negative bias helps ensure turn-off.

Tim

[Gazello]

Hello Tim,

is the turn on not done when the upper transformer leg is positive and directly connected to the gate? , the 56 ohm is in the path for turn off ( turn on the npn)

i tried to measure the diode from the good side off the h bridge , but i get results off 1mA 2V , 5mA 2.5V , 20mA  3V .  I ordered some 2.4 and 2.7 and compare it.

 

[T3sl4co1l]

Ah good, measured at current, so it's very likely 2.7V zener.  Gotcha.

Turn-on with upper leg positive, yes, through 56R, diode and zener.  Zener loses 2-3V (they're fairly soft diodes as you've seen, so the figure is approximate) and diode another 1 or so, hence for Vge(on) = 15V, more like 19V is required.  Turn-off through 56R+820R into BJT (with "speed-up" cap so it turns on as the transformer voltage starts to fall, before it even reaches zero voltage or reverses).  Not sure why the B-C diode (Baker clamp), maybe they wanted to avoid storage time into turn-on.

Tim