Totem pole gate driver unsuitable for SMPS?

Also the 2N2907 C E are reversed, it can work like that but this circuit shouldn't be taken as an example for good design.

Also the 2N2907 C E are reversed

Click to expand...

Thankyou, woops sorry that was a mistake..please find it edited and corrected now in the top post.. (attached).
Thankyou very much Vbase...well spotted!

Also, the BJT totem pole will never be as good as a gate drive IC such as eg the MIC4423, which ensures fast switching right up to the rail of the driver, due to the combo of FET and BJT drivers...surely the plain totem pole BJT is just not good enough really, specially not for high power SMPS?

MIC4423

Discrete push-pull gate drivers are used in many designs. Transistors need to be selected according to gate-drive currents, fast high current drivers use typically 5 - 10A rather than 500 mA transistors.

When the load is inductive the gate drive is deliberately slowed for the purpose of using the mosfet as snubbing for high frequencies.
To my opinion this gate drive is weak but good enough. I only suspect that there is another error in the drawing and the gates should be connected to the cathodes of D5, D2.

I only suspect that there is another error in the drawing and the gates should be connected to the cathodes of D5, D2.

Click to expand...

Thanks, the gate drive bit is drawn on a separate page to the fets, so its impossible for me to know if thats the case.

The gate drive schem labels the gates "N26" and "N23".
But nn the main schem, one gate is labelled "N23", but the other is labelled "N22"....so its not at all clear unfortunately.

Driving a PFC FET gate through essentially a series 1u cap sounds a bit unusual though.....so i kind of tend to doubt that they are doing that?

The attached sim in LTspice shows the gates being driven through the 1u cap

On switching the gate on the 1uF increases the charge current of the Ciss for a duration of few microseconds, that switches on the mosfet faster. Maybe it decreases switching loss.

The gates are definitely connected to the zener diode clamps, not the BJT emitters. That's the only way the circuit makes sense, anyways.

Personally, for these sorts of circuits I would want the transistors to be specified for the peak current they need to deliver. 2N2222/2N2907 don't have peak current ratings, so hard to say if they're suitable here.

The gates are definitely connected to the zener diode clamps, not the BJT emitters. That's the only way the circuit makes sense, anyways.

Click to expand...

Thanks, I see what you mean, I too have been trying to look at it from the “makes sense” viewpoint

On switching the gate on the 1uF increases the charge current of the Ciss for a duration of few microseconds, that switches on the mosfet faster. Maybe it decreases switching loss.

Click to expand...

Thanks, I too thought this may be it, and then I simulated the circuit with the 1u cap in circuit, and then shorted out (as per LTspice sim attached) …and the one where its shorted out does not manage to bring the gate voltage down to zero, whereas the one with the series 1u cap sucessfully brings the gate voltage down to -0.7V....which is an improvement....so i am guessing that for this type of totem pole driver....the series cap allows you to bring the fet gates definetely down to zero.....

...of course, the fet gates could have been driven down to 0V if a proper gate drive IC had been used...but thats another subject........though it does make me wonder why people use these BJT totem pole things at all for FET gate driving?.....i mean SOT23-5 gate drive IC's are cheap, abundant and plentiful.......even if one went nil-stocked, there would always be another variety on the shelf somewhere that fitted the footprint like-for-like.