Transistor "ringing" or oscillation

[ACharnley]

I developed this driver, it works well apart from the choice of transistor for Q1/Q5 has an oscillation that I've not been able to squash. Different transistors make a large difference but it's always there. Any help appreciated.

The attached LTSpice uses standard components, no changes should be necessary to run.

 

[ronsimpson]

L1 seems to cause the problem. Change "0.1" to "0.1n".

 

[crutschow]

What is the purpose of, or reason for, L1?
0.1H is a large inductor.

 

[ACharnley]

It's part of the power supply characteristics. Deleting it would be nice I know!

 

[ACharnley]

I spent a few hours on this yesterday trying to replace the VOM1271's which are very hard to come by (and also bloody expensive).

When a non-isolated gate voltage is used a zener has to protect the high fets gate from being pulled too low. Now when Vgs goes positive it tries to flow through it so a reverse diode is needed (D1, D2).

The fetDC goes up to ~18V but in AC over-voltage the gate can now be 100v - zener, D7 prevents backflow now as Q5 can't handle > 7V.

By introducing D7 the transistor to ground (not shown on diagram but was under Q5) can't drain the gate. Therefore R5 must be lowered into an acceptable range.

At 100v - zener there's quite a flow through a low value R5.

And I'm in a loop!

So isolated drivers are definitely the way to go, I just don't want to rely on an almost impossible to source chip to make it work.

 

[MrAl]

It's part of the power supply characteristics. Deleting it would be nice I know!

Hello,

There is a huge inductor in series with the power supply output lead? That does not sound right.

Step one would be to eliminate that inductor, then see if it works better. If it does then you know you have to find a way to mitigate that inductance. If it does not change anything, then you have to look elsewhere.

 

[ACharnley]

It seems to be related to the zeners. I removed them but I'm fairly sure the low side FET's can't be driven from the same supply so I'll probably go back to the classic resistor/zener for those (I don't need high frequency driving).

All being said, I'm still going with VOM1271 if I can get ahold of any.

 

[crutschow]

It's part of the power supply characteristics.

No supply I know of has a 0.1H inductor directly at the output.
The supply output should also have a large capacitor to ground (100uF or greater) for ripple and noise suppression.

 

[ACharnley]

No, the inductance and resistance are part of the supply. The V1 doesn't accept inductance characteristics.

 

[crutschow]

No, the inductance and resistance are part of the supply. The V1 doesn't accept inductance characteristics.

Sorry, but that makes little sense.
What exactly is that power supply?

 

[ACharnley]

A bicycle dynamo hub.

 

[MrAl]

A bicycle dynamo hub.

Hi,

Makes more sense now