Anybody Know How to put a Snubber on Surf PI?

[FatBob]

I'm building a Surf PI and read that longer TX pulses will cause the FET to avalanche and heat up.

I thought a snubber could help but most are configured for the N type FETs. Can anyone show me how
to use one on the P type?

Schematic;

[ODM]

If the snubber is built parallel to the coil, you can use the same type of snubber in either top- or bottom switched type of circuit, when taking the current direction into account.

In bottom switched the N mosfet is usually switching the current flow from common rail to negative rail; in top switched the P mosfet is usually switching the current flow from positive rail to common rail. The flyback current is opposite to this current direction.

If you aren't sure, show a schematic with the snubber you wish to add to your SPP.

[FatBob]

Well I have this one that was on a similar detector;

[kt315]

...

[green]

Well I have this one that was on a similar detector;

[ATTACH]34975[/ATTACH]

My thoughts. Reverse diodes D101 and D102 when using a P channel fet. C104 controls ripple volts(220u probably a lot larger than needed). R103 controls snub volts if sample rate and peak current stay the same. If changing peak current or sample rate replacing D101 with a zener string suggested by Mickstv in another thread might be a better way.

[Altra]

Here is one that I have been using for p-channel. It is basically a shunt regulator and not frequency dependent like
some snubber designs. You can use zeners in series to adjust the the shunt voltage. I have mine set for about 180v
now. What I would like to do next is derive the minus 5 volt supply from this and eliminate the dc to dc converter.

[ODM]

Well I have this one that was on a similar detector;

That one doesn't look like anything strange. Just add it the other way around for your pmos switched coil, so flyback flows through the diode, and electrolytic cap polarity is correct.

[FatBob]

I really need to see a drawing to understand so would it look like this?

[ODM]

The FR307 diode should be connected to coil, in opposite polarity. If you think of how it works:
* Mosfet switches current to flow through mosfet diode and coil.
* Snubber diode won't conduct during TX pulse.
* Mosfet cuts off current. Flyback voltage appears because inductively stored coil current still flows.
* Snubber diode conducts due to appearing reverse voltage, coil current to capacitor.
* High voltage during coil current damping helps damping speed.
* Once flyback voltage is below capacitor voltage, diode stops conducting.

[FatBob]

Yes as I looked at it after I posted I was thinking one of the diodes was backwards. So it would go something like this;

[ODM]

Looks reasonable. What's with the 12 ohm resistor shunting the coil though, looks like a typo?

I would recommend a basic electronics book for confidence in modifying existing designs. Do you have an oscilloscope, one might be very useful for tweaking the snubber until it works right for your coil and design? The cap and resistor value might need tweaking for different operating voltage, coil, and pulse width, repetition rate.

[FatBob]

Thanks for your help!

Yes the 12 ohms is a mistake.

I have access to a scope so can fine tune it.

I just get a bit confused using a P type everything is backwards!

[ODM]

Well if you just think of things in terms of voltage polarity and current direction it's pretty straightforward.

[green]

Trying to understand capacitor selection for the snubber. How would a 250v, 220u aluminum, 250v 1u X7R ceramic or a 250v 1u film capacitor compare in the above circuits? Include a 250v, 1u aluminum in the comparison.

[FatBob]

Ya this snubber has 100n cap...