Decent Schottky diodes for bootstrap - 404

[dick_freebird]

Working on a project that needs a high voltage Schottky for
bootstrapping the high side driver, I'm not finding the combo
of high (200V) voltage and low (<<10nS) reverse recovery
time. I think this is because the HV devices have a parallel PN
guardring. And of course I'd also like a nice low reverse leakage,
which is what the guardring gives you. And a SPICE model
would be nice.

Going through Digi-Key and Mouser I came up empty handed
by the time I got through my gotta-haves.

Anybody run across a really superior Schottky device or
a vendor that stands above the rest? Maybe a distributor
who sells modest quantities in die form?

 

[Electro nS]

why donot you try Silicon carbide diodes . i heared those diodes are High voltage and high speed devices ..

just my humble opinion.

- - - Updated - - -

why donot you try Silicon carbide diodes . i heared those diodes are High voltage and high speed devices ..

just my humble opinion.

 

[schmitt trigger]

Second the motion on Silicon Carbide schottky diodes.

Cree produces devices in the 600 to 1700 volt range.

 

[FvM]

I think, regular "ultrafast" silicon diodes like US1M are best suited for HV bootstrap purposes.

 

[Electro nS]

As far as i know , anything under 50ns is considered ultra fast recovery diode ,isn't 10ns is very low ?
is it really nesseary to have such Trr in high side drivers ??!!

- - - Updated - - -

As far as i know , anything under 50ns is considered ultra fast recovery diode ,isn't 10ns is very low ?
is it really nesseary to have such Trr in high side drivers ??!!

 

[dick_freebird]

I'm looking to go to an insane fSW on a PWM buck, and
make it work up to 90% duty (so, a very low pulse low
time for the bootstrap to do its thing). Even 10nS is a
lot.

Maybe this is one of the reasons nobody does it. But
it's my job to try.

 

[mtwieg]

Silicon schottky diodes won't work out then. At high voltages, schottky diodes just tend to lose their favor vs silicon diodes.

SiC devices might, be a decent option, if you can find them in a low inductance package.

I'm wondering if your driver can even work properly with such high slew rates...

 

[dick_freebird]

Well, that's the whole point of the exercise - very high
slew rates are possible, in this (RF) technology, provided I
can keep it fed.

What I see in digging, and have experienced back when
I worked at a place that let people design their own device
structures as they pleased, is that higher voltage Schottky
diodes need a PN guardring, and the guardring ends up
carrying a fair bit of the current once Vf in the actual
Schottky region rises up, leading to charge storage that
a real, plain Schottky wouldn't exhibit (if you could in
fact make one that behaved acceptably, without GR).

Radius of curvature is just too small in a bare Shottky
junction, to support very high voltage.

Looking at GaN and SiC Schottky diodes, their Vf would
eat up all my low voltage supply headroom.