[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Shottky Diodes in SSTC ---> Not A Necessity



Original poster: "Justin Hays by way of Terry Fritz <teslalist-at-qwest-dot-net>" <pyrotrons2000-at-yahoo-dot-com>

Hi All,

Thank you for the valuable information given, that's the stuff!!

Also Terry, what you're doing with the ultra-high power switching is
incredibly cool. I wonder how different these devices would perform
under LN2!! Die heating is the devices main limitation, right?

About not using zeners/transorbs from Gate to Source: I completely
agree that not using zeners leaves your MOSFET's wide open. Unless
you live on Mars and the local Radio Shack is out of parts
(someday!), use these diodes. Dan/The Captain: If I was using $50.00
MOSFET's, you can bet that there would be 10 zeners, 10 transorbs, an
overvoltage comparator, and...an RC snubber thrown in for good
measure ; ) Not really. I paid less than $0.25 a piece for my
MOSFET's, so I had to at least try the no-zener approach.

---> Below 2 snips from Steve C., regarding parallel ultrafast diodes
across MOSFET body diodes <---

 > The difference might not be very great, but the dynamic impedance
 > of both diodes is very low, so the external diode will take the
 > vast majority of the current.

I totally agree......

 > However, you can't guarantee that it'll work unless you use the
 > series Schottky.

I would say, that it will always work, provided that the correct
antiparallel diodes are used. By "right" I mean:

1). Trr is less than 100nS
2). Significantly lower Vf than the MOSFET diode it's in parallel
with.
3). Package can handle the I^2(R) dissipation

Some diodes I've successfully used: EGP30K, MUR4100B, FEP16JT,
NTE577, NTE576

I recently visited with an engineer friend of mine at NASA, and
reviewed a schematic of a motor controller design that flew in
several space flight missions. Keep in mind that a design like this
bears a paper trail from here to Pluto, and the design itself is
tweaked and turned until it ends up close to perfection. There were
no series shottkys in this design. Just antiparallel fast diodes,
like everything else. Granted, SSTC's are very different from motors,
but inductive kickback issues are addressed quite similarly.

---> Below snip from Ken Herrick <---

 > How do you handle crossover in your MOSFET bridges?

I have been using a single gate drive/isolation transformer, with one
primary, and as many secondaries as there are MOSFET's in my power
module (2 for half-bridge, 4 for H-bridge). All gate signals come to
0V with nearly unmeasureable (with Aron's TDS210 scope) precision.
Not that that's a big deal. Amidon FT-114-77 toroid core. 7T's on
pri, 14T's on secondaries.

---> Below 3 snips from Marco Denicolai <---

 >> "What is "testing"? If testing is watching with a O-scope that
MOSFET Id and Vds are OK, there is no shoot-through, etc. then that's
OK to me.

The waveforms were nearly identical with and without the series
shottky diodes. Nice square wave, with the normal diode reverse
recovery looking stuff at the edges. This stuff had a slightly
different look with and w/o shottkys, but not much different
(probably junction capacitance stuff here).

 >> Build hundreds of them, try each one hundreds of times, will it
break? That is the right perspective, IMHO.

This is the best way to do it, I agree. I certainly do not think that
this *has* to be the method used to prove a given design adequate for
the task at hand.

 >> You might want to investigate on the cause of these diodes
failure.

(Marco was speaking of my series 1N5822's). Overcurrent/overtemp!!! I
was no doubt exceeding the average forward current rating (3A) of the
diodes. They got hot enough to boil water, but I didn't think they
mattered, so I let them fry. And they didn't matter!

Once again, your thoughts and contributions are/were much
appreciated.

Regards,

Justin Hays
KC5PNP
Email: justin-at-hvguy-dot-com
Website: www.hvguy-dot-com

_