Sorry to hear of the shortage.
But those Vicors are insanely excellent.
I personally wouldnt ever do a product with one, purely because of the nil-stock issue of them.
Some big co's somewhere will be getting them.
You need to be speaking personally with the sales guys, and best if you can get the US guys to come over and speak to you and promise you stock...otherwise
, assume you wont get any.
____________________-----------------------__________________________
Another point, i am sure you will know about,
is that Vicor's , due to their high Fsw and resonant nature
, allow you to produce really small , superb SMPS solutions. The Vicor,
Sensibly, does not incorporate the in and out filters....we have to add
them. Now...supposing you are making a "general OffTheShelf SMPS", for sale to all takers...
(whoever they are and whatever is their use case)
.....Then we have no idea what will be the output LC filter of any SMPS that gets cascaded upstream of
it, and no idea of the amount of input capacitance that whatever gets connected to its (..the vicor's)
output will have. We also have no idea if any smps cascaded up or downstream of it
will have a control loop frequency close to that of the vicor.
As such....using a vicor in
this situation, may not be particularly great...because to "honour" all these occurrences,
you will need to use a big output capacitor bank both upstream and downstream of the vicor...
..and as such, it doesnt have a small filter...so you loose all your "small size advantages"
...and in this case....the vicor module is no longer quite as wonderful for you....and in fact,
probably cheaper to just use whatever standard solution involving paralleled dicrete stages
etc etc.
So as you know, the moral of the story is....a small, high frequency switching module,
may not result in a particularly small overall solution...a cheaper discrete solution using much
lower Fsw may be just about the same size as the vicor solution.
High frequency switchers, with their (often) super high control loop bandwidth, can be a problem....
....the input filter must always have a Zout which is <0.1(f[control loop])....and that means there's
a wide range of frequency there where you will have to be sure no LC input filter resonances mean a high Z peak...
..which would cause instability in the switcher........often, its better to have a low control loop frequency so that
there is less chance that whatever gets connected upstream doesnt bring about input filter based instability.
If an LC gets connected upstream of an SMPS module, and its (the LC filter's) F(resonance) is 5kHz say, then a smps with a control loop bandwidth
of 8khz may go unstable due to that...but a smps with a 1kHz bandwidth wont even "see" the 5kHz...and wont go unstable due to it.