AW: [sdiy] Having a hard time finding the chips you want - roll y our own!
Czech Martin
Martin.Czech at Micronas.com
Wed Aug 29 09:13:39 CEST 2001
Just a garage and the equipment will not be enough.
-first of all, garage won't do, clean room does, it's not only yield,
but also "contamination poisoning the wafer issue"
-clean room and equipment given, you need a process, ie. the recipy.
The problem is that the recipy for wafers is a bit more complicated
then for cake. Even a simple process may require 100-200 steps,
and each step has to be taken with greatest care, sometimes
depending on the tolerances of previous steps to make up for
that.
-clean room and equipment and recipe given, you need to "run in"
the process. Ie. if you buy a perfectly good recipe, you will
notice that your first results will be crap, even if you follow
the instructions very carefully. Why? Because of tolerance.
You may not have the same equipment as the recipe supplier, and
even if, then your equipment is still slightly different.
We are talking about a few fractions of um or ppm only here.
Even with engineering aid from the recipe supplier you will
need several 100s or 1000s of wasted wafers until the process is
run in on spec. It's now the point when you can actually design
something, cause still some parameters will be different from
the original process.
-all run in you have to continuously watch and maintain the
process, very little changes in equipment and environment as well
as incomming chemicals can turn you back to the 100% crap
where you started. Basically this means that the wafer fab must
run day and night 360 days a year (5 days for cleaning).
A major halt can mean another run in.
You have to watch for each and everything.
Just a few examples: paper from one supplier may be ok, another
product will contaminate your wafers. 100% crap because you changed
the notepad you usually use.
Waferboxes behaved very well until your supplier changed something.
Again, contamination or static will turn your wafers into crap.
The make up your wife uses (we aggree that you need some help)
sometimes leads to contamination, as well as your smoking
outside the clean room in the breaks. And your beard and hair
cause trouble as well, as do little skin particles which you
distribute in millions all over the clean room.
So, running even a little wafer fab in a moderate feature
size process is not an easy task, and I doubt that the expertise
of one man allone is sufficient to get anything other then crap
out of it.
I see it here at www.micronas.com every day, and I have seen
it at the little fab at the chair for microelectronics
at Dormund University, the later was actually run by few
people a few days in the year thus the results where, mmmmhh,
difficult.
m.c.
> -----Ursprüngliche Nachricht-----
> Von: Magnus Danielson [mailto:cfmd at swipnet.se]
> Gesendet: Dienstag, 28. August 2001 23:59
> An: synth-diy at dropmix.xs4all.nl
> Betreff: [sdiy] Having a hard time finding the chips you want - roll
> your own!
>
>
> Dear all,
>
> I constantly hear people looking for this and that chip, but have a
> hard time finding it. Since we are all more or less into the mad
> scientist level, it is just logical that we eventually will go for the
> Chip-DIY trend.
>
> Here is only one place to look:
> http://www.universitywafer.com/
>
> There's more...
>
> Actually, I'm not very serious about this... but if we only pretend
> for a moment, just how much stuff would be necessary for a not overly
> complex analog design, say up to 20-40 transistors, about the equalent
> numbers or resistors and a few caps and a bunch of diodes. I think
> I've heard the voices (keyboards actually) of a few sufficiently
> knowledgeable people.
>
> Sure, there are some pretty hefty chemicals in use...
>
> But if someone think he could spare a garage-place and has been pretty
> lucky in both gamble and love (I think an understanding wife is
> required if one has one).
>
> But anyway, it would be fun to just pick up some of the knowledge in
> this field! Please speak up!
>
> Cheers,
> Magnus - has no spare space, no spare money and no wife
>
More information about the Synth-diy
mailing list