[DSP-10] Re: dave garnier's message

[Michael Lodico]

Dave I am glad that I could help with some input.

>
> Mike,
>
> Very interesting stuff!  I am beginning to appreciate all of the work
> that goes into the making of a crystal.  Years ago I had a tour of a small
> crystal company in Burlington Wisconsin, it was a quick 25 cent tour
> because the owner was also a ham, (his company's pride and joy was
> a specialized a crystal housed in a octal tube envelope manufactured for
> HP but that's another story.
>
> Your last statement suggests one could recondition a crystal element by
> cycling the unit through long periods of constant temperature extremes
> while cycling the drive to the crystal's.  Correct??? (I take it, in the
> caseof temp controlled units one would disable the oven part, right?)  ;-)

When a crystal becomes unstable you cannot recover them.  We used this to sort our
components i.e.: age them before manufacturing the oscillators.  But if you have a
pre built oscillator and want the best stability then you can do the aging with
out power applied to the heating unit.

> Mike, any idea or theory on what�s going with the crystal element during
> this cycling period?

What happens is that the holder i.e.: the spring clips that hold the crystal blank
expand and contract and you create a set repeatable pattern of pressure on the
blank.  Now as for the drive on the blank when they are first made you will have
some of the plating on the blank fall off over time until it reaches a point that
what is going to stay bonded to the blank will be stable.

>
>
> Mike I have some other questions about the manufacturing process,
> how is the crystal blank tweeked during it�s path through manufacturing,
> some sort of etching?  Anything to do with the silver plates deposited on
> either side of the blank?  Are crystal blanks ever polarized by a high
> voltage in the process of manufacturing, is there a need?  Interesting
> stuff,

I never got involved with the actual manufacturing of the crystals themselves as
this is a specialty and we had a crystal engineer for that end.  But I can tell
you what I know of it.  When the man made crystal blocks would come in the
engineer would have a small slice of it sawed off so that he could determine the
characteristics of the block.  We used primarily AT cut crystals and if we we
going to do a run of say 100 10 MHz OCXO's I would let the crystal engineer know
what I was looking for as an operating temp for the oven lets say 80 degrees C and
what the crystal loading would be in this example.  Now he would probably make 150
to 200 10 MHz crystals with say an apex of 80 degrees C.  As the crystals were
made they would plate the crystal blank and mount it in it's holder but not put
the can on it.  They would then go into a testing oven and a computer would record
the resonant freq. of each crystal.  We used trays that held 50 crystals at a
time.  After they had aged a bit in the ovens they would then take the printout of
each crystal and hand etch it to bring it to freq. testing it in a jig
constantly.  I guess the magic of the whole thing was they knew exactly how much
the crystal would shift in freq. from putting the cover on it and cold welding the
holder plus the back filling with an inert gas.  If you had good crystal blanks
and an engineer that was good you would probably have a fallout of 15% plus we
always want a few spares kicking around for engineering samples.  As far as
polarizing I suppose you could say that occurred when the plating was flash
applied but I'm not sure a crystal engineer would call it that.

>
> Mike, thank you.
>
> dave garnier  - wb9own

Like I said Dave I hope this sheds a little light on crystals and could be of some
help to people.

Mike  Lodico - K1EG