If anybody was wondering why a usable set of those “holy grail” NOS transistors is so rare...

JayGee

http://www.thevalvepage.com/trans/manufac/manufac1.htm

Well I thought it was interesting anyway!

jacksparrow

That was a good read Thanks,  The complex process of manufacturing seems like a ball ache.

springhead

Nice, thanks for that.  Any desirable types - OC44, AC128, OC75, OC81 etc. have been scoured for by many individuals and booteek pedal builders for at least a couple of decades now.  What's left is generally the dross - out of spec for gain, leakage, noise etc. that large electronics manufacturers rejected as unsuitable for their products.   And reject rates were pretty high for Ge devices.  So any that test ok for our purposes command a price.  For dirt pedal use I don't think the writing on the metal can matters much, for home builds anyway, providing they have whatever characteristics you need.  For the booteek builders it does matter - "Rangemaster clone with real Mullard OC44" etc. etc.

ICBM

With the 'sorting' that's gone on with NOS transistors now, it may even be better to scour vintage radio/hi-fi type equipment for them - at least you'll know they were good examples to begin with or they wouldn't have been used when supplies were plentiful. I wouldn't scrap a decent radio or something else worth keeping, but it may be possible to find something of no commercial value to raid.

That said, there may well be enough vintage-type germanium fuzz pedals in the world already. [/heresy] And I like fuzz...

springhead

ICBM said:

.

That said, there may well be enough vintage-type germanium fuzz pedals in the world already. [/heresy] And I like fuzz...

Ooh, no no no no!  I've got a Mick-Ronson-in-a-box Tonebender Mk 1 and 'cocked-wah' combined pedal in build at the moment.  3 x OC75 or equivalent required for that one!  Just when you think you've got all the fuzz pedals you need something like that tickles your fancy and off it goes again.

ArchtopDave

Who wants to buy my soot then?? I live in Northumberland.

JayGee

It was the amount of hand work and manual test processes involved which surprised me, along with the (relative) crudity of some of the processes. What’s described and illustrated in that article looks much more like laboratory work than factory production and I imagine that this goes a long way towards explaining the low yields, the variability between even nominally “good” in-spec samples, and the need for individually selecting/matching transistors.