Replicateing integrated circuits using discreets......comments ?


Leaving bad site ASAP!
Hi All

I have an interesting question, would you replicate an integrated circuit with perhaps 30-50 transistors in discreets and if so why ?

Would it be for the theoretical challenge ?
Would it be to learn more thoroughly how the IC functioned ?
Would it be because the IC design was not quite right for what you had in mind ?
Would it be to save cost ?
Would it be to improve reliability ?
Would you ever consider going beyond simulation and actually making it ?

Comments welcome :)

John :)

chris holloway

New member
When I was in school, a buddy of mine had an idea to breadboard a "741" op amp out of discrete components. So we scrounged up dozens of 2SC2274 transistors, and whatever glass diodes that were available to us and set out to do just that. We breadboarded it using the schematic diagram from the datasheet as our template. Of course, the circuitry more elaborate than would've been otherwise required as the chip used many extra transistors to avoid the on die real estate cost involved with using resistors, but we actually attempted it. So, after spending nearly the entire school day hooking it up, debugging and testing it as a standard inverting op-amp and you bet it actually worked quite well. But then we couldn't think of any practical use for it beyond that and so we returned all of the transistors and diodes we had scrounged up back to the lab. That was way back in 1986-87, we were both complete electronics junkies back then and we were always looking to challenge ourselves in that way. My friend went on to have a brilliant career in engineering. I suppose today you could simply emulate such a thing, but it wouldn't be nearly as glamorous though, at least for us, ...what fond memories ... sigh.