Signals 001

THAT Corporation announced the end of a substantial group of analogue audio ICs: InGenius input stages, microphone preamplifiers, line receivers and drivers, VCAs.
The reason was not the circuits. It was the process. The devices were made on an older 4-inch wafer line, and keeping that line running had become too difficult and too expensive. The line stopped. The parts went with it.
That is one of the realities of analogue audio. A circuit can survive for decades on paper. The component it was designed around may not.
And parts are not always interchangeable. A different transistor changes behaviour. A different transformer changes the circuit around it. A replacement IC may do the same job and still not be the same thing.
Sometimes that matters. Sometimes it does not. The useful part is knowing the difference.
Specialised analogue parts tend to live in a strange place. They are rarely produced in enormous quantities, but they often remain useful for a very long time. A good balanced input stage does not suddenly stop being a good balanced input stage because a newer process exists. A well-designed VCA does not become irrelevant because the market around it has moved on.
The problem is that manufacturing economics do move on.
Semiconductor processes disappear. Tooling becomes expensive to maintain. Foundries close older lines. Components that still solve a problem perfectly well can vanish for reasons that have very little to do with the problem itself.
For designers, that changes things.
Sometimes a circuit can simply be adapted to a newer part. Sometimes the change is small enough not to matter. In other cases the component was part of the character of the design: electrically, mechanically or sonically. Then substitution becomes redesign.
That is especially true in audio, where specialised parts often sit at very specific points in the signal path. Input stages, line drivers, VCAs and discrete amplifier modules are not interchangeable in the same way as a generic logic function.
My response to the THAT announcement was simple: I bought a small stock of the devices I know I will use.
Not enough to preserve them forever. Enough to keep building with them while they still make sense.
There is no romance in that. It is just practical. If a circuit works, is understood and uses a part that is still available, there is little value in redesigning it early simply because that part may become difficult to find later.
When the stock is gone, the design can move on.
Maybe the next solution will be better. Maybe it will be more elegant, easier to source or easier to manufacture. That is part of engineering too.
But a good part does not become obsolete because the factory that made it disappeared.