Thermal linking of transistor pairs on Novo

Oh no! You don't want to get me started on Ebers-Moll modelling of transistor thermal properties? All that fiddly awkward Boltzmans constants....

If the transistors of a differential input or in this case of a current source are thermally matched they will work better... in theory.

It has been known with differential inputs used for extremely small signals that thermally connecting the transistors reduces noise - in fact some manufacturers make transistor pairs on the same slice of silicon for that very purpose.

As for hi-fi at line level, I have yet to be able to find a way of measuring this, and having worked for a couple of other manufacturers the same goes for them.

Many of these fussy differential input stages using thermal coupling do so because they are expected to work well instantly in places (like outside broadcast) where the temperature isn't a nice comfortable constant.

My thought is that if they're as close together as in the Novo, and that they're used in comfortable (as opposed to a blustery outside broadcast) surroundings, both transistors will adopt exactly the same temperatures save for self warming due to current differentials in the case of a current source, but that should equalise after a few hours as one transistor heats the ambient by maybe a difference of a degree.

However, as you know I know, differences that you wouldn't think of, or seem not to exist, can make an audible difference. And even if I can't detect it, you may be able to do. So give it a go and let us know.