[sdiy] Matched transistors - critical parameters?

[Mattias Rickardsson]

On 8 February 2012 10:37, Matthew Smith <matt at smiffytech.com> wrote:
>
>> Heat from one transistor's own power dissipation, on the other hand,
>> gives a *much* greater temperature gradient in the chip, causing a small
>> temperature difference between the two. Even if it's much much smaller
>> than when gluing single transistors together, and unproblematic for
>> synth applications, I guess different constructions & fabrication
>> processes could differ by some order of magnitude in this respect.
>
> Thank you. I don't see this as a problem, but it does pick up the gap in my
> reckoning, of where Δt could get into the works!

In other words, if temperature matching is extremely important, it's
worth ensuring that the surrounding circuit design allows the
transistors to dissipate as little power as possible in the
application - or at least to not vary its power dissipation much over
time. :-)

/mr