[sdiy] LM394 vs. MAT02 vs. ?

[harrybissell]

Dana Scott wrote:

> Here we go, ARP questions. Now it's my turn.
> The ARP PNP/NPN  pairs were never electrically matched at all. Just take two
> transistors and stick them together.

Actually I asked Bob Pease how this might be done, and he gave a detailed
answer.  My question was actually "why the npn/pnp approach" rather than the
more common op-amp servo method.  His answer was "because the npn/pnp
was probably good enough"

I will try to find that e-mail (hope I still have it) and post the details.

>
> The ARP PNP matched pairs used in the later style VCF's were matched for
> similar current gains at 2 points. The LM3900 has current inputs so  the Vbe
> doesn't matter.

The 2SA798 is a natural for the ARP filter...

>
> The clips were solid copper, not merely copper-colored.( Like that matters)!

oh but it DOES.  Copper has a lot of mass and excellent thermal conductivity. In
order
to change the temperature of the transistor, you need to move the COPPER...and
that
cannot happen very fast.  This would equalize the temperature WAY better than
thermal
grease, or even epoxy.

This clips USED to be common, heatsink companies made them. I'd look at Aavid,
Thermalloy,
Aham-Tor... if any of these exist anymore.

H^) harry