[sdiy] EH frequency analyzer

[Paul Perry]

>From the designer of the LM13600 & LM13700:

" After layout,design management decided to just go ahead and make it a
product. First silicon work even with very unusual beta. This last fact
explains why both the  LM13600 and LM13700  came out together. Control of
beta was not tight at the time and all transistors needed to be biased up
with currents well above the leakage current levels in the process. One way
to do that was to use area scaling to force the beta of the output
transistor in the darlington to 50. That what was used to the LM13700.The
other method involve feeding some of the control current back to the
darlington. At the time, this enabled the  LM13600 to have usable dynamic 
range well above 96dB."

I think the NE5517 is a LM13600 knockoff. Certainly seems identical to me.
And it's staying in production, because Philips asked On Semi to make them,
after the Philips plant burnt down.
It's kind of annoying, because On Semi is a spin off from Motorola, so
obviously they have the skills & gear to make all the other stuff missing
from this era, but not enough people give a shit.

But in any case, the best thing currently available to multiply A & B
together is an AD633, so far as I know. Certainly it beats a LM13600.
At least, you don't need a noise gate with an AD633.

paul perry (Frostwave) Melbourne Australia