Identifying tempcos

Grant Richter grichter at
Wed Nov 10 05:59:03 CET 1999

A dim light bulb went on about this from some trade magazine
article in the 60s. Something like "Transistor circuit improves
rejection of Radar ground scatter" or something?

I am guessing the uA726 was used in analog military radar sets
for logarithmic conversion. I am not sure if this was for signal
compression or because taking the FFT of a log converted signal
gives a time axis rather than a frequency axis? Somewhere the
time (t) gets flipped from the bottom of the equation to the top
so you get period instead of frequency using a log converter.

Something about multiplication being the sum of two logarithms,
so filtering a log converted signal and then taking the exponent
gives you what?

That is H(t) * F(t) = (Exp (log H(t) + log F(t)) which is the basis
for multipliers like the RC4200. But something funky happens
mathematically when you filter a log converted time varying
signal before exponent conversion since you dealing with an
addition rather than a multiplication of two functions.

Could this maybe be useful for music? Anybody?

> From: Sean Costello <costello at>
> To: Harry Bissell <harrybissell at>
> Cc: dougt at; synth-diy at
> Subject: Re: Identifying tempcos
> Date: Tuesday, November 09, 1999 9:04 PM
> Harry Bissell wrote:
> > Nothing but a synth or a Log converter is likely to have a real
> > tempco... unless you guys have access to old guided missle parts in
> > which case we'll have to kill you (to get your tempcos of course....)
> Did missiles make use of the uA726? I know that this rare beastie was
> used in some military specs.
> Sean Costello

More information about the Synth-diy mailing list