[sdiy] More about the Aries VCO

[jhaible at debitel.net]

Zitat von Don Tillman <don at till.com>:

>    > From: "JH." <jhaible at debitel.net>
>    > Date: Tue, 15 Jun 2004 00:04:18 +0200
>    > 
>    > This CA3080 based schmitt trigger is very clever! 
> 
> Hey JH,
> 
> Oh, it's not *that* clever.  :-)
> 
> In the CA3080 data sheet there is an example of a triangle core VCO
> that uses a pair of CA3080's, one as a schmitt trigger.
> 
> Here's the CA3080 data sheet:
>   http://www.intersil.com/data/fn/fn475.pdf


Hi Don,

my emphasis was on "this", not on "CA3080 based".

I've seen a lot of 3080 schmitt triggers, used them myself as well,
and mostly the bounding of the trigger output and the amount of
feedback is defined by Iabc and a resistor at the 3080's 
(current) output. This gives very well defined thresholds in theory, but in 
practice the high impedance this implies is not so pleasant.
I've also seen (and used) diodes instead, which have a lower
dynamic resista´nce, but which also have the usual temperature
dependence. One can play to get some compensation of the various
temp effects from Iabc, the diodes, and (as you need some overdrive to
get the comparator going) even the differential pair.
Not so easy as it looks at first glance! (I remember an EN design
with a comment of BH that it works nicely despite the theoretical
temp dependence of diodes, but he doesn't know exactly why. 
- I'm paraphrasing.)

The Aries trigger appears to be quite different, as the bouding
function is not accomplished by a combination of bias current
and load resistor, but - as in a "real" comparator - with rather
high open loop gain and a discrete bounding network, where the Vbe of
transistors are temp compensated by diodes. Few components, fast
and discrete, and can drive remarkable load currents. At least it
looks like this to me. Haven't run it thru Spice yet, but sure will.

JH.


-------------------------------------------------
debitel.net Webmail