ARP VCO

[Joachim Verghese]

On Mon, 21 Apr 1997, Haible Juergen wrote:

> Mostly the scaling is done by braking a resistor into 90% fixed value and
> 10% rheostat trimpoti, so the tempco of the whole thing would still be 
> dominated by fixed resistor.

Absolutely right. The above concept surely works well in most cases.
It's just that temperature drift isn't always specified for trimpots,
so it's difficult to know for sure.

BTW, in the Minimoog schematics the VCO trimmers are labelled "WW".
I've been wondering, does this mean that they are wire-wound types?
Aren't such trimmers rather drifty?

> Now, as I had the Odyssey docs in front of me, something else caught
> my attraction: The feedback path from the reset pulse over a low pass filter
> back to the expo converter. Now what is this ?!
> Looks like some strange HF-tracking circuit.

That's correct. At low oscillation frequencies the voltage across the
0.01uF cap is around -5VDC. This voltage is used to sink a current from
the reference node of the exp. converter.

At high frequencies, the sawtooth reset pulse is integrated, causing
the cap voltage to decrease, and less current to be sunk from the exp.
converter. Thus there will be an increase in control current and
oscillation frequency, i.e. we've got HF tracking compensation.

> Just  wondering why they have done it *that* way ... The lp corner is 
> at several hundred hertz, so it would even change the waveform at low
> frequencies, wouldn't it?

The sawtooth reset pulse is only 2 us long, so there's very little
ripple across the integrating cap. Seems to work ok in this particular
VCO design (PNP/NPN exp conv, active pull-up reset pulse).

> Does this circuit have any advantages compared with other HFT schemes?

The only advantage I can think of is that the saw amplitude stays 
constant throughout the useful oscillation range.

-joachim