[sdiy] SAW core VCO flyback time

[Tom Wiltshire]

On 30 Aug 2016, at 22:41, Donald Tillman <don at till.com> wrote:

> 
>> On Aug 30, 2016, at 10:40 AM, mskala at ansuz.sooke.bc.ca wrote:
>> 
>> On Tue, 30 Aug 2016, Richie Burnett wrote:
>>> BTW, I think it's fascinating how the analogue designers strive to achieve
>>> perfection in the form of "instantaneous" sawtooth resets, low-distortion
>>> sinewaves, and symmetrical triangle waveforms without the little nipple at the
>> 
>> I was holding off on saying it, but if you're serious about analog design,
>> do you even *want* a super-fast reset?
> 
> The high C on a traditional piano is 4186 Hz.
> 
> A 2uSec reset delay would pull that down to 4152 Hz, for a 35 Hz beat. 
> 
> A 7uSec reset delay would pull it down to 4066 Hz, halfway down to the next lowest note.  That would be unplayable.
> 
> The effect is worse when you're playing along with stringed instruments with stretched turning.
> 
> Besides the capacitor discharge, the delay would include the comparator sensing the limits.
> 
> So the concern about reset time is not for the timbre.  I mean, hell, sawtooth waves are so boring already that any timbral variation would be welcome.  Amiright?
> 
> The ARP/Rhodes Chroma VCO had a unique charge pump VCO core that had a long reset time, but the reset didn't introduce a tuning delay in the cycle.

Ok, so the point of worrying about the reset time is because of the potential effect on the tuning, right? But that's not insurmountable, and reducing the actual reset time is only one of a number of ways to go at the problem. In fact, given that you're not going to be able to reduce that reset time to zero, there will be a need to compensate it somehow, in which case, it doesn't seem to matter much how long it is (to me, at least). Just compensate more, or less - whatever.
It seems like the effort would be better spent on the compensation methods rather than the reset time, since shortening the reset time isn't going to eliminate the need for compensation - just make it less serious.

On a similar topic, I've recently been looking at CEM3340 implementations for various synths, and noticed that the Sequential Prophet 5 and the T8 don't even use the CEM's high frequency tracking trim (the current out from pin 7). This trim is for the exponential convertor base resistance rather than the reset time, since it's a triangle core oscillator, but it's a similar thing. Presumably their autotune routines compensate for this instead.

Tom