[sdiy] Reverse Sawtooth

[Neil Johnson]

Hi,

Matthew Smith wrote:

> Quoth Ingo Debus at 25/01/11 03:31...
>>> * strip the software to bare bones and time the system  
>>> externally,  using INT0.
>> Wouldn't this cause a lot of jitter?
>
> I think that it would depend on the clock speed of the device being  
> interrupted. There is obviously a certain window of time in which  
> the interrupt has to be recognised - I would have thought that this  
> would decrease with greater clock speeds. I'll be running the  
> TINY2313s at 12MHz in the first instance - that being the fastest  
> crystals I have on hand, without trying overclocking.
>
> But, as Harry points out (albeit jokingly,) how is this worse than  
> a jittery analogue system for musical purposes. (As opposed to  
> instrumentation grade signal generators or whatever.) After all,  
> exactly how many analogue systems are actually in *tune* ;-)

But you're comparing two different-sounding things.  In an analogue  
oscillator, you will see jitter in the timing of the reset circuit.   
Happens once per cycle, and *may* give rise to a "more analogue  
sound" (...hype meter at 5 and rising...).

But in the digital domain you're talking about jitter *per sample* -  
a much different beast altogether.  And really NOT desirable at all.

Neil
--
http://www.njohnson.co.uk