[sdiy] BBD help

[harrybissell]

OK now lets get serious...


> I want to simulate a SHORT delay, at LOW FIDELITY. How short? Why low? Well,
> I started with the premise that guitar feedback is related to transmission
> delay (cord length + amp circuitry) and a low-fidelity interaction between
> sound waves and guitar strings.

The delay in the amp and cord may be disregarded for all intents and purposes...
if you
want to simulat the delay in a amp, especially a tube amp with little negative
feedback...
probably a simple RC filter would give you 'that' much delay.

More important is the speed of sound in air... roughly 1.1' per ms.

The feedback from the amp to the body / strings is much more a function of the
coupling
as you noted.


> Now, I'm not just thinking of using this as
> a guitar feedback simulator circuit to be built into a tube amp simulating
> distortion unit; I think it might sound good for all sorts of sounds I
> create in my modular as well. Besides which, I already own a real tube amp.
> I am thinking it might make a nifty headphone amp, or just a neat guitar
> stompbox, but I'm not aiming for any one thing here, just playing with a
> concept. As a result, I might want more delay than is in a real
> guitar->amp->guitar feedback loop, but nowhere near enough that the delay is
> really a chorus effect.

What might suit your fancy even nore would be a number of bandpass filters,
closely
spaced. They have done these as resonators to simulate strings, voices, etc.
When you
put a guitar through them, feedback becomes almost inevitable.

I've gotten a really good guitar feedback simulation with a MutronIII (envelope
follower)
set to High range, sweep down... in front of a BigMuff II fuzz.   It seems that
as soon as
the follower sweeps past the actual string, the resonance locks it in place
pretty well.  Gets
that Jimi Hendrix sound for people who (like me) really can't play quite that
well... ok can't
play within three orders of magnitude that well...

>
> So...let's say I need 5ns-5ms of grungy delay...what is the simplest,
> easisiest, and/or cheapest solution? Or 10ns-1ms? etc. Are there certain
> values we can plug in there that yield optimally easy/cheap solutions?
>
> Aside from all that, another use of extremely short delay might be to design
> a circuit that emulates transformer/rectifier "sag". In that case, you might
> use a HIGH fidelity delay, and run it at full volume...but process the
> original signal (split off before delay) and mix it in after potentially
> nasty heavy filtering. Sort of a reverse BBD :)
>
> I have this feeling that at short enough durations (measured in ns), there
> is a solution that doesn't involve special IC's....something that is
> probably buried in simple electronics theory, but outside of the beginners
> texts that I've read so far. Earlier threads about chains of
> inverting/non-inverting integrators come to mind, but I really couldn't
> follow it enough to know what to do with all those taps...tie 'em together?
> Nah, it can't be that simple.

ns delays could be just sending the signal through an opamp buffer. ns is WAY
too short to get an effect.

H^) harry

>