[sdiy] Question about LFOs

[Pete Swarbrick]

Hi David
The reason I've stuck with the FETs is more to do with innate
conservatism than anything more techie: ie if it ain't broke don't fix
it.
 This flies in the face of much of my other s-diy noodlings where like
many of us,I will fiddle about for ever trying to build a
bigger/better/quieter/ simpler/faster/easier/louder/sweeter gizmo but
here's the LFO history: I built the first one on stripboard as a
straight copy from the MS-20 (so long ago I can't remember exactly when
!) but substituted a couple of JFETs borrowed from my then employer,
BBCtv. They were probably BF244As since I know we had lots of those!. It
worked very well and I think ended up in a Transcendant 2000 that my
colleague was building. Some time later I laid out an expanded triple
PCB using the then state of the art method we had available at the Beeb:
double-size-drafting-paper-and-tape-photographic-reduction-UV-photo-neve
r-works-first-time-round-unpleasantness. Eventually I redid the board
using Easy-PC pro (that was REAL state-of-the-art back in the day...)
and then used the same layout to produce the two triples I currently
have. 
Were I to make some more (or rather WHEN I make some more, as I surely
will) I will do the same thing, they are quick, simple, repeatable,
cheap and above all USEFUL. Actually at these quantities I find pricing
of the two components largely irrelevant; looking at Farnell here in the
UK although BF244As are a massive 34 times more expensive than 1N4148s
they still only cost 30p each!
I agree with Magnus that diode mismatch would affect frequency over the
range of the waveshape but one of my JFET versions does that too...
maybe I should fix that some day but it's never bothered me - it is
"only" an LFO after all.

Enough memory lane mutterings

Pete



 -----Original Message-----
From: David G. Dixon [mailto:dixon at interchange.ubc.ca] 
Sent: 12 March 2009 17:52
To: Pete Swarbrick; synth-diy at dropmix.xs4all.nl
Subject: RE: [sdiy] Question about LFOs

> I have a total of six MS-20 style LFOs in two triple modules. After
some
> time using them I beefed-up the variable ramp output from +/- 2.5V to
> +/- 5V on all LFOs in one module just to see how often it would be
> useful and ended up doing the other one too. Love the circuit, it's
> simple, repeatable and flexible plus I use the variable ramp pretty
much
> ALL the time and would be lost without it.

Thanks for the excellent response (you too, Magnus)!  Now, I'm
wondering: Is
there some reason why you've chosen JFETs over diodes for the
wave-shaping
integrator?  Both circuits behave equally well in simulation.  Is there
some
problem lurking in the latter which my little noobie brain hasn't
thought
of?


checked by CanIt4