[sdiy] PT2399 site with sound files

Scott Stites scottnoanh at peoplepc.com
Sat Aug 24 20:55:16 CEST 2002

Batz Goodfortune wrote:

> Y-ellow Scott. 'n' all.
> I'm glad you did. I'd seen the Holtek chip before but not the Princeton(s)
> They look very encouraging. They have a number of interesting chips I'd
> like to look into further. It was also an interesting hypothesis put
> forward in there somewhere that the sound of an analogue/BBD chorus unit
> has much to do with the harsh filtering. It was stand to reason that it's
> mainly the low portion of the spectrum doing all the combing. I guess I
> could mess with this with one of my other DDLs and do an A/B when I get
> some time. But I'm kinda interested in that 800ms delay possibility with
> the Princeton chips. Depending on how awful it sounds at that length I

Hi Batz,
I've never tried the Holtek chip.  When I was just getting back into this a
few months ago, I found a reference to it in the synth DIY archives about
too much clock bleed-through.  The PT2399 was rarely mentioned, and readily
available to me, so I purchased it.

The PT2399 does seem to strike a good balance between BBD's and digital
delays.  In my own humble setup here, it has been surprisingly clean and
noise free.  Longer delays do have a certain amount of clock noise at higher
freq's, but experimentation with a LP VCF seems to clean that up and allow
for darker long delays.  I'd say the one drawback of the PT2399 is the
minimum delay time - a little too long for true flanging.  That being said,
I've goosed the minimum delay time to around 17 mS (if my scope ain't
lying - this is contrary to the data sheet), which puts it into the range
for chorus type effects.  Also, I do get some comb effects that could only
be described as "granular".  I've gotten really cool psychoacoustical
effects on longer delay times by just barely goosing the delay time with a
sample and hold fed with white noise.  So subtle, you don't really notice it
unless you remove the S&H and listen to the difference.  Of course, it
doesn't have to be so subtle ;-)  I'd love to try it with a Wiard Woggle Bug
or a Buchla Source of Uncertainty.  The PT2399 also can impose a very sweet
vibrato on an unmodulated signal.

Anyhoo, as far as the other Princeton chips go, the PT2395 uses external RAM
and also has a variable delay based on a resistance value. It's been pointed
out to me that just using a few of the address lines might put it into the
range of flanging(?).  I've got a few PT2395's, but have yet to try them
out.  You all just keep coming up with such cool stuff, I don't know what to
do next!

> Thanks again for posting that.

It was my pleasure.  I've gotten so many good schemos and ideas from
everyone on this list (including the Electric Kitchen), that it almost feels
like stealing.  It'd be nice to contribute at least one thing that may be of


More information about the Synth-diy mailing list