PT2399 Noise Generator - Help with schematic to vero.

Started by spectreman, September 09, 2021, 10:36:35 AM

Previous topic - Next topic


Wow Ducky! You've gone a long way with that one.
I see you've avoided the shortest delay range - it is just a buzz no matter what's in the loop.
I've got mine built & boxed. Hasn't locked up once. Delay knob and gap switch are it's most useful tools. Doesn't really need a volume knob if you plug into something that already has one.
I've scoped the audio. Tones are square waves, noise portions are trains of pulses either positive or negative.

I like the idea of a piezo impulse input. It should do the opposite of my Gap switch. "Musical" input probably won't live long enough although it should be useful if it's only modulating the delay time.


updated - still not finished.

I can't prove it, but the size of the load cap might influence the noise, so you might be able to use a sprung both ways centre off rewrite toggle (hasn't everyone got one?) and a same-size and larger-size to get different bumps, but I don't know if they would discharge in a reliably re-usable manner.

I've added the smooth toggle - this is used (waggle it, then on) on loading, as the smaller cap produces a more off/on signal at that pin - which produces a different tone, odd phasery outputs, and maybe different drive for the opto led.

the step switch values are tricky, but this is used for both loading and running, to give steps. the sweep switch now switches an opto transistor shunting some/more of the delay string. it works, not terribly convincingly, the resistor values are tricky.

AJ - any pointers on adding the piezo? or even maybe an electret mic? I should be able to use the OP2 for gain and then take output from OP1, no? I might be able to wrangle the piezo as modulation somewhere along the string, too, maybe. my intended build will have a wooden front panel, I was planning on a toothbrush drumstick.
the circuit output is not the input. the reverse holds true for the circuit input.


Hmmm now - you've used the spare opamp (13-14) but pin12 can drive output without it - or do you like the extra comparator fuzz effect the opamp brings? Whatever, a piezo pulse can probably hit one of the inputs directly only if it happens when the thing is already in full clip it may not be noticed. Piezo is capacitive. It may need a high >1M bias resistor from pin2 but it wouldn't need a blocking cap unless you find it helps with the pulse shape. Then again you might just hook the piezo to pin16 and get something useful. It will have signal on it though from the 1M - I don't if it will be audible from the disc like that. Or you might split the 1M in two and feed the piezo in the middle or if the 1M was a pot to the wiper?

Electret setup is easy. 5v supply and pick a bias resistor to the mic+ that gives *2v; just like setting a JFET drain which is what it actually is. However, the thing full clips so the mic needs a preamp and if you allow it to be monitored "live" it will probably howl by itself with nothing to do with the delay. It might be best if fed to the delay input and only be heard after delay.
*2v as that's what I saw in a Panasonic electret datasheet long ago.

John Lyons

Interesting tweaks Duck!
Here's what I get out of the original schmatic as outlined
in my post above (touching different pins somewhat at random
to pin 16). I added a strong low pass tone knob to get more thump
or sizzle, depending...

Sound Clip
Basic Audio Pedals


John, your audio is exactly the rubbish I've been listening to all week. except more musical. I want that tone control. I've found the step switch is very useful for getting from here to there, and back again.

I think I've sorted the sweeping bit of the delay string, in my head at least. switching resistors problem, it seems.

as for piezo - I did standard proper opamp, and the bloody thing wouldn't go sick properly. one thing - whacking the piezo and opening the smooth switch in playback created the exact sound of a pushbike bell, with variable tone. the problem with the piezo tho is the tiny short burst goes in, then gets time shifted into a bip, and nothing results. piezo whacks into pin 2, however, does like a load/bump, depends on how hard.
the circuit output is not the input. the reverse holds true for the circuit input.


Awww.... mine sounds nothing like Johns. I suppose I will have to post a clip.

A slow LFO to pin 2 might create PWM/Phasey goodness since it's the reference for opamps working like comparators - depends on what the audio is doing - square waves won't change much.

John Lyons

Post a clip of what you guys have going.
Jim. Yours seemed to get more airy swooshes
and softer noises in the one I breadboarded
from your first schematic.
Basic Audio Pedals

Try this.
It starts fading in on a loop I had left running for 20mins or so.
I've got an spdt gap switch that is spring-biased one way, latching the other. Outer lugs are joined so it can be flicked for short gaps or held for erase. Funny thing - erase still makes a click. I suspect I'm getting a pulse in the chip when the memory address count bits all flip back to 0 - but that's a guess. Also, have a power switch. It's cool the chip has a soft start so there are no big thumps when you switch power. Oh, and this has no LED on pin7.
Apologies for the length of the clip, but at long delays, it can take a while for things to develop. I've actually swung it about faster and more often than I would for myself.




I've versioned this 1.0, so it must be buildable.
[removed.replaced - another one down, see later post]

some values changed. I've sorted the sweep string and the action makes sense now, is possibly a little extreme. also added a somewhat brute force topcut, at least till John shares his. [tone controls is not my best thing.] I tried AJ's version as well, I thought much the same as JL, softer than mine.

it's a fun thing to do a bump/load and leave the delay set at one or other end of travel, when the led lights you wind slowly to the other end of the pot. zounds!
the circuit output is not the input. the reverse holds true for the circuit input.

John Lyons

I need to get my head around that schematic Duck.  :icon_redface:
My Hi Cut control is just a 50k pot in series with a .02 cap to ground
just before the volume pot.
If you get a chance I'l love to hear some of the sounds you're getting.
Basic Audio Pedals


here, see, don't be scared, it's V1.1, probably the final, some more value changes.

JL - seems we both used the same tone method. I added R7 and C10 in an effort to reduce the tiks - actually just muddied them a bit.

and, for a short time only, here's some short sound bites of the version above in action. first and second are just switch mashing, demonstrating the step and the sweep functions - no movement on the delay pot at all. third one is the sweep switch and the delay pot tuning bird whistles and bell ringings. -
skwrp.too.mp3 -
skwrp.tree.mp3 -
the circuit output is not the input. the reverse holds true for the circuit input.

John Lyons

Basic Audio Pedals

rutabaga bob

Impressions: some remind me of the noise you hear when you dial a fax number instead of a phone number.  Others sound like flea-sized jackhammers on metal.  How are those for recommendations?  ;-)
Life is just a series of obstacles preventing you from taking a nap...

"I can't resist a filter" - Kipper


yes indeed, you get a lot of modem noises.

for those without optocouplers wanting to sweep, you could use a photofet thing probably, I've never had one, or a vactrol, but the ldr lag will make for not much effect. the part diagram below shows how you could [probably, I haven't tried it, can't see why it wouldn't work] wire an ordinary transistor. just aboot any silicon npn small signal/switch would suit. a small pF value cap added base to ground might add a little laggy swoop to the action.

the circuit output is not the input. the reverse holds true for the circuit input.


has anyone else built this thing into a boxed and finished pedal?

my soundclips above were done on the breadboard, and I'd been noticing how much the delay pot setting was being modulated with mains hum everytime I touched it. but I thought it would clear up when boxed. my box is wood, so I slug-taped it to within an inch of its life. a stray 9V connection fried the tiny brain of the IC on the breader, and I ended up making a second board [don't ask why second]. all buttoned up and tested, it was sparse and lifeless, not much tones.

as the screening was only connected to the circuit ground by one wire, I snipped it, and retested. much much better. not as good, but better than the lifeless paperweight it had been. so the question - has anyone else built it and noticed a change in the tonal mess, and the way the controls work?
the circuit output is not the input. the reverse holds true for the circuit input.


I have not built your circuit but I think it kind of needs some noise to live. Mine is in an unscreened ABS box (recycled 'natch) and doesn't seem any different from the BB experiment.
You might get some benefit from your OCD slug taping by grounding it via a small cap, so only RF gets shielded?

I'm wondering about that chorus hack that Rob DA announced nearby. It may need a more dynamic signal (like guitar) going on though, as it seems to be a kind of envelope mod of the delay.


unscreened box, you say. now where would I get one of those from ......

and it seemed to me, after wrapping the thing all ready for the post, that the easiest thing to mod in my build would be isolating the delay pot from the screening. too late now, tho.
the circuit output is not the input. the reverse holds true for the circuit input.