PT2399 Noise Generator - Help with schematic to vero.

[spectreman]

Hey everyone,

I recently dug up this schematic for a noise generator using the PT2399 chip.  You can see it in action here:

https://www.youtube.com/watch?v=k810qG5nKXs&t=7s

To me, it sounds great, and I was hoping someone with more advanced skills than myself could make a compact vero of the schematic.

Check it out and let me know what you think!

[anotherjim]

That thing's crazy. Me like.
Vero pretty much designs itself doesn't it? Have a strip above and below for ground, link them together. Cut tracks between all chip pins and add parts & links pretty much as drawn on the schematic. That's how I'd do it -  I never plan layout on paper first. Go on - give it a go!
You do need to think about how to get the 5v power. USB power-only cable comes to mind.

[moid]

That sounded cool - does the PT2399 need to be connected to IC3 to generate a sound input? Or can IC3 be omitted (if it is required what is it, because a vero layout is no use without it). A 7805 could be placed in the layout so you could run standard 9V to the 7805 and then 5V to the PT2399. I could probably draw something in diylc on vero if you need some help, but I would need to know what IC3 is first (if it is required)

[mdcmdcmdc]

This thing sounds cool as heck, I did a little snooping and found the following:

Its part of a larger circuit that I built at a workshop with an artist called Ewa Justka. I'm sure she won't mind if I share it... Ignore the big line that goes to the left. That connected to an oscillator but the box in the video has no input, it just works from self oscillating feedback

[spectreman]

Oh nice!  I was wondering where IC3 went, but it looks like it can be omitted from this layout.  I'm not sure how to set up the regulator for 5V, so I figured a nice little vero layout would clear & clean up everything. 

Thanks for the continued help on this, it's appreciated!

[duck_arse]

sounds good to me. however, as pin 14 on the PT2399 is an opamp output, it will be sitting at 2V5. the circuit needs a DC blocking cap between pin 14 and the outside world, and a pulldown resistor on the end of that wouldn't hurt.

[blackieNYC]

Don't people report that the chip can "lock up" if pin6 is brought to ground?  That pin usually has a 1k in series with the pot, right?

[anotherjim]

Yes, the speaker needs DC blocking and a minimum resistor in pin6 to the pot.
Not sure if everyone would want a feeble onboard speaker, a volume pot to output jack could be better.

Is it required to work off 9v DC pedal power then?

[Jdansti]

^I made a noise maker for one of my nephews and I put a switched speaker out jack that would disconnect the little internal speaker when a plug was inserted.

[spectreman]

^I made a noise maker for one of my nephews and I put a switched speaker out jack that would disconnect the little internal speaker when a plug was inserted.

Very nice!  What noisemaker did you build?  I love this kind of stuff...

[Jdansti]

^ It was a conglomeration of effects stuffed into one box with a bunch of knobs you tweaked to get all kinds of sounds. The front end was an oscillator using a 555 chip. I ran that through a tremolo, then flanger, then delay and finally through a little 386 amp. Just a bunch of small circuits chained together that could be active or bypassed. I had a switch that let the user choose continuous noise or manual control. This caused the unit to only make noise when a momentary button was pressed. I also had an input jack which disconnected the oscillator and let you play an instrument through it. The delay was set up for infinite repeats so you could get some really cool crashing sounds.

I almost didn't give it to my nephew!  I plan to make another one for myself.

[moid]

Hello spectreman

I had a go at a vero layout for you



Don't build any of it until someone with more knowledge than me has checked it! I'm no expert at this sort of thing. I added duck's suggestion about the DC cap and pulldown resistor, but I'm not sure if I've put them in the right place, and I also added a minimum 1K resistor for the pot as suggested by anotherjim. I made the output to an audio jack. The power is a standard 9v like you use on any effects pedal, but it goes through a 7805 to turn it into 5V for the IC.

Edit: I think I can get rid of the long jumper on the right of the board and make the board smaller... I need some sleep first, it has been a long day...

[mdcmdcmdc]

I could be wildly mistaken, but I think the LM7805 needs a few caps attached to make it function as a regulator per the datasheet?

[anotherjim]

I knew Moid would have a go at it! Ah, the universe turns in its expected order.
The output cap shouldn't go to 0v though, and you don't need that r/h long link.
If you cut the track below pin 9 (K8) then move the 1M pull-down to the bottom track bridging that track cut (K7-K9), the output wire also moves down to the bottom track (K14). The polarity of the output cap is negative to the bottom trace.

If you make the long link under the chip in 3 pieces, it can call in at pins 3 & 4 and you don't need the 2 links on the left.

This is a noise machine, so noise and audio quality are not on the agenda. All the nF value caps can be cheap ceramic discs or mlcc as they can better fit over adjacent tracks.
Any constant supply noise/hum from a less than perfect PSU might be objectionable. Another 100uF cap between the top 2 tracks would deal with that.

[duck_arse]

I don't think we should take the speaker symbol to mean connect a speaker here [it shows no impedance, either], but rather this is where you connect for output to something else, not a speaker.

moidy, you have a DC bypass cap, we want a DC blocking cap - they go in series with signal. 100uF is probably way too much, maybe 100nF would suffice from all the screeching sounds in the clip. and 1M could be 100k, or an A100k or A50k pot. all this will become clear once you breadboard it, tho.

[spectreman]

Amazing!  I'm super amped that you all are into this!  Thanks for all the input & work (thanks moid, for the initial layout!).  I'm excited for the future posts...looks like we're getting real close to a working layout.  Awesome!

[John Lyons]

In my experience with this circuit (as drawn) I get almost none of
the sounds in the video. Shorting pin 16 to 15,9,10 briefly, not all at once
but randomly glitches the repeat to get some of the sounds though.
It's hit or miss depending on the knob setting. Wildly variable and unpredictable...

[moid]

Hi everyone, round two!



I've tried to add everyone's suggestions, so please point out anything I missed or stuck in the wrong place.

I could be wildly mistaken, but I think the LM7805 needs a few caps attached to make it function as a regulator per the datasheet?

I think you could be wildly correct Thank you for reminding me that drawing layouts at midnight is not the best plan I ever had!

I knew Moid would have a go at it! Ah, the universe turns in its expected order.

Wow, I had no idea that I had that level of reliability Sadly it won't last, work is picking up for me now and will be stupidly crazy until Christmas, which means my chances of fixing that Sentinel Reverb are looking slim... but I haven't forgotten it!

The output cap shouldn't go to 0v though, and you don't need that r/h long link.
If you cut the track below pin 9 (K8) then move the 1M pull-down to the bottom track bridging that track cut (K7-K9), the output wire also moves down to the bottom track (K14). The polarity of the output cap is negative to the bottom trace.

If you make the long link under the chip in 3 pieces, it can call in at pins 3 & 4 and you don't need the 2 links on the left.

This is a noise machine, so noise and audio quality are not on the agenda. All the nF value caps can be cheap ceramic discs or mlcc as they can better fit over adjacent tracks.
Any constant supply noise/hum from a less than perfect PSU might be objectionable. Another 100uF cap between the top 2 tracks would deal with that.

Thanks anotherjim. I've incorporated all your suggestions (I think). Would you mind checking I put them in the right places please?

moidy, you have a DC bypass cap, we want a DC blocking cap - they go in series with signal. 100uF is probably way too much, maybe 100nF would suffice from all the screeching sounds in the clip. and 1M could be 100k, or an A100k or A50k pot. all this will become clear once you breadboard it, tho.

Well luckily my head is still screwed on or I'd forget that too! Thanks Duck - I've changed those values you suggested too. Haven't added  the pot yet, because I'm not sure if spectreman wants that. Spectreman, let me know if you want another pot; I'm not sure what it will do, but Duck knows his stuff so it'll probably something fun.

Amazing!  I'm super amped that you all are into this!  Thanks for all the input & work (thanks moid, for the initial layout!).  I'm excited for the future posts...looks like we're getting real close to a working layout.  Awesome!

That's no problem, this does sound like something I'd like to build at some point and I would do it on vero anyway. It's doubtful I'll get the chance to make this until Christmas though So if you build it I would like to know if it works. It does sound like you should probably breadboard this first if John's comments are anything to go by.

In my experience with this circuit (as drawn) I get almost none of
the sounds in the video. Shorting pin 16 to 15,9,10 briefly, not all at once
but randomly glitches the repeat to get some of the sounds though.
It's hit or miss depending on the knob setting. Wildly variable and unpredictable...

I wonder if the missing IC3 is actually an opamp or other IC (a PLL?) that generates a loud tone to feed into the PT2399 to get it oscillating? So it might work without it, but not very well as you have found? I wonder if just piping guitar audio into pin 16 might be enough?

[John Lyons]

In my experience with this circuit (as drawn) I get almost none of
the sounds in the video. Shorting pin 16 to 15,9,10 briefly, not all at once
but randomly glitches the repeat to get some of the sounds though.
It's hit or miss depending on the knob setting. Wildly variable and unpredictable...

I wonder if the missing IC3 is actually an opamp or other IC (a PLL?) that generates a loud tone to feed into the PT2399 to get it oscillating? So it might work without it, but not very well as you have found? I wonder if just piping guitar audio into pin 16 might be enough?

There has got to be more to it, I agree.
My PT2399 theory is week. I need to look
a some schematics and figure out some stuff...
This look like a good place to start.
https://www.electrosmash.com/pt2399-analysis
A little more info here.
https://www.facebook.com/chrismenessound/posts/2012014219074621

[anotherjim]

^Latest layout looking good I think.

This thing has been "designed" at a circuit bending workshop. Therefore there is a lack of proper development. It raises concerns & questions...
Why is the feedback resistor only 22R? Pin 16 is a hiZ input, that resistor there may not do anything - and 22R is more like a value you might choose to allow an 8R speaker to run off an output pin. Is it in the right place?
Without the resistor we've added to the delay pot, the chip can lock up & ceases working. Could that be why some report failure with it as it was?

Somebody needs to breadboard it!