Frequency Central's "Little Angel" chorus got me thinking hard about the PT2399 and the possibility of even more as-yet undiscovered functionality.
I believe that the caps to ground from pins 7 and 8 are used to synthesize switched-capacitor resistors, and form integrators when used in conjunction with OP1 and OP2.
One possibility would be use the output of OP1 (pin 9) as an audio signal. I think the signal here is basically an inverted and lowpass-filtered copy of the signal coming in pin 16. But it's cutoff frequency is controlled by the clock, which means it's voltage controllable.
Then there's the possibility of mixing this output with the delayed output. For short delays this would be most interesting, because depending on how you mixed the signals, you could get a combination of chorus and phasing.
OP2 is the same as OP1 except that its input comes after the delay line. However, by utilizing the CLK_O pin and some additional switched capacitor trickery* you could split the signal at pin 7 and send the delayed signal out of the chip while freeing pin 11 to accept an independent audio source. By this method one could build many different integrator-based filters, and the CLK_O pin could be used to cascade multiple sections and build higher order filters. Meanwhile the delayed audio is still available to be used for something else...
*I'll make a schematic to illustrate this when I get the time.
The old EHX Attack Delay used a BBD so that envelope information could be extracted properly and reliably before being applied to the actual signal.
Interesting ideas. I seem to remember someone, possibly R.G. mentioning that one of the pins has the output of the AD converter on it, so so that might be useful for something.
With a belton brick for a predelay?
Removing the modulation on Little Angel and instead using the regular electro cap to ground, with a 100R resistor from pin 6 to ground, yields metallic reverb if you create a feedback loop. As the PT can handle up to 6.5v, I'm still wondering if delivering more voltage than the accepted 5v would result in an over-clocked VCO and even shorter delay times.
Quote from: frequencycentral on August 13, 2010, 12:58:29 PM
As the PT can handle up to 6.5v, I'm still wondering if delivering more voltage than the accepted 5v would result in an over-clocked VCO and even shorter delay times.
Interesting. :)
I'll rummage around for a 6v regulator. :icon_mrgreen:
Quote from: culturejam on August 13, 2010, 01:39:25 PM
Quote from: frequencycentral on August 13, 2010, 12:58:29 PM
As the PT can handle up to 6.5v, I'm still wondering if delivering more voltage than the accepted 5v would result in an over-clocked VCO and even shorter delay times.
Interesting. :)
I'll rummage around for a 6v regulator. :icon_mrgreen:
You can use a 5V regulator and a diode or two to up the output of the regulator.
Instead of connecting the middle pin of the regulator to ground, connect it to the anode of a diode, and connect the cathode to ground. It will lift the output voltage by the forward voltage drop of the diode. So, one diode would be what... about 5.6V - 5.7V? Two diodes in series between the regulator ground pin and ground should push the output to 6.2V -V 6.4V.
:)
Mike
Neat trick! Thanks for sharing.
Quote from: .Mike on August 13, 2010, 05:27:47 PM
You can use a 5V regulator and a diode or two to up the output of the regulator.
Instead of connecting the middle pin of the regulator to ground, connect it to the anode of a diode, and connect the cathode to ground. It will lift the output voltage by the forward voltage drop of the diode. So, one diode would be what... about 5.6V - 5.7V? Two diodes in series between the regulator ground pin and ground should push the output to 6.2V -V 6.4V.
:)
Mike
Wow, thanks for the tip, Mike. I would have never thought of that. :)
You're welcome, guys. I wish I could take credit, but I picked it up somewhere along the way. ;D
Quote from: frequencycentral on August 13, 2010, 12:58:29 PM
As the PT can handle up to 6.5v, I'm still wondering if delivering more voltage than the accepted 5v would result in an over-clocked VCO and even shorter delay times.
actually, I can answer that. I tried that a couple years ago and according to my notes
with pin 6 grounded,
5V supply = 24ms
6V = 18ms
6.5V = 16.5ms
7V = 15ms
7.5V = 14ms
8.5V = dead chip
also current draw nearly doubles from 28mA @ 5V to 58mA @ 6.5V
hope that helps
Perhaps Princeton designed their PT2399 datasheets (even the more in-depth one) to be so useless intentionally, to encourage creative thinking in the DIY community...
...that is to say, subscribed. Always interested to see chips being repurposed and abused.
Damn, thanks Luja!
So maybe shooting for 6v is a good idea to sit in between delay shortness and mad current draw?
Quote from: J. Luja on August 13, 2010, 09:14:25 PM
Quote from: frequencycentral on August 13, 2010, 12:58:29 PM
As the PT can handle up to 6.5v, I'm still wondering if delivering more voltage than the accepted 5v would result in an over-clocked VCO and even shorter delay times.
actually, I can answer that. I tried that a couple years ago and according to my notes
with pin 6 grounded,
5V supply = 24ms
6V = 18ms
6.5V = 16.5ms
7V = 15ms
7.5V = 14ms
8.5V = dead chip
also current draw nearly doubles from 28mA @ 5V to 58mA @ 6.5V
hope that helps
Yes thank! That's some great information. 6v would seems to be the way to go, still comfortably within tolerance.........
I've been thinking of this way to use the PT's built-in high frequency clock to synthesize a LFO waveform. Use of the LM3900 "Norton" amp IC allows for a simpler circuit than if you tried to do the same thing with a more general purpose amp.
(http://www.aronnelson.com/gallery/main.php?g2_view=core.DownloadItem&g2_itemId=43553&g2_serialNumber=1)
The RC filter going into pin 2 is just to smooth out the MHz ripple of the LFO's output.
This is based on this circuit (http://www.seekic.com/uploadfile/ic-circuit/20096242569513.gif) from a National appnote, but the key feature here is that the LFO drives the PT2399's Vref pin, which in turn modulates the clock frequency. This feedback loop changes the triangle-like output of the LFO to become hypertriangular.
I would suggest using this with a very short-delay type of effect like the Little Angel chorus. Hypertriangular LFOs often sound good with flange-type modulation, so I imagine it would also work well for chorus.
If ya can get the delay low enough, you can put some mad low filtering on the delay side, and put a heavy tone control on the straight signal. I imagine this would work best for bass, unless you can get REALLY short times. Basically, you make a Sonic Maximizer by delaying the high frequencies so there is more apparent bass. I did this on accident when I was breadboarding Little Angel trying out different filtering to see if I could enhance the sound for bass with simple mods. All I did was put a SWTC before the wet signal return.
Controls:
Mix (Process)
Tone (Low Contour)
I'm thinking the LA running on 6V with the second amp wired as an active mixer/post amp should do the trick... as long as it will handle running that long at that clock speed, lol... would suck to lose a chip in the middle of a badazz solo grind!