PT2399 Delay Time?

[paku5535]

Hello,

What is the maximum delay time I can achieve on a single pt2399 chip design. I have tried the "SEA URCHIN" from madbean. I am getting maximum ~350 ms delay. I have also changed the delay pot to 100k.

Setting:
Delay--maximum
feedback--minimum
mix-maximum

Is the delay time only depend on The Pot value and the series resistance?? or anything else??

Any suggestion is highly appreciated.

Thanks
Paku

[R.G.]

The longest practical time is 350-400mS. Longer than that and the inherent noise of the digitizing process gets bad.

The critical item in determining the delay time is the current pulled out of the pin where the delay resistor sits. A resistor is OK for most purposes, but a current sink gives better control.

[slacker]

A 100k pot plus what ever other resistance there is between pin 6 and ground should give way more than 300ms. In my experience a 50k pot gives about 700ms.

[deadastronaut]

^ yep,   i used a 100k pot and put a 150k resistor on its outer lugs and squeezed a bit more time out of it before break up....

max time around 2:20 on vid..

http://www.youtube.com/user/100roberthenry?feature=mhum

[Mark Hammer]

Whether it is digital or analog, you can always squeeze more delay time if you are willing to restrict bandwidth enough to make the artifacts unobjectionably audible.  hell, you could probably squeeze a second out of an MN3205 if you were willing to live with 500hz (or less) bandwidth.

So the question is really more "How much time can I squeeze out of it, yet still have it do something musically useful in the direction I'm aiming for?".

In some respects, the PT2399 was intended to bump the Holtek HT8955 out of the market.  The 8955 provides up to 800msec of delay time with usable bandwidth.  I think expecting the 2399 to offer 600-700msec of usable delay, with approximately the same audio quality as the 8955, is not unreasonable.

[rring]

I was able to get 1 sec with a 8 pole switch cap filter -but one factor to consider is the mix level. My baseline is  wet = dry   level as the max mix setting. Also your high end rolloff starts to creep near the very edge of the highest guitar note fundamental.  As decsribed in the last post, 700 - 800mS is the practical limit but is very doable.

[paku5535]

thanks all of you for your feedback. I am using 100k pot in series with 10k resistance from pin 6 to ground (Pin 4). Then Why I am getting small delay.

Another Question I want to ask whether to connect pin 4 of Pt2399 to ground ?( Somewhere I have seen Pin 4 is internally connected to Pin 3 via 10E res).

[Maik]

Hi Paku,

900ms! Check out
http://circuitworkshop.com/forum/index.php?topic=1112.0

posting 6

[rring]

I experimented with various (pin four) terminations. My best results were had when the regulator ground, analog ground and digital ground (pin 4) are all bonded together at one point. Any kind of long trace or ground with these connections  going along one after the other in a line is not optimal. I would cram the regulator right next to the PT2399.

Have you measured your pot? is it really 100K? Something either is defective about your pot, the IC , etc. Its got to be a simple problem - because properly functioning PT2399's (for me)have been very consistent with regard to the pin 6 resistance to ground  and the clk rate. What you are describing is not typical behavior at all. I get approximately 700mS with 50k.

[paku5535]

Yea its 100k Lin. I am suspecting whether my Pt2399 at all original!!!

[R.G.]

I was able to get 1 sec with a 8 pole switch cap filter -but one factor to consider is the mix level. My baseline is  wet = dry   level as the max mix setting. Also your high end rolloff starts to creep near the very edge of the highest guitar note fundamental.  As decsribed in the last post, 700 - 800mS is the practical limit but is very doable.

If one is willing to play games with the high frequency rolloff, as Mark said, you can get longer delays. But it gets to being a really noticeable tradeoff. As the OP said, he wants a single chip setup. The practical limit for that version, with the built-in opamps used for filtering, tends to be about 400mS, depending on the individual's tolerance for the noise and sampling artifacts.

Sampled delays all have the same issues and tradeoffs of sampling frequency and alias/noise reduction filtering. Getting the high end rolloff down near the guitar note fundamental, no harmonics in the delay, is one compromise, but it tends to be a big one. Still, some people might be OK with it.

It's the sampling frequency that sets this kind of limit. In the 2399, the sampling frequency and sampling delay are tied together in the chip in a way that can't be uncoupled.

For stunt guitar playing, delays of 1S or more are neat, and that's kind of become the de facto standard needed to advertise and sell a delay pedal. But it is rare for a delay of more than half a second to be generally useful on stage according to a lot of guitarists I've talked to.

It's down to the Zen of Motor... er, Delay Pedal Maintenance - delay time versus high frequency filtering is a personal preference.

[deadastronaut]

it is rare for a delay of more than half a second to be generally useful on stage according to a lot of guitarists I've talked to.

true,  i'm happy with 400-600ms.

[~arph]

  Yeah any longer and they tend to become distracting

[Mark Hammer]

Delay has a number of qualitatively different uses.  If the intent is to use it to mimic physical space, then yes, a 4096-stage BBD or a digital device capable of 400ms or so is sufficient to mimic most familiar space.  Remember that sound travels at 1126ft/sec, so a sound that repeats after 400msec has travelled:

(1126 * 0.4) / 2 ft.  or 225ft there and 225ft back.

...which is pretty much as big a physical space as we are accustomed to hearing or being in.  We have an auditory memory of being in an indoor parkade, or stairwell, or gymnasium, and the memory of that allows us to associate the electronically delayed signal with a physical space.  We don't really have much experience with reflective  physical spaces that are 500ft or 800ft across.

Delays noticeably longer than a half second tend to be implemented for other purposes; not to mimic physical space, but rather to mimic additional musicians.  Some years ago, I tried to get at this here by asking members what the natural breakpoints were for them in delay time and what they attempted to do with delay.  I suspect the "natural breakpoints" are a big part of what has shaped the commercially available devices.  The Princeton and Holtek chips aim for no more than 800msec, and neither Panasonic nor Reticon ever produced an analog device with 8192, or more, stages, or provided convenient solutions for cascading more than 8192 stages.

[R.G.]

Delay has a number of qualitatively different uses.  If the intent is to use it to mimic physical space, then yes, a 4096-stage BBD or a digital device capable of 400ms or so is sufficient to mimic most familiar space.  Remember that sound travels at 1126ft/sec, so a sound that repeats after 400msec has travelled:

(1126 * 0.4) / 2 ft.  or 225ft there and 225ft back.

...which is pretty much as big a physical space as we are accustomed to hearing or being in.  We have an auditory memory of being in an indoor parkade, or stairwell, or gymnasium, and the memory of that allows us to associate the electronically delayed signal with a physical space.  We don't really have much experience with reflective  physical spaces that are 500ft or 800ft across.

Delays noticeably longer than a half second tend to be implemented for other purposes; not to mimic physical space, but rather to mimic additional musicians.  Some years ago, I tried to get at this here by asking members what the natural breakpoints were for them in delay time and what they attempted to do with delay.  I suspect the "natural breakpoints" are a big part of what has shaped the commercially available devices.  The Princeton and Holtek chips aim for no more than 800msec, and neither Panasonic nor Reticon ever produced an analog device with 8192, or more, stages, or provided convenient solutions for cascading more than 8192 stages.

Mark,  you're in serious danger of being seduced by the Dark Side and becoming an engineer.

[Mark Hammer]

...which would make my late father VERY happy. 

He was a tool and diemaker (bequeathed me his calipers).  When I was 12, he took me to an engineering department open house at one of the local universities.  I had just seen Goldfinger, and there in front of me was an actual laser!    The grad student doing the demo defocussed it and encouraged me to pass my hand through the beam.  Then he focussed it, asked me to stand back, and zapped a hole through a couple of razor blades.  In Liz Lemon fashion, I thought "I  want  to  go  to  there", and ended up spending a quatrter century connected to univerisities.  As it turned out, *I* never went into engineering, but our older son did, in that very same faculty.

But hey, I thought I was already ON "the Dark Side" by being in government?    There's an even darker side?

[R O Tiree]

There's an even darker side?

Buying dept?  Or QA?