Small Time: PT2399 Delay with tails

[Gandalf_Sr]

Running the values of the Multiple-Feedback (MFB) filter around pins 15 and 16 through the useful tool at:

http://sim.okawa-denshi.jp/en/OPtazyuLowkeisan.htm

..shows that the filter has a gain of x-1.8. So there's a bit of boost on the way in. I haven't tried the output filter on pins 13 and 14, but if that's similar, that'd probably explain it.

It looks to me like those filter values have been tweaked to ensure that you can get oscillation even with the fairly low cutoff (2.7KHz according to the tool).

HTH,
Tom

Thanks, I tried playing around in that calculator and could get the values you saw but that's as far as I got.  On the actual circuit, I tried adding a 40k resistor at the top feed into P1 which has made the tails sort of work, the Repeats pot doesn't drive into doom loop tails until I'm about 2/3 round the pot but the tails are not particularly clear - I was hoping to play a note and have it repeat and gradually fade into nothing (with the right settings of course).

Can you make any specific suggestions as to changing component values?

[ElectricDruid]

Depending on the delay time setting, the tails aren't *going to be* very clear. This is the PT2399 we're talking about! It's not noted for its fantastic audio quality at longer delays.

The filter values have probably been chosen to limit the crunchiness when you get to longer delays. This also has the often-desirable side-effect of making it sound more like an analog BBD delay, since they frequently used 3KHz filtering too.

Further down the page on the Okawa filter tools is a form where you can do the filter design "the other way around"  e.g. instead of putting values in and seeing what cutoff and Q you get, you can put the cutoff and Q in and get it to give you the values required.

Sticking "5KHz" and "0.707" into that gives a first-pass attempt, and tweaking the C values improves things to give us the attached filter. This should cut back the runaway feedback a bit more, and will also open up the sound slightly.



Of course, "more open" also means "more noise", so try it and see what you think.

HTH,
Tom

[anotherjim]

When I did my Soft Delay...
https://www.diystompboxes.com/smfforum/index.php?topic=114973.0
...I got a side-effect that with regen just short of "Doom loop" (great name for it) bypassing raised the regen level and the tail would start feeding back.
I do find that having tails isn't all that useful, to be honest. If you just want to bypass to play normal over the tails, they appear to die out too soon under the now dry playing because if not strong enough to cause Doom Loop, they can't help decay in level each time around. You'd need some kind of AGC (*) in the regen to maintain the level of the repeats perhaps?
If you want to stop playing and let the tails fade off - there's no need to bypass.
But, of course, YMMV.
(*) I have an idea to use the AGC in a voice recorder chip (can be used without the recording part) to AGC PT2399 repeats. Not tried it yet.

[Gandalf_Sr]

Depending on the delay time setting, the tails aren't *going to be* very clear. This is the PT2399 we're talking about! It's not noted for its fantastic audio quality at longer delays.

The filter values have probably been chosen to limit the crunchiness when you get to longer delays. This also has the often-desirable side-effect of making it sound more like an analog BBD delay, since they frequently used 3KHz filtering too.

Further down the page on the Okawa filter tools is a form where you can do the filter design "the other way around"  e.g. instead of putting values in and seeing what cutoff and Q you get, you can put the cutoff and Q in and get it to give you the values required.

Sticking "5KHz" and "0.707" into that gives a first-pass attempt, and tweaking the C values improves things to give us the attached filter. This should cut back the runaway feedback a bit more, and will also open up the sound slightly.



Of course, "more open" also means "more noise", so try it and see what you think.

HTH,
Tom

Tom,

Thanks for your input

I changed to the values you suggested and the doom loop effect is totally gone.  I can get obvious tails now by cranking P1 fully clockwise.  I can make a recording for anyone interested but how do attach that here? Can I upload a .mp3 file and, if not, what's the best practice to share audio or video here?

Updated schematic with correct component values attached.

Ted