PT2399 based delay. Decrease delay output

[merlinb]

The profundis uses an annoying arrangement where the repeats pot is 'backwards' compared to the conventional arrangement. I would arrange the pot normally (like the mix pot) and tweak from there. That way it has less complex interation with the MFB. More like this:

[nonost]

Yes. I have it wired like in your Small Time. My voltage divider changes a lot the Feedback pot sweep. I'm using a voltage divider cause I don't want to end up with a MFP of Q = .96. But I can't stand the Feedback pot working so badly.

Any tricks?

[merlinb]

Yes. I have it wired like in your Small Time. My voltage divider changes a lot the Feedback pot sweep. I'm using a voltage divider cause I don't want to end up with a MFP of Q = .96. But I can't stand the Feedback pot working so badly.

Where is your divider, before or after the pot? Can't you just put a series resistor immediately before the pot (so the pot forms the lower leg of the divider)? A smaller pot resistance would also help, e.g. 10k instead of 50k.

[nonost]

I tried both.  Before and after the Feedback pot. It reduces gain properly but the pot doesn't behave well. I only get three repeats with the pot at noon.

I can feel I'm doing this wrong, I know. I'm just trying stuff.

[merlinb]

You should post a diagram of what you actually have built, or we're flying blind from here.

[nonost]

A 15k resistor before a 10k Feedback pot did the trick! Now the feedback pot works great. In the first  third of the rotation I can get between 0 and 5 repeats. In the last third there are plenty of repeats where I can dial pretty precisely. This last range is great, I can manage the "infinite" repeats accurately, from a few seconds of repeats to infinite ones. I really like this broad last range.

Thanks a lot!

I've also changed the Mix pot to 10k, the 50k one doesn't work well (bad sweep, everything in the last 20% or so) with the new 10k Feedback.

BTW, do you know why when the effect is bypassed the repeats get louder and louder? It's a problem when the feedback is close to infinite repeats, because when you turn off the pedal, the repeats start to get louder and eventually they oscillate and become annoying. I have this issue since time ago and I can't figure it out why.

Thanks Merlin!

[ElectricDruid]

BTW, do you know why when the effect is bypassed the repeats get louder and louder? It's a problem when the feedback is close to infinite repeats, because when you turn off the pedal, the repeats start to get louder and eventually they oscillate and become annoying. I have this issue since time ago and I can't figure it out why.

How have you got the bypass wired up? If it's true bypass, you shouldn't even be able to *hear* the echos when the effect is off, so I don't understand what you've done.
It is possible to arrange a delay pedal so that the remaining repeats can still be heard. This is often known as "delay trails", since you hear the delay trailing away instead of being switched abruptly off. But the schematic you posted earlier doesn't show anything like that.

[nonost]

Hi Tom. I've built it with tails. It's actually the same as the one in the Merlin Small Time showed above.

I'm using a J112 and the repeats get louder when buffered bypass.

[anotherjim]

There is a tendency for repeats to grow louder once dry input has stopped. I think it's because there's more headroom in the delay as the dry audio signal is no longer adding to the input mix. In a crude way, the dry level ducks the repeat level. Doesn't the same thing happen if you mute your playing in normal use?

[PRR]

....we're flying blind...

[nonost]

Hi Jim. Mmm, nope. When the effect is ON and I set it to infinite repeats, it doesn't matter if I play or not in that regard. Repeats are under control. It's when I turn it off when repeats somehow get louder and start to oscillate after a few seconds.

[ElectricDruid]

I don't know. I don't see why disconnecting that input would make any difference. I don't understand it, sorry.

[anotherjim]

It can happen if the disconnected feed presents a load on the feedback path when it is connected. Need to see a schematic!

[nonost]

I tried another FETs and PT2399 without luck. I can't post an schematic because there isn't one... At least an exact one. Right now its pretty close to the Small Time. But I've used a bigger cap on pins 9&10, so I had to decrease output level at the repeats and mixer. I'm also using a FET input buffer. I think it could be the problem. It's a caveman FET buffer. I will post a pic Jim.

[nonost]

The input buffer:

[anotherjim]

Right, I think I see it now. So with the FET on, the repeats can see the low impedance source from the input opamp. Actually, I don't know enough to say if the inputs mixing to the PT2399 input filter network like that are proper virtual ground mixing (as the wet/dry mix to the output opamp is). I do suspect though that the input opamps output is loading and attenuating the Repeats. I don't know what you can do about that apart from arranging another JFET to connect the input to AC ground when the bypass is engaged so it is always loading the repeats, bypassed or not.

[nonost]

Would an opamp input buffer work instead of the FET? Like the one in the Small Time posted before. Or still the same problem?

Thanks for taking your time!

[anotherjim]

Now I thought you said it was like the Small Time?
Ok, can you photograph a sketch of the circuit? Save it as a .jpg and drag the file onto this page...
https://postimages.org/
Then copy the direct link it gives you and paste that link into a reply.

Yes, an opamp buffer after the bypass JFET so it disconnects the dry signal but leaves the buffer "holding" the dry mix path 10k into the PT2399 input.

[nonost]

Ohh...I think I can see what you mentioned earlier. At least I can grasp a bit.

So the problem it's not the buffer at the input of the circuit (either FET or opamp) but the FET bypass little circuitry. In that case I think adding a FET would be easier!

Another FET: drain to the previous FET(bypass) source, gate to the previous FET(bypass) gate, and source to the 100nf + 10k path. Is that ok?

I hope I got it right!

[anotherjim]

Won't be that easy adding another JFET. It would have to switch on whenever the existing bypass JFET switches off and vice-versa so cannot have the same gate connection.
What type of bypass/tails switch are you using? If it has changeover contacts, it might make it easier to control the second JFET.