Delays and sea urchin distortion

[hooperharp]

Hello guys. As for my sea urchin build, it sounds ok. not great because of one thing I'll explain in a second and see if someone can help. I've built a rebote 2.5, a cave dweller and a sea urchin. Each has it's own features and sound. The thing with my sea urchin is the first repeat. It's waaaaay too distorted. Yet this delay has great things. I love the way it oscillates (and if I set the delay just before oscillation I get a shimmery delay I love). Also how the mix control gets the delay to "take over" the sound. As for the distortion on the first repeat, it happens on all settings, from short to long delays. The next repeats are great, it's just that first one. It's not noisy, but distorted. It's like echo on max on the cave dweller and strumming really hard. This is something I don't have an issue with in the other delays. How would I go around reducing this distortion? I've been looking at the schematics and other delays (jon's hamlet, grind custom's, zero points) but I'm really not having time right now (and for some time in the future unfortunately. law student. what was I thinking!) to breadboard and try replacing components. I'm using a tlo72, yet I replaced it with 4558s and lm358s, same result, so I guess it has more to do with the design of the delay. I guess also that the components I would be changing are around pins 9 to 16 of the pt2399 (please correct me if I'm wrong). I also added modulation (led+ldr combo). Any help would be of use. It's just that first repeat that needs to be tamed, filtered, or something. Or how would I go modding the rebote (or any pt2399 delay) to make the repeats sound louder than the clean signal (like in the sea urchin)

[hooperharp]

Just read something I'll try on the rebote in another topic, "Topic: Less Delay in Mix?", raising the resistor value after the mix pot reduces mix, so I'll reducing and see if I get more mix. Still if someone knows how to reduce the first repeat's distortion on the sea urchin or a post were this has been discussed I'll appreciate it

[jkokura]

It's possible you're hitting the input too hard. Are you sure all your cap and resistor values are correct? That's the first thing I'd suspect.

Knowing which PCB/project docs you used and some pictures of the board might help us determine if there's an error or if you're just experiencing an oddity.

Jacob

[gtr2]

Yeah, check out the values in the first gain stage.

[kothoma]

Hm, if the input of the PT2399 is hit too hard wouldn't all repeats necessarily be distorted?

[hooperharp]

I used the layout by Brian on the sea urchin topic. I'm posting the schematic. So, I reduced the resistor before the output mixing stage on the rebote and I got what I wanted, more mix, louder repeats than dry. One thing solved and one more thing I understand. Now, if I wanted to reduce gain getting to the delay, would I reduce gain on the "delay in" in the pt2399? Looking at the rebote and the sea urchin I see the resistor at the input of the circuit is higher on the rebote. Is this resistor limiting the input on the entire circuit? Thanks for helping guys. Delays are like overdrives, a personal thing as to what the "right" sound is

[kothoma]

I reduced the resistor before the output mixing stage on the rebote and I got what I wanted, more mix, louder repeats than dry.

This could probably explain a distorted first repeat... (if it only happens at high mix levels)

Now, if I wanted to reduce gain getting to the delay, would I reduce gain on the "delay in" in the pt2399? Looking at the rebote and the sea urchin I see the resistor at the input of the circuit is higher on the rebote. Is this resistor limiting the input on the entire circuit?

The input stage of the PT2399 around pins 16 and 15 is a low pass filter formed by R8, C6, R9, R14, and C7.
Changing any value will also change the cutoff frequency...
You can play with different values here: http://www.cepd.com/calculators/lowpass_ana.htm

But the gain should be unity in the passband anyway.

(Another thought: instead of boosting the wet signal you could attenuate the dry signal by increasing R4.
Or you change the gain of the input stage of the circuit, gain is -R3/R2, i.e. -2 at the moment. The minus just says that it is inverting.)

[midwayfair]

Need to eliminate some variables.

An easy thing you can steal from the Hamlet is this:

Solder a green LED from pin 7 of the PT2399 to ground. Anode toward the chip.

This will cap the signal in the PT2399 to 2V without clipping.

If that doesn't fix it, then your problem isn't that the PT2399 is distorting from the size of the input signal and it's something else.

I'm suggesting this first because it only takes a minute to add the LED and you can deal with the fallout from that later (mainly that the repeats volume is capped and might not get above unity).

I'll tell you right now that you can't get the output of the PT2399 to distort the second op amp unless something's wrong. The PT2399 runs on 5V and thus can't output anything above 2.5V. The op amp runs on 9V and the second stage does not further amplify the output of the PT2399.

Keep in mind that ALL PT2399 delays have harmonic distortion problems. That sounds like a soft, static-y, trashy distortion on the repeats. It's possible you're just not noticing it as much until the repeats are very loud.

[hooperharp]

Problem solved. I used the values from the zero point micro. 220k input resistor, 220k for gain at the input mix, and 22k (20k in the zero point but I only have 22k) for the gain on the output opamp. No noticeable distortion on the first repeat.
I can tell that repeats on pt2399 delays aren't clean, but the first repeat just was unbearably ugly. Now it's flowing (best description I can give haha).
I'll definitely breadboard a delay and try your suggestions. I'll also try the Led from pin 7 and see for sonic  improvements.
The rebote and the zero point have the same value for the input resistor and the gain on the first opamp. The sea urchin has half the value for the input resistor than the gain on the first opamp (180k and 360k respectively).