PT2399-based Echo Noise Floor Issue

moodsystems

moodsystems

New member
Hey all! This is my first post on this marvelous forum. Lots of brilliant minds and valuable information here. Hoping I could get some help with a PT2399-based echo circuit I've been tinkering with.

Essentially I have taken the Deep Blue Delay/Sea Urchin circuit and slightly modified it to accommodate a Cornish input buffer (which I use in practically ALL of my builds), higher cap values, a "LONG" footswitch which increases the resistance within the Delay Time network, and an oscillation footswitch. See the attached schematic to see everything at play. With the Cornish input buffer, I have essentially defeated the usual 2x Gain op-amp input stage and have moved the op-amp further down the dry path, using it only for the summing amplifier and an inverting output buffer.

The issue I am having is the overall noise floor with the circuit. When I turn the pedal on even with the Mix knob completely off OR with the PT2399 chip removed, there is a slight hiss present and I can't seem to get rid of it despite using shielded cabling and higher value caps all around. When I attach an alligator wire at the junction of the 10k mix resistors and connect the other end of the wire directly to the output jack, the noise seems to go away. So I'm suspecting there's something going on with the U3 output stages? Should I try lowering the value of the resistors surrounding the op-amp (including the R3/R4 voltage divider)?

I've replaced the TL072 with an LM833N in an effort to employ an op-amp with a much lower noise specification, but that didn't really change anything. Running out of troubleshooting ideas, hence I've come here for some help!

Any thoughts?



SCHEMATIC:

EugenesEchoREV2.png
 
Sort by date Sort by votes
Great discussion on noise floor issues with the PT2399! Have you considered how different power supply noise could affect performance? It might be worth exploring dedicated power options for cleaner sound.
moodsystems said:
Hey all! This is my first post on this marvelous forum. Lots of brilliant minds and valuable information here. Hoping I could get some help with a PT2399-based echo circuit I've been tinkering with.

Essentially I have taken the Deep Blue Delay/Sea Urchin circuit and slightly modified it to accommodate a Cornish input buffer (which I use in practically ALL of my builds), higher cap values, a "LONG" footswitch which increases the resistance within the Delay Time network, and an oscillation footswitch. See the attached schematic to see everything at play. With the Cornish input buffer, I have essentially defeated the usual 2x Gain op-amp input stage and have moved the op-amp further down the dry path, using it only for the summing amplifier and an inverting output buffer snake.

The issue I am having is the overall noise floor with the circuit. When I turn the pedal on even with the Mix knob completely off OR with the PT2399 chip removed, there is a slight hiss present and I can't seem to get rid of it despite using shielded cabling and higher value caps all around. When I attach an alligator wire at the junction of the 10k mix resistors and connect the other end of the wire directly to the output jack, the noise seems to go away. So I'm suspecting there's something going on with the U3 output stages? Should I try lowering the value of the resistors surrounding the op-amp (including the R3/R4 voltage divider)?

I've replaced the TL072 with an LM833N in an effort to employ an op-amp with a much lower noise specification, but that didn't really change anything. Running out of troubleshooting ideas, hence I've come here for some help!

Any thoughts?
Click to expand...
It might not be the PT2399 at all—the Cornish buffer is changing how the op-amp interacts with the rest of the circuit, which could be where the hiss originates. Anyway, curious to see how your mods affect the noise floor overall
 
Upvote 0 Downvote
Why the big resistor values at R23/R26? (e.g. how were they chosen - something that's
worked before in another design, or ...?) My non-expert intuition is that you're cutting
a ton of signal going into it and then amplifying the tiny remaining signal a great
deal with the large feedback resistor. (amplifying any noise that comes along with it,
or introduced by the large resistors)

I think that your intuition that you need to cut these down is correct. You might just
try bridging them both entirely (just jumper both of them) and see what happens to
your noise floor.
 
Upvote 0 Downvote
There's no reason for R23 and R36 to be any higher than around 10k. They make noise. The higher the value, the higher the noise. Don't just jumper them, as suggested above, or the circuit won't work at all.

In several places, you have RC filters that will significantly roll off the lows (C6-R6, C5-R20). Is this intentional, or just values pulled from a hat?
 
Last edited:
Upvote 0 Downvote
@z2amiller @JTEX

Decreasing those 1Meg resistors (R23 and R26) to 10k has significantly decreased the noise! Thanks so much for the advice and insight. The 1Meg values were there due to that op-amp stage circuit being used for a high input impedance phase/polarity inversion switch in a different project. The 10k did the trick!

The 0.1uF input cap on the Wet path (C6) was lowered to that value (from 1uF) to lessen some of the low-end present on the echo repeats at the attack of the note. I didn't like that bit of "bump/thump" that was protruding slightly in the way of the dry signal lead playing. Decreasing that cap helped to keep the echo trails smooth and contained underneath the dry signal, if that makes sense. The 68nF cap (C5) was increased to maintain a rigid bass response within the dry path but I am now realizing that it being in parallel to R18/R6 can allow the opposite to happen. Will play around with that cap.
 
Upvote 0 Downvote
You must log in or register to reply here.
Back
Top