SOLVED Gatekeeper muffled and weak, won't oscillate

[Fama]

Gatekeeper from Effects Layouts, schematic here: http://effectslayouts.com/wp-content/uploads/2022/05/GateKeeper-Build-Doc.pdf

I get a delay, but it's very weak and muffled. With tone at full it's like a bit crappier version of the dry signal, but quickly devolves to mud, while with the tone lower it's just mud to begin with. Mix at full is ~unity or a bit under.

I've reflowed some suspicious joints, they still look wonky, I'll try to maybe fix them up more and use a solder sucker to make them look nicer too. But looking at the board with a magnifying glass, I can't see anything that looks obviously bad.

I've double checked every cap (and their orientation), double checked all the pot values, and checked about half of the resistors with a DMM until I hit R18 which gives me 44k when the resistor itself is 220k, I think it's just getting a lower resistance through ground maybe? Or is this some sort of an issue? It's identical with the rest of the 220k's in the bag, and those measure 220k. I also tried another TL072. I don't have spare voltage regulators or PT2399's (the JRC4558 is in the modulation part I think, and I haven't even tested that yet).

One substitution, a 330nF and 150nF caps in parallel instead of the 470nF in the BOM, both are just in the same holes.

The PT2399 was the wrong way around at first, but apparently that shouldn't damage it?

There's also a constant buzz when my guitar volume is on, but could maybe be because it's not boxed? Not sure if that's another clue.

Sound sample, mix, tone and repeats are at full here

https://soundcloud.com/user-913671490%2Fgatekeeper-sample%2Fs-8Ure49vecXO

 

[BuddytheReow]

If the pt2399 is supposed to be facing up, your chip is upside down. I don’t think you’ve fried it, but do you have extras?

 

[Fama]

If the pt2399 is supposed to be facing up, your chip is upside down. I don’t think you’ve fried it, but do you have extras?

I'm pretty sure it's supposed to be facing down, it's both on the build docs and also on the silkscreen. Sadly I don't have any extras of that or the voltage regulator - although I guess the voltage regulator should be fine since the PT2399 works at least somewhat?

 

[BuddytheReow]

You can always test the regulator. Should be 5v on one of the pins or close to it

 

[Fama]

You can always test the regulator. Should be 5v on one of the pins or close to it

Maybe I'll do that, compared to ground? Or apparently is the middle pin ground too?

Also using an audio probe was suggested, the test rig has one, do I just start probing things to see where it starts sounding muffled or how would I go with something like this? I think I understand how it usually works if you don't get sound, I haven't ever used one.

 

[Paradox916]

R18 should read 220k because it’s only 44k is a little concerning I would definitely replace it, (would be a quick shot gun) because it’s right there next to the tone stack the lack of resistance to ground, it is kind of suspect to your problems of dark tone and low volume, could be wrong but I bet it’s dumping some of your signal to ground, with one test lead on ground one side of the resistor should read zero and the other side 220k, I can’t imagine a situation where it would read otherwise if everything was correctly installed, and yes your pt2399 if correctly oriented. The sound clip isn’t too far off of what it sounds like the disaster transport is kind of a darker sounding delay.
It’s not the best representation because there is a lot going on there but here is a demo with the one I built https://forum.pedalpcb.com/threads/effects-layouts-vorticity.16651/

 

[Fama]

I didn't replace R18 yet, but that's probably the next step. I did solder in input/output jacks, the power jack and both footswitches to eliminate any weirdness with those. The buzzing went away, so that was probably related to input or output ground, but no other change.

I also measured the voltage regulator, it reads around 9V on one side, and like 5.12V on the other side, so I think that's fine? Next suspects is R18.

 

[Fama]

The resistor was indeed 220k, it was correct. One side of the hole reads 0 ohms, one side reads a few dozen kOhms. And now I can't get a new resistor in one of the holes because it has solder in it. So this was just a big step backwards.

Edit: It now stands one leg on Solder Mountain, so that has been solved. Other leg is through the hole.

 

[Harry Klippton]

And now I can't get a new resistor in one of the holes because it has solder in it.

Do you not have a solder sucker?

 

[Fama]

Do you not have a solder sucker?

I do, that's how I got the other hole clean. This one had the solder inside the hole, so I think I couldn't quite reach it from either side? In any case it just wouldn't come out with the solder sucker. But the new resistor is there, so this is not an issue any more.

Currently I suspect it's maybe just a bad PT2399, since that apparently happens (at least with Tayda), so I'll order a bunch more and try them out. A domestic electronics store has them too, but they're 5€ a piece (+ shipping, and other parts aren't that cheap either, for the most part) so I think I'll just get 10 of them from Tayda - should work for any other PT2399 too if I decide to build them in the future, Magnetron looks cool for one.

 

[Paradox916]

I would really double check all your work and audio prob it before just rolling the dice on a delay chip, if you don’t confirm that’s the issue you are going to be upset if you find out it’s not.

 

[Fama]

I would really double check all your work and audio prob it before just rolling the dice on a delay chip, if you don’t confirm that’s the issue you are going to be upset if you find out it’s not.

I can try the audio probe still - the issue with checking all the resistors is that measuring them is not that easy for me (because the measurements aren't always same as the BOM says), and I don't know the color bands well enough to be able to read those at all quickly. But maybe I can compare with the other resistors I have to check the bands.

I've gone over all the soldering so many times I'm pretty sure that part is just fine.

Cybercow mentioned the Tayda PT2399's were inconsistent here: https://forum.pedalpcb.com/threads/psa-more-things-to-avoid-from-tayda.16888/#post-198183

And someone on the Mad Bean forums mentioned that bad PT2399's can be muddy in a bad way and with noise (which I didn't mention, but there's some white noise in the repeats now after I fixed the buzzing issue), also mention of delay time, which I didn't measure. https://www.madbeanpedals.com/forum...e1fbdac097f9b&topic=24772.msg241419#msg241419

 

[Fama]

Goddammit, turns out C13 was 22nF and not 2.2nF. I didn't check that one well because it was one of the WIMA caps where you can't see the value from the top, only on the side which was hidden, and I was fairly sure it was the correct value. Well, at least it works now as it should! Plenty of volume (goes over unity), goes into oscillation and sounds nice.

There's a filtered white noise sounding thing going on with every repeat, but it's not so bad that it's unbearable... but I think I didn't have that at one point, so not sure what could be causing it.

 

[Paradox916]

Good job seeing the trouble shooting process through, it can be a tedious pain in the ass but it pays off.