Weird Noises – Boss CE-2 Based Project

[Juan Wayne]

I tried to be short, but I don't want to waste people's time by omitting half the stuff, so please bear with me here. Hopefully it'll be a fun one to solve. Audio clips of the pedal and noises in the end.

I breadboarded a regular CE-2 clone and started modding it, mostly your garden variety tweaks. It sounded amazing except for the usual breadboard + loose-wires-everywhere noise, so off to PCB it went.

Once assembled and floor noise being gone, I discovered one mod was giving me a couple issues. The mod in question is a secondary vibrato line, identical to the main MN3102 + MN3207 one with the only difference that the LFO comes in inverted, meaning when both are used together one goes up in pitch and the other goes down, so it's even more chorus than normal but with less oscillating-out-of-pitch sound. I love it.

Schematic (before some required fixes):



***First issue*** is both MN3102s seem to "bother" each other somehow, as in the clocks become unstable unless I completely take one MN3102 away from the board, doesn't matter which one.

This causes two problems:

1- The change in pitch ramp is weird and glitchy when the LFO hits a high or low point and starts going the other way. It doesn't happen at the LFO itself, but rather at the clocks. This gets reasonably solved by changing the value of the capacitor on the self-oscillating circuit associated to one of the MN3102s, kind of "detuning" both MN3102 circuits from each other as much as possible, but I can only set them apart so much before one vibrato begins to sound weird and becomes unusable. Besides, the point was for both BBDs to be doing the same thing to begin with.

2- There's a whistling, oscillating noise that changes along with the LFO, plus some hiss because why not? The only solution is to get rid of one MN3102, not the BBD, not de-activating one vibrato, not preventing the LFO from reaching the MN3102, nothing. Only removing one MN3102 allows the other one to behave normally, which puts me back in square one, with an impeccably sounding CE-2 clone, but just that.

***Second Issue*** was a lot of hiss, which I kinda solved but not really. Initially, dumb me summed both BBD lines right into the post-delay reconstruction filter as you can see on the schematic. Buffering both MN3207s at their outputs gets rid of a lot of said hiss, but not all of it. I know two BBDs equals twice the noise of one which I can live with, but it makes no sense to still have extra noise when the second line is disabled. Yet it was bleeding in even when that vibrato line was not engaged.

Apparently the only way to remove it is to fully "lift" the outs (pins 7 and of the second MN3207 when not in use. I don't know why but only that does the trick, even separate power supply lines were useless on this case (or any of the following issues). I'm still puzzled with this, but a fix was achieved. I'd love some ideas though.

***So now the audio clips***:

1- Normally sounding CE-2, mid settings for LFO and depth (one MN3102 is out of the circuit):
https://soundcloud.com/230_official/01-chorus/s-rGy2d

2- Same settings but the other MN3102 is present (turn it up, some noise can be heard, especially in the end). Also, I still think the first clip sounds better, regardless of the noise issue:
https://soundcloud.com/230_official/02-chorus-with-noise/s-IBjv2

3- Both vibrato lines engaged, same settings, not so much noise but more hiss:
https://soundcloud.com/230_official/03-dual-chorus/s-xfKWp

4- The noise alone. This is what happens as soon as you add some juice to the amp:
https://soundcloud.com/230_official/04-noise/s-JXi60

5- For the curious, a shorter delay setting and a long (and noisy) setting. I thought those sounded cool.
https://soundcloud.com/230_official/05-bonus-chorus-sounds/s-YxES4

Sorry for the long winded explanation but I'd hate to waste your time by only providing half the story.

Thanks in advance!

[Scruffie]

Welcome to the horrible world of heterodyning, search for the term and you should find ample information on what it is, what causes it and how to prevent it.

[Juan Wayne]

Oh boy... I hate you, but I knew something along those lines would be coming, haha. Thanks for pointing out the right term, it's a huge thing for me being a Spanish speaker. I studied all my electronics in Spanish, so any new concept I want to learn I gotta find its name first or somehow describe it to the browser and hope for the best (googling anything technology-related in Spanish stinks, there's nothing about anything!)

Guess I'll have to dig deeper, but I'm still praying for that one Hail Mary of a chance someone faced the same problem before here and solved it. I wouldn't be surprised. One of the reasons I don't comment much here is someone always seems to know all the answers. It's truly amazing for those of us who just stumbled in by accident.

[Scruffie]

Yeah, I'd hate me too

What you're trying to attempt is basically the BOSS DC-2, can I ask why you don't want to go that route or have you not considered it? The hard work is done there if you find a good layout.

[Juan Wayne]

Actually, I had never seen one before. Living down here on the depths of South America I miss out on a lot of things, especially anything outside of the mainstream common toys. I'll check it out, the answer might be in there!

[ElectricDruid]

There was a thread here recently about building a cabsim using delay lines and someone was having the same problem with that. There were some useful pointers in there:

https://www.diystompboxes.com/smfforum/index.php?topic=120147

DrAlx pointed out that you need to separate the *inputs* as well as the outputs, which I hadn't come across before. So there's various things you can do to improve the situation.

HTH,
Tom

[Juan Wayne]

That's definitely helpful, and a very interesting thread too. I hope I can patch the circuit without having to redo the whole thing again.

Thanks a lot for the tip!

[Juan Wayne]

So I went on and tried all the things that came up from your suggestions. Very informative stuff! Temol's cabsim project looks awesome.

It does seem to be and issue with some degree of heterodyning, which I had never dealt with before on pedals. Sadly no quick fix seems likely to do the job. De-tuning clocks to frequencies separated well enough from each other as to not produce and audible effect makes one of the delay lines unusable, while trying to balance the outputs of the BBDs with a trimpot only makes matters worse; seems like they're already balanced enough.

I also tried to filter the inputs of the BBDs more aggressively to see if some noise was being created at the inputs from sampling, but to no avail. Same thing with completely different power supplies.

So, it seems like despite my initial efforts to keep everything away from everything on the PCB, there was simply not enough room and now I'm stuck with a partially useless circuit, while the rest performs as a marvelous CE-2 clone with mods. Actually at this point, I think it's for the best, and I can see why the DC-2 wasn't all that big of a hit (2 NE570s on a stupid chorus pedal with no rate or depth control? Wow!).

I will keep my hopes for a magic solution if someone happens to think of it. Thanks all for your help.

[DrAlx]

You need separate LPFs on both BBDs at both input and output.  See link in red for the circuit where I first learnt how to solve these heterodyne issues. It's a long thread but the final circuit in the link is what finally worked for me...
https://www.diystompboxes.com/smfforum/index.php?topic=107353.msg984407#msg984407

Note that the 2 "digital" VCOs/BBD sections have supplies decoupled. That applies to both the positive lines and the ground lines.

EDIT. Also look at the layout in the build document. I kept the two digital sections as far from each other as possible on the board.

[Juan Wayne]

Holy crap that one hell of a project! And it sounds incredible too. I really like the subtractive mode. That was one of the reasons I didn't want to give up on the two delay lines. On some of the patching I did to the circuit I inverted one of them and the chorus effect I was getting is awesome.

Either way, now that I see the precautions you took it seems so reasonable to take a few extra steps to prevent those issues! I thought I could just add the essentials to the base circuit and maybe get away with it. Probably my initial mistake was cutting corners so I could fit everything inside a small enclosure, which led to simplifying to much.

I guess it'll have to wait for the next project and it'll have to be from scratch, but now I know exactly what too look at. And what the hell, at least I have a great sounding chorus.

Thanks a lot for your help! I'll keep a close eye on the NZF when I have to re-think this one. Cheers!