Hiss - How To Diagnose?

[Ginsly]

I'm pretty used to fuzzes being generally noisy (mostly from picking up electro-garbage from LED lights etc), but wanted to ask for future troubleshooting - if you're getting a good amount of background noise, what is the best way to see if there's an actual problem?

I just built an Executive and added a voltage starve. The hiss happens no matter where the Gain or Starve pots are set, and doesn't change as I change position with my guitar. It's definitely hiss, not buzz. Other than that, the pedal sounds excellent.

I know Bosstones can have some background noise, so this may be normal. Trying to figure out a good step-by-step to help determine that.

I socketed the transistors and diodes, so I plan on trying different transistors. What else should I look at? Would love to reduce the hiss if I can.

Much appreciated!

 

[Ginsly]

Any insight into this? I think some circuits just have a higher noise floor, but it would be nice to know how other people hunt down (and hopefully reduce) noise. Are funky transistors the usual suspect? Caps? Sometimes it's just EMI being picked up and amplified, but in this case it's background hiss. If it was a constant buzz I'd probably look for a grounding problem in the wiring, but that's not the case here.

 

[jwin615]

You can try a small cap between B and E of the various stages.
10pico-300pico. Whatever sounds right without killing the top end.

 

[JTEX]

In my experience, most high gain guitar effects are not optimally designed for lowest practically achievable noise. I'm talking about circuit hiss (white noise), not various environmental noises picked up by the pickups,such as buzz, computer chatter or transformer hum. Designing for low noise requires a holistic approach. One needs to consider the inherent noise floor of the source (pickups), the self noise of every active stage in the signal chain (transistors, op amps) relative to the signal level at that stage, and especially the very often neglected self noise of all the resistors surrounding the active components. A single 10k resistor in series with the guitar input will permanently degrade the overall noise floor by about 3dB right from the start of the signal chain, on its own (assuming a 10k pickup source impedance). Similarly, a single TL072 op amp, while not particularly noisy, will still degrade the theoretically achievable noise floor of a 10k source by a few dB. It all adds up in the end.

 

[Ginsly]

You can try a small cap between B and E of the various stages.

Interesting! Is this something that's typically done to reduce hiss? I'll certainly give it a shot. Q1? Q2? Both?

@JTEX This is a really good point, and one I'll keep in mind - thanks. It seems like there are SO many variables with noise, and many types of noise too - usually mine is a product of a city environment filled with EMI - huge transformers close by, LED lights are a HUGE source of it, computers, etc. It's usually not the pedal itself, but it amplifies all of the garbage coming in through the pickups. Some pedals are just noisy no matter what, though (Big Muff, Death By Audio, etc). I use good Truetone power or simply batteries, so it's never really been a power issue.

I've gotten to the point of needing a transistor Bias or overall Gate knob in most fuzzes I use, just because I can dial out the background chatter and buzz.

Still - as I'm learning to build I really do want to do my best to limit/reduce noise as much as I can when it comes to component choices. For instance, I know some transistors can be noisier than others (or become microphonic..?) but I'm not sure if that would present itself with hiss, buzz, degraded signal, etc. Same thing for bad caps, etc.

Kind of a separate topic, but when it comes to external EMI,would shielded wire for the Input and Output help? Ferrite beads? The whole circuit is in a metal box, so I'm not sure how much that'll help.

 

[JTEX]

Some pickups do a better job of rejecting EMI than others. Also, getting the pickups closer to the strings will raise their signal-to-EMI ratio, at the cost of more magnetic pull on the strings (lower sustain and possible intonation problems if you get too close).

If your circuit is in a grounded metal box, using shielded internal wiring won't help with reducing noise -- only with reducing internal crosstalk. This can manifest itself by loud squealing in some cases, if you have high gain and the output wiring is too close to the input and unshielded.

 

[Ginsly]

@JTEX Gotcha. It's good to know that shielded wire wouldn't really make a difference.

I'm still not quite sure how to tell if transistors I try in any given circuit are extra-noisy, microphonic, or just "bad"... Given that fuzz is inherently prone to noise, I'm not sure what the telltale signs of a bad transistor are.

 

[Ginsly]

@jwin615 I think I'd like to try that small capacitor idea for reducing hiss - can you tell me in practical terms what that entails? I socket all transistors, and I'm not sure how/where I would be attaching a small cap so that it's functionally between Base and Emitter. Still learning the basics. Thanks!

 

[jwin615]

@jwin615 I think I'd like to try that small capacitor idea for reducing hiss - can you tell me in practical terms what that entails? I socket all transistors, and I'm not sure how/where I would be attaching a small cap so that it's functionally between Base and Emitter. Still learning the basics. Thanks!

Hey Ginsly. Missed the first reply. My bad.
There's a great thread here about the executive. Note mod number 3.
Also, note the comment about using a low noise Q1(Chick recommended BC109C).
Since you socketed the transistors, I would try rolling some to start(that's the whole point of sockets, right?)
IDK what I read before my initial comment but rereading now, I'd try the linked above route first.
If you want to toy with the caps, and have some jumper wire to go from your sockets to a breadboard, you can try that as well. It's non- invasive and reversible.
Jump the 3 legs of Q2 to a breadboard, each on their own row.
Install your transistor, correctly orrientated across the 3 rows.
Then, test a few different pf caps bridging B and E. Try the 82-120pf range.
Since E is tied to ground, you're actually bridging the cap from B to ground.
Capacitors have a variable impedance, depending on frequency.
Low freq AC will ignore the capacitor and skip on by while higher frequencies will drop to ground.
This is AKA a treble bleed...
Often, but not always, when you see ~100pf caps in these simple circuits, they're to bleed HF or prevent oscilations.
This can also be useful on fuzzes that oscillate at high/max gain.
Hope that helps.

Edit-fat fangers.

 

[jwin615]

To add on to the why, simply put.
To tiny little caps like 100pf, who have very fast charge/discharge rates, lower frequencies look like DC, which caps "block"
Since the cap is to ground here, it blocks it from bleeding to ground.
There's a whole relationship between cap values(charge rate) and frequency wavelengths.
That's why different cap values "block" different frequencies.
Because our cap is tiny, it only lets super HF pass.
Now, there's are equations to work this out, but we don't know how many Joules are present at that junction, but we do believe from history and heresey that 100pf +/-20pf tends to work well. Use your ear. The bigger the cap, the lower the HF cutoff will be, and the quieter your hiss+your upper range.
And do you play himbuckers and single coils, be sure to test multiple guitars before commiting.

But as @JTEX stated, there's going to be some noise. (Some)Newer pedals have filtering in place but in the time the bosstone was made, tape and amplifier noise were the noise floor.
With gain pedals with hard clipping, you are always compressing the signal and raising/expanding the noise floor.
Good tests:
-putting amp in standby-
Unplug the guitar. What % of noise went away?
Unplug the input of the pedal(assuming it's not switched). What % went away?
It's not always the pedal, it's just that the pedal amplifies it.

 

[Ginsly]

@jwin615 Wow, thank you SO much for the detailed explanation!

That link to several Executive mods is excellent, and it's bookmarked!

I get LOTS of EMI, and while I avoid what I can there's always going to be some. I've really gravitated towards Fuzz pedals with transistor bias knobs, as that helps me kinda gate it out; plus, I like those textures anyways. What I'm hearing from the Executive (and other Bosstones) is more a constant hiss which seems to be the circuit more than EMI (though it's sometimes hard to separate the two).

I think I'll try some other transistors in Q1, since it sounds like 2N2222A's are known to be kinda noisy. I did add a Voltage Starve (which is excellent!), but I think I'd like to get crazy and also add a transistor bias knob (for Q2 I assume?). Not exactly certain how to do that to the Executive, though.

I like the breadboard idea for the filtering cap, and although I'm still learning the breadboard (I'm only two months into building) I can easily do what you're suggesting.

Let's say I've found a small value cap I want to add- would it just be to Q1 or Q2 as well? Where would I physically be connecting each end of the cap? Would I just kinda cram both the Base and Emitter legs into their sockets WITH a capacitor leg? I'm just struggling to picture how to properly insert it between B and E.

As a beginner, I really, really appreciate builders like you taking the time to help me understand what to do and why. It's slowly making sense, and I'm able to put together modded PPCB circuits as well as Vero builds.

 

[jwin615]

Do you still hear the hiss with nothing plugged into the circuit, just output to amp, or with your guitar volume rolled to 0?
EMI shouldn't be a huge issue within the circuit with the circuit enclosed as the aluminum enclosure has a faraday effect. It's not perfect but works pretty well. Now, EMI vs your pickups/guitar is a different story.
Just want to be certain the hiss IS front the circuit and not just being amplified by it.
Since you have both in sockets, move them both to a breadboard and try a cap in both, one at a time. If you have a lot of noise while doing this, try turning nearby florescent and led lights off(tons of EMI) and other mains electronics. Need to clear out extra noise so you can hear the circuit.
Once/if you find a setup that helps and sounds good to you, you can solder the caps to the pot side of the board. I wouldn't try to cram both in a socket. Just space and trim the leads and tack it onto the socket pins. You can use tape or blue tack to help hold it in place if needed.

 

[jwin615]

Ceramic/mlcc caps are best used for this btw.

 

[Ginsly]

@jwin615 Gotcha! That makes total sense, and I wholly expect there to be some hiss even with no cable connected- Bosstones just kinda do this unless you tweak em, it seems. I'll check.

Ahhhh you solder the cap to the BACK side of the pcb! Totally forgot you could do that.

And now that you mention it, the ceramic/MLCC thing has come up before when I researched filtering. I have greenies and box caps, but I need to stock up on MLCCs. I have a handful of nice TDK blues from Mouser, but need to fill some gaps.

A kit from Amazon would be fastest, but I'm not sure of the dielectric with those. Is it safe to assume that most/all MLCCs in low pf values would be C0G/NP0, no matter where you're sourcing them? I'm not sure what the cutoff usually is with those, and if you really need C0G/NP0 to be spelled out for those smaller caps. It sounds like X7R wouldn't be an issue in the grimy fuzz pedals I'm building, but... still.