Pointers for a noisy first build?

[charlesj]

Hi everyone. I recently completed my first overdrive pedal (not my own design, but just to get some of the building skills down). It was super rewarding and I'm excited to get started on this hobby.

I just had a few questions I was wondering if anyone could answer. The pedal works, but it's a lot noisier (kind of a white noise/hiss) compared to another cheap overdrive pedal I have, and sounds less articulate as well (flatter, less dimension, lower resolution?). I know that this could be from a range of issues, but I was wondering if anyone could give me some pointers on how to do the best build possible.

For instance, I've read that low quality parts can cause a noisy pedal but can anyone elaborate on this for me? I know op-amps can be sources of noise in overdrive pedals, but are all low quality parts potential sources of noise? Why?

What about degradation of the signal quality?

Does the positioning of off-board wiring play any role?

I'm just looking for some pointers along with explanations for why they're important for a high quality pedal. Thanks!

[thermionix]

What circuit did you build, and what "cheap overdrive" are you comparing it too?  Are you using a battery or wall wart?

[charlesj]

Thanks for the reply. I built a landgraff dynamic overdrive and I'm comparing it to a Joyo "sweet baby".

Obviously not the same, but the sound quality difference was big enough that I figured it must be something I'm not doing right.

I'm running it from the wall.

[EBK]

Welcome to the forum!

Do you have your circuit in a metal enclosure?

[charlesj]

Thanks!

Yep sure do. It's a 1590b style enclosure.

[thermionix]

Some wall warts are really noisy (SMPS).  Some pedals filter that noise out very well, some do not.  The high-pitched whine is not what I would call hiss or white noise, but maybe you would.  One way to eliminate that is to hook up a battery, even if rigged up temporarily, as batteries aren't noisy.

[EBK]

Some wall warts are really noisy (SMPS).  Some pedals filter that noise out very well, some do not. 

The circuit in question is a tweaked tube screamer.  The op amps should, in general, handle the power supply noise somewhat decently.

[Mark Hammer]

There are a few sources of noise to be considered.
One is certainly poorly-regulated power, although that tends to generate hum rather than anything described as hiss.
A second is poor layout, particularly if you have a lot of wires stuffed inside the chassis, running all over the place.  Having wires carrying output near wires carrying input will pose a noise risk.
A third is poor choice of semiconductors.  Some are inherently noisier than others.  What did you use?
A fourth is too much bandwidth.  Higher gain devices not only provide some noise of their own, but also amplify any noise entering with the audio signal.  If the circuit aims for widest bandwidth, it will keep all that inappropriately-produced hiss.  You may be able to tame it with an additional, or higher-value, cap.  Note that many guitar amp speakers roll off above 6khz, so if you are using speakers with wider bandwidth, their "virtue" may be in letting you hear things you'd rather not hear.
Finally, tolerance for noise is an individual thing.  Some folks will hear the same hiss as another person declaring "Dear gawd, what IS that noise?", and think nothing of it.

Note that none of these precludes any of the others, and pedal noise can result from any combination.

[EBK]

Could you post some pics of your build?

[charlesj]

Hello everyone. I really appreciate the suggestions. Below I posted some pics and here are some other details.

1. It doesn't sound like a hum (it's non-tonal) to me. I'd call it white noise.

2. Regarding layout, in the pics you can see the off board wiring I used. I definitely didn't put too much care into addressing it. Are there any specific rules of thumb or techniques I should take care to use? The layout of the circuit itself is from effectslayouts and can be found here: 

http://effectslayouts.blogspot.com/2017/10/landgraff-dynamic-overdriveclay-jones.html

3. Sorry I'm still really new to this, but when you say semiconductor are you referring to anything that provides the clipping? For the hard clipping diodes I used 1N4148s. The soft clipping are generic 3mm red LEDs. The transistors are both BC546Bs.

4. Again, a bit of a novice question but for trying to tame the incoming noise, I just want to clarify that you mean by increasing the value of the 22 nF capacitor in a high pass filter configuration at the beginning of the circuit. Is there a value that you suggest trying?

5. My soldering is definitely messy but I'm sure it will get better with more experience. I know there's a correct way to do it, but if the right connection is made could this introduce any noise or decrease signal quality? I was thinking that at the worst a connection might be higher resistance than desired, but wouldn't think this would really have any effect on sound "quality" other than a volume drop. Please let me know if there's something else I should be aware of.

Thanks!

[EBK]

I have to confess that I got a bit distracted by the text in those pics....  (I'll stop there with this particular comment). 

[EBK]

Are you insulating or securing that board to make sure nothing is shorting out when it is in the box?

Also, the soldering is a bit heavy.  Double check your work against a picture of the board layout you used to make sure you didn't bridge something with a solder blob.

There seems to be a lot of corrosion on your board (the black stuff)....

[Mark Hammer]

1) Your builddoesn't look all that much different from any of my own.  The only difference is that I like to slip a piece of heat shrink over the pot solder lug and last 1/8" of the wire connecting to it.  It serves as both strain relief and insulation, so that when pots get twisted around during installation, the wires don't fracture, and the solder lugs don't short out against nearby things.

2) "Semi-conductor choice" refers to any semi-conductor in the circuit, whether transistor or op-amp.  Transistors used for input and/or output buffers are generally not an issue, since they are used in unity-gain mode, making any differences between designated low-noise and non-low-noise transistors moot, even though differences might be quite apparent in something like a much higher-gain mic preamp.

For me, the poster child for this is my Behringer acoustic modeller pedal.  Essentially a clone of the Boss AC-2, Behringer elected to use the noisier (but lower current-draw) TL064, instead of the identical pinout TL074 that Boss used.  A battery lasts longer in the Behringer unit, but in this era of pedalboard power supplies that's a moot consideration.  The hiss, meanwhile, is a royal pain.

In the case of the traditional TS9 circuit, there isn't the sort (max = 114x) of gain that might risk aggregious hiss.  The Landgraff version applies much greater gain (max > 1000x).  So even though a JRC4558 doesn't pose much noise-risk in a normal circuit, the much greater potential gain in the Landgraff implementation makes greater demands on the op-amp to NOT be noisy.

3)  The 22nf cap on the input is the same value as on the TS9, and really doesn't need any changing.  IF, however, you have omitted the 51pf feedback cap in the clipping stage, or misread the value and installed something smaller, you won't be getting the normal treble rolloff as gain is increased.  That cap, in conjunction with the feedback resistance setting, forms a "sliding" lowpass filter.

4) Another possibility is that your tone section is in error in some manner.  For instance, if you had inadvertently used 22nf instead of 220nf for the tone cap, you would be getting a boost of content in the treble range without the normal accompanying mid boost.  The result would be more hiss than signal.

5)  Consider strapping a 22nf cap in parallel with the 1k feedback resistor in the tone stage.  That will form a lowpass filter rolling off around 7.2khz.  You should hear a reduction in hiss.

[thermionix]

Your 4558 is a fake.  The real ones have somewhat different markings, with "JRC" all the way over to the right.

A google image search brought this result from AliExpress:

https://www.aliexpress.com/item/100PCS-JRC4558-4558-4558D-DIP-8-Integrate-IC-Chip-DUAL-OPERATIONAL-AMPLIFIER-Original-and-NEW/32856168315.html

Another fake with (surprise!) the exact same date code.

Get a real one from Small Bear, Mouser, etc, and I bet your noise problem will be solved.

[charlesj]

Awesome that's a lot to work with. That's a great catch on the 4558 I just ordered it from eBay and never would have known. Based on the suggestions I've ordered some 4558s from Small Bear and plan to:

1. Double check my caps in the clip and tone stages.

2. Replace the 4558 with a real one (also ordered a couple TL072s to try - I read that these might be a little quieter but any other suggestions?)

2. Add a 22 nF cap in parallel with the 1k feedback resistor in the tone section if it's still noisy.

Thanks again everyone. I'll let you know what happens!

[diffeq]

Awesome that's a lot to work with. That's a great catch on the 4558 I just ordered it from eBay and never would have known. Based on the suggestions I've ordered some 4558s from Small Bear and plan to:

1. Double check my caps in the clip and tone stages.

2. Replace the 4558 with a real one (also ordered a couple TL072s to try - I read that these might be a little quieter but any other suggestions?)

2. Add a 22 nF cap in parallel with the 1k feedback resistor in the tone section if it's still noisy.

Thanks again everyone. I'll let you know what happens!

NE5532 is considered as relatively low noise opamp. NJM4580 too, if I recall.