I have a fuzz face type of pedal that I build from a schematic using prototype board. I notice when I use one of my power supplies that has only 500mA, I get noise from the pedal’s output. (Almost sounds like night crickets in a forest.) But, when I use my other power supply that has 1700mA, the pedal sounds fine and is noise-free. Why is that? And is there something I can do within the build to ensure a noiseless power supply experience?
I know many pedals have a “clean power” section and never fully understood what that is all about, but wondering if this is the type of problem those power sections are resolving.
One thing I’ve picked reading this board and for you to try too is adding a 100nF cap (ceramic/plastic) from power jack socket positive connector to pedal housing with leads as short as possible. Best to ensure secure connection with bolt and lock washer. My OneSpot whines more than cheap Mr.Power one, but isolated Vitoos 12 (sold as fender engine room, behringer iso-12 etc.) is quiet even when powering digital and analog pedals with it.
Buildwise the placement of components on circuit and to each other, component lead lenght, wiring placement and orientation and quality of soldering give their effect to overall noise level with decisions on circuit topology and component choice. Not all components are equal (not even in the same batch) so compensation might be needed somewhere f.ex. on transistor biasing or IC slew rate adjustment. Tweak, adjust, tinker and lose your sleep! Maybe we could think populating a pcb with BOM components as navigation with GPS and you honestly don’t know if it misleads you without knowledge about the route. Then breadboarding could be thought as navigation with a map and compass, you’re free to choose how you get where you want and can make changes in the plan when needed to really get there.
Maybe something is misbiased or your beefier supply just has beefier AC filtering before transforming AC to DC. Dunno.
I once had a pedal which was fine when I played it at home but had that whine when I used a cheaper power supply in a music store. The trick in that case was to use the 1N5817 polarity diode. For whatever reason that cured the whine. Usually using the diode inline works, occasionally I need to add a 100R resistor in series too. Some power supplies aren't as well filtered as others I guess. It also seems that adaptors with transformers can be quieter than switching adaptors.
I tried to think what a diode would do, but cannot make any proper statements with current level of knowledge. But if you combine 100r resistor HamishR suggested and 100nF cap from me you get this:
Lowpass filter on DC and cutoff frequency at ~15.9kHz.
I haven’t tried it yet, but will if I can’t get it fixed on the board. I’m trying to think of what the difference would be that a PedalPCB board might have in its treatment of power that neutralizes the issue.
Part of me wants to say “oh well, use a different power supply”, but since I want to make good stuff I’d like to figure out why my circuits are doing it and not the ones I order from PedalPCB.
I think what you suggest above is similar to this part of a PedalPCB schematic. Am I correct? (See attached.)
I will try something like that next. (It’s not that I don’t want to try it, just that I’d rather fix it in the schematic.)
A 100 ohm resistor followed by 100uF (or more) to ground (maybe in parallel with a 100nF, which works better at higher frequencies) would kill all but the most egregious noise coming from a power supply. Just as long as your pedal doesn't draw a lot of current, in which case the 100 Ohm resistor might cause too much of a voltage drop. Values are highly negotiable as long as they form an RC filter with a low enough cutoff. I've even used 10 ohm and 2000uF when I needed more current.
Do you happen to know of a way to test how much current a pedal needs? I have a nice multimeter, but not sure if it can test that. (I think the Electra circuit requires very little but I don’t know.)
Do you happen to know of a way to test how much current a pedal needs? I have a nice multimeter, but not sure if it can test that. (I think the Electra circuit requires very little but I don’t know.)
Set your meter to mA DC and (somehow) insert it in series with the power supply.
OR:
If your pedal already has a series resistor (such as 100 ohm, or 22 ohm or whatever) right after the DC input jack, you can measure the voltage across this resistor (V) and compute the current using Ohm's law: I=V/R
For instance, if you measure 0.5V across a 100 ohm resistor, you have 0.5/100 -> 5/1000 -> 5mA.
A 100 ohm resistor followed by 100uF (or more) to ground (maybe in parallel with a 100nF, which works better at higher frequencies) would kill all but the most egregious noise coming from a power supply.
This seems to have solved my issue. The result was the same with or without the parallel 100nF cap so I omitted that part of it. Here is a video of the result. Thanks.
In peepeecb schematic there’s no 100nF capacitor to ground/shield for gnd ref and for HF/RF noise rejection. Some circuits include it parallel between positive and neg leads after first filter electrolytic filter capacitor (values vary from 22-220uF). Or you do it like in my sketch and 100nF cap is from positive lead to chassis/enclosure as close as possible on pwr input jack. For what i’ve understood it’s most effective applied so. But you have to understand this 100nF cap isn’t mandatory, most here go without it.
Here’s a approx calculation what this 100nf-to-shield cap does if we combine it with 100 ohm resistor series in power input. It’s a low pass filter for power!
As a sidenote earlier today I played for a while and I know my pickup is bit microphonic, but it got to whole next level as I was using my cheapest PSU. Funny thing is from my experience this cheap PSU has treaded me well with pedals build with digital circuits. From now on that psu won’t get near my analog pedals.
And also one important thing for you to consider. Does your PSU&pedal whine when there’s no input from guitar, guitar volume is down or if you short circuit’s guitar input jack? And bonus round; does the circuit whine change with single coil vs humbucker?
Btw, here’s some octave fuzz circuit’s power section I had saved:
Edit. I was missing 100ohm input resistor completely earlier, but you guys got there while I was writing this message parallel IRL duties!
This seems to have solved my issue. The result was the same with or without the parallel 100nF cap so I omitted that part of it. Here is a video of the result. Thanks.
Cool! The rest of the noise you're getting could probably be fixed by shielding your circuit. Try a sheet of aluminium foil under the breadboard (make sure you ground it). Maybe put a sheet of paper between them to avoid any shorts. I use this method a lot when I measure my circuits, so the measurements aren't affected by noise picked up from the surroundings.
I live at the base of a radio tower, so I've always had issues with noise. Originally I thought it was just a noisy power supply, but after moving things around my house/trying different properties, playing with faraday cages, etc I found something super weird - I had RF noise from the power supply coming in on the input lead to the pedal (but dead quiet on the power leads) AND I had environmental RF noise coming in on the power leads. Once I figured out I had two separate sources of noise that were getting demodulated into noisy funk on higher gain pedals, I trial and errored a stack until I came up with this (just a schematic from my switching PCBs that are universal across everything I build).
Right off the DC jack, there's a schottky for reverse polarity protection. Drops .3v in most situations, which means basically nothing. From there I've got a ferrite bead feeding parallel caps - while RC worked fine, I found LC in this setup was significantly better at reducing modulated RF noise and didn't result in losing an additional 0.3-0.5v. 470uf is electro, 100nf is ceramic, ferrite bead is one of the little axial jobbies from Tayda. You absolutely could add in a <100R resistor after the ferrite bead if you didn't mind losing a bit of extra voltage, but some pedals I make require clean headroom so wanted to keep supply as high as I could.
I've also added in is a RC filter going to the PCB of 10k/100pf. It's so, so far out of the audio range it has no impact on sound even with some bad input/output impedances, but made a noticeable difference in noise (especially using a couple of lower quality leads I've got). You can definitely increase this to 22k/220pf or so, but I found 10/100 were components I always had thousands of.
There's also a connection for a Pogo pin on board - basically it's a little spring loaded brass pin I use for the single connection to the enclosure (using shielded jacks). I have experimented with AC coupling the pogo pin to the enclosure (instead of connecting directly to the enclosure, it goes through a 100nf ceramic), and found that on some pedals it worked better than DC coupling, but on others it was worse.
Nothing here is particularly innovative or rare, but all together they solved something I deal with for months!
