Noise when guitar volume is rolled off?

[tommycataus]

Hey guys,

I'd like to start by saying that my problem is definitely to do with my power supply.

I've never had this problem before, but I only noticed it since I've finally decided to use a voltage inverter to power fuzz faces. When playing with any of my guitars through a number of pedals powered by filtered wall wart, or the two fuzz faces I built powered by battery, all sounds good on high gain with the guitar volume rolled off.

When using a voltage inverter to power the fuzzes from the mains, all sounds sweet as can be until the volume is rolled off. The noise is pretty loud when volume is off fully. As I said this phenomenon occurs with both the LP copies and strat copies on any configuration.

One of these fuzzes is on vero and the other is still on bread board.

The voltage inverter sections seem fine - one was a kit from Chromesphere's store complete  with MAX1044 and pcb, and the other was built from the vero layout on tagboard effects with a 1660 from Tayda. Both sound the same on both fuzzes, but are good with batteries.

Like I say the only problem is when I roll off the volume which immediately led me to believe it was the guitar,  then the fuzz...

I'm using a cheap wall wart filtered with a separate circuit featuring a 9v regulator along with a 1, 000 uf capacitor and a 47 uf capacitor as seen in Smallbearelec's Small Wart. This setup has served me well for the most part and currently works well alongside all other pedals, built or bought.

So, my questions are:

-Given the checks I have done, could this be a result of a lack of filtering on the power supply?
-If so, why does it only affect the circuit when the guitar volume is rolled off, and why only with positive ground circuits powered by an inverter?
-Am I right in thinking that the negative side of the circuit has better filtering than the positive?
-If so, what sort of additional filtering should I use and where?

I'm at work at the moment so can't offer any pics/data but will do so gladly if required. Help would be much appreciated as I've been scratching my head over this one, thanks in advance!

Tom

[tommycataus]

Also No other pedals were either in the signal chain or attached to the daisy chain adaptor when testing, it was just straight guitar-fuzz-amp with the fuzz being the only thing connected to power.

[R.G.]

Any of the issues you mention are possible, but from reading your description, there is another possible cause. The switching voltages and currents in an inverter are large compared to the other voltages and currents in the pedal. It is possible that this results from wire routing issues in the pedal.

There are some other much less likely possibilities, but try moving the wires around inside the pedal with an all-wooden probe like a chopstick and see if the sound changes any when one or another wire is moved.

[tommycataus]

Thanks for the heads up RG, my wiring leaves a lot to be desired, I will admit...

Nevertheless, poking around and pulling all wires aside, it makes no difference to the sound. Not a zip. I just don't understand why it sounds fine in every scenario except with the volume rolled off on the guitar and I want to get my head around it...

Here's the charge pump I'm using:

http://tagboardeffects.blogspot.com.au/2012/05/negative-voltage-inverter.html

Advice is to use either a MAX1044 or an ICL7660S. The ones I've used are a MAX1044 or an ICL7660CPAZ, the latter of which has not been recommended. Nevertheless in this application it makes no difference.

[karbomusic]

Every fuzz I've built does this horribly while on the bread board (with no inverter) but disappears by the time it's built and boxed. I thought it might be related to my volume pot though since I use the same one on the bread board each time but not sure. I have one on the bb doing it right now.

[tommycataus]

Thanks Karbo, that's interesting. In that case I might just go ahead and build. Incidentally I kinda have 3 stages to my builds. Stage one involves all parts wired up to the breadboard (the only constants are the switches and jack sockets), the next is to populate the pcb/vero (testing ins and outs plus power via the breadboard again), and the final stage is boxing it up.

I have been waiting on some supplies for my first etching adventure, so perhaps I should just hold fire and go with it once the enclosure is ready. Having looked closer, the 9v+ doesn't have filtering, whilst the 9v- side does. Not sure whether this may be significant...

[karbomusic]

As an FYI, my noise is pure hiss not buzz if that helps.

[amptramp]

Once you turn down the volume pot on the guitar, the output impedance of the guitar increases from the roughly 7000 ohms in series with 4 henries of inductance in parallel with a pot to a much higher resistance near the midpoint of the pot and dropping as you turn it lower - but nobody uses the low settings.  A unity-gain buffer with a low and constant output impedance may offer some relief by driving the pedal chain with more available current to swamp out any induced currents.

[italianguy63]

Advice is to use either a MAX1044 or an ICL7660S.

Tommy, Try a TC1044SEPA.  I get them for about a buck from Mouser.  (Note, the TC1044SCPA from Tayda does not seem to be the same quality, and the price at Mouser is actually better).  Next time you make an order try one.  It is a drop in replacement for the MAX.  MC

[tommycataus]

Great suggestions, thanks guys. I've just put an order in from Tayda and I'm going to have a small break from building until that arrives, but will definitely try these out. Great heads up on Mouser, too!

Thanks

[pinkjimiphoton]

tommy,
i've been fighting this demon for years too. fuzz faces don't like some pedals... whammy pedals are one, and poor power supply decoupling can be an issue too. i finally bit the bullet and use a battery in my DA. but the one i built works fine on an isolated power supply .... but is npn.

i anxiously wait to see if ya find a solution. good luck bro...

[karbomusic]

I have four Fuzz Faces on my pedal board but I don't put anything before them ever, they are required IMHO to interact directly with the guitar circuit (for anyone who doesn't already know). I even bias them against the guitar I use most with it. Do that, place a 100uF cap across the power rail, use a good adapter (I have a converted one just for PNP fuzz faces) and I have zero problems outside of the expected gain noise with no issues with the volume knob rolled down. The one I mentioned above I soldered yesterday and no problems. I never use converters though since it's much easier to just convert a wallwart to PNP and split it out for my PNP pedals, for me anyway.

This is the one that was going bananas with the volume down until I dropped it onto this PCB I made yesterday morn:

[tommycataus]

Thanks Jimi, I have a small number of Ge NPN transistors and it had crossed my mind, but I have a bunch more PNP's and I know I'll be using them

I think maybe it's a case of experimenting. I'm convinced that there is a solution and when my order comes in I'm going to shut myself away and get to the bottom of it!

Nice one Karbo, it's really heartening to know that someone has beaten the problem. I've read about the electro across the power rail although I didn't actually try it, despite reading about it before making this post. I'll add it to my list of possibles and keep exploring

You guys rock!

[karbomusic]

I have a bunch more PNP's and I know I'll be using them Wink

I think maybe it's a case of experimenting.

That's for sure with GEs. I've tossed a few that were just too noisy to even use so you might have to hand match a couple.

[merlinb]

The fuzz face is a shunt feedback amplifier where the input resistor is actually the guitar itself. When you turn down the guitar's volume, the input cap is grounded, leaving a 90 degree phase shift in the feedback loop and making the circuit probably close to instability and very sensitive to picking up interference. A cheap solution would be to put a resistor in series with the input cap (try a 10k trimpot and adjust until the noise abates).

[karbomusic]

The fuzz face is a shunt feedback amplifier where the input resistor is actually the guitar itself. When you turn down the guitar's volume, the input cap is grounded, leaving a 90 degree phase shift in the feedback loop and making the circuit probably close to instability and very sensitive to picking up interference. A cheap solution would be to put a resistor in series with the input cap (try a 10k trimpot and adjust until the noise abates).

Would that change the way the FF reacts/sounds when using the volume knob on the guitar? I've never had to solve the problem above (always goes away once I box it up) but curious.

[merlinb]

Would that change the way the FF reacts/sounds when using the volume knob on the guitar?

Everything affects everything, so yes. But not necessarily very much.

[karbomusic]

Would that change the way the FF reacts/sounds when using the volume knob on the guitar?

Everything affects everything, so yes. But not necessarily very much.

Thanks, I'm quite attached to the control I get with the interaction of the guitar volume knob and fuzz faces so wanted to ask.

[pinkjimiphoton]

tried it. you can go as high as about 50k without messing with the tone too much, or the interaction, but it may require a slightly different biasing to get the same tone... which could make the noise come back in spades.
i too agree nothing can be between the fuzz and guitar for it to work right.
i've tried doing the 100u cap across the rails, including full gospel power supplies with series resistance, diode protection, pf bleed caps, the whole 9. the answer isn't in the power supply tho.
merlin's answer is closer. but still not an absolute. 
i don't care about the "blow" from the fuzz itself. i can let my noise gates deal with that.
the problem only seems to happen when certain other pedals are present, usually digital... i have a feeling they dump a lot of hash into the power supply... and pnp. i've never had this issue with an npn fuzz, ge or si.
i even went so far as to try a charge pump to run my fav, the venerable original ff i made.
trying to keep noise down, i went with HUGE caps to filter the supply... i forget, something like 220u or 470u for the + and - rails from the charge pump (+9 side to run the LED, -9 side for the fuzz, and "ground" in the middle) and ended up with a very quiet fuzz... but with a problem. the dang thing won't come on or work until the caps are charged. about 3 minutes it takes, on battery or wall wart. you can tell it's ready when the led comes on. sound has been very consistent, and the bias stays stable despite temp changes... this thing used to need to go in the freezer sometimes during the summer, and i used to carry a heat gun with me in winter to blast the inside with thru the jack if it got too cold.. but now it's fine no matter where i throw it down. UNLESS i try and use my whammy pedal or xp all. then suddenly, the noise is back only when i turn it down.  and this is the standard kinda charge pump dealio RG turned us all onto years ago (other than the giant caps) and having a power supply stage independent before THAT.

ay yi yi.... make your head hurt.

i finally gave up on it... i have an NPN FF on my board running off the regular power supply, and use my DA first in line with a battery. no more issues.... well.... no more unexpected issues, anyways...

tommy, let us know how ya make out... i still am not really set up to do any electronics in my workshop, been too busy doing guitars (god, over 40 of them in my shop right now... i lost count 5 or 6 ago) and amps and crap.

karbo.... i hear ya. the guitar must interact correctly, or the circuit ain't worth poo

rock on brothers... peace  out

[tommycataus]

Update: I tried the cap across the power rails. 100uf made a slight difference, 220uf a bit more... 470uf was better. I decided to put a very large (3,300uf) across the rails in my power supply. Bingo!

Any changes will probably make a slight difference like you guys have mentioned. My other pedals sound fine. My thinking is that all Ge's are different and therefore there would be variation either way; so as long as it sounds good (which it does to me) then there's no problem. I can't think that it's economical to add a large electro to every fuzz.

So, can anyone see any potential problems with this?