Phase 90 build noisy with radio signals

[ItsGiusto]

I just recently finished modding a phase 90 build with mods. After swapping a couple of pots out (1m width -> 2m width and .5m depth -> 1m depth, with appropriate adjustments to the adjacent resistors), it seems that it's a lot more noisy and picks up radio frequencies. It's possible these problems were there before and I didn't notice, but I don't know if they were.

At any rate, what might this be caused by? Is there any way for me to track down the source of the noise and eliminate it? I tried connecting the unused lugs of each new pot to the middle lugs, since I was not doing that before, but that didn't seem to help the noise.

[ItsGiusto]

Ah, forgot to mention, I'm also getting a light thumping noise that thumps faster or slower depending on where the speed knob is set.

[italianguy63]

Probably RFI interference.  Is it in a plastic or metal box?  (or still not boxed?) MC

[Mark Hammer]

One of the problems the original never anticipated was DC in the feedback path.  You will note that the feedback from the last to 2nd stages is direct, via a simple fixed resistor.  Although each phase-shift stage is supposed to be unity gain, component tolerances being what they are, teeny bits of gain can be added over 4 stages, Because of that, a tiny bit of DC added to an otherwise ostensibly ground-refenced-zeroed signal, can introduce some DC in the feedback path unless there is a cap blocking it.

I built myself 6-stage and 4-stage versions with additional controls, including the option to add more feedback than stock (or less).  I get thumping that covaries with rate when the resonance/feedback gets turned up.  I'll need to stick a DC-blocking cap in there.  One of today's projects.

[ItsGiusto]

Probably RFI interference.  Is it in a plastic or metal box?  (or still not boxed?) MC

Metal box

[R.G.]

If it's in a metal box, the RFI is getting in by conduction from the cables, nearly always from the input.

If the box is already "grounded" by the jacks, put a 100R in series followed by a 47pF to 100pF >>ceramic<< capacitor to signal ground at the input.

A simple capacitor across the input jack may work - but it can also tune in a different station without some series resistance.

[ItsGiusto]

If it's in a metal box, the RFI is getting in by conduction from the cables, nearly always from the input.

If the box is already "grounded" by the jacks, put a 100R in series followed by a 47pF to 100pF >>ceramic<< capacitor to signal ground at the input.

A simple capacitor across the input jack may work - but it can also tune in a different station without some series resistance.

Thanks, I'll try this!

How do I know if the box is "grounded" by the jacks? What I currently have is this:
The input jack is a TRS. The sleeve goes to the 9v input ground, and the ring goes to the circuit ground.
The output jack, I left the sleeve lug with no wires attached - I was told that the contact with the box would ground the sleeve, and that adding a wire to ground would only cause ground-loops. Does this sound alright?

Also this is the build I did (though I modified it further):
http://tagboardeffects.blogspot.com/2014/05/mxr-phase-90-with-script-switch.html?m=1

There's already a 10k resistor at the input. If I just add the ceramic cap to ground after that one, will that do the trick?

[GibsonGM]

It seems that you probably do have the box grounded...you can check by using your DMM set to continuity (the beeper)...one probe on box, one on the battery " - " terminal, with a jack plugged into the input to be sure it is all connected.  Should beep.

A lot of people don't connect the output jack, and hope the box grounds it...the logic is there, but...if that jack ever rotates loose, all hell is going to break loose.   BBBRRRAAAAAAAPPPPP!!!!!!! until you kill it and try to figure out what's wrong.

I add the wire across to the input jack's ground, for that reason, and if there is some 'trouble' that occurs, then I would remove it.  Tho I have not had any trouble at all doing it, even on high-gain pedals...the circuits we build just seem too small to really develop 'ground loops' this way.

Your choice how you do it.

I'd try the cap to grnd from input jack first (right across the lugs is ok, that is what many ppl do) and see what happens.   If not good, remove input wire from jack, add 100 ohm resistor, then small value cap to ground....the input wire continues on to the board input...so the 100R is in series with the input wire, but that cap goes from where they connect to ground, see?

Input >>100R>>>wire to board input
                       ||
                     47p - 100p cap
                       ||
                      Gnd (can be the jack ground)

[ItsGiusto]

It seems that you probably do have the box grounded...you can check by using your DMM set to continuity (the beeper)...one probe on box, one on the battery " - " terminal, with a jack plugged into the input to be sure it is all connected.  Should beep.

A lot of people don't connect the output jack, and hope the box grounds it...the logic is there, but...if that jack ever rotates loose, all hell is going to break loose.   BBBRRRAAAAAAAPPPPP!!!!!!! until you kill it and try to figure out what's wrong.

I add the wire across to the input jack's ground, for that reason, and if there is some 'trouble' that occurs, then I would remove it.  Tho I have not had any trouble at all doing it, even on high-gain pedals...the circuits we build just seem too small to really develop 'ground loops' this way.

Your choice how you do it.

I'd try the cap to grnd from input jack first (right across the lugs is ok, that is what many ppl do) and see what happens.   If not good, remove input wire from jack, add 100 ohm resistor, then small value cap to ground....the input wire continues on to the board input...so the 100R is in series with the input wire, but that cap goes from where they connect to ground, see?

Input >>100R>>>wire to board input
                       ||
                     47p - 100p cap
                       ||
                      Gnd (can be the jack ground)

Thanks, I'll try that out.

Interesting to note that the circuit only has rf interference when not in bypass. When in bypass, there's no radio or noise.

[GibsonGM]

Yeah...there are a few things at play when the circuit is engaged...there is the input impedance/reactance going on too...if you get resonant and happen to have the right RF hanging around, you make the circuit 'appealing' to it...then it is amplified too.   

The length of wires, parts crossing each other, poor grounding, etc...all work to make the right conditions to make your 'thing' a good radio!    The resistor/cap is just a low-pass filter, which will send really high frequencies to ground at the input.   If you made that cap progressively bigger, you'd send lower and lower frequencies to ground as you did so.   Those low pF values will keep your discrimination up in the RF zone...

In bypass, that junk just goes thru, unamplified, or is shunted to ground easily. 

[italianguy63]

I used to have a Fuzz Face that liked to pick up the local AM Mexican channel... way creepy and funny too.

[reddesert]

If it worked before and is noisy now, I would also look for a poor solder joint or a connection that got damaged when being taken in/out of the box. It sounds like the mods you made were to the LFO, which wouldn't easily explain why the audio path got noisy.

I don't believe that ground loops are a particularly serious problem in effects powered by 9 V DC, and think that it is a good idea to ground both jacks. There are some pieces of pedal lore that appear to have come over from tube amp practices without accounting for the differences in the operating conditions.

[ItsGiusto]

If it worked before and is noisy now, I would also look for a poor solder joint or a connection that got damaged when being taken in/out of the box. It sounds like the mods you made were to the LFO, which wouldn't easily explain why the audio path got noisy.

I don't believe that ground loops are a particularly serious problem in effects powered by 9 V DC, and think that it is a good idea to ground both jacks. There are some pieces of pedal lore that appear to have come over from tube amp practices without accounting for the differences in the operating conditions.

Unfortunately, the cap and resistor didn't actually help too much with the noise. Maybe with the radio itself.

Could the pots themselves I put in have been the cause of the noise? They're probably pretty cheap.