Figaro noisy on PS

[stallik]

I built a Figaro (Galileo) today and it appears to work well except it picks up noise from the PS. Tried it with a 1 spot and my own build (which is actually quieter - go figure) No other pedals in the chain and its very quiet on battery.

Would someone above my paygrade mind casting their eyes over the schematic and letting me know if you think this circuit should be prone to this behaviour? I have many other high gain pedals which all remain quiet on the same supplies

[antonis]

Although schematic is a bit blurry, I think you need a +9V LPF, Kevin..
Resistor & Cap placed AFTER (upstream) Q5 output buffer with +9V from power jack connected on Q5 Drain..
(no need to "polute" high gain stages with current hungry Source follower fluctuations..)

Like below..


P.S.
Reverse polarity protection diode should also be trasfered to +9V point, of course..

[tootsMcgee]

What's the size of C12 and C18?

I'm a newbie but I *think* that putting a resistor like 100R in line with 9V before it hits the cap is important since it gives the power filtering (low pass filter) a known value...otherwise it depends on the series resistance of the power supply which can vary. This is totally not something I read and am regurgitating blindly. It does kinda make sense though, at least in my head.

p.s. Really nice clean wiring!

Edit: yeah, what antonis said!

[ElectricDruid]

I was wondering about C12 and C18 too, Toots.

And if it was me, I'd put the LED connection *outside* the LP filtering, direct from the 9V point.
The filter serves to separate the audio part of the circuit from the LED current (a bit, at least), and it should be more directly connected to the power supply to minimise clicks when switching it on or off.

PS: It *is* a very neat build, Kevin. Nice work.

[antonis]

I'd put the LED connection *outside* the LP filtering, direct from the 9V point.

Eagle eye, Tom..
(mine was restricted on reverse polarity diode, only..)

[stallik]

680n & 47uf are the cap values.

Antonis, thanks for your suggestion, I'll probably understand it better when I have a screen larger than this phone 

Tom, I've actually wired the led directly to the +9v supply ignoring the provision for it on the pcb.

[antonis]

680n & 47uf are the cap values.

Not bad but, as already said, no bueno without series resistor..
(resistive part of LPF leans upon power supply output impedance which can be both small - as it ought to be for a self-recpecting power supply- and variable, as well said by tootsMcgee..)

[ElectricDruid]

Tom, I've actually wired the led directly to the +9v supply ignoring the provision for it on the pcb.

Oh good, it's not just me who gets twitchy about stuff like that then!

[stallik]

Tom, I've actually wired the led directly to the +9v supply ignoring the provision for it on the pcb.

Oh good, it's not just me who gets twitchy about stuff like that then!

Not so much twitchy, I usually do it for practical (OCD) reasons. Placing led where I want, getting to the resistor easily, disconnect to test for issues, habit.

Think I prefer twitchy...

[stallik]

Progress. In fact a solution really.

The series resistor may be missing from the schematic but it's present on the board. Decided to try an extra filter attached to the dc socket and stumbled on the
Huminator
by Beavis. The original article appears to be gone but pinkjimiphoton posted his interpretation.

Works great! PS now as quiet as battery 

Must find out why it works but am thinking that I'll build another Huminator to keep in the tool kit

[PRR]

why it works

It is a plain R-C filter with extra trimmings.


R1 C1 do the basic filtering. Frills: C2 extends filtering to supersonics (radio). D1 tries to kill a reverse-connected power supply. R2 D2 tells you you got power. (With LEDs from the 2020s you can go up to 10K at R2 unless you play noontime shows outdoors.)

100r versus 100uFd is "only" 17Hz. Low, but it is a very slow cut-off. It will cut 60Hz about 7:1, which is nice, but not "wow!". Also 10mA load in 100 Ohms is a 1 Volt drop, and many pedals suck more than that or change tone at 8V.

[stallik]

Clear explanation PRR thanks.
Rather than asking why this circuit works, I should have asked why the original r-c circuit on the board did not.
Perhaps r c values or perhaps a build error on my part

[ElectricDruid]

Rather than asking why this circuit works, I should have asked why the original r-c circuit on the board did not.

Well, as Antonis pointed out, the far-right transistor stage is wired into the power supply *before* the filter, so it's seeing no benefit whatsoever. Which somewhat nullifies the whole thing. So it could be that...

[stallik]

Try as I might to trace this board in order to implement Antonis's suggestion, the 3 pots on the back of the board obscure too much and I'm loathe to start bending them out of the way to see where the traces go and desoldering them is going to be painful. Even if they were removed, the board is still quite densely packed.

From what I can see, the board does not follow the schematic exactly - starting with the 100r at the 9v+ input...

I'd like to get this one right so I've ordered another board to make tracing easier. Seems a little excessive perhaps but the pedal is a keeper and I'd like to get it right rather than just fixed.

[ElectricDruid]

Well, the circuit would certainly make more *sense* if that schematic were wrong and the filter is actually the first thing after the power input, along with the LED.

But given how PCBs are designed these days (from the schematic in a CAD package) I'm not sure I understand how you'd have a PCB and a schematic that were different.
I suppose there are ways - I've sometimes done a "display copy" of my schematic, which might include the 3PDT switch and jack sockets to make the diagram clearer, although on the version that the PCB is designed from, those are simply off-board connections.

[stallik]

The copyright date on the pcb is 2013 so perhaps it predates modern techniques? Or, as you say, it's a display copy or, most likely given how many must have built this without complaint, I've made a mistake which I've yet to spot

[stallik]

Today, the new pcb arrived. This time, it's completely white making it rather difficult to see where the traces are going. Done my best, now to compare it to the schem. My first time doing it so it'll be slow going and I can't be at all sure that I haven't made any mistakes with this image

If it's wrong to show such an image of a commercial pcb, please let me know and I'll delete it

[stallik]

Thanks guys! Of course, you were right as was the schem. What was amiss was indeed the resistor rendering the filter useless.

Before hacking up the board, I simply inserted 100r in series with 9v+ at the dc socket and silence prevails.

[antonis]

Before hacking up the board, I simply inserted 100r in series with 9v+ at the dc socket and silence prevails.

Good but you should insert that resistor AFTER LED's CLR..
(no need for extra voltage drop due to LED's current..)

[stallik]

Thanks Antonis but I didn't use the pcb led provision. Instead, I wired the led directly to the 9v socket via a separate CLR. Does that count?