1980 EHX Small Stone - Power Filter Issue?

[clarisso11]

Hi everyone,

I'm currently working on an EHX Small Stone Phaser, made ~1980. I think this is close to the "Issue J" circuit, with "EH1048" branded ICs - schematic attached.

The problem:
On battery, the pedal works perfectly. As soon as the 3.5mm power adapter is connected, noise is introduced. It sounds like some kind of electrical noise, scratching like a bad contact, but seemingly random.
I'm using an isolated 9V power supply (Cioks DC7).

Tried so far:
- Replaced power jack with brand new one
- Used different power adapter
- Tested all electrolytic capacitors with Peak Atlas ESR meter, all look good
- Confirmed that noise is independent of input of output jacks

Question:
I'm struggling to understand why this is quiet on battery, but not the power adapter. Is this something to do with a problem in the power filtering? What components should I be testing?


Many thanks in advance, any help would be greatly appreciated!

[Helmholtz]

Are there additional components (diode, cap) in the external power wiring from jack not shown on the schematic?
Do you get effect with ext. power or only noise?

For a test you could connect your ext. supply to the battery clip.

[clarisso11]

Hi Helmholtz,
I get effect with the external power supply connected, sounds the same as with battery, but with the added noise.

After your message I decided to take a photo of the components side, and noticed a discoloured resistor. Couldn't entirely make out the bands, but saw from photos online that it should be 47R.
The resistor still reads 47R, but I replaced it anyways - no improvement, but maybe sign for damage somewhere else, maybe due to a wrong power supply connected at some point?

Will try connecting external power to battery clip later, I suspect it will introduce the same noise.

[Mark Hammer]

This may be a terrible underestimate of your knowledge, but I hope you realize that adapters ending in a 3.5mm phone plug use tip as positive and shaft as negative. This contrasts with 2.1mm barrel plugs that use shaft/outside as positive and tip as negative/ground. Both types of plugs require the ability to push a contact out of the way to switch from internal battery to external power. For mini phone plugs, it needs to be the tip, whereas for barrel plugs, it needs to be the shaft.

If it were my pedal, I would enlarge the hole used for the external power jack and simply install a "proper" 2.1mm jack so that supplies like the CIOKs can be used without having to employ adapters or worry about polarity. The old style EHX folded sheet-metal boxes are stubborn to machine, because they are steel and not aluminum, but they will provide enough room for the new jack, regardless of whether you use an "innie" (attaches from the inside) or an "outie" (nut is attached from the outside).

[clarisso11]

This may be a terrible underestimate of your knowledge, but I hope you realize that adapters ending in a 3.5mm phone plug use tip as positive and shaft as negative. This contrasts with 2.1mm barrel plugs that use shaft/outside as positive and tip as negative/ground. Both types of plugs require the ability to push a contact out of the way to switch from internal battery to external power. For mini phone plugs, it needs to be the tip, whereas for barrel plugs, it needs to be the shaft.

Always good to check these things! I have had this and a previous version before, they can be powered with a regular 9V power supply, using the positive tip 3.5mm adapter. If it was reversed, I doubt it would produce any effect at all, but it sounds as it should - just with added noise.
I agree that it's good to swap in the modern socket, but would like to keep it as original as possible.

[Mark Hammer]

When a "vintage" device is modified in some manner that alters its functioning, and makes it hard to know what "the original" was like, that's one thing. But there are tens of thousands of these things floating around. I see little reason to want to maintain what may have been necessary 40 odd years ago, but is going to be a nuisance for the foreseeable future. I doubt we shall ever see a day again where players have the necessary adapters just "hanging around". Go 2.1mm and be done with it. You'll never have to worry about inserting the power plug before applying power to the adapter ever again. My only reluctance would be if you lack the tools or mechanical skills to install the new jack properly.

[clarisso11]

When a "vintage" device is modified in some manner that alters its functioning, and makes it hard to know what "the original" was like, that's one thing. But there are tens of thousands of these things floating around. I see little reason to want to maintain what may have been necessary 40 odd years ago, but is going to be a nuisance for the foreseeable future. I doubt we shall ever see a day again where players have the necessary adapters just "hanging around". Go 2.1mm and be done with it. You'll never have to worry about inserting the power plug before applying power to the adapter ever again. My only reluctance would be if you lack the tools or mechanical skills to install the new jack properly.

I guess it just comes down to personal preference.. In any case, wouldn't my issue persist, even if I changed the power socket? Since the circuit would see the same 9V, just through a different socket.

[Helmholtz]

Does your ext. power jack actually disconnect the battery?

[clarisso11]

Does your ext. power jack actually disconnect the battery?

I've just checked this with a continuity tester, inserting the external jack breaks the battery-board connection as expected.

Could it be that whatever overheated the first resistor damaged a component down the line? I might have to try replacing some of those next, although I'm not sure where to start..

[Helmholtz]

Do not assume, try the battery clip test.
We need to know if the ext. power supply is clean.

Also the schematic obviously is not complete.

[clarisso11]

Do not assume, try the battery clip test.
We need to know if the ext. power supply is clean.

Good point. I found a battery clip adapter and tested it, same noise is present. It's like a constant very fast ticking that doesn't change speed with the rate knob, combined with the occasional scratching noises. Effect is loud and clear.

[Helmholtz]

Good point. I found a battery clip adapter and tested it, same noise is present. It's like a constant very fast ticking that doesn't change speed with the rate knob, combined with the occasional scratching noises. Effect is loud and clear.

Try adding a 100µ cap between +9V and ground.

How is the 47R wired?
If it is in series with the positive supply, the 100µ cap should be connected after the resistor.

[Mark Hammer]

I guess it just comes down to personal preference.. In any case, wouldn't my issue persist, even if I changed the power socket? Since the circuit would see the same 9V, just through a different socket.

My suggestion stemmed from an assumption - perhaps erroneous on my part - that the pedal was working with a battery, but not working with an external supply simply because the external supply may have been providing positive to negative terminal. If the non-functioning with an external power source has to do with issues other than polarity, feel free to ignore my suggestions.

Your unit is the later 5-chip issue. The earlier 6-chip issue had a unique and clever quirk. Part of what makes the Small Stone so musically pleasing is that it uses a "hypertriangular" LFO waveform, that goes sinusoidal as it approaches the lowest part of its sweep and goes triangular as it ascends, making movement of notches in the lower part of the spectrum more noticeable. The "Color" switch does several things at once. It not only adds more feedback but alters the overall sweep rate AND shape of sweep, such that one position is more resonant, slower, and hypertriangular in sweep, while the other is somewhat faster, less resonant, and more triangular.. Essentially optimized for faster and slower sweeps. Having tinkered with it, I will say that hypertriangular is not only "better" for very slow sweeps, it is downright annoying for faster bubbly sweeps. The neat trick that the earlier 6-chip issue included was a resistor/capacitor pair on the LFO output that would begin to add lowpass filtering to the LFO waveform above 1.6hz, such that things would get automatically more "triangular" and sweep width reduced a bit, when you turned the sweep rate way up in either Color setting.

[clarisso11]

Try adding a 100µ cap between +9V and ground.

How is the 47R wired?
If it is in series with the positive supply, the 100µ cap should be connected after the resistor.

I connected the 100µ cap between +9V and ground, before the 47R, hadn't seen your question about the resistor yet.

It completely removes the noise, problem solved - thanks a ton!

Should I still go ahead and connect it after the 47R? Pretty sure the resistor is wired in series.
Edit: I tried it both ways, noise only disappears when the capacitor is connected before the 47R resistor.

[clarisso11]

My suggestion stemmed from an assumption - perhaps erroneous on my part - that the pedal was working with a battery, but not working with an external supply simply because the external supply may have been providing positive to negative terminal. If the non-functioning with an external power source has to do with issues other than polarity, feel free to ignore my suggestions.

Your unit is the later 5-chip issue. The earlier 6-chip issue had a unique and clever quirk. Part of what makes the Small Stone so musically pleasing is that it uses a "hypertriangular" LFO waveform, that goes sinusoidal as it approaches the lowest part of its sweep and goes triangular as it ascends, making movement of notches in the lower part of the spectrum more noticeable. The "Color" switch does several things at once. It not only adds more feedback but alters the overall sweep rate AND shape of sweep, such that one position is more resonant, slower, and hypertriangular in sweep, while the other is somewhat faster, less resonant, and more triangular.. Essentially optimized for faster and slower sweeps. Having tinkered with it, I will say that hypertriangular is not only "better" for very slow sweeps, it is downright annoying for faster bubbly sweeps. The neat trick that the earlier 6-chip issue included was a resistor/capacitor pair on the LFO output that would begin to add lowpass filtering to the LFO waveform above 1.6hz, such that things would get automatically more "triangular" and sweep width reduced a bit, when you turned the sweep rate way up in either Color setting.

Thanks for this explanation, very interesting stuff!