Rotary switch for passive tone filter?

[TonyAm]

Hello,

I am making a passive tone filter for electric guitar. The rotary switch is used to switch between different values caps. Simple circuit where the guitar plugs in directly (unamplified) to the circuit then out to an amp.

Like a passive preamp? However the rotary switch I use creates a scratching sound when switching. Could anyone recommend a "smooth action" rotary switch that does not add noise to the circuit?

Greatly appreciated,
TonyAm

 

[rayma]

Is your switch break before make? That may be the problem.

 

[TonyAm]

Hello,

I'm not sure about the terminology (make/break, etc.) but it is this one:
Radio Shack 2-Pole, 6-Position Rotary Switch Catalog №: 2750034 40293985608 | eBay

Thanks for you help.
TonyAm

 

[PRR]

We usually apply a lot of 10Meg resistors to bleed-off leakage charge on the capacitors.

Post your schematic before we all go crazy with off-beat answers.

Also switching at the LOWest signal level is more prone to stray racket. We more often see the switched capacitor trick after the first tube.

 

[TonyAm]

Hello,

Yes, the 10M resistors. I have seem circuits similar to mine where 10M resistors were used however, I have seen those resistors connect to the signal line, such as here: Varitone Wiring - Gibson ES - Gibson Brands Forums

I always thought that the "bleed off" resistors should go to ground?

Here is the circuit I am working on:
The Craig Anderton 'Passive Tone Control' | Telecaster Guitar Forum

Thanks again
Tony

 

[Mark Tillotson]

However the rotary switch I use creates a scratching sound when switching. Could anyone recommend a "smooth action" rotary switch that does not add noise to the circuit?

Not a mechanical switch - they are always intermittent on switching and create noise like this especially switching caps which retain whatever charge they had from before.

One approach is using a JFET as a switch, with slow switching over several ms to reduce transients. However that requires active circuitry, but its a common technique in mixing desks etc.

Can you not use a circuit where the resistance is varied, so a pot can be used?

 

[PRR]

...I always thought that the "bleed off" resistors should go to ground?...

Or to a signal wire which eventually flows to ground.

 

[steveu]

Most likely the problem is not the switch. All mechanical controls suffer some bounce. But they only generate noise if there is DC bias across them. Be sure there is no bias voltage from the source nor from the following input. This usually requires DC blocking capacitors and bleed resistors on both sides of the control.

 

[TonyAm]

Or to a signal wire which eventually flows to ground.

Yes, I understand that now. I'll experiment with connecting the bleed off resistors to either signal or ground.

Sincere thanks,
Tony

 

[Mark Tillotson]

Most likely the problem is not the switch. All mechanical controls suffer some bounce. But they only generate noise if there is DC bias across them. Be sure there is no bias voltage from the source nor from the following input. This usually requires DC blocking capacitors and bleed resistors on both sides of the control.

If you switch a signal suddenly, there will be a crack unless you happen to time it fortunately at a zero-crossing. Mechanical switches tend to bounce and the crack may be a crackle as a result. The fast transient of switching puts energy all across the higher audio frequencies and this is very audible (especially if the signal has mainly LF energy itself). Some specialized audio switch chips and digipots exist which implement zero-crossing detection to time things without generating transients.

 

[rayma]

it is this one:
Radio Shack 2-Pole, 6-Position Rotary Switch Catalog №: 2750034 40293985608 | eBay

That is a NON-shorting switch, so it could indeed cause noise when switching.
Can you post the schematic?

 

[steveu]

If you switch a signal suddenly, there will be a crack unless you happen to time it fortunately at a zero-crossing. Mechanical switches tend to bounce and the crack may be a crackle as a result. The fast transient of switching puts energy all across the higher audio frequencies and this is very audible (especially if the signal has mainly LF energy itself). Some specialized audio switch chips and digipots exist which implement zero-crossing detection to time things without generating transients.

Modulation of the audio is going to generate a small tick at a level well below the original audio vs switching bias which is likely 20dB+ over the audio. Gating a sine wave at 90 and 270 degrees is hardly noticeably different than gating at 0 and 180. The derivative of 0 and 180 gating is just as abrupt. So zero crossing switch is a bit misguided. Probably less of an issue than absolute polarity.

Bias on a selector switch or POT is a classic issue and the result is a transient that clips the following amp.

 

[Mark Tillotson]

Gating a sine wave at 90 and 270 degrees is hardly noticeably different than gating at 0 and 180

No, the spectrogram is very different in fact, the former puts out lots of HF energy, the latter much less. This is because there is a discontinuity in the signal, whereas with zero-crossing there is only a discontinuity in the derivative of the signal. Compare the sound of a sawtooth wave with a triangle wave to hear the difference if you don't believe me. Extremely noticably different!