Zero Signal audio switching

[gtrdude]

Hi there,

I have been reading Kevin O'Connor's "The Ultimate Tone" which is a guide to building and modifying tube guitar amps. There is a nice little section on audio switching. It details switching using various components like relays, CMOS, Jfets, mosfets etc. I can understand the parts about series and shunt switching, but I'm not quite sure about zero-signal switching.

Is this just a combination of series and shunt switches to achieve a higher level of headroom between on and off states?

**broken link removed**

What is a "zero-signal point"?? as you can see they are marked with a "0" in the diagrams above. I don't understand the significance of this.

Also what is a "virtual earth"? I'm having trouble understanding that also.

If anyone can also point me to information regarding how to achieve very clean audio switching such as in high end hifi systems I would appreciate it. I am just a beginner so be gentle! :lol: Thanx

 

[crust]

I could very well be mistaken, but judging by the look of the circuit, "virtual earth" seems to be the same idea as "virtual ground." Take a simple op-amp inverting amplifier circuit. In this case, the + input of the amp has a path to ground. The idea of an op-amp is that 0 (or about as close to 0 as possible) current flows into the - terminal of the amp. This puts the + and - terminals at the same potential since the + terminal is grounded. This is referred to as a virtual ground.

 

[Roff]

Virtual earth, or virtual ground, is a concept that is essential to the understanding of feedback amplifiers. The drawing below illustrates the concept for a typical operational negative feedback circuit. The resistor values and gain value are just examples. Different op amps have different gain values, but they are all very high. Remembering that the input current is very nearly zero (another characteristic of op amps), try different values of voltage for V(-). You will find that the one on the schematic is the only value that will work.

Note that the closed loop gain here is almost exactly -R1/R2, in this case, -2. If the gain of the op amp changed by 50%, the closed loop gain would still be -2, and V(-) would still be virtually zero. The fact that the closed loop gain is almost independent of op amp gain also means that distortion will be nearly zero.

With that as a basis, consider an analog (solid state) switch in series with R2. If you put it on the left (source) side, it will have basically the full signal swing on both terminals when it is ON. If you put it on the right (virtual ground) side of R2, it will have almost zero volts on both terminals when it is ON. The ON resistance of analog switches is generally not constant over the allowable input signal range. If the resistance varies as a function of the signal, as it will when the switch is on the source side of R2, the gain will vary as a function of instantaneous signal voltage. Since Vout=-R1/R2, this will cause distortion. If you instead put the switch on the virtual ground side of R2, distortion will be minimized, since the switch will have almost no signal voltage on its terminals when it is ON.

 

[spuffock]

The idea of zero point switching is to eliminate clicks and pops caused by suddenly switching a signal at a random point. If you take the audio signal and apply it to a comparator or an op-amp without feedback. the result will be a nice squarewave with its edges at the zero point of the audio. This squarewave is applied to the clock input of a D flipflop (4013) . and the gating signal to the D input. The signal appearing at the Q output of the flipflop will be a version of the gating signal with its edges delayed to coincide with the audio zero crossing, positive going. Using this signal to drive the audio switches will eliminate clicks caused by random switching, and thumps caused by residual dc components of the gated bursts.

 

[Roff]

Spuffock, I agree. I just couldn't see the relationship between the zero-crossing switching that you described and the schematics that gtrdude posted. That's why I wrote the epistle above. Both techniques would be useful in a high-performance audio switcher.
To achieve popless switching, wouldn't zero-crossing switching need to be used when dropping a signal as well as when adding a signal? I guess your description takes this into account.

 

[spuffock]

You're right, Ron. The flipflop delays all edges until the next time the audio crosses zero, going positive. Negative going edges at the clock input do nothing.

 

[gtrdude]

Ok, I think I'm starting to understand. There is some circuitry that controls the switching. It determines when the next zero-crossing is, and when that occurs it cause the switches to change channels.

Could somebody please explain the operation of this setup in more detail please? An example schematic would be wonderful :lol:

BTW, thanx for the replies so far. Great help!

 

[Jaw174]

reverse

how would i flip the above circuit to have one input and 2 outputs, using opamps

 

[audioguru]

An opamp can have one or many inputs.
An opamp can have one or a few outputs.

 

[Nigel Goodwin]

You're asking about switching in high quality amplifiers?, a high quality amplifier will use a mechanical switch - NOTHING ELSE - nothing else is needed, and would only lower the quality.

 

[Jaw174]

no pop

but i would like to have them switch with no "pop" noise... its a line level signal, but im sure there wil be an electrical difference between them... how can i do it without the pop

 

[audioguru]

No switching pops:
1) Make the inputs and the switch all have exactly the same DC voltage with coupling capacitors and a resistor to ground.
2) Fade the existing and new signals down to zero, switch the input then fade-up the new signal.

 

[Nigel Goodwin]

but i would like to have them switch with no "pop" noise... its a line level signal, but im sure there wil be an electrical difference between them... how can i do it without the pop

As Audioguru suggests, as long as there is no DC voltage where you switch(and you shouldn't be switching where there is!), then there will be very little 'pop' as you switch. Making it completely pop free requires a brief mute as you switch, but any electronic method is really going to reduce quality.

Going back a LONG time, the old HiFi manufacturer Leak continued to use rotary switches on their amplifiers, they didn't consider push button switches of high enough quality for their products.