Multiplexing audio sources?

Just to clarify

so what will the gain of C be when A is on, will it be muted or at unity gain, while A has a different gain

martin

My interpretation of what you want is called a "priority" system rather than a multiplexed system. As I understand it, you have four audio sources called A, B, C, and D.

You only want to hear D if there is no audio present at A, B, or C.

You only want to hear C if there is no audio present at A or B.

You only want to hear B if there is no audio present at A.

You always want to hear A.

Therefore, Channel A is always connected to a summing port on your amplifier.

Channel A also needs a detector of some sort to tell whether or not audio is present and output a digital signal if there is audio. Generally you can use a comparator with the trigger level set to whatever you say is "audio present" and then some sort of rectifier-capacitor with a time constant you select to say that there is no more audio present on this channel.

Channel B needs the same thing, but it also needs a gate in series with the audio. Google "CMOS" and "4066" to see a typical analog-digital switch. The digital signal from channel A needs to be connected to gate this channel off when signal is present at A.

Channel C needs the same thing, but you need an OR function from both channels A & B to gate the signal. It can be as simple as two diodes or you can use digital logic.

Channel D needs no detector, but it needs a gate that is OR by channels A, B, and C.

Jim

I have a need for one of these myself. I would suggest investigating "energy detect" circuits used in voice band modems. My recollection of how this was done in the analog days was to rectify the AC coupled input signal, follow it with a Schmidt trigger, then the averaging circuit of your choice. It can be as simple as taking the digital signal from the Schmidt trigger and RC filtering it. Then use one more Schmidt trigger after this averaged circuit to make the final decision.

Please post what you come up with. You are really just gating the signal, not amplifying them. The amplification, or more accurately level equalization, should be done prior to the gating circuit.

You might also find relevant information on "ducking" circuits, though ducking only attenuates one of the signal based on priority.

You should also consider how to do the switching without creating clicks as the muxes are switched. I don't think this is a simple project if done well.

Thanks, Jim!

I actually saw a sample ducking circuit that used NOR gates. This gave me an idea.

I'm going to try and draw a block diagram of what I had in mind, and hopefully, post it soon for some more critiques.

Muchos Gracias!

RST Eng> My interpretation of what you want is called a "priority" system rather than

Hey miso,

Thanks a lot!

I got plenty of ideas by looking up your suggestion of ducking circuits. That's the first time I've heard of it.

snipped-for-privacy@sushi.com wrote:

Hi martin,

If a higher rank signal is present, then all other sources are muted.

Thanks!

mart> >

inputs (A,B,C, an D)

So, for each channel you need a detector (amplifier, rectifier, threshold detector) feeding a retriggerable monostable (pulse stretcher to keep quiet passages from dropping the signal).

Send all the logic-out from the monostable to a priority encoder (CD4532 is a CMOS eight-input chip that does this), which performs the which-is-highest-priority function and has binary number (3 bits) output.

Then use the 3-bit output to address an analog switch to select between the inputs (CD4053 is an eight-bit switch that can do this).

Since ALL inputs might have no signal, it would be worthwhile having a resistor to ground so as to keep the quiet-noise to a minimum.

Cool idea; I might use it. But, I'm not sure what priorities I'd give that would be livable; maybe DVD first, CD second, TV third, FM fourth?

inputs (A,B,C, an D)

I was thinking of using several scanners (radios) with a priority system. If you ever ran a scanner, it is a bit annoying if it locks on a signal that is kind of interesting, but not your favorite signal. The scanners have a priority mode, but what that does is interupt what even the scanner is locked on to sample the priority frequency, which might not be active. So you would park say 4 scanners on 4 different frequencies, but hear them in the order of your priority.

Umm... I gibberished that a bit; the CD4051 is the part, and it's an eight-way switch (an eight-to-one multiplexer analog switch).

As for the quiet input ('cuz I just HATE the hum that a high impedance on the input seems to generate), you can just ground the lowest-priority channel, and strap its detection bit to always ON. That loses one channel (so it's a seven-way solution). You can also use the group select output on the priority encoder; if no input is detected, it goes low, and you can use that to turn on a (for instance) photoresistor-type optocoupler to ground the output and mute the non-signal.

Thanks for the info all! I haven't managed to fully grasp most of the other tips. But I was wondering if something like this can work:

I figured I can just redirect the other signals to ground via a BJT transistor if a priority signal is present. Also, I've been playing with the PIC microcontroller before, and I think I'll be able to use it along with the MAX4562 clickless audio switch.

Thanks again!

a bit; the CD4051 is the part, and it's an

that a bit; the CD4051 is the part, and it's an

I think you need a Schmidt trigger after the rectifier to do the energy detection.

A uP seems overkill for the application, and it will generate noise.

Dumb question here, but thinking outside the box, can't you make an audio mux using an inverting op amp in the summing configuration? What you would do is replace the input resistors with two resistors in series, then use an analog gate to pull the unwanted signals to ground at the center tap of the resistors. What you are trying to do is set up a circuit that doesn't make a click due to charge injection.

Hi miso,

You are right about the microprocessor adding noise to the circuit. That's why I'm going to have separate ground planes connected at one point near the supply. I figured the microprocessor will give me access to other functions.

I don't see anything wrong with a summing amplifier. I was actually thinking of using it after the MAX 4562 device.

Thanks!

that a bit; the CD4051 is the part, and it's an

the energy