# 4066 and audio

#### docb

Feb 11, 2010
131
Hi,

Do I need buffer ICs, or input caps or anything if I just want to switch mic level audio on and off with a 4066?

#### (*steve*)

##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
You must ensure that your audio signal is between Vss and Vdd (the power supply).

If the signal is generated by your curcuit, it will already be (most likely) between the supply rails, so it may be DC coupled to the 4066.

If the signal is coming from an external source, you may wish to have an input capacitor, and then bias the input of the 4066 at half the supply using a pair of high value resistors.

For similar reasons, you may also need an output capacitor (or not).

#### docb

Feb 11, 2010
131
I don't fully understanding all you are saying, so please correct me if I'm wrong.

Vdd is 5v, and audio is guitar, or possibly line level, so all well under 5v.

By DC coupled, you mean direct, less the cap. What value is the norm for this AC coupling cap? Is a .01 disc cap acceptable for mic/line/guitar Z?

I don't really follow you on the biasing. Since the 4066 is just a switch, would I bias *only* to reduce clicks in the audio? If so, how do I do bias to 2.5v? Two resistors as a voltage divider to get 2.5v, on the IC side of the cap, and the audio input from the other side of the cap?
One circuit I saw uses 1meg to Vdd on the chip input, and then a 1 meg on the other side of the cap, at the input.

Thanks!

Last edited:

#### BobK

Jan 5, 2010
7,682
An audio signal is typcially going to go above and below ground. So the positive part of the signal may be < 5V, but the negative part is below 0, putting it out of range. The biasing brings the signal up to the midpoint (2.5V) when the signal is 0, so it allows the signal to go either positive or negative by 2.5V.

Bob

#### docb

Feb 11, 2010
131
An audio signal is typcially going to go above and below ground. So the positive part of the signal may be < 5V, but the negative part is below 0, putting it out of range. The biasing brings the signal up to the midpoint (2.5V) when the signal is 0, so it allows the signal to go either positive or negative by 2.5V.

Bob

And the value in this is that I can put higher signal level without distortion?

#### jackorocko

Apr 4, 2010
1,284
And the value in this is that I can put higher signal level without distortion?

no, it is so the switch will work over the full range of voltage swing in the audio. As bob said, right now the switch would be off for any voltage lower then ground potential, which is half the audio.

#### docb

Feb 11, 2010
131
While I agree in theory about the negative waveform, in practice the 4066 seems to switch line audio just fine, unbiased. I have it working on a breadboard, so that's a bit confusing.

#### (*steve*)

##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
The datasheet says this:

For proper operation, Vin and Vout should be constrained
to the range VSS <= (Vin or Vout) <= VDD.

And whilst you can feel at liberty to ignore the advice, many people don't.

#### docb

Feb 11, 2010
131
It seems easy enough to just add two resistors and and a cap to the input, and then a cap on the output, so I will.

I think line level is only about 2v, so with Vcc at 5v, I still have some headroom even after biasing to 2.5, yes?

#### (*steve*)

##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
A lot depends on what sort of 2V you have.

If it's 2V peak to peak, then you will have plenty of headroom.

2V RMS would be 5.7V peak to peak (for a sine wave) and that signal would be too large.

#### docb

Feb 11, 2010
131
Just regular line level audio.

#### (*steve*)

##### ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
OK, line level should be under 5V peak to peak, so you should be fine.

Replies
21
Views
8K
Replies
31
Views
5K
Replies
4
Views
4K
Replies
3
Views
2K
Replies
16
Views
4K