4066 switching

[WhenBoredomPeaks]

Hi guys.

I am experimenting with using a 4066 switch as a sequencer with 4 outputs. So far i could make it work with leds, but it have problems with audio signal.

The first is the usual stuff: clock noise, but i could try to fix it later.

Second: when i pluck the strings quite hard i get horrible saw wave fuzz like sounds but only at the peak/"attack" of the guitar signal. The crazy thing is that it makes it both on and off. (i mean the internal state of the switch in the ic)
It's like the internal resistance of the switch is so low that higher signal can get through it. (my guitar pickups are low/med output single-coils with only the guitar connected so the peak of the signal is can't bee higher than about 1V and the datasheet specifies switching of signals up to ~15V)

Here's a simplified "layout" of what i am talking about:




(other interesting phenomenon: when i connect the 9V to the input and a LED going to ground at the output, my LED survives that. My common sense tells me that there should be 9V on the LED after i triggered the switch open with the control pin but there is actually 2-3V there. Maybe my IC is defective?)

[R.G.]

I am experimenting with using a 4066 switch as a sequencer with 4 outputs. So far i could make it work with leds, but it have problems with audio signal.

The first is the usual stuff: clock noise, but i could try to fix it later.

Second: when i pluck the strings quite hard i get horrible saw wave fuzz like sounds but only at the peak/"attack" of the guitar signal. The crazy thing is that it makes it both on and off. (i mean the internal state of the switch in the ic)
It's like the internal resistance of the switch is so low that higher signal can get through it. (my guitar pickups are low/med output single-coils with only the guitar connected so the peak of the signal is can't bee higher than about 1V and the datasheet specifies switching of signals up to ~15V)

Here's a simplified "layout" of what i am talking about:

(other interesting phenomenon: when i connect the 9V to the input and a LED going to ground at the input, my LED survives that. My common sense tells me that here should be 9V on the LED after i triggered the switch open with the control pin but there is actually 2-3V there. Maybe my IC is defective?)

Welcome to CMOS switching. The chip is probably working as designed.

Some basics of working with CMOS switches for those who do not read geofex.

CMOS switches can only switch between their power supplies. Put ground on the bottom side and 9V on the top side, and the switch outputs cannot go beyond those. In fact, it can't go to the power supplies. There are internal protection diodes which prevent the inputs and outputs (and these are the same thing on the analog switched pins) from being pulled beyond the power supplies by more than a diode. This fact causes your interesting phenomena of distortion on pick attacks. It's also one reason why (as shown in my article at geofex on CMOS switching with the CD4053) audio should be biased to near the midpoint of the power supply for the CMOS chip before being applied to the chip in/out pins; otherwise, the protection networks or power supply range can cause distortion on peaks.

The other reason for biasing to near the middle of the power supply is that CMOS switches consist of one P-channel and one N-channel in parallel as the switch elements. The P channel conducts the signal when the voltage is near ground, the N-channel conducts when it's near +v. Both have some clock feedthrough. But near the center of the power supply, the clock feedthrough approximately cancels, leaving you a much smaller mess to clean up.

The internal resistance of the 4066 is a few hundred ohms, and it varies by how close to the power suppies the signals are. I'll have to look at your circuit in more depth to know what the LED issues are, but be aware that an LED itself represents a 1.5V - 3V diode clamp. This is suspiciously like the issue you're seeing, so it may be that you're measuring the LED voltage.

[flo]

And for those that want to read Geofex:  
"Bypassing and Switching with the CD4053 CMOS Analog MUX"
http://www.geofex.com/Article_Folders/cd4053/cd4053.htm

[WhenBoredomPeaks]

Thanks guys! I had a thought that i should check your site before asking this but that LED thing was so "bizarre" to me that i thought that maybe the IC is bad or i am really wrong about the supply voltage of the ic.

from your site:

"People who have tried these before have often complained that they pop or distort, and have then moved on and forgotten them."

Yeah that was me before your and flo"s post.

[R.G.]

(other interesting phenomenon: when i connect the 9V to the input and a LED going to ground at the output, my LED survives that. My common sense tells me that there should be 9V on the LED after i triggered the switch open with the control pin but there is actually 2-3V there. Maybe my IC is defective?)

Actually, your IC is keeping your LED from burning up.

LEDs have a 2-3V forward drop. The current through them can go to destructive levels at that same 2-3V if it's not limited by something external to the LED. In this case, the IC may be preventing the current from getting so high that it burns out the LED. If you replaced the LED with, say, a 10K resistor, chances are you'll see 9V on the resistor.

[WhenBoredomPeaks]

(other interesting phenomenon: when i connect the 9V to the input and a LED going to ground at the output, my LED survives that. My common sense tells me that there should be 9V on the LED after i triggered the switch open with the control pin but there is actually 2-3V there. Maybe my IC is defective?)

Actually, your IC is keeping your LED from burning up.

LEDs have a 2-3V forward drop. The current through them can go to destructive levels at that same 2-3V if it's not limited by something external to the LED. In this case, the IC may be preventing the current from getting so high that it burns out the LED. If you replaced the LED with, say, a 10K resistor, chances are you'll see 9V on the resistor.

now it makes sense, thanks!

tomorrow i gonna try the biasing and i gonna try some filter caps hera and there