Red LLama CD4049 how does it work?

[mmaatt25]

Hi,

I've been building these cos I love the sound they make and I want to build a Hot Harmonics as that looks like it takes things even further. I've been looking at the schematics for these and thw data sheets for 4049's but don't understand how they produce the fuzz/distortion I love.

So whats going on??

Thanks

Matt

[PRR]

http://www.fairchildsemi.com/an/AN/AN-88.pdf

[mmaatt25]

Thanks PRR.

I see the answers in there I just need a bit more knowledge. I'm reading the art of electronics, I didn't realise how little I know!!

Thanks again.

Matt

[PRR]

Amplify a signal too much, it distorts, "fuzz".

All "logic" gates _are_ high-gain analog amplifiers. They are optimized for extreme on/off signals, but if you put in small audio you get large audio out.

CMOS logic chips have very high gain, high input impedance, and fairly soft clipping.

The AN-88 paper (originally from RCA or National; dunno how Fairchild got it) details several forms of linear amplifiers based on CMOS.

The paper uses 74C04 but the 4049 is now much more common.

[WGTP]

Thanks, interesting read.  Is there also crossover distortion? 

[Paul Marossy]

All "logic" gates _are_ high-gain analog amplifiers. They are optimized for extreme on/off signals, but if you put in small audio you get large audio out.

CMOS logic chips have very high gain, high input impedance, and fairly soft clipping.

That explains a few things. I have also wondered how you get distortion out of something like a CD4049.

[Electron Tornado]

http://www.fairchildsemi.com/an/AN/AN-88.pdf

Something interesting in Figure 5 in that app note is that the voltage gain is larger for smaller voltages, while at the same time the bandwidth becomes narrower as higher frequencies are attentuated. Seeing that got me wondering. I understand that distortion pedals made with CMOS chips tend to be noisey. Is that noise broadband or can it be decreased somewhat by using 3 volts for Vcc?

A second question comes from Figure 2. What is the relationship between Rin and R1? Do they have something to do with the gain in that stage?

[PRR]

> What is the relationship between Rin and R1?

"Inverting op-amp gain stage".

The difference this or a TLO72 is that in CMOS the "+ input" is internal and floats to about half the supply.

Yes, playing with the power voltage has effects. I think it would be quicker to just-do-it than to speculate.

CMOS just IS noisy. MOS gates are such good insulators that tiny specs of electric charge can be heard. On top of that, inverting stages always hiss worse, and six high-gain amps in one chip begs for over-over-kill gain out the wazoo, and anything amplified a million times will hiss to beat the band.

[Mark Hammer]

I stick a cap of sufficient value (220-390pf) in the feedback loop of the two cascaded gain stages to take out the unneeded highs and hiss, and that does it for me.  It ends up being a pretty quiet overdrive, despite the gain and "natural properties" of the 4049.  I am also given to understand that some chips are a little noisier than others, and that one ought to swap out chips and listen for the quietest one.  No particular brand, really, just a matter of tolerances and such, same way that people making noise generators with a 2N2712 transistor are encouraged to try out a bunch and listen for the noisiest one.

[Electron Tornado]

Thanks, Mark. That explains why those caps are used there.

Here's another question. Looking at Figure 6 from the above app note, I don't recall anyone building a pedal with that sort of circuit. The ones I recall always have a feeback resistor for each individual inverter. What are the pros and cons of cascading stages and using a single feedback resistor?

[PRR]

> pros and cons of cascading stages and using a single feedback resistor?

Higher accuracy. Lower distortion. NOT what we want here.