New CMOS Booster Topology

[Peter Snowberg]

I built Andertons tube sound fuzz in the late 80s and I completely blown away that a digital IC could get used for audio. Digital logic and audio? Nah... couldn't be. :?

That started me tinkering quite a bit in the past with CMOS linear stages and always loved them, but I knew there had to be more out there. Then it happened... I found the data sheet for a CD4000 Dual 3 input NOR with inverter. The cool thing about this chip was that not only was it made unbuffered, but now you can bias the inputs differently to affect operation!

The typical CMOS unbuffered NOR has two N channel devices in parallel that short the output to ground while there are two P channel devices in series to connect with power. If you bias one of the inputs high, the NOR stage is less able to pull the output up and that stage gets it's output loaded down at the same time. Using the 3 input gate, I'm feeding two inputs with feedback biasing while input 3 sits at a static voltage keeping behavior in check.

CMOS 4K Booster

I've used the National CD4000C for these experiments. When configured as a plain inverter, this chip sounds like junk. Add some new bias and....

With the right values, this circuit sounds really nice and clear at guitar volume 3-7 and gets really fuzzy above 7. You will have to play with C1, C2, C3, R5, and R7 to your taste. You may also want to check out using two seperate 500K gain pots instead of the dual gang unit. For starting values... C1=C2=C3=.022uF, R7=220K, and R5=0 ohms. (click on the home link under the pic for more) Caution.... this thing BOOSTS! R7 should be higher I'm sure, but it then need a bypass cap to restore the highs.

Have fun!

-Peter

[petemoore]

I'm qiute tempted to build one of these.
 Great job on the Page. Schematic looks real good .!

[aron]

Very cool.  I will have to try this.

[Marcos - Munky]

I'm having trouble to see the page .

[gez]

Thanks for sharing Peter!  I've tried wiring up a 4007 in the past as a NOR gate to get some asymettry, but it sounded abysmal.  I thought it might just be the 4007 as the intermodulation distortion with these chips isn't that good, so I've always wanted to try a 4001UB.  Trouble is, unbuffered 4001s are as rare as Scottish Tories over here.  

By the sounds of things it wouldn't sound that great anyway, I'll have to do some experimenting with your idea.  Thanks again!

[Vasiliok]

I'm playing a lot with digital ICs for audio!
Tried CD4047, CD4069 and it's russian subs.

THE SOUND IS GREAT!!!

Very close to big amp's sound!
Sort of tube amp emulation!

You MUST try digital ICs guys!!!!  

[gez]

This is the only place in the UK I've been able to find an unbuffered 4000!  I haven't used this firm before, but I notice they also do the 4001UB...hmmm, I think I'll order both and see what the deal is...

http://www.viewcom.force9.co.uk/data/4kpage1.htm

[petemoore]

Would be [if I had to choose 1] the werd I'd ue to describe the [biggest/main?] characteristic difference between the CMOS ckt and all the other types I have...FF BF Boosters, Oa clippers etc.
 And I like notes that Bloom.
 I hope I'm not alone or off base on this observation.

[Nasse]

I build a preamp for me about 20 years ago with 4001 s (in the good old days propably they all were unbuffered). I have somewhere old circuits in an old magazine that included a basic preamp with baxandall tone control in feedback loop of 4001 and bipolar transistor emitter follower for lower output impedance, some thought about powering cmos in linear mode, and poor man´s "surround sound" matrix with sum and difference of left and right channels, all made with cmos inverters, with so much negative feedback that gain was unity. I built it (even designed a pcb for it), and it worked quite nice and had a warm and pleasant sound quality. The level of my cd player might be too much, but for my old cassette and recco player and FM tuner it was ok. Maybe I have time next weekend and dig my archive, and try some scanning and posting...

Few years later I found Graig A´s Tube Sound Fuzz circuit in local library...
In my experience this circuit is quite good for distortion, even if you add some negative feedback it still sounded like tubes...

But finding unbuffered 4000 and 4001 chips may be difficult, last time I mail ordered 4049UB´s I got buffered ones...

[gez]

Maybe I have time next weekend and dig my archive, and try some scanning and posting...

That circuit sounds really interesting, if you could scan it sometime that would be excellent!

But finding unbuffered 4000 and 4001 chips may be difficult, last time I mail ordered 4049UB´s I got buffered ones...

I had the opposite experience recently, for the first time ever I wanted buffered 4049s and I got sent UBs...sod's law!

[Tim Escobedo]

Looks like a good idea. Unfortunately, I've never come across a CD4000. The
internal schem of the NOR gate is interesting.

The issue of buffered and unbuffered CMOS gates is difficult. Some manufacturers don't seem to indicate the status. I personally have had very mixed results using any type of CMOS gate in linear mode other than 4049, 4069, 74HCU04 (that were specifically marked as unbuffered).

I've been trying to find a two input gate that can have each input biased linearly with the eventual aim of getting some kind of ring mod effects. Unfortunately, all the chips I have that fall into that category don't like linear biasing at all.