An overdrive/distortion.... A little different, this one.

[brett]

Hi
this circuit is guaranteed good fun.  Not a 1-trick pony at all IMO.

It uses three of the six inverting amplifiers in a CD4049UBE.
Although looking superficially like the tube-sound fuzz and its variants (Mark Hammer in particular is an expert concerning these), it is somewhat different.  It starts with a JFET gain stage that generates just a tad of distortion. 

One of the keys is that the 4049's supply voltage is reduced to about 6 V by the 220 ohm resistor (and polarity-protecting diode).  This helps give a nice "compressed" feel and ensures that the onset of distortion is subtle and progressive, only getting severe when pushed hard.  If you don't want the subtlety, the resistor should be changed to a low value (e.g. 10 ohms).  Also note that the J201 JFET nominated here will give more gain than most other types, so try others if you want cleaner tones or lower gain.

The "drive" pot is shown as a linear type, but I actually use a LOG type that gives greater control over the "overdrive" end of the gain spectrum, but not much control over the "maxed out distortion " end (which happens as a mad rush between 3 and 5 o'clock). 

The "drive" and "filter" pots interact in a similar way to those controls on a "Rat".  At high gain, you'll probably want plenty of filtering to control excessive "sharpness".

Thanks to RG for the suggestion about unexplored magic in CD4049.
That's about it. Have fun.  All feedback welcome. 

http://aronnelson.com/gallery/albums/BinOfBrett/Forty_niner.jpg

[Mark Hammer]

Looks good, and I'll bet it sounds good too.  We tend to look at those little triangular boxes and go "Ho hum" then get all excited about discrete FET or MOSFet-based units, all the while forgetting that inside those little triangular boxes on the schematic are MOSFets.

The supply-line trick is worth thinking about.  If you haven't already done so, take a look at the Charles Fischer project from EM that JD Sleep has posted over at GGG.  Fischer used a current-control IC to vary power to the 4049 in his design in order to achieve variations in "splat".

If I can make one small suggestion.  Consider altering the Filter control pot arrangement to the SWTC arrangement (http://hammer.ampage.org/files/SWTC.gif).  This will give you the same control over tone, but without the interaction that normally occurs between filter setting and volume setting.

And just out of curiosity, I note that you use U1f essentially as an inverting buffer.  It is not uncommon to find use of several paralleled invertor sections in other kinds of designs as a means of creating a higher-current output.  Since you have 3 invertor sections just hanging around, I'm wondering if there is any tactical advantage to making use of them in that manner.  Such a musing is directed at those with more electro-savvy.  Is there something one could do more easily with several parallel invertor sections on the output?  That could be as simple as being able to use a different value output pot or whatever.

[Dragonfly]

very cool, Brett...ive been playing around with 4049's and 4069's over the past week, and theyre fun little chips....i'm woring on a combo boost / overdrive / distortion / tremolo off one 4069 currently.... 

[MartyMart]

Nice Brett, there's mileage in those 4049's and 4069's, I have a couple that I messed with in a similar
way, Jfet front and the 220Ohm limiting resistor, I seem to have about 1 in 2 of my cmos builds "failing"
or at least not working quite as expected !
Perhaps I'm a bit full of "static" !!

Marty.

[Dragonfly]

Perhaps I'm a bit full of "static" !!

Marty.

You're full of "something", thats for sure !   

[brett]

Hi
thanks for the support and suggestions.

Consider altering the Filter control pot arrangement to the SWTC arrangement (http://hammer.ampage.org/files/SWTC.gif).  This will give you the same control over tone, but without the interaction that normally occurs between filter setting and volume setting.

This is an interesting idea.  I found the interaction OK.  I guess that is because the filter pot is only 25k and the volume pot is 100k log.  Given that the output is usually large (say 3V p-p), the volume pot will usually be set only a few k from ground.  I *think* this minimises the interaction.  For example, if the volume pot is set at 10k from ground, there's typically 90k of resistance "upstream", so changing the filter pot by 5k, and adding 5k to the 90k will have negligible effect on volume.  (ie 10/90 is almost equal to 10/95)  Is that right?? 

The limit to this approach of using a small filter pot and a large volume pot is that the output impedance of the previous stage must be much smaller than the filter pot.

And just out of curiosity, I note that you use U1f essentially as an inverting buffer.  It is not uncommon to find use of several paralleled invertor sections in other kinds of designs as a means of creating a higher-current output.  Since you have 3 invertor sections just hanging around, I'm wondering if there is any tactical advantage to making use of them in that manner.

I've used parallel buffers before (in a 4049-based sine-wave generator), and was thinking of using them here.  However, it messed up my layout, and I couldn't see any advantages from extra current capacity. So I just used one inverter as both a buffer for the previous stage and driver for the filter.

If someone knows a cool use for a circuit like this that could push about 100mA, then 3 or 4 parallel buffers would certainly do the job.

thanks again for the feedback 
PS thanks Marty for putting up that soundclip of the Hidrosis - really nice work

[MartyMart]

I have a 4049 circuit using 2 stages, 470k on the first with a 68pf, 150k on the second with a 120pf and
20k between, 22n caps ... does low to medium gain quite well and sounds good.
Front end is pretty much a fetzer valve.
Might be well to add a TC and then volume recovery using a third stage ?
How did it go with that transformer Brett ?

[brett]

Hi Marty
still working on that valve project....

Here's an interesting thing about 4049 circuits...  I think that many people have been smothing out the natural harmonics then adding artificial ones back in with more clipping in later stages.

Looking at Marty's resistance and capacitance in the first feedback stage (470k and 68 pF), the roll-off frequency is between 2 and 2.5 kHz.  That's low.  The reason why Marty's circuit sounds good is that the signal is clipped by a later stage with a rolloff at 8 to 10 kHz.  This adds back a whole lot of top end.  But only when the clipping is quite hard.

My circuit uses small caps relative to the resistors (22pF with a 1M resistor and 100pF with a 100k resistor) so that when clipping is slight or moderate, the natural highs of the guitar come through.  Hence designs with large caps like the TSF are very "fuzzy" (ie synthetic sounding) and Marty's with intermediate sized caps will be "distortiony", but this one has small caps and is therefore capable of very mild overdrive.  To make it work for distortion I needed the "filter" control to cut out the highs when the gain is turned up.  It's a bit of a nuisance having to adjust both controls, but there is a good range of tones available.

Anyway, those are my 2c theories...

[Mark Hammer]

That's essentially why I ramble on constantly about how one needs to spool out highs in multi-stage clipping devices.  One has to think about what it is that subsequent stages have available to clip.  Keeping tight reins on which harmonics find their way through to a 2nd or third stage allows for an eventual output that I think many find more musical because it has more growl than rasp.

Of course, the notion that one should gradually loosen the reins on high end is entirely separate from how much to do it and at what frequencies.  That's where trial and error and "taset-testing" comes into play.

Your comments about whether the filter/volume control interaction is problematic are quite cogent.  There are plenty of circumstances/settings where no audible interaction should occur.  I guess the reason I lean towards the SWTC is that it means you simply don't have to worry/think about interaction at all.  You pick a tone-pot and volume pot value that don't impose too much default attenuation (e.g., you wouldn't want a 250k tone pot and 10k volume pot), pick a rolloff range and appropriate cap, and away you go.

Finally, I have found that sticking a bypass cap in the "Drive" pot (think "bright" switch) provides some nice bite to lower drive settings.  Kind of "instant Rickenbacker".  I leave it to you to figure out what suits your taste, but starting out with .01uf  might be useful.

[Ben N]

Thanks, Brett, cool roject. Thanks also, Mark, for the interesting comments and suggestions.

Uh, Brett, I don't mean to cause trouble, but I am sure you were not aware that Mark had 4049-based project called the "Forty-Niner," right?

http://hammer.ampage.org/files/FortyNiner-1.zip

Ben

[puretube]

sorry, folks,
but I don`t see an inverting buffer in U1f:
just a dead short between in & out;

I see two (shunting) FET-diodes going from the (center-biased) output of U1a to the respective rails. ("clipping"...)

[pott]

Ooops nevermind that...

[Mark Hammer]

sorry, folks,
but I don`t see an inverting buffer in U1f:
just a dead short between in & out;

I see two (shunting) FET-diodes going from the (center-biased) output of U1a to the respective rails. ("clipping"...)

He's right.  What self-respecting electron would go through U1f when there is a perfectly good piece of wire to pass through?

Um, just how DO you get something like a unity-gain inverting stage with a 4049?

[puretube]

2 equal (1series input-R & 1feedback-R) resistors...

[gez]

Is there something one could do more easily with several parallel invertor sections on the output? 

Drive heavier loads.  A year or so ago I posted a schematic that parallels a few inverters to drive headphones.  Not exactly low distortion (I don't meant fuzz here, I mean fidelity), but it works well enough.  Downside is the current consumption (ridiculously high)

[Mark Hammer]

2 equal (1series input-R & 1feedback-R) resistors...

...excluding a zero-ohm input and feedback resistor, of course.

Thanks.

Is there something one could do more easily with several parallel invertor sections on the output? 

I posted this query because I remember discussing use of piggybacked op-amps with Mike Irwin, and he indicated that in some instances the behaviour of diodes might change with higher current drive feeding them.  I didn't completely understand his point, but I figured that perhaps there might be some sort of context where extra current drive from paralleled invertors might conceivably be useful.  If not, so be it.  I figured that if they're just sitting there doing nothing anyways.....

[Ben N]

Being mosfets, can inverters be used as clipers?

[puretube]

Mark: zero ohm input-R = OK;
         zero ohm feedback-R: No.

Ben: see Brett`s schemo, 3rd inverter... (IMHO).

[Ben N]

OK, Ton, that third inverter is being overdriven, right? What I meant was a clipper in the same sense as the way diodes are used in a TS or Dist+, to ground after gain or in the FB loop. Or does that make no sense for an inverter?

Also, w/re the question about paralleling inverters for better drive capability--would that not make sense ahead of the tone control? (Just based on the article that Aron linked to in another thread: http://www.diystompboxes.com/pedals/jo4049.html.)

TIA for your responses,
Ben

[puretube]

check John`s link, fig. 12a :
in Brett`s circuit, Q1&Q2 are like antiparallelled diodes in a "clipper-to-ground" (after the gain of the previous inverters);
the fact that one goes to plus and the other to ground, doesn`t matter signal-(AC-)wise.