Discrete TubeLess...

[puretube]

looks like somebody ( ) needs to hook up 3 ampere- and 4 volt-meters plus scope and amp to the circuit,
and make up a datasheet... 

(gez: former post is add-on edited...)

[gez]

I tried a single stage this afternoon and it biased up just as you said it would (trebles all round)!

I used a couple of medium power MOSFETs that can take up to an amp of drain current; all run from a 12V supply.  Have only tested it with my function generator and scope so far.  With source bypass caps the signal goes straight into hard clipping.  However, once removed I got the soft clipping you normally associate with inverter ICs.  This pretty much backs up my findings from the last time I investigated all this.

Although the chips - they're little 4pin DIL MOSFETs - I used can source a fair bit of current, gain was killed with a heavy load (even with the bypass caps) due to the output being taken from the drains: impedance is too high.  I'll stick a little audio transformer on the end tomorrow and see if I can get the circuit to drive headphones.

[Elektrojänis]

Petri: more food... from R.G., this time.

Oh... Somehow I have missed that even though I have read much stuff from geofex during the last 5 years (or maybe 10 years). Thanks again. I think I will print these all as I need some reading for a holiday trip I'm going.

That 4007 chip seems handy for this... I wonder how it sounds compared to separate discrere mosfets...

[Isaiah]

Does anybody have any recordings of this circuit with bass guitar please?

[gez]

Thanks for the inspiration Ton!



For a one stage circuit, it's not bad: like an amp just on the verge of breaking up.  My scope seemed to suggest some limiting of highs and this was confirmed when I plugged a guitar in (slightly muffled sound, but nothing too dull and it meant there was no need for a tone stack).  Probably gate capacitance; I'll have to check the data sheets.

[puretube]

welcome, gez! 
(did you try a second "boostchannel", or "fuzz" with one or both Rs switchably bypassed?)
(yes, I know it pops while switching...)

couldn`t get my hands off the basic circuit myself,
and, inspired by a comment in another post (reply#38)
I came up with the: "CSPWMB" *:


(block schemo)

the sounds of which in turn inspired me towards a new song...


*: "Can`t Stop Playing With My Baby"

 

[puretube]

the following pic shows, that a "cleaner" ordinary frequency-tripleing(as is well known from other circuits...)

Hey, keep out of my waveforms!
Andrew

I`m doing an initials-collection:
in reply#21 it was the "MM"
a little while ago I had a very nice "Z"-sign 
and recently, a quite realistic looking (fastood-)"M" logo... 

Today`s special is the "VW" initials:



it was drawn with the follow-up circuit of the aforementioned "CSPWMB",
whose inspiring reedy, raspy, and brassy  as well as its cut-membrane and mis-biassed-amp sounds has lead
to another modification, the:

"WORN-SECTION".


To me it sounds that terrific, that even Sir Stephen would let it slip through the fuzz-box barrier... 

[gez]

Messed around a little with headphones this morning.  I know from previous experiments with CMOS headphone amps that you need to treat them like Class A amps and get the drain current right up.  Using 15 ohm source resistors (with suitable power rating) I got about 120mA.  Slightly short of what I calculated would be needed to drive 320mW 32 ohm phones (both cans in parallel), but it worked.  The chips ran hot, but according to the supplier's calatogue they can take 1A and don't require heatsinks.

As a headphone amp it sounded better than a stand-alone effect, and was a lot of fun (almost like having a tube amp in your cans: clean, but with a little breakup here and there).  Not exactly efficient, but less hassle than a toob.

[puretube]

Gez: did you use a "hard" power supply, or rather the "saggy" type?


btw:
the comment mentioned in reply#45
was about the fact that these inverters are class AB devices...

trying to exploit some possibilities concerning crossover distortion, I thought about swapping the "inline"-resistors from the sources to the drains, for the "CSPWMB", but with no result;

Instead, I then put some resistors in series with the drains additionally to the circuit with the "source"-resitors...
to try out all possibilities, I substituted these resistors with a pot (P2), whose wiper constitutes the new output,
and parallelled another pot (P1), in order to be able to blend out the effect of the drainresistors altogether:



(both 100k, but 50k or 25k will do, too, depending a bit on the sourceresistors values...)

[puretube]

Playing a bit more with the "Worn-Section", it was able to create an even less sharp "TriFold" than in the "WavePly":


(@ ~25mV input amplitude)

but the most fascinating effect is the phasey changing interaction of the harmonix,
when the louder signal gradually descends from the "Fivefold":


(@ ~100mV input)

down through above trifold to the single fundamental, as it decays...

 

[The Tone God]

but the most fascinating effect is the phasey changing interaction of the harmonix,
when the louder signal gradually descends from the "Fivefold":

down through above trifold to the single fundamental, as it decays...

Now I am again going to have to ask you to step away from my waveforms.

Andrew

P.S. A forth coming design does something similar.

[puretube]

You ain`t heard nothing yet...

ps: I`m not aware of your waveforms... 


 

[The Tone God]

ps: I`m not aware of your waveforms... 

The tripling is something Finish Line does and a newer version goes into this phenomena further hence me bugging you to "keep out of my waveforms".

Andrew

[gez]

Gez: did you use a "hard" power supply, or rather the "saggy" type?

Hard/stiff: variable-voltage, regulated power brick (running things off 12V).

Interesting that you're experimenting with drain resistors.  I'm wondering if this isn't a better way to go in reducing current as the inclusion of source resistors generates a little distortion of its own.  An alternative would be to bias the gates either side of the centre point and leave out all source resistors.  I'll do a little testing after I've finished experimenting with the headphone amp.

[Isaiah]

I haven't had a chance to get any MOSFETs yet - can anybody tell me how this circuit sounds with Bass guitar, please?
Or better yet, does anyone have any samples or Youtube videos of the above, please?

Thanks,
Alex

[puretube]

ps: I`m not aware of your waveforms... 

The tripling is something Finish Line does and a newer version goes into this phenomena further hence me bugging you to "keep out of my waveforms".

Andrew

Sorry, I didn`t wanna anticipate future circuits/waveforms of yours... 

(arrgh - and I was so proud of the linear+exponential sawtooth generation of the "Worn-Section" with that sharp little dent at the bottom, during attack...)

(maybe this then is the right thread to note that I happily delivered some of my more advanced IC-based "dancing harmonics" tri-and-more-folders to a safe place, during MusikMesse Frankfurt, in April 2005) 

hopefully, the following 2 waveform movies don`t interfere with anyone`s plans:

The 2 "OverCrosser" pots described for the "CSPWMB" work as follows:

P1 controls the amount (percentage) of crossover-distortion, and can vary the waveform horizontally, when rotated,
while
P2 controls the position of the crossover point in the waveform vertically, when rotated.


another one for Isaiah, who may have missed the other clips hidden in this thread... :
"Devolution" shows the decay from multiple (sharp sounding) peaks to more or less pure sinewave, when the signal gets weaker (from the "Waveply").

(haven`t tried bass, yet).

[The Tone God]

The 2 "OverCrosser" pots described for the "CSPWMB" work as follows:

P1 controls the amount (percentage) of crossover-distortion, and can vary the waveform horizontally, when rotated,
while
P2 controls the position of the crossover point in the waveform vertically, when rotated.

And I answer with Blade.

But now that I bring that up I got some new ideas.

Andrew

[gez]

hopefully, the following 2 waveform movies don`t interfere with anyone`s plans:

The 2 "OverCrosser" pots described for the "CSPWMB" work as follows:

P1 controls the amount (percentage) of crossover-distortion, and can vary the waveform horizontally, when rotated,
while
P2 controls the position of the crossover point in the waveform vertically, when rotated.

That's clever!  I didn't understand the schematic at first but it finally clicked. 

I've been experimenting a little too.  The Kink Tank works well as a headphone amp from a 12V supply, 15 ohm source resistors (bypassed) and a larger drive pot (the thing was too loud).  With headphones loading the output, clipping was soft with a test signal.  In the real world, the distortion (when it kicks in) has a slight rattle too it, which I'm not that keen on.  The sound is reminiscent of 4007 circuits, which never sound as smooth as 4049 circuits IMO.  Not bad, just different.  For a while now I've wondered whether the n and p devices in the 4007 aren't as well matched as in 4049 circuits.  The scope images I'm getting with discrete MOSFETs are similar to those I've got with 4007 chips.  Who knows...

I tried biasing both MOSFETs either side of the centre point.  There comes a point where drain current drops dramatically.  The further away from the centre the gates are biased, the more the mismatch between thresholds becomes apparent: the outputs no longer bias up nicely and the 'symmetry' pot ends up being used as a 'bias' pot to get things to sit in the middle again.  I wonder if my earlier comments about Ron being more consistent at higher drain currents is the reason?

[puretube]

Dear My Tone God:

in order to get out of Your waveform-territory,
I implemented a change of topology concerning the output-circuitry of the "Worn-Section"...
(while the complete rest of the circuit stays absolutely the same, ie.: 2 cascaded "SanseValves". );

according to the saying:
"as well as every input can be used as an output, every output can be used as an input"
(or at least something similar to that has been said by a wise (wo-)man a while ago...)    ,

out comes a blendable (rotating the blend-knob*) from "clean" (well, err..., cmos-distorted fundamental)
saw-wave, to twice the frequency,
(without the transitional wave-folding going on in those other aforementioned circuits):

http://www.youtube.com/watch?v=hmyplRWeYXM

actually, the clip starts with a spaced sawtooth @ the middle of the blend knob`s rotation,
turning to the (distorted) fundamental at ccw, then rotating cw for full duplex (=octave up),
and back and forth, again.

Yes, I owe You and Gez 2 extra-schemo details by now...

(PM-club...)

btw: the manual blending can be done remotely (in a followup circuit...) by expression-pedal, or dual VCA.

[The Tone God]

Thank YOU!

BTW I was kidding about playing around with "my" waveforms but it did get me back to thinking about a few different things which I will play with later. Right now I have to work on my FX-X entry for this month which I have yet to start.

Thanks again for the videos and circuits.

Andrew