introducing SIGGY the sigmoid shaped distortion thingy...

[Quackzed]

    I've been playing around with this idea for a while now. i've built a few variations of it and it's an interesting 'at least to me' in the way it softly clips the signal.
   it sounds good, has a nice amplike drive to it, and retains alot of dynamics, which is unusual for diode type clippers. thought i'd share as its a different take on creating distortion though it has some things in common with amz saturation controls...



   I'm calling it Siggy, due to the sigmoid shape of the wave after its been clipped. I have a few 'saturation' plugins that i use on live tracks and like the way they compress tracks without being too overbearing with their compression, but this is for guitar.
   any thoughts good or bad are welcome. i used a basic 'to me' muffer type stage for each section though I considered op-amps due to their larger signal swing abilities... but this seems to work pretty good with 2 stages...

[Quackzed]

i have a falstas sim of it, if anyone is interested in seeing it in action...
let me know.
I had originally posted the falstad sim, but its too big and so i drew it up. 
i was being lazy. its pretty cool though and could probably be altered to do a 'triode like' clipping response.
diode ladder clippers are cool!  and i havent seen anything like this for guitar though its similar to saturation type controlls, so its original in that sense.

[NPrescott]

Ach! I saw your previous post for the sim and came back to check it out since I can't breadboard anything at the moment 

Do you mind posting it again? Looks interesting.

[Jdansti]

Nice. Maybe i missed it, but which diodes are you using?

[Quackzed]

heres the sim, diodes are 1n4148's

$ 1 5.0E-6 1.1208435524800693 19 5.0 43
t 512 272 624 272 0 1 -0.21279629677951295 0.5666968606921885 200.0
w 512 272 512 224 0
r 512 224 624 224 0 470000.0
w 624 224 624 256 0
r 624 224 624 160 0 10000.0
r 624 288 624 352 0 3330.0
r 512 272 512 352 0 470000.0
c 512 272 432 272 0 1.0000000000000001E-7 1.9069624077009015
g 624 352 624 384 0
g 512 352 512 384 0
g 784 336 784 368 0
O 784 272 832 272 0
R 128 272 64 272 0 1 900.0 0.5 0.0 0.0 0.5
R 624 160 624 112 0 0 40.0 9.0 0.0 0.0 0.5
c 656 224 688 224 0 5.000000000000001E-7 4.805454095545709
d 720 320 720 352 1 0.65
d 704 352 704 320 1 0.65
d 704 320 704 288 1 0.65
w 720 256 704 256 0
r 704 224 720 224 0 6800.0
r 704 352 672 352 0 3300.0
g 672 352 672 368 0
d 720 288 720 320 1 0.65
w 704 288 704 256 0
r 704 320 672 320 0 6800.0
w 672 320 672 352 0
w 720 288 720 256 0
w 784 224 752 224 0
w 784 272 784 224 0
174 416 272 416 368 0 100000.0 0.005 Resistance
g 416 368 416 384 0
w 432 320 432 272 0
w 656 224 624 224 0
w 688 224 704 224 0
w 720 224 752 224 0
w 720 352 704 352 0
w 720 320 704 320 0
t 224 272 288 272 0 1 -6.605872388377678 0.4986037812782687 100.0
w 288 256 288 224 0
r 288 224 288 144 0 10000.0
R 288 144 288 96 0 0 40.0 9.0 0.0 0.0 0.5
r 288 224 208 224 0 470000.0
w 224 272 208 272 0
w 208 272 208 224 0
r 288 288 288 368 0 1000.0
g 288 368 288 384 0
c 288 224 352 224 0 3.0E-8 4.6473117750839865
w 416 272 416 224 0
r 208 272 208 368 0 470000.0
g 208 368 208 384 0
c 208 272 160 272 0 1.0000000000000001E-7 1.0255063013794243
w 384 320 368 320 0
w 384 352 368 352 0
w 384 288 384 256 0
w 336 320 336 352 0
r 368 320 336 320 0 6800.0
w 368 288 368 256 0
d 384 288 384 320 1 0.65
g 336 352 336 368 0
r 368 352 336 352 0 3300.0
w 384 256 384 224 0
w 384 256 368 256 0
d 368 320 368 288 1 0.65
d 368 352 368 320 1 0.65
d 384 320 384 352 1 0.65
r 352 224 384 224 0 12000.0
w 416 224 384 224 0
w 736 256 720 256 0
w 752 224 736 256 0
r 784 336 784 272 0 100000.0
174 128 272 128 368 0 1000.0 0.005 Resistance
g 128 368 128 384 0
w 144 320 144 272 0
w 144 272 160 272 0
o 11 1 0 34 1.1692013098647223 9.765625E-55 0 -1

[Jdansti]

Thanks! I've experimented with different diode clipping techniques, but haven't seen this one. I'll have to give it a try.

[Quackzed]

you might want to lower the 2nd transistor emitter resistor, i went all the way down to @ 220ohms but at that point the signal is getting squared by the rails and has less of the character of soft clipping, but it sounds nice and gainy with the 220ohm. more heavy dist less 'soft clipped' with the 220...
something to try anyway...

[PRR]

Fine design. Gain, gentle bending, controllable small gain, gentle bend again.

I see a rather low input impedance (35K) which will trim the pickup's treble above 500Hz-1KHz. The 33nFd caps shave bass below 400Hz. That may be a good thing for this application. Reduce intermodulation of chords and harsh treble hash.

[Quackzed]

thanks prr, i appretiate the response.
it wasn't really my intention to cut so much low end/high end, but really just due to the resistors i used in the diode ladder, maybee i'll scale it up a bit and see, though it probably is a good idea for the reasons you mentioned, i'll try and see how more high end /low end i can feed the stages and the results.

There are a few things i noticed that seem a bit odd to me. for one, it seems to have a sort of strange intermodulation phenomenon happeining, at least its my guess that its intermodulation distortion. there seems to be some verry low, well below the fundamental, tones as it distorts the wave. its faint, but there is at least the perception of these very low frequency 'wolf' tones. they're barely there.and seem unrelated harmonically, not loud but just barely discernable beneath the notes.
i figure a quick fix would be to use a high pass after the last diodes to ground, or in a recovery stage after the diodes. strange tho.
also the intent was to keep the signal small enough to get a soft clip without any hard clipping so i had to be carefull about too much gain and stay within the power rails. still, as i mentioned in a previous post, the second stage, and possibly the first may need some tweaking for more gain if using low output pickups, or perhaps normal pickups/ not sure... its hard to get a solid answer for how big a signal a guitar pickup puts out and this circuit  rides the line of less gain for no hard clipping at the rails vs. not enough output to hit all 4 pairs of diodes for a decent rounding action...
these are my general thoughts on the problem, but as is it seems to do what i intended. so far so good. if i get anywhere i'll post an updated schem.

thanks again for takin' a look.

[Govmnt_Lacky]

Gil,

Is there any coorelation between the Emitter resistors on the transistors? Converse or inverse perhaps??

I was thinking about a dual 5K pot either wired parallel or inverse. 

[Quackzed]

only in that the first one has more gain due to the available headroom, whereas the second stage has less available headroom due to the first gain stage amplification and so needs less amplification to not hit the ceiling and hard clip. but each stage is affecting both halves (top  + bottom) equally but the second stage is @2x as gentle. due to the 6.8 vs the 12k...
so yeah, a dual pot would let the 1st stage soft clip gently or square off the tops, depending, then the second stage could round the hard clipped edge or  just add some more gentle clipping of its own... go from very gently rounded to slammin' in parallel, in inverse it might not go to those extremes , not sure.
try it.!!! 
btw the 12k and 6.8k series resistors/ respectively affect how strongly or gently the clipping action is. so a 20k would be closer to normal hard clipping with only a modest amount of signal geting past the first diodes in the ladder, where a 1k would very gently clip the signal and allow it to pass through both thresholds (to the ceiling) with only a very small amt of attenuation at each step in the ladder.
if you put it into falstads its easy to mess around with values and see what you get.
try it out, and dont forget to share! inquiring minds want to know...