question about diode clipping

[TimWaldvogel]

ok, so if i put more diodes in the feedback loop of an op amp than it can clip, all it does is raise the output a bit right? for instance most people would not put 3 sil. diodes one way and then two sil diodes the other way for asymmetrical clipping in the feedback loop of a ts808. but IF you do wont it clean up the signal significantly?

here is my idea. i want to build a modified ts 808 clone or tube reamer... idk yet, but instead of doing the traditional diodes in the feeback path, i want to put 2 or 3 which raises the voltage at which they get pushed into clipping...i may even use mosfets ot 5.1v zeners.... and up the pot to a 1M for a little more output. i instead want to put diodes to ground afer the op amp a passive tone control.

is this practical ? has this been done before?

[Quackzed]

sure, you could do that , and yes it's been done before.

In the schematic above, you can see hard and soft clipping diodes. Soft and Hard clipping are described above in the Creating Distortion-part. Soft and Hard describe quite well the sound they produce. Soft clipping softly rounds out the signal peaks while hard clipping roughly cuts the signal peaks. See the oscilloscope view below to get the idea what I mean.

but it's a different sound than a tube screamer sound, and basically it will be more of a distortion + type circuit and sound
the ts's unique sound is due to the clipping in the feedback loop 'soft clipping'
and the distortion + type sound is created by diodes to ground after the op amp' hard clipping'

check out 'clipping diodes' here
http://www.generalguitargadgets.com/richardo/distortion/index.html

[TimWaldvogel]

well i was thinking i want to smooth out the waveform before it reaches the diodes to ground. i think i might use yellow\Red LEDs in the feedback loop and asymmetrical sil. 1n4001s going to ground after it. and actually what i was thinking about with the ts808 is that i would get the general frequency. response of a tube screamer but with the hard clipping to be more like the ibanez\maxon sonic distortion. i am curious about only moving the clipping diodes as well. i have no problem with the distortion plus other than i dont like how it changes frequency as you move the drive. i recently got inspired to make my own distortion idea loosely based on brian wamplers distortion plus\dod 250 mods in premier guitar.

[TimWaldvogel]

how can you make the tube reamer at ROG push two Red LEDs into clipping harder in the feedback loop?

[TimWaldvogel]

i am also looking for nice sound clips for LEDS in tube screamers and tube reamer sound clips outside of the ones from runoffgroove if ou can help

[Quackzed]

well if you wanted to use high threshold clippers like leds in the feedback path of a ts type clipper, it would be to raise the threshold so that the signal will clip less, but if you then want to raise the signal level hitting the clipping diodes so it will clip more, you sort of end up right back where you started. but i can see that if you have some diodes to ground after, youll want enough signal level coming out of the feedback loop stage to be clipped by the diodes to ground.
the gain of an op amp stage is basically r9/r10 in the pic above so as it is 10k divided by 1k =gain of 10 . so if you made r10 500ohms it would double the gain to x20 or if you made r9 20k (doubled) you would double the gain for x20 as well.either way depending on what type of diodes to ground you have after and their threshold to clip you may not need that big a signal.
        in other words, the bigger the signal is when you clip it with the diodes to ground ,the more you will hear just the diodes to ground clipping. so with a big signal youll have to raise the threshold of the diodes to ground as well to keep some of the other flavor audible. so you can see that as you raise the signal of one stage, you may need to raise the threshold of each succesive stage as well, so that the last stage doesnt clip off all your hard work.
after all it doesnt matter what you do to the signal peaks, if they will be clipped off in the end.

[rousejeremy]

   From my experience, Germanium to Silicon to LED clipping basically sounds like a volume difference. Germanium and Silicon clip at a much lower voltage rate and give the   impression of higher gain. Germanium especially. Check this out http://www.youtube.com/watch?v=wnxTn1OE6_Q
   The actual tone of the distortion is pretty much unchanged. The overall "gain" or perceived amount of distortion is the result of the clipping diodes. Don't be fooled by fancy switches and the like. Use your ears and stick with what you like. There's a reason production pedals don't have 7 knobs and switches to choose between every possibility, a reason that goes beyond dollars and cents.
    I've begun to doubt the whole "clipping" diodes obsession lately if you haven't noticed.

[TimWaldvogel]

I disagree I found some videos of tube screamers with all 3 on a switch and the texture of each is noticably differnt. Red LEDs always gave a more Marshall crunch kinda feel.
And thanks for the info. So I might try two Leds back to back in the feedback loop and change the cap/resistor values accordingly

[Quackzed]

with a soft clip then hard clip setup like you mentioned, it might be cool to have a few different types of diodes to ground to switch in/out. a high threshold leds to ground will let more soft clipping from the first stage be heard more.. whereas 2 ge diodes to ground will most likely clipp all that soft clipping off and give you a hard clipping only sound...i think the boss os-2 did something like this, though via a different 'parallel' setup...check this out for a general idea of the difference in sounds and how they mix...

http://www.youtube.com/watch?v=8JOY_F4STX4

[petemoore]

  Where's the ceiling in relationship with the waves...simple like water.
 Lower the ceiling and the wave peaks...more of 'em gets chopped off every wave cycle.
 Which is about the exact same thing that happens when the volume of the wave increases [say you added boost or just turned up the gain], all of a sudden more of 'em wave peaks get lopped again.
 Whether boosting the signal input to diodes, or lowering the threshold of the diodes, the effect of controlling the proportion of the waves being clipped is had, both/either way. Are they exactly the same? Imagine a wave and a channel and a 'low' ceiling to the channel...water wave peaks hitting and being flattened by the ceiling, exactly what happens except to the signal waves.
 Are the waves short [high frequency] long [low frequency], are you taking the 'foam' [fizz-fuzz sound] off from what the 'ceiling' [diode threshold] leaves by putting a small capacitor across the clipping diodes ?
 Whatgoesintacomesouttas.
 I did much building of clipping diode/boosting inputs, LP and HP filterwork...boiling...boiling...choosing to work with a DIST+, preboosted, tone control after it [100k > .1uf to ground at diodes [IIRC] as far as diodes...that's me, my stuff etc., history with that box also helped me gravitate..
 +The TS finally succumbed to "high temperature exposure switch syndrome failure" [last of the batch that survived the blaze], and I haven't gotten back around to it.
 Generous gentlemen: one Joe Gagan, and friends helped inspire me to do some BiG Muffwork, since that circuit finally succumbed to the inevitability of becoming 'lit', the 'ol FF barely fits on the pedalboard anymore.

[TimWaldvogel]

I have trouble breadboarding things lol. I am gonna make the tube reamer and add a Passive tone control. I am gonna do asymmetrical LEDs in the feedback loop and change the drive to 1m... see how it works out.   

[Quackzed]

...for assymetry you could try 1 led and 1led+1ge back to back as 1 led / 2leds is very assymetrical and one side(2leds) may not clip much at all ...
though that may be good too... use your ears really... if it sounds good thats what matters.

Whether boosting the signal input to diodes, or lowering the threshold of the diodes, the effect of controlling the proportion of the waves being clipped is had, both/either way.

yeah, its relative. if you like the 'marshal crunch' of red leds, its because of their threshold in 'relation' to the signal size. if you fed red leds a signal that was 4x as big, they would sound suprizingly similar to germanium diodes. just as if you fed them a tiny signal, they would sound clean, never getting anything large enough to clip. its all relative... not to say they each dont have a certain 'flavor' all else being equal, but its a lot more to do with how much signal is getting clipped, and thats the signal amplitude/threshold relationship.

[teemuk]

  From my experience, Germanium to Silicon to LED clipping basically sounds like a volume difference. Germanium and Silicon clip at a much lower voltage rate and give the   impression of higher gain. Germanium especially. Check this out http://www.youtube.com/watch?v=wnxTn1OE6_Q

Well, in that video the Ge diodes do introduce more distortion (resulting into a slightly fatter tone) - but that's mainly because they clip the signal relatively more due to their lower forward voltage. That's why they also attenuate the signal a lot. All that volume difference you hear there is due to Ge diodes clipping that much more. It's still almost the same sound (because the overall circuit is EQ'd certain way and it makes a lot more difference). Furthermore, if you would equalise the levels driving both diodes (so that both would clip the signal as much) there would be only really minimal differences in the clipped tone and I seriously doubt they would be audible. The Ge diodes really do have a slightly less steep slope into conduction than LED diodes but you can always compensate that with a series resistor on the LEDs.

The only real issue is that LED requires a lot more input signal to overdrive as much as a Ge diode, and voltage swing is a sparse thing in SS circuits, especially when powered by a 9V battery.

I just recently wrote to one of the several "clipping diodes" threads about this and even posted some graphs that illustrate the whole thing about the alleged differences of Ge vs. LED diodes.

[petemoore]

  There's only two ways it works right.
  1 is my way, and the other is my way too.
  Ok...maybe there are three or even more ways to do diodes up and get something mean, or nasty, or sweet or...

[TimWaldvogel]

I have been giving it thought and I may just do one red LED and one silc. 1n4001. Cause if I do that it may still be quite asymmetrical do to each diodes threshold being different. And I really love the sound of both diodes. Maybe I should place a 8 pin ic socket in the loop so I can have the option of back to back, asymetrical, stock, whatever cause I'll have 4 holes on each side 

[Quackzed]

thats a good idea, leave yourself a 'test bed' to try out different things. that way you can have a lab to try diode ideas/sounds.

[TimWaldvogel]

i think will sound good. has anybody successfully built a tube reamer pedal ? what diodes did you prefer?

[amptramp]

ok, so if i put more diodes in the feedback loop of an op amp than it can clip, all it does is raise the output a bit right? for instance most people would not put 3 sil. diodes one way and then two sil diodes the other way for asymmetrical clipping in the feedback loop of a ts808. but IF you do wont it clean up the signal significantly?

here is my idea. i want to build a modified ts 808 clone or tube reamer... idk yet, but instead of doing the traditional diodes in the feeback path, i want to put 2 or 3 which raises the voltage at which they get pushed into clipping...i may even use mosfets ot 5.1v zeners.... and up the pot to a 1M for a little more output. i instead want to put diodes to ground afer the op amp a passive tone control.

is this practical ? has this been done before?

Gettimg back to the original questions, putting diodes in series does raise the clipping level.  Everything below this level is amplified linearly and everything above it is clipped.  You have the option to increase gain to bring the clipping level down (referred to the level of input at which clipping starts).

The TS808 uses clipping diodes in the feedback loop of a non-inverting amplifier so that even when the signal is being clipped, the amplifier still has a linear gain of 1 for input signals.  Thus, some of the original signal always appears at the output.  Clipping diodes to ground have no gain above the clipping threshold and therefore sound more extreme.  The latter case would be like the DOD 250.  It is linear until the clipping level is reached than the output stops having much reation to the input.

If the diodes are the same in both directions, there will be mainly odd-order harmonics added.  If the diodes are different in each direction, both even- and odd-order harmonics will be added.  The latter case is usually considered more musical.

[TimWaldvogel]

K am Also contimplating putting a socket for experiementing with hard clipping and tone caps etc