please help me understand clipping diodes

[evanj1969]

 OK, I'm a new guy needing help, or at least an understanding of clipping diodes.
I've spent the last 6 months or so reading, and trying my best to learn. Been reading here almost every day, and have exhausted my searching trying to find an answer to this dilemma.
For my first project (breadboard), I put together a Tube Screamer "ish" thing. No input or output buffers / and the first side of the op amp is a stock Tube Screamer, and the second side is configured like Timmy. Simple enough. I even got the stereo in / mono out / and bypass switch hooked up without to much trouble. (guess working on computers all day helps with learning to diagnose problems ??)

I don't seem to be getting much difference with or without clipping diodes. I've tried several variations (Ge, Si, Led's), with no difference in sound / feel. I have even rebuilt the circuit twice hoping I would find my error, but with no luck. Series, parallel, asymmetrical, symmetrical, and yes ... no diodes at all.

I know the sound thing is very subjective, I've been playing for 20+ years, I can hear and feel subtleties that most could care less about, so .......
Am i missing something? Is this a very subtle thing used for shaping the distortion?? The gain control, volume, and tone all work fine. As I said, even without diodes, I still get grit, crunch, and distortion, but with them in, in any configuration, or completely out, I can tell no difference.

What should I be expecting from the diodes that I'm not getting without them?

Thanks for any help, as well as those who have helped so much in the past.

Evan

[shimster]

If I were you I would start with a boosted clean signal then add a diode to ground after it and see if you hear the difference.  If you're getting a good amount of clipping and distortion before adding the diode, it is very difficult to say how much adding diodes can add to your sound.  In one of my circuits, a single clipping diode added a nice high frequency clipping to my already slightly clipped signal, basically adding a hard clipping on top of soft clipping to give it more hair.  And this effect became more pronounced down the line when I sent the signal through several more gain stages.  I hope this helps.  Good luck!
-Dan

[Vitrolin]

more hair???

[blester]

This link is a great place to start: 
http://www.geofex.com/Article_Folders/TStech/tsxtech.htm

[shimster]

more hair???

Ya know, I once worked at a summer camp teaching music and at the beginning of the session we would perform for the whole camp.  I would usually play some Van Halen or Ozzy on my guitar.  Anyway, one of the campers, a little girl, came up to me after one of those shows and asked "Why does you guitar sound so furry?" .  I thought that was cute.

[evanj1969]

 Thanks for the fast reply guys.

The Technology of the Tube Screamer is the first place I started, days before I even thought about posting here.
There is still allot I still have to learn. Some of R.G.'s stuff  IS a bit beyond my comprehension though. After reading through it several times, I still don't have a clue what I'm missing.

I have even tried the diode to ground after the first op amp gain stage, but again, I didn't get any results. I fear I just am missing something simple (just like at work).
I AM getting overdrive out of the circuit, and it's not unpleasant at all. In fact, I'm kinda diggin it, especially considering it's my first project, but I still get no difference with the diodes in or out.
As for the diode to ground, I'm not entirely sure "where" it should be placed. I tried it right after the first op amp, as well as right before the output, and again, no difference.

Thoughts??

[Earthscum]

You are running into the same confusion I am, I think. If you want to play a little bit with diode clippers, try the bazz fuss. It was interesting to screw around with. What I noticed with the BF and other fuzz circuits is that you can apparently get away with only one diode. I didn't hear any difference with the second diode in. If you run one diode in the wrong direction, it totally screws up the sound... cuts it out and adds nasty scratchy distortion. That's about all I have for ya in the way of experimentation. I tried a booster circuit and all the diodes did was attenuate the sound, didn't give it any distortion. From my observation, diodes (to ground, at least) just seem to shave off the top of the waveform where all the stacking harmonics are at and cleans up the top (for visual reference, picture Bart Simpson's hair as the top of a heavily distorted signal after diode clipping, and Lisa's hair before the diodes).

Am I correct, or at least close?

I still don't understand the mechanics of the feedback path, so no comment on that one.

[valdiorn]

As for the diode to ground, I'm not entirely sure "where" it should be placed.

this comment tells me you're not getting some fundamental key concept of circuit design... sorry dude, don't know what I can say to be of assistance.
the diodes should be placed from the signal path to ground, as simple as that, I can't really clarify it any more than that

Also, if you could post some pics of what you're trying to build (circuit schematics) it would help alot (one picture says a thousand words)

[Earthscum]

Hey, check out my thread, basically concerning the same thing... I have a feeling you are missing the same thing I am, or at least going to be once you get to playing around with them some more. GibsonGM gave me a good explanation that helped out a BUNCH. Basically, you may be getting distortion, but the signal isn't big enough to clip across a diode, or the higher signals are getting clipped, but you don't hear them missing (like a hiss or something). Read his explanation, and I think you will understand what I'm talking about.

http://www.diystompboxes.com/smfforum/index.php?topic=76607.0

[blester]

This link will show how the waveform is changed in the basic TS circuit:
http://www.bteaudio.com/articles/TSS/TSS.html

[JKowalski]

Key points:

Diodes only conduct current one direction. They do this whenever the signal on the side of the > part of the diode is around 0.6 volts or higher then whatever is on the | end of the diode. So lets say, that you have a sine wave, with an amplitude of one volt. So it oscillates from one volt, to 0v, then -1v, and so on. Alright -  if we put two diodes, going opposite directions, from ground to the signal, what happens? When the sine wave reaches that critical voltage (0.6V), the diode starts conducting the signal to ground. The same for the negative side of the wave - the OHTER diode starts to conduct at -0.6v. So what happens is that each side of the wave, positive and negative, cannot go above 0.6V or below -0.6 before the diode starts to conduct, so what we get is a wave that abrubtly cuts off at those voltages. Whenever the wave tries to go above those critical points, the diodes stop it, and leave it as a flat voltage at 0.6V until the signal goes below the points again.

Notice that if you have a signal that is BELOW 1.2V peak to peak (0.6V amplitude on either side of ground), then the diodes WILL NEVER CONDUCT. You will just have your original wave. In that case, you have to FIRST amplify your signal, so that it can reach those critical points.


Now - the tube screamer is a more complicated circuit. It uses a particular trait of diodes. When you increase the voltage across a diode, current flows through it in a logarithmic manner. Look here:



This is what is known as the VI curve (voltage V, Current I), it shows you how much current flows through the diode at different voltages across it. You see how the current sharply goes up at around 0.6V? Thats our "Critical point"! As you can see, below the 0.6V, hardly any current at all flows, it's almost 0. That's why we have to reach the signal level of 0.6V in order to cut off the top of a wave.

The tube screamer does not simply cut a wave off. Instead, it is a LOGARITHMIC AMPLIFIER. Basically, it uses the interesting VI curve of a diode to SMOOTHLY cut off the top of the wave. How exactly this works might be something you need to take a good long time to look into - it's alot of stuff. But heres what you really need to know:

Even with the logarithmic amplifier, you STILL have to reach around that critical point! Since the bulk of the change in the logarthmic curve occurs around 0.6V, thats where the bulk of the change in your waveform will be! You have to reach the 0.6V in order to do anything substantial to your waveform. Also, note that log amp clippers, once they go a certain distance above the main log curve, the current becomes linear. Now, when you have a linear current in a situation like this, your voltage is linear at the output. So, you have a flat wave top, with the VI curve happening at either ends of that flat top. So - as you increase the AMPLITUDE of your input, the output goes from unchanged wave, to more rounded, to square wave with rounded tops, then to square wave with rounded transistion points.


This is alot to take in, I didnt realize. But!

- You have to reach the diode curve to clip! ALSO NOTE: Silicon diodes have their points at 0.6V, but other diodes have different ones! LEDS have something like 1.5V, GE diodes have something like 0.3V, etc.





ALSO - YOU CAN'T PUT THE DIODES TO GROUND IMMEDIATLY AFTER THE GAIN STAGE. This is because - Your circuit uses Vref. What that is is a votlage halfway between your ground (0V) and your battery voltage (9V). You need this because: When you amplify a signal, if you do it at ground, then that signal has to be negative and positive. Well, you cant add more negative votlage to your signal because all you have available is the voltage range from 0-9V! No negatives! SO instead, you force the signal to be around 4.5V instead of ground. That way, you can go above or below the signal votlage, and you have voltage to spare! What you should do, is think of the 4.5V as the "fake" ground.

If you do that, then what you pretty much have is a -4.5V, 0V, and 4.5V. See what I mean? That way, you can make, say, a 1V/-1V signal 2/-2V, or even up to 4.5V/-4.5V. You cant get voltage from nowhere, basically.

No - what does this have to do with the doides? This: If you connect two diodes to the ACTUAL ground, and the signal is oscillating round 4.5V, then what happens is that the diode is conducting ALL THE TIME. You see? You have a signal going from, say, 3.5V, 4.5V, 5.5V, 4.5V etc. Obviously, this voltage is 0.6V above 0!!! SO what you need to do is connect those diodeds to your FAKE ground. Now, you have your signal on one side, and 4.5 on the other.

Remember - voltage across a diode is simply the smaller voltage subtracted from the smaller. So say, the signal reaches 4.5V(Vref)+0.6V. Well, that means the voltage on the signal side is 5.1V. THe voltage on the OTHER side of the diode is 4.5 still (vref), so 5.1-4.5=0.6. There's 0.6 volts across it!!! We have clipping!

I hope you can get something at all out of this. I wrote it really fast, im afraid i might have effed up alot in explaining.

[shimster]

Best clipping diodes explanation EVER! Thank you so much for this!

[evanj1969]

this comment tells me you're not getting some fundamental key concept of circuit design... sorry dude, don't know what I can say to be of assistance.

Maybe I should have asked, "where in the signal chain would you suggest". I have tried it in several locations with no effect on the sound whatsoever. nothing. I know what it means and I suppose how the do that. A diode coming off the signal chain to ground.
At this point I chose to rebuild the circuit on the breadboard piece by piece to hopefully track down my problem.
Now on the 8 pins of the op amp, I'm getting voltage readings all over the place.

This may have been part of the problem all along as I am using the same setup for the voltage supply/bias (if I am using the correct terminology). When I say the same, I mean that part of the circuit hasn't been moved or rebuilt, so that may be the root of the problem? remember I'm new to all this, and I never checked it before.
The battery reads 9 volts and some change.


Regardless, I appreciate all the great answers that everyone has thrown at me.
If anything comes to mind after reading about my voltage issues, please speak up, as it is getting frustrating at this point.

Thanks again guys

[evanj1969]

and by the way JKowalski , that was a great explanation of diode clipping.... Indeed!
unfortuanly, it didn't seem to help resolve my problems.