What wrong? where is the distortion?

[Le québécois]

If my guitar have and output of 350mv (that what I have normally) and I feed a TS like clipping section with two germanium diode in the feedback loop and a gain set to 214 I should have and output of 214 * 035 mv = 74.9V. That is way over my 9volt battery. Distortion should be unbearable!!!

Believe it or not I have almost no distortion coming out! If I put 1N914 instead of the Germanium, Bam! there it is distortion as with all my TS build. Why that is not the case with my 0.3v GE diode? What it is that I don't get about clipping?

Has a first stage in a very high gain pedal will this be a usefull initial boost? I mean, is there something usefull about GE in the feedback loop despite the absence of obvious distortion? Does such signal is interesting for the next gain stage or just a waste of time?

thanks.

[runmikeyrun]

check the obvious stuff first- are the diodes in the correct way?  Is the diode good?  Sometimes even "new" items are bad.  don't forget that ge components (transistors, diodes) sound changes with temperature.  It's possible you're heating the diode when soldering it in, changing the sound it's making.  Keep the iron on the diode's leads for as short a time as possible, and clip an alligator clip to the lead between the board and diode to act as a heat sink.  It's even possible you cooked the diode when soldering it in.

So, try a new diode, use a heatsink, and don't solder too long.  Also, give it a minute after soldering it in just in case it overheated.

EDIT:  Forgot to add- if you solder in a couple of sockets ,you won't have to worry about all.  Plus you can change diodes very quickly and easily.

Good luck.

[Le québécois]

Thanks, for the answer but It's not a debugging issue. The thing work. It just that I have no distortion with Ge diode and I can't figure out why?

It is more a clipping bad understanding maybe?

edit: according to your post, you think I should ear distortion with the Ge. I have it with 1n914. The GE work I've just test them to be sure.

[runmikeyrun]

hmm.  Well, not all components work with every circuit.  Maybe it's just not a good fit. 

[Johan]

GE diodes have a softer knee than silicon diodes making the transition from clean to overdrive smother and once they open, the internal resistance of the diodes goes down and therefor the gain goes down.
J

[Le québécois]

GE diodes have a softer knee than silicon diodes making the transition from clean to overdrive smother and once they open, the internal resistance of the diodes goes down and therefor the gain goes down.
J

Once GE open, they short out my gain loop? hence no distortion? This would mean that the opamp do not lack headroom in order to boost my signal up to 0.3v?
If this is right, does decreasing my input signal (with series high resistor value) would be more demanding to the opamp and make it bang the supply voltage to reach the 0.3v of the diode? The resulting distortion (if any) will then be the result of the opamp reaching the supply voltage and not from the diode clipping the signal?

[Minion]

GE diodes will cause you to have a much lower output than a SE diode , about half the output of a SE diode and 1/4 the output of a standard LED ....... The distortion of a GE is a lot smoother and less brittle than a SE so it might not seem that there is a lot of distortion compared to a SE ....... Try an LED , they are my faves ......

[jk-fm]

Once GE open, they short out my gain loop? hence no distortion? This would mean that the opamp do not lack headroom in order to boost my signal up to 0.3v?

The Tube Screamer is a non-inverting opamp circuit, like the Boss OD. In this configuration, once the diode conducts and shorts the feedback resistor, it doesn't clip the signal - it runs at a gain of one. This is what give is a more "overdrive" sound rather than distortion. By using low-voltage diodes, the opamp is is running at gain 1 through most of the cycle.

If you want a distortion sound, put the diodes from output to ground like in the Boss DS. Or, change the circuit to an inverting opamp configuration, in that mode it will act like a logarithmic amplifier

[ashcat_lt]

The TS just won't bang the rails.  As soon as the output signal goes above the forward voltage of the diodes it turns into a buffer.  The distortion you hear is from the sides of the wave form getting steeper and steeper, but the peaks stay rounded.  It's really a form of crossover distortion.  This seems counter intuitive, but I guess an LED would give even more distortion for a given input x gain.

Negative feedback loops are like Bizarro world.  Everything is backwards!

[earthtonesaudio]

I have a huge assortment of Ge diodes of various types and vintages... none of which work.

[Le québécois]

So if I have distortion with SI diode this is mainly because the op amp is never able to give an output higher than 0.6v and start to bang the voltage rail? Otherwise the diode would short the gain resistor.
I know that the op amp is a big deal in TS because it is mainly responsible of the distortion we ear. This is how?
TS is therefore op amp clipping alone not diode clipping at all? One diode, a LED, two LED or a mountain of LED would not matter because we could have the same thing by lowering the supply voltage alone?

The goal of all this was to have distortion out of the TS but not a signal to high in order to add more clipping with diode to ground after the TS stage. I want both process at the same time (no switch) just more distortion added one after one along the signal chain. If the feedback loop is to strong, we won't ear the diode to ground effect and the other way around.

Maybe it's impossible ? and my conception of diode clipping to ground is wrong?

thanks for your answers

[ashcat_lt]

 

The opamp won't bang the rails!  The distortion comes around the zero crossings.  If the diodes have low enough forward voltage that region is very small, so it gets in and out of distortion very quickly, and it becomes difficult to hear.

Diodes to ground work the other way.  The signal passes by unaffected until the toward voltage is reached, and then gets clamped down.  The zero crossing region isn't distorted, but the peaks are squared off.

In order to combine the two, the feedback diodes need to be bigger than the diodes to ground, so that they allow enough output to actually exceed the forward voltage of the diodes to ground.

[Le québécois]

The opamp won't bang the rails!  The distortion comes around the zero crossings.  If the diodes have low enough forward voltage that region is very small, so it gets in and out of distortion very quickly, and it becomes difficult to hear.

So this is the zero crossing thing that I don't get. The way I see it (correct me please) is: let say the sine wave get in and vary between 0 and 0.5 V at the op amp input. The gain is apply and produce and output that may vary from 0 and up to X volt. From 0 to 0.3 volt output, the diode is close and the op amp apply the selected gain to reach the desired level and as soon as the signal reach more than 0.3v, or 0.6 for Silicone, the loop is broken and the thing start to act as a buffer with a gain of 1. Before reaching 0.3V the opamp need to bang the rail or distortion will come from nowhere!!! 

[PRR]

> I have no distortion with Ge diode

Are your Ge diodes good?

[ashcat_lt]

Before reaching 0.3V the opamp need to bang the rail or distortion will come from nowhere!!! 

No!  The rails sits at something like 3V for a typical opamp.  Can't possibly get there because the diodes won't let it.  But, with a sine wave input and a lot of gain you get straight sides up to rounded tops, rather than a nice gradual curve.  The slope of a sine wave near the zero crossing is 1 (a 45 degree angle).  The slope of the output near the zero crossing will approach infinity (a 90 degree angle).  Output doesn't look like input only larger, the waveform has been distorted to something different.

[Le québécois]

@ ashcat_It : since a picture worth a thousand words.... is this correct  http://dl.dropbox.com/u/43158392/Clipping.tif
sorry for the sketchy paint work!
If this is correct increasing the gain will make the slope steeper but the top will always remain round up to a point?


@ PRR:  To be completely honest, there is a tiny bit of distortion but nothing I expected from a TS. I did check the Fwd V of the diode and they work.


with the explanation above I start to understand why distortion was not at the rendez-vous.

[ashcat_lt]

Yeah, I think you've pretty much got it there.  The thing with the "TS followed by GE to ground" is only really accurate if the TS diodes have lower forward V than the GEs, and of course real components can't make perfectly straight sides or super sharp edges, but I think you've got the theoretical bit.

[Le québécois]

The thing with the "TS followed by GE to ground" is only really accurate if the TS diodes have lower forward V than the GEs, and of course real components can't make perfectly straight sides or super sharp edges, but I think you've got the theoretical bit.

Without really understanding it this is why I was trying with Ge in the TS (initial reason of this thread!). I wanted to put Si to ground after the TS and have both type of clipping simultaneously BUT you also said that :

In order to combine the two, the feedback diodes need to be bigger than the diodes to ground, so that they allow enough output to actually exceed the forward voltage of the diodes to ground.

I conclude that my goal is not possible with that design  

edit: thanks for your help by the way!!

reedit: Well, before someone suggest it, It is possible and actually working on my breadboard with LED in the loop, 2 SI to ground and 2 GE in the signal line (very similar to the B**S HM2). That is high gain!!!
Now I have to get rid of the high pitch noise and I will be happy!