Tube Screamer Clipping

[GibsonGM]

Measure the voltage drop across the resistor, that's Vf 

Not enough coffee, Mike?

There is NEVER enough coffee, Antonis..... 

It definately isn't, Sir ...

I can claim it using the fact that you answer to me quoting Jim's post....!!! 

Ha ha!  Coffee and then in a hurry.  I have other duties besides stomp boxes, but I prefer DIY over them!  Time stamp suggests I was being called to help the wife with dinner!    Food is almost as important as coffee 

[ashcat_lt]

I like the difference in dynamics with more or less diodes. It makes a difference in the "feel".

The difference is how far into or past the "knee" you are hitting at a given gain setting.  Changing gain before this point has the same effect.

[Mgt280y]

If anyone is following I have got round to starting this
Found a diode function on my DMM, what is the units it measures in across my yellow led pair opposed reads 1790 both ways
And
3 asymmetrical 1n914 1138 one way 590 the other

Not decided on the last yet

Just a thought I am wiring as the diagram at the beginning but putting the diodes on a little perf not straight on the switch now depending which way I wire the asymmetric ones with either clip more bottom or top is it worth adding a switch between to flip the order

[EBK]

is it worth adding a switch between to flip the order

No

[Mgt280y]

Thats an easy one then 

3 stage im going to try the zendrive
Can someonme check I have the orentation/jumpering of Q1, Q2, D4 and D5

[PRR]

> what is the units it measures in across my yellow led pair opposed reads 1790 both ways And 3 asymmetrical 1n914 1138 one way 590 the other

It is "milliVolts".

Divide by 1,000 (push the decimal point over 3 places) to get Volts.

One (each way) red LED we expect to be 1.6V, yellow a little more, so 1,790mV (1.79V) is about right.

One plain Si diode we expect 0.6V, so 590mV (0.59V) for one and 1,138mV (1.138V) for two series (569mV each) is perfectly reasonable.

"Hot" guitar level may be 500mV peak. The simple Si diodes limit the level to a "hot guitar". The 2-diodes' 1.2V and the LEDs' 1.8V is "really hot!" at the diode, but there may be a way to turn-down after the diodes.

Note that the different voltages suggest different turn-down to get a "usable" level out of the pedal (into your amplifier). When you do this, there is not a lot of difference from one diode to another. They all obey a fairly general physical rule.

[ashcat_lt]

Note that the different voltages suggest different turn-down to get a "usable" level out of the pedal (into your amplifier). When you do this, there is not a lot of difference from one diode to another. They all obey a fairly general physical rule.

Yeah, we talked about that earlier.  In fact, I think it was part of the OP, but I'm not going back a page to check. 

In the TS, though, the position of the gain pot affects the action of the filters, so turning up gain and turning down volume to run LEDs at the same relative threshold as Si should have a noticeable effect on the frequency response.  Not really thanks to the diodes, though.

[Frank_NH]

Interesting topic.  There is a nice post on the effect of have a non-inverting op amp gain stage (versus an inverting stage) at this link:

http://www.bteaudio.com/articles/TSS/TSS.html

Basically, the tube screamer gain stage mixes the unamplified input signal with a phase-shifted clipped signal which is frequency dependent.  So there are more aspects of the design that effect the sound than just the clipping diodes - and, hence the myriad of different tube screamer variants.

Speaking of tube screamer input/output levels, I have a question.   A typical guitar signal amplitude is, as mentioned above, around 0.1 - 0.5 volts.  Normally, this signal level would be fine for the input buffer of a tube screamer.   But what if you are stacking pedals (e.g. the Visual Sound Double Trouble)?  Wouldn't a hot pre-amp or overdrive cause the input to be overloaded, or at least asymmetrically clipped?  I suppose that can sound good ("saturated" distortion), but maybe the input buffer should be redesigned to permit a hotter signal?  I'm wondering aloud here because I'm playing around with a tube screamer circuit on my breadboard, and thinking that a simple JFET input buffer with the gate biased at half supply (4.5V) would respond differently when presented with a hot pre-amped input signal than the usual BJT buffer of the tube screamer.  I guess I'll test this out on the breadboard...

[anotherjim]

I guess a JFET input buffer ought to give it some margin of asymmetric soft-clipping versus the BJT one. If there isn't already some boutique versions out there doing just that, I would be very surprised. Most designers, I  think, shy away from features that depend on something else pushing the front end, but if it's yours, you can have whatever you want.

If, because of higher Vf in the clip network, you increase available gain by upping the gain pot to 1M, it will deepen the LPF formed with that 100pF feedback cap to around 1.5Khz compared to 3Khz with the 500k pot. If it looks like you will normally be running with high gain and if the distortion sounds too dark, you can get back to the original filter range by changing the 100pF to 47pF. With Gain pot settings of 300k or below, the filter roll-off with that 100pF is above guitar cab bandwidth.
Interestingly, the Boss SD-1 has a fixed low pass in the second op-amp of 1.5Khz and no feedback cap at all in the first op-amp.