different clippers in feedback loop (SD, TS, fulldrive,... style): too subtle?

[Steben]

I didn't swap anything yet. I always tweaked gain and freq response.
I heard fulldrives before and wasn't impressed by the MOSfet clippers in the feedback loop. To me, they sound the same as Zeners ar LED's.
The reason is quite simple I guess: the circuit forces the gain to unity once the clippers start conducting. this means the clipping onset, internal resistance, etc... is less important.
When used in clippers to ground (rat, dist+, DS,...) I guess MOSfets make way more difference, right?

I'm thinking about putting switchable MOSfet clippers to ground after the gain stage, measuring the treshold and choosing the appropriate clipper configuration in the feedback loop (fe with Zeners). this way, you can have overdrive (zener loop) and extra switchable MOSfet distortion if desired.

[bioroids]

I'm not so sure about that. Check out some graphs I made here http://www.dedalofx.com.ar/bioroids/experiments/e06_distortiongraphs.php

I think I made a mistake by not normalizing the graphs, I'm considering repeating the test anyway

Greetings

Miguel

[Steben]

I'm not so sure about that. Check out some graphs I made here http://www.dedalofx.com.ar/bioroids/experiments/e06_distortiongraphs.php

I think I made a mistake by not normalizing the graphs, I'm considering repeating the test anyway

Greetings

Miguel

I think your graphs make it very clear the curves are very alike for example the 2 leds and the 2 MOSfets.
It's strange to say, but actually a lower treshold with rough generic devices (like the two standard Si diodes) tends to have smoother high gain settings. (This is of course as long as you compare full gain with silicons with a gain 3x as much on the MOSfets (scaling). With the same gain, there will be always less distortion with the higher treshold.)
Why? because the feedback loop circuit adds the basic signal above the treshold (unity gain) to the amplified signal till the treshold (high gain). With infinite gain you would have a square wave around the treshold with the basic round signal added on top (house with round roof). With a treshold of zero, you would have the basic signal with no distortion at all. The higher the treshold, the smaller the added "roof" looks compared to the steep rise.
2 Germanium devices are reported to sound "muffled" in a tube screamer. That's because the rising part (walls of the house) is so low (+/-0.2V) it can be as low as the added signal on top, resulting in a very unnatural signal like cross-over distortion (as small signal in octavia).

With diodes to ground there is no round addition on top the signal. It's plain cut off (like the solo clipping opamp). Here the detailed transition into clipping is very distinctive. I bet if you make some graph comparisons there as well, you'll see bigger differences!

[Mark Hammer]

The reason is quite simple I guess: the circuit forces the gain to unity once the clippers start conducting. this means the clipping onset, internal resistance, etc... is less important.

Well they probably would move towards, and reach, unity gain if the signal hitting the diodes came from somewhere else.  Keep in mind that as the diode starts to conduct and the gain starts to move towards unity, the signal is now no longer amplified quite as much and stops forcing the diode into conduction.  That is why circuits like the TS have compression; because the resulting signal dynamically changes the gain, just like a compressor does.  That's also why sticking more diodes, or diodes with a different forward voltage in there has the impact it does.  If the diodes don't start conducting until higher signal levels are reached, then greater dynamics are preserved.

In contrast, the gain of something like a Dist+ or Rat or DOD250 is completely independent of the action of the diodes.  Imagine a dinner party where you let a big dog into the room while people are there.  People will adjust to the presence of the dog to keep the chaos to a minimum.  Now imagine that all those people have eaten their fill and left the room, and THEN you let the dog in and close the door.  Without the adjustment of people to the presence of the dog, that same dog has a different impact on the orderliness of the room.  That's the difference between sticking diodes in the feedback loop and after the output of the op-amp.

You know, all of this kind of makes me wonder what a BMP would be like if the diodes went to ground.