Are we fooling ourselves when it comes to diode softness?

R

rossbalch

Active member
I've been reading up on diode softness and curve tracing and all that kind of stuff. 2 threads in particular have made me question some of our understanding of practicalities of measuring the softness of the clipping/conduction knee.

Take this thread for instance on MOSFET clipping, in this case the more interesting Gate to Drain type, not the Gate to Source type where it's essentially just a regular diode.

On the graph the MOSFET looks like it would clip softly, except actually the only holds true at high current, which we might not often actually ever encounter in a pedal. At pedal currents, the clipping is actually quite steep. https://www.freestompboxes.org/viewtopic.php?t=32532

Another thread I've been reading suggests that if you normalise the curves, many that appear to be different actually aren't. Someone also stated, that contrary to what I thought was popular wisdom, putting two of the same diodes in series does soften the curve (as well as increasing vf), it actually stays the same. https://www.diystompboxes.com/smfforum/index.php?topic=121964.0

Would love any theory people to weigh in on their thoughts on all of this.
 
JTEX said:
You can play around with a hard clipper and soften it to your heart's content by simply adding a small resistor in series with the diodes on their way to ground. Just tweak and listen until it sounds right to you.


View attachment 99444


View attachment 99445
Click to expand...
Absolutely!

I guess what I'm saying is, there seems to be some conventional wisdom that actually doesn't hold true according to some testing.

For instance this statement: https://www.diystompboxes.com/smfforum/index.php?msg=1148986

If you take a single diode and two of the same diode in series and perform the normalization you will conclude the softness is the same.
Click to expand...

I've heard many times that stacking diodes not only increases Vf (which is easy to demonstrate), but also softens the knee (which looking at the data in that post, doesn't happen when you're careful about processing the data).

It seems like a lot of times, things we assume might behave a certain way, actually don't when you normalise the data, or in that MOSFET example, if you actually focus on the part of the data that represents the working conditions in an actual circuit, and it's interesting for those who design circuits to think about that.
 
I think it's like the tone wood debate - does it make a difference? Sure. Are there a million other things going on that could have as much, if not more, of an impact on what you're hearing than a single characteristic? Absolutely
 
drew.spriggs said:
I think it's like the tone wood debate - does it make a difference? Sure. Are there a million other things going on that could have as much, if not more, of an impact on what you're hearing than a single characteristic? Absolutely
Click to expand...


Also very true, given how much signal amplification is happening before diode clipping, the softness or not probably has diminishing effects on the actual output.
 
JTEX said:
You can play around with a hard clipper and soften it to your heart's content by simply adding a small resistor in series with the diodes on their way to ground. Just tweak and listen until it sounds right to you.
Click to expand...
Definitely true! A great example is the Fuchsia Fuzz (BJFE Pink Purple Fuzz) with a variable "Clip" pot as the series resistor.

1753229647085.jpeg
 
A common modification of the TR-909 Bass drum uses a similar trick: Here the fixed resistor R52 is
replaced by a 10k poti to introduce more distortion. Pitch the bass drum as high as it gets and
you've got instant Teckno.
 
You must log in or register to reply here.
Back
Top