1N4001s vs. 1N4148 As Clipping Diodes - What's The Difference?

[anchovie]

Not questioning your figures, but I'm surprised a 51pf cap would cause a frequency roll-off that low.. especially as some effects use caps as big as 1000pf in a feedback loop.
Can you elaborate on the formula used to calculate that please ?

The frequency roll-off is affected not just by the cap, but by the value of the feedback resistor too (in brett's example the gain pot set at 450K). If you've seen 1000pF used then there was most likely a smaller value resistor.

Formula is 1/(2*pi*R*C).

[Greg_G]

Not questioning your figures, but I'm surprised a 51pf cap would cause a frequency roll-off that low.. especially as some effects use caps as big as 1000pf in a feedback loop.
Can you elaborate on the formula used to calculate that please ?

The frequency roll-off is affected not just by the cap, but by the value of the feedback resistor too (in brett's example the gain pot set at 450K). If you've seen 1000pF used then there was most likely a smaller value resistor.

Formula is 1/(2*pi*R*C).

Thanks anchovie.. I've had a good think about what's going on and I believe i have a better understanding now.

I wonder if the diodes in the loop have any effect, compared to a feedback loop with just RC and no clippers ?

[Paul Marossy]

I wonder if the diodes in the loop have any effect, compared to a feedback loop with just RC and no clippers ?

Do you have a breadboard? You could put something simple on it and do a quick A/B comparison.

My guess is that you looked at it with a scope you would mainly see far more clipping with the diodes and there would be more odd order harmonics present than without the clipping diodes.

[Greg_G]

Paul

I was really wondering aloud about any interaction between the effect of the diodes and the capacitor... as opposed to the capacitors effect when no diodes are present.

[Plexi]

Reborn old thread! 

As I said in another thread, I'm building the Lovepedal Karl fuzz.

It use 1n4148 and BAT46 for clipping.

Are 1n60 or 1n5819 good substitute for BAT46?

[bluebunny]

Probably, but depends on your value of "good".    Try it and see!

[bool]

Reborn old thread! 

As I said in another thread, I'm building the Lovepedal Karl fuzz.

It use 1n4148 and BAT46 for clipping.

Are 1n60 or 1n5819 good substitute for BAT46?

1n60 (the silicon vesion) maybe.
You could try BAT85 or a BAT41

[Plexi]

Thanks!

I mean by "good" for the most close result to the original one

I used 1N60.. but.. I thought that they was schottky germanium

[ashcat_lt]

Come on guys!  The difference is obvious just from looking at them.  The 400x is black and boxy, while the 4148s are more colorful and transparent.  Duh. 

Anywho, the cap in the feedback loop parallel to the gain pot is actually a shelving high pass.  It acts like a low pass in this case because it's in the bizarro world of the negative feedback loop where everything is upside down and backwards and attenuation becomes gain.  So, you figure the corner frequencies like this:

Fz is going to be basically your "cutoff" frequency, everything above Fp is going to be unity, and the gain below Fz will equal 1/a...

...but the value of R1 is actually the parallel resistance of the gain pot and the effective resistance of the diodes.  That resistance changes depending on the voltage across them, which of course changes depending on the gain of the circuit which of course you can't calculate until you know that effective resistance, all of which is dependent on where the signal actually is in its signal swing at any given instantaneous moment...

So it's already pretty complicated.  The cutoff frequencies literally slide around continuously.  You can sort of squint and estimate what it would be when the signal is very small and the diodes are basically open, and also what it would be when the signal is very large and the diodes are kind of like a short...except they're never actually 0 ohms and their real value is difficult to determine.

I have to admit that I don't actually know whether diode capacitance will appear to be in series or in parallel with that effective resistance.  If parallel, it's kind of easier (given the above complexity), if series, who the hell knows?

[pinkjimiphoton]

do a search on my old "transmission overdrive"

it uses a pot to sweep between a 4148 clipper and a 400x clipper.  works great, you can get some interesting sounds thru the sweep. the 400x's clip later and are very smooth sounding, the 4148's clipped faster and had a more "normal" distortion sound. you could use the pot to blend between them.

[bluebunny]

I'm surprised no-one has pointed out that the difference is 147.   

[Plexi]

do a search on my old "transmission overdrive"

it uses a pot to sweep between a 4148 clipper and a 400x clipper.  works great, you can get some interesting sounds thru the sweep. the 400x's clip later and are very smooth sounding, the 4148's clipped faster and had a more "normal" distortion sound. you could use the pot to blend between them.

Nice one..!
Like the "clip" pot in the WH Swollen Pickle MKII, that pan between two diodes.

[bluebunny]

do a search on my old "transmission overdrive"

it uses a pot to sweep between a 4148 clipper and a 400x clipper.  works great, you can get some interesting sounds thru the sweep. the 400x's clip later and are very smooth sounding, the 4148's clipped faster and had a more "normal" distortion sound. you could use the pot to blend between them.

This reminds me that I need to box up my transdrive board that I stuffed sometime last year...