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

[Paul Marossy]

Does anyone know which one of the two clips softer? I ask because sometimes I see pedal mods where a 1N4148 clipping diode is to be replaced by a 1N4001 diode. I'm not sure if that is to make it clip harder or a little softer...

[Lurco]

the comparison test costs you less than my 10 cents.

[cctsim]

1N400x smoother, fuzzier with a bit less output ~0.5V.

1N4148 crunchier a bit more output ~0.6V-0.65V.

I personally prefer the 1N400x type.

[oliphaunt]

1N400x smoother, fuzzier with a bit less output ~0.5V.

1N4148 crunchier a bit more output ~0.6V-0.65V.

I personally prefer the 1N400x type.

Odd, I find that 4001s always have more a good bit more output than 4148 or 914s (to my ear, not to a meter).  The 400x is another clipping diode I have yet to find a good use for.  I always try them alone or in combinations but never seem to like them that much.  I'm I will find a good use for them someday.  I just got some BAT46, 1N100 and 1N60 to try, looking forward to that!

[Paul Marossy]

I was under the impression that 1N4001s sounded a little smoother, but I wasn't positive about that.

[oliphaunt]

I think they break up a little more harshly than 4148 and not as quickly, about 40% of the way towards the sound of an LED.

[Paul Marossy]

I think they break up a little more harshly than 4148 and not as quickly, about 40% of the way towards the sound of an LED.

Hmm... that would make them in between a 1N4148 and an LED then. I'm looking for something that clips softer than a 1N4148, that's not a 1N34. I hate the way those sound. I wonder if some MOSFETs as clipping diodes might get me what I want?

[dubiousss]

i couldnt hear the dif when i tryed it on my distortion+ but it sounds over the top anyway

[brett]

Hi
the clipping voltages of 1N400X and 1N4148 are within a smidgen of each other (they are both conventional Si diodes, and nobody can avoid the laws of atomic physics).  The gold doping of the 1N4148 provides a *slightly* larger transition to conduction (the "knee").  I've always thought that this should be less obvious in a clip-to-ground circuit than in an op-amp feedback arrangement (but I've never simulated it). 

In any case, the differences seem to be at the limits of audibility.

For much much larger differences, try Ge (relatively large knee) and LEDs (relatively small knee).  The size of the knee is very similar on a variety of diodes, so the larger the Vf, the smaller the size of the knee in relation to Vf.
cheers 

[Paul Marossy]

I've always thought that this should be less obvious in a clip-to-ground circuit than in an op-amp feedback arrangement (but I've never simulated it). 

I guess that I should have mentioned that I am thinking about this in the context of clipping diodes in an opamp feedback loop.

[ayayay!]

There's a good little list of voltages on page 17 of Brett Miller's Build Your Own DS-1 pdf document.

http://www.diystompboxes.com/smfforum/index.php?topic=76021.0

[Paul Marossy]

There's a good little list of voltages on page 17 of Brett Miller's Build Your Own DS-1 pdf document.

http://www.diystompboxes.com/smfforum/index.php?topic=76021.0

Voltages are part of the equation, but not all of it. One of the first pedals I built was Joe Davisson's "Amber Boost" which used a pair of 2N7000 MOSFETs as clippers in the feedback loop of the opamp. They are said to have a softer knee, but I don't know at what voltage they clip at. I think I'm going to try them in this circuit I reverse engineered and am now tweaking. That's what prompted this question in the first place - my client wants the first gain stage to sound smoother, and the clipping diodes are a part of how the tone is shaped. I didn't know how 1N4001s fit into the scheme of things, so that's why I asked. My intuition tells me to try 2N7000s in this circuit, so I will see what results that yields.

[mac]

Imho, as both have similar forward voltage drop differences can be capacitance and how fast they react. I'm guessing here but 400x are designed for high voltage and current levels, maybe 400x feel more happy with larger signals than 4148.

mac

[slideman82]

I think they break up a little more harshly than 4148 and not as quickly, about 40% of the way towards the sound of an LED.

Hmm... that would make them in between a 1N4148 and an LED then. I'm looking for something that clips softer than a 1N4148, that's not a 1N34. I hate the way those sound. I wonder if some MOSFETs as clipping diodes might get me what I want?

400x and 4148 sound different because of the "barrier capacitance" (I think that's its name), which is larger in 400x (making 4148 a faster diode), so maybe this capacitance sucks more highs.

Have you tried Schottky diodes, like 1N5718 (I think that was the code number)? They have about 270 mV of forward voltage. And mosfets and diodes sound great too!

[Paul Marossy]

400x and 4148 sound different because of the "barrier capacitance" (I think that's its name), which is larger in 400x (making 4148 a faster diode), so maybe this capacitance sucks more highs.

Have you tried Schottky diodes, like 1N5718 (I think that was the code number)? They have about 270 mV of forward voltage. And mosfets and diodes sound great too!

I have never tried Schottky diodes. But I was thinking about trying the MOSFETs in conjunction with Si type diodes, though. That's a project for tomorrow morning.

[DSV]

You can also try to add a resistor (< 1k) in series with the diode clipper, in my opinion it sounds a little smoother than just with diodes. I have tested this only with LED clippers though.

[Paul Marossy]

You can also try to add a resistor (< 1k) in series with the diode clipper, in my opinion it sounds a little smoother than just with diodes. I have tested this only with LED clippers though.

Huh, that's another interesting idea. Might try that, too. 

[brett]

Hi
regarding op-amp feedback circuits, the diode does two things: (i) forward conducts, and (ii) adds capacitance.  I hadn't thought the capacitance was too important until now.

Most diodes have a fixed capacitance and a frequency-dependent component to their capacitance.  Simple Si diode have a larger fixed capacitance and small frequency-dependent component.  Schottkys and I *think* LEDs have more frequency-dependent capacitance. 

A 1N4148 has about 5pF of capacitance, while a 1N4004 has about 15pF.  In a tubescreamer, 15 pF is a significant addition to the 47/51pF feedback cap.  At about 3 o'clock on the drive control (450k feedback), the standard cap gives rolloff at 7 kHz, but an extra 15pF lowers this to 5.4 kHz.  That's half an octave difference.  Many people would be able to hear that.  If you put two diodes in parallel, it would probably tame the distortion quite a lot.

Schottky diodes (e.g. 2N5819) have low and frequency-dependent capacitance, because they don't have minority carrier injection and stored junction charge (the junction is a semiconductor to metal).  Even more than the 1N4148, the capacitance is probably insignificant.  LEDs have high capacitance at high frequency (e.g. 60pF at 1 MHz), buy I'm not sure low low it is at audio frequencies (e.g. 10 kHz).
cheers

[Greg_G]

Hi
regarding op-amp feedback circuits, the diode does two things: (i) forward conducts, and (ii) adds capacitance.  I hadn't thought the capacitance was too important until now.

Most diodes have a fixed capacitance and a frequency-dependent component to their capacitance.  Simple Si diode have a larger fixed capacitance and small frequency-dependent component.  Schottkys and I *think* LEDs have more frequency-dependent capacitance. 

A 1N4148 has about 5pF of capacitance, while a 1N4004 has about 15pF.  In a tubescreamer, 15 pF is a significant addition to the 47/51pF feedback cap.  At about 3 o'clock on the drive control (450k feedback), the standard cap gives rolloff at 7 kHz, but an extra 15pF lowers this to 5.4 kHz.  That's half an octave difference.  Many people would be able to hear that.  If you put two diodes in parallel, it would probably tame the distortion quite a lot.

Schottky diodes (e.g. 2N5819) have low and frequency-dependent capacitance, because they don't have minority carrier injection and stored junction charge (the junction is a semiconductor to metal).  Even more than the 1N4148, the capacitance is probably insignificant.  LEDs have high capacitance at high frequency (e.g. 60pF at 1 MHz), buy I'm not sure low low it is at audio frequencies (e.g. 10 kHz).
cheers

brett

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 ?

Thanks

[manson]

What about the ultrafast UF4007 diodes? Would those sound any different?