LEDs as clipping diodes, diffused vs clear, is there a difference?

[pappasmurfsharem]

Are diffused LED's only aesthetically different from clear? I suppose their are other factors as well such as what the LEDs are rated for.

I am building the Son of Clay Jones OD and planned to try the recommended red LEDs then also try some 4148's, but I don't have diffused LED's I only have the RED waterclear ones.

[Seljer]

the more obvious differences are between different colours (and to a less extent between the normal/ultrabright varieties).

If you want to get down to raw numbers, get two multimeters, wire a potentiometer meter in series with the LED and measure the actual voltage-current characteristic of the LED.

[Mark Hammer]

The cosmetic aspects of the LED, like shape, size, diffuse-vs-clear, etc. have no bearing.  Neither does whether they actually glow in a visible way or not.  Different colors have different forward voltages, with red being the lowest of the LEDs at around 1500mv.

Seljer suggests that ultra-brights may be different from so-called "garden variety" of the sort you'll see as status indicator LEDs in Boss pedals.  That may well be true, but my guess is that any hypothetical differences simply won't show up in our context.

Note that a 1500mv forward voltage provided by a single LED should not really sound any different than the same forward voltage provided by a trio of 4148s, strung end to end.

[pappasmurfsharem]

The cosmetic aspects of the LED, like shape, size, diffuse-vs-clear, etc. have no bearing.  Neither does whether they actually glow in a visible way or not.  Different colors have different forward voltages, with red being the lowest of the LEDs at around 1500mv.

Seljer suggests that ultra-brights may be different from so-called "garden variety" of the sort you'll see as status indicator LEDs in Boss pedals.  That may well be true, but my guess is that any hypothetical differences simply won't show up in our context.

Note that a 1500mv forward voltage provided by a single LED should not really sound any different than the same forward voltage provided by a trio of 4148s, strung end to end.

That's the answer I was looking for, Thanks Mark

[Mike Burgundy]

Note that a 1500mv forward voltage provided by a single LED should not really sound any different than the same forward voltage provided by a trio of 4148s, strung end to end.

Mark, I do remember seeing a scope graph of different diode types having different clipping characteristics. I imagine the knee in the characteristic might have some influence here - it's not perfect devices so I would not be surprised to see - and hear -  differences. Anyone remember that graph?

[Mark Hammer]

I am open to being corrected (i.e., smacked upside my smarmy little head), but my sense is that the knees of most diodes pertain to frequencies well above those relevant to guitar noises.  If diode X takes 100 nanoseconds to conduct, and diode Y takes 300 nanoseconds to conduct, that is most assuredly important to folks designing microprocessor-controlled stuff, but those times, as big as they are for switching purposes, are simply way out of the  usable range for music.  Consider that, even though they are measureably different, a tweeter that is flat clear out to 40khz doesn't really sound very different than one which is flat out to 26khz, because our ears stop caring well below that.

But, like I say, if there are knees that transition slowly enough to have some bearing on the relevant part of the spectrum, I'm certainly not going to pooh-pooh their possible audible impact.

[guitar.mod]

Even-thou this topic is not active for a while - I want to correct Mr. Mark Hammer.

"Note that a 1500mv forward voltage provided by a single LED should not really sound any different than the same forward voltage provided by a trio of 4148s, strung end to end."
"If diode X takes 100 nanoseconds to conduct, and diode Y takes 300 nanoseconds to conduct, that is most assuredly important to folks designing microprocessor-controlled stuff, but those times, as big as they are for switching purposes, are simply way out of the  usable range for music. "

The start and the end of diode conduction (the "knee" of diode forward transfer characteristic) does matter, and it matters especially when we are playing at "low gain" settings (at this "knee"), or listen how overdrive "cleans out".

Also different diodes have different leakage current (reverse current) it could be hundreds of microAmpers (some typical Leds) to some nanoAmpers (specialized LRC diodes). This current is also nonlinear and in some situations (for example 100uA reverse current) acts for example as 100-300kOhms nonlinear resistor (depending on reverse voltage). So if we do not want to "hear it", the parallel resistor to these diodes at negative feedback (if the clipping diodes are at the active elements negative feedback path) should be at least 10 times smaller 10-30 kOhms.

[Digital Larry]

I'm not a semiconductor expert er nothing, but AFAIK the published I-V characteristics of diodes don't have an explicit frequency dependence.

Diode switching frequency spec has more to do with switching from forward biased to reverse biased as in a full wave rectifier and involves the speed at which charge carriers can be removed from the junction.  So, roughly speaking, faster diodes tend to be smaller and have lower reverse breakdown.

The impact of the knee on the sound depends on the circuit topology.  Series resistor with shunted back to back diodes to ground will exhibit more "knee" effect on the output than will back to back diodes in an op-amp feedback loop.

Not that it has anything to do with this discussion, but I was messing with a millenium bypass circuit last weekend and had some LEDs which I've had for years that JUST BARELY came on.  I swapped in a new one and was blinded.  I put the two in series to make sure the current being drawn was the same.  New green one still blinding, old red (or green) one barely making a peep there.

[amptramp]

DO NOT mount LED's used for signal clipping on the outside of the enclosure or ambient light will modulate the leakage current, which in a LED is much higher than in silicon diodes.  That means artificial lighting may induce a hum that you cannot avoid and outdoor daylight use may mean you have to adjust the clipping values.

[marcelomd]

DO NOT mount LED's used for signal clipping on the outside of the enclosure or ambient light will modulate the leakage current, which in a LED is much higher than in silicon diodes.  That means artificial lighting may induce a hum that you cannot avoid and outdoor daylight use may mean you have to adjust the clipping values.

Everyone forgets, me included, that LEDs also act as photovoltaic cells. There are a lot of projects that use LEDs as energy harvesting units to power them. Neat stuff. Once I saw a low low power embedded system that was misbehaving. I don't remember the details, but the culprit was the power indicator LED, that was supplying enough energy to upset the internal logic of the microcontroller during reset cycles.

[amptramp]

A company I was at once tried to make a touch panel using green LED's to avoid the high infrared levels in the application (avionics).  It worked but drifted all over the place with temperature because of the leakage and photovoltaic effect (we were using them as photodiodes with them being reverse-biased.)  We abandoned the effort but it left us with the idea that any diode can be a photodiode or photocell and you might not be able to tolerate the effect.

[samhay]

Another thing to consider is that diodes have some intrinsic junction capacitance.
This is on the order of some 10's pF for an LED and varies with the bias voltage. This capacitance is large enough to consider when working alongside high resistances - e.g. in a feedback loop of an op-amp.

The capacitance varies with LED size (larger for 5mm LEDs than 3mm than SMD LEDs...) and must depend somewhat on colour (due to difference in forward voltage) too.

https://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-works-in-a-650-mV-LRC-circuit

[marcelomd]

The capacitance varies with LED size (larger for 5mm LEDs than 3mm than SMD LEDs...) and must depend somewhat on colour (due to difference in forward voltage) too.

https://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-works-in-a-650-mV-LRC-circuit

I thought 5mm leds were 3mm leds with more plastic around... You learn something new every day...

[samhay]

 〉I thought 5mm leds were 3mm leds with more plastic around... You learn something new every day...

You're probably right some of the time.
I recall measuring capacitance of a (small) sample of  LEDs and generally finding it to be smaller in 3mm vs 5mm. packages. Probably not statistically significant, so I could be quite wrong.

[guitar.mod]

Another thing to consider is that diodes have some intrinsic junction capacitance.
This is on the order of some 10's pF for an LED and varies with the bias voltage. This capacitance is large enough to consider when working alongside high resistances - e.g. in a feedback loop of an op-amp.

The capacitance varies with LED size (larger for 5mm LEDs than 3mm than SMD LEDs...) and must depend somewhat on colour (due to difference in forward voltage) too.

https://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-works-in-a-650-mV-LRC-circuit

Also larger pn crystal has lower opening voltage at the same current. It also affects the knee and the sound. For example 1n4148 and 1n4007 diodes I-V characteristics (same principle is for LED's): http://www.indiastudychannel.com/attachments/Resources/160424-2405.gif