Clipping LEDs

[chris84]

As far as LEDs as clipping diodes go, anyone had any success with Superbright torch LEDs, or little tiny ones? That's quite vague really...

[Minion]

I believe you should probably use Low Voltage/Current LED"s (1.5-2v) because with the Higher Voltage LED"s it takes much more signal to achieve clipping and Clipping might not even be achieved if the Voltage of the LED is too high......I generally don"t like the sound of LED"s as Clippers but when I do use them I usually use the Small RED 1.5v 10-20ma Ones.....

[Electric_Death]

I've got some really bright red ones..something like 2000 mcd's, maybe even a bit higher. They take a lot of juice to get them clipping so basically they're useless unless you wanna make a laser pointer activated by your guitar playing.

The one thing about the reds however is they seem to be the favored color as far as which colors sound best.

[dxm1]

...which colors sound best.

 

[raulgrell]

It's not mojo, different couloured LED's have different foward voltages, giving them a different sound... I'm sure there's some variation in other aspects as well, but I'm not learned enough to tell you about them...

[wampcat1]

my ears tell me that different colors do sound a little different, but not drastically. I find that one germanium diode to the next has more variation than different color LED's and MCD rating (generic vs. ultrabrights). Some guys do like the ultrabrights better though, and they definitely look cooler when use in something like the ds-1 seeing eye mod.
bw

[Electric_Death]

I'm under the impression it's the clear but colored LED's that sound different since the dielectrics used are responsible for what color they produce. Like I always say, tungsten will sound different than copper etc.etc.etc. When I've surfed around online, most guys swear by reds and blues, tell you to avoid greens and a couple others.
I wonder what they have to say about the I.R.'s

I bet there's some voodoo to them that needs to be explored.

[johngreene]

I'm under the impression it's the clear but colored LED's that sound different since the dielectrics used are responsible for what color they produce. Like I always say, tungsten will sound different than copper etc.etc.etc. When I've surfed around online, most guys swear by reds and blues, tell you to avoid greens and a couple others.
I wonder what they have to say about the I.R.'s

I bet there's some voodoo to them that needs to be explored.

LOL
Yes, different diodes will sound different. Germanium diodes sound different that silicon. 1n4148's sound different than 1n4007's. Red LEDs sound different than Green LEDs. All for the same reason. I wouldn't go so far as to say some are just 'better' than others. It all depends on how they are being used. LED's are diodes. And just because an LED doesn't light up doesn't mean it isn't conducting. They just require a certain level of current to produce light. Most distortion circuits will not provide that much current. At least hopefully not, because battery life would be really, really bad.

--john

[dxm1]

That was kinda my point with the smileys... LED are semiconductors. They have physical properties that control their behavior in a circuit. Color is simply a byproduct. If one wanted to catagorize an LED's behavior in a circuit in poetic terms, perhaps 'High Efficiency' or ' Ultra Bright' would make more sense. It all comes down to Voltage, Current, and Resistance.

[culturejam]

The one thing about the reds however is they seem to be the favored color as far as which colors sound best.

Most of the LEDs I've seen get their color from the plastic bulb and not anything intrinsic to the component. You know, the crap LEDs you get from RadioShack?
I can't see much difference being made by a different colored plastic.

But I do have a few LEDs that have clear plastic with colored light.

[dxm1]

Read this:

http://en.wikipedia.org/wiki/Light-emitting_diode

Pay attention to the table listing the foward DC voltage (forward DC potential) for different recipes. Different dopants (impurities) added when the crystal is grown control the emitted wavelength (color) as well as the threshold voltage when the diode starts to conduct.

This voltage is what is important for 'Cliping Diodes'.

Manufacturing technology has advanced to the point to where we can now house any color LED in a 'Water Clear' epoxy glob and get good color generation. The old days of adding pigment to the 'plastic' covering to force the color we want are pretty much gone.

The datasheet is your friend. If you want to try different clipping behavior, grab a bunch of diodes (and LEDs) with different threshold voltages. One tenth of a volt will make an audible difference at higher gains. Audible, not visual.

In the end, it's math, not color, that determines how the clipping happens. Too bad, it sounds a lot better to say 'Red Rocks,  Green sucks'. Unfortunately, that may not always be the case...

[oldrocker]

Well I for one only use Radio Shack LED's.  They're the best I found for clippers and LDR combinations when other brands of the same color wouldn't work right.  Some other brands I bought had different specs even though the LED color was the same as their RS counterparts.  IMHO brand of the LED's is just as important as color.  Read the specs before you buy.   I've used the RS yellow LED's with much success with many builds that were using red. I never use the green except for light indicators.  In that area it's a matter of opinion.

[wampcat1]

I always buy my LED's by the truckloads from this fellow:
http://stores.ebay.com/Chi-Wing-LED-product-shop

they are very nice - and as brights as you can stand it! Plus, they are cheaper than anywhere else that I've found...
bw

[johngreene]

Well I for one only use Radio Shack LED's.  They're the best I found for clippers and LDR combinations when other brands of the same color wouldn't work right.  Some other brands I bought had different specs even though the LED color was the same as their RS counterparts.  IMHO brand of the LED's is just as important as color.  Read the specs before you buy.   I've used the RS yellow LED's with much success with many builds that were using red. I never use the green except for light indicators.  In that area it's a matter of opinion.

I'm working on a giant LED sign currently at my day job and the word I get is that there are only a few LED 'die' makers. Many of the LED companies use the same dies and put them in their own unique packaging. Unfortunately there is no good way of telling if LED company 'A' uses the same die company as LED company 'B'.

--john

[JDoyle]

I bet there's some voodoo to them that needs to be explored.

There is no voodoo. There is no mojo. There is only Mother Nature and man's attempt at controlling her.

The difference in the sound of different diode combinations is determined by the threshold voltage of the individual diode, which itself determines the relative magnatude of the clipped output signal, and the V/I transfer curve of the individual diode as it goes from off to on, which is the factor that accounts for the differences between diode types (and individual diodes of the same type) in the harmonics created by the act of clipping.

The reason claims of 'voodoo' or 'mojo' arise stem from the fact that the reported results of any 'qualitative' testing are by the vary nature of what we are doing subjective, because they relate to sound and thus are entirely determined by personal preference, and therefore are essentially useless.

Beyond that, those two factors, Vth (the threshold voltage that determines the clipped output signal's relative magnatude) and the V/I transfer curve (the rate at which the diode goes from fully 'off' to fully 'on') are both dependant on the instantaneous voltage across, and the instantaneous current through, each individual diode (the Vth actually does change as you push more and more current through the pn junction; the transfer curve doesn't change in any significant way, it's more that different magnitudes of input voltage and current cause the signal to travel through the transfer zone at different speeds changing the shape of the 'corner' of the clipped output signal). In other words, both of those variables are circuit dependant.

So going further, I would contend that if a person were to find the exact diode combination they liked in a stock TS circuit, and then that same person were to scale the resistance in the feedback network so that the ratio of resistance, and thus the overall range and amount of gain, were to remain the same, but the current through the network were to say double or triple (or be cut in half or a third), there is more than a passing chance that person would not find the original prefered diode combination to be the ideal for the new circuit.

And continuing even further, if you factor in part tolerances and all the other variables that occur because of natural variation between the same circuit, two people, especially if they both have 'golden ears', would not be able to REALLY objectively judge the difference, and thereby the difference in their individual prefrences, unless they were using the EXACT same circuit and using the EXACT same diodes (one box and batch of diodes that they trade back and forth).

So you can put as much stock in other's opinions of the qualities of using certain types of diodes (or colors of LEDs) as you want, but I really think that unless you are using those opinions as a guideline to calibrate your preferences with those of the holder of the originial opinion, and are both referencing the EXACT same circuit in which the clipping will occur (and even then there is a lot of gray area), any and all opinions on the relative qualities of different clipping diodes are simply useless.

Put the circuit on a breadboard and try them out for yourself. It's the final arbiter of whether or not you yourself like it...

Regards,

Jay Doyle

[JDoyle]

Pay attention to the table listing the foward DC voltage (forward DC potential) for different recipes. Different dopants (impurities) added when the crystal is grown control the emitted wavelength (color) as well as the threshold voltage when the diode starts to conduct.

This voltage is what is important for 'Cliping Diodes'.
...
In the end, it's math, not color, that determines how the clipping happens.

In the end it is math, but while you are doing a great job of following the doctor's maxim of 'when you hear hoofbeats, it's most likely a horse and not a zebra' (meaning that in most cases, its the mundane, not the exotic, that is the root cause), I think you chose the wrong horse.

Vth is Mother Nature; silicon is 0.6V, germanium is 0.3V, etc. That is why textbooks can state categorically that those voltages are true. It is intrinsic to the substance with which the diode is made. But in the end it is only a determinant of the relative output size of the clipped signal and neglibly important as to the tone and/or harmonic content of the clipped wave. It is the transfer curve that is the cause of the difference.

The transfer curve is determined by the doping formula, the width, length and shape of the depletion zone, etc. and each of those factors is decided by the manufacturer to optimially create the type of diode they want.

For example, 1N914s and 1N4001 diodes sound different as clippers. But they are both silicon and both have a threshold of 0.6V (most datasheets for the Vth of the 1N400x series show it as 0.9V to 1.1V but note that the datasheet is measuring that value with 1 full ampere through the diode, vastly more than we will ever hit a diode with) yet they sound different. How is that?

It is because the slope of their individual transfer curves are different so as the signal starts to turn each of them on, the rate and shape of the initial clipping (the corner of the waveform as the diode JUST starts to clip but before it flattens out when the diode is fully 'on') is different for each diode because the electrons travel through the diodes in different manners. So even though that curve occurs at the same voltage for each diode, because the curves are different, the resultant sound is different.

Regards,

Jay Doyle

[Mark Hammer]

Not knowing nearly enough about diodes, here's a question.

Although an LED is merely a diode with a given forward voltage and light as a byproduct, I'm curious as to whether there is any difference in the conductive behaviour of the diode when and if it reaches a point where visible light is produced.

Or is it the case that it has a linear response in terms of the current/light tradeoff, and always produces light, just not always enough light to be visible above the ambient light conditions?

Finally, I have some either UV or IR leds I bought a year or two ago for remote-control building.  Unfortunately, I failed to label them and they sit in a plastic bag, unusuable because I don't know if they give off light or respond to light.

I can still visually identify which end is cathode and which anode.  Are these usable for clipping purposes, or are we talking impractical forward voltages?

[johngreene]

Not knowing nearly enough about diodes, here's a question.

Although an LED is merely a diode with a given forward voltage and light as a byproduct, I'm curious as to whether there is any difference in the conductive behaviour of the diode when and if it reaches a point where visible light is produced.

Or is it the case that it has a linear response in terms of the current/light tradeoff, and always produces light, just not always enough light to be visible above the ambient light conditions?

Finally, I have some either UV or IR leds I bought a year or two ago for remote-control building.  Unfortunately, I failed to label them and they sit in a plastic bag, unusuable because I don't know if they give off light or respond to light.

I can still visually identify which end is cathode and which anode.  Are these usable for clipping purposes, or are we talking impractical forward voltages?

Yes, there is a difference. The 'sweet spot' with diodes is the knee when it first starts to conduct. Especially when it is used in a feedback arrangement. This is because very little current flows, or needs to flow, through the diode to change the response of the opamp. When the LEDs go to ground, the low impedance drive of the opamp will dump as much current as it can into the diodes (unless limited by a series resistor). To put it simply, I suspect you would notice much more of a difference in diodes when used in a feedback arrangement than you would if they were going to ground.

IR LEDs have a voltage drop similar to other LEDs, 1.5-1.7 IIRC, so they will work just as well as any other.

IMO, if your clipping LEDs are lighting up, you are using them in their most boring range. 

--john

[JDoyle]

Not knowing nearly enough about diodes, here's a question.

Although an LED is merely a diode with a given forward voltage and light as a byproduct, I'm curious as to whether there is any difference in the conductive behaviour of the diode when and if it reaches a point where visible light is produced.

Or is it the case that it has a linear response in terms of the current/light tradeoff, and always produces light, just not always enough light to be visible above the ambient light conditions?

The cost of the voltage drop acoss the diode when it is conducting is a release of energy as the electrons cross the depletion zone. In diodes made from some materials, like silicon and germanium, that energy is given off as heat. In diodes of other materials, such as those found in LEDs, that energy is a release of photons and the velocity of the photons determines the wavelength and therefore the color and because it takes different amounts of energy to cross depletion zones made of different semiconductor materials, you can have different colors.

So you are right in thinking that it always produces light; when it first starts conducting the current through it isn't sufficient enough to release enough photons to make it visable, but as you increase the current, more photons are released making the light brighter and brighter. This is why different values of the current limiting resistor before or after an LED allow for different levels of brightness; you are putting a cap on the amount of current through and therefore the amount of energy released.

Finally, I have some either UV or IR leds I bought a year or two ago for remote-control building.  Unfortunately, I failed to label them and they sit in a plastic bag, unusuable because I don't know if they give off light or respond to light.

Actually, I'm pretty sure that all LEDs do BOTH. They give off light and they also respond to light shone upon them.

Another way to think of the 'sweet spot' John was talking about is this:

Mother Nature decreed that for a diode made of silicon, when the voltage across it is less than 0.6 Volts, no current will pass through it; and when the voltage across it is greater than or equal to 0.6 Volts, all current passes through it.

If you think of Voltage to Current output curve of this diode, you have a flat line along the voltage axis at zero current until the voltage axis reaches 0.6 Volts, then there is just a straight vertical line to infinity current at exactly 0.6 Volts.

This is an ideal silicon diode. If you were to put a pair of these ideal diodes in a Tube Screamer, the output wave form would be squared off; the output would ubruptly and absolutely stop at, and continue to be, 0.6 Volts when the output is sufficiently large (0.6V or larger) to turn on the diodes. I don't think this would be sonicly pleasing as the sharp corner would create a lot of high harmonic content.

Unfortunately for engineers, but fortunately for us, man hasn't been able to manufacture this ideal diode yet and there are defects that take a 'real world' diode away from the ideal. Some of these are fairly harmless if obided by, such as, there is a maximum amount of current that you can pass through a diode before you have to buy a new diode. But others are more of a hassle and these can be leakage current below Vth, parallel capacitence, or the fact that the the output curve does NOT instantly change over from zero current to full current but in fact changes so that at some point at or before Vth a little bit of current starts and the diode FOR A VERY VERY SMALL VOLTAGE RANGE acts like a resistor with a non-linear relationship to current. If we put it in terms similar to Ohm's Law then Vd=ZdId where Vd is the voltage across the diode, Id is the current through the diode and Zd is the transfer curve of the diode. This curve is equal to zero when Vd << Vth, is equal to the maximum the diode can pass when Vd >> Vth and has a unique curve and relationship between Vd and Id that is dependent on the diode's unique physical construction when Vd is a little above and a little below Vth; with the amount 'a little' being determined by the diode's unique physical construction as well.

It is this Zd, this device-dependant relationship between voltage and current when the diode is just starting to conduct, that gives each type of (and even individual) diode it's own sound and response.

The reason that diodes can be made of the SAME material yet still sound different in the same circuit is that the inventors of each diode after the first one created were trying to eliminate different defects, such as parallel capacitance, but as a diode is essentially just a depletion zone, they would have to change the shape or doping of that zone to affect the change in performance they needed. So in improving the characteristic they wanted to, they also changed the depletion zone and with it the transfer curve at and around the point of conduction.

I hope that made sense,

Jay Doyle

[johngreene]

This is an ideal silicon diode. If you were to put a pair of these ideal diodes in a Tube Screamer, the output wave form would be squared off; the output would ubruptly and absolutely stop at, and continue to be, 0.6 Volts when the output is sufficiently large (0.6V or larger) to turn on the diodes. I don't think this would be sonicly pleasing as the sharp corner would create a lot of high harmonic content.

Actually in a Tube Screamer circuit it would follow the input once the diode began conducting. It is basically a short after the .6 volts which would make the opamp a voltage follower.

Don't mind me, I'm just picking nits....

--john