Do different sized LEDs have different voltage drops?

[Bishop Vogue]

I mean LEDs of the same colour.  A 3mm vs a 5mm vs a 10mm - same voltage drops?  I've done a bit of research to figure this out but have found conflicting answers.  Most sources say it does not change depending on size, but others seem to contradict that.  Figured I'd ask the collective brain here.  Any help is appreciated. Cheers!

[Marcos - Munky]

Didn't thought about that. I just did a test with 5 3mm clear red leds and 5 5mm clear red leds. Got 1.86V-1.87V voltage drops from the 3mm ones and 2.09V-2.12V voltage drops from the 5mm ones.

[ElectricDruid]

I've never seen any suggestion that there's any link between LED size and forward voltage. Mostly it seems to be related to colour and technology/composition, but I suppose size may have a relation to that - I don't know. I doubt it, frankly.

My guess would be you can get many different forward voltages in all the different sizes and there's no real correlation between them. I doubt anyone has much of a dataset to prove it either way, and even if they did, it would probably only be "statistically true" (e.g. "it is more likely that...") rather than a solid rule for all cases.

Why does it matter, btw? Find something with the voltage drop you need and use that, no?

T.

[amptramp]

I used an LED as a voltage regulator on one circuit I built and the voltage drop was 1.88 volts.  It is defined by the semiconductor alloy used.  The earliest ones were gallium arsenide.  Later ones may add different materials and dopants.

The main distinction is the colour.  The visible range will necessarily be on the order of 1.7 to 3 volts and the UV ones may be higher than that because photons of these wavelengths correspond to energies on the level of this number of electron volts.

The discrepancy between different sizes may indicate different alloys or dopants and possibly differences in wavelength since not all reds are the same wavelength.

[Bishop Vogue]

Thanks guys, I guess we can say that there's no difference.  @Druid, the reason I asked is that I built a little LED selector box to facilitate my designing sessions and I did put 2 of each colour, at 3mm and 5mm, into the thing.  Guess I can go ahead and remove half of them. Cheers, all.

[PRR]

I would suspect that at the same current and color, it's all the same die with more or less plastic around it. Like Ford makes the small Escort but also a big-ass trade-van on the same mechanicals. (I may have the wrong model but the right idea.)

[erio tedesco]

A long time ago (and for many years), I had worked at one of the major players of electronic components producers. For each color, indepedent of the size (3 or 5 mm) and final form of the LED (normal, "arrow", square, etc), the chip was the same. But the variation in electrical parameters varyed a lot from chip to chip, even from the same waffle, causing different visual perceptions. At that time, a big problem to suplly HiFi system manufacturs with their multi-led´s displays .

[R.G.]

The forward drop of any diode depends on
(1) the semiconductor material and its resulting physics
(2) the dopant materials
(3) how much dopant is driven into the material
(4) how abruptly the doping profile changes in the material
(5) in operation, the current density in the junction - how much current you're pushing in makes the voltage go up exponentially at first.

In the case of LEDs, color is set by the exact ingredients of the "brew" in the semiconductor material. The dopants and other things down to current density have smaller effects; the biggie is the brew, as evidenced by the color. It is possible to get the same color by more than one set of ingredients, I suspect, but mostly it's the exact color that sets the junction physics, and that chooses the forward voltage drop per current density.

I say "current density" because if you use a 5 nano-acre chip, it takes less total current to get it up to a certain voltage than a 200 nano-acre chip. So it's current per chip area (or maybe volume...or maybe that's the same thing).

If the chips are the same color, and all the same size in various plastic packages (and I think they are too, agreeing with the earlier posters) then the forward voltage at a given current will be the same.

You do need to be cautious of one verbal trap. Semiconductor junction diodes do not have a single forward voltage. Their forward voltage is an exponential curve, shading into a linear "resistance" region. So without some careful plotting, it just looks like they conduct at a certain voltage. They don't, it's just very close to turning on as opposed to an exponential up.

And you don't want a sudden turn on, probably. It's the shifting voltage per current that's not just a switch closing that gives diodes a softer clipping than the absolutely flat clipping of something like an opamp.

[EBK]

Today, I learned about nanoacres.  Thanks, R.G.! 

Just to make sure that I understand correctly, nanoacres are always used facetiously or at least arbitrarily (i.e., for a generic comparison where the specific area doesn't matter to the discussion), and they aren't an actual unit semiconductor folks use in a serious and precise context, right?

[R.G.]

They're semi-humorous.

An acre is 43560 square feet.
One nano of that is 43.56 micro square feet.
This is 0.00627 square inches.
An inch is 25.4mm so you multiply that by 25.4 twice to get square millimeters.
That's 4.047 square millimeters, or an area 2mm x2mm.
Silicon wafer area for semiconductors of all kinds is measured in mm squared, so 0.25 nano-acres is 1mm squared roughly.

It's accurate-ish, but it is used humorously. Or was some years ago when I was hearing jokes about logic chip sizes on wafers.

I hope I did the math right. 

[PRR]

> I hope I did the math right.

Or you could Google it....

https://en.wiktionary.org/wiki/nanoacre
nanoacre (plural nanoacres) - (humorous) A unit of surface area equal to 10-9 of an acre, or 0.00627264 square inches. {You got it right!}
Usage notes - Used humorously by microchip designers with regard to available physical space on the chip.

http://www.catb.org/jargon/html/N/nanoacre.html
nanoacre: /nan�oh�ay`kr/, n. -  A unit (about 2 mm square) of real estate on a VLSI chip. The term gets its giggle value from the fact that VLSI nanoacres have costs in the same range as real acres once one figures in design and fabrication-setup costs.

too long link
How big is 1 nanoacres?
   
It's about six times as big as a Ballpoint Pen Tip.
In other words, 1 nanoacres is 5.80 times the size of a Ballpoint Pen Tip, and the size of a Ballpoint Pen Tip is 0.170 times that amount. The most common size of ballpoint pen "a medium-tip" measures 0.170 nanoacres. A typical ballpoint pen can write more than 8,500 m (28,000 ft) before running out of ink.

It's about one-twentieth as big as a Nailhead.
In other words, 1 nanoacres is 0.05206 times the size of a Nailhead, and the size of a Nailhead is 19.210 times that amount. (for 11-gauge wire nail) An 11-gauge nailhead measures 19.210 nanoacres. Nails have been used since ancient times; archaeologists have discovered Ancient Egyptian nails that are over 5,000 years old.

It's about one-sixtieth as big as a Dime.
In other words, 1 nanoacres is 0.01607 times the size of a Dime, and the size of a Dime is 62.230 times that amount. (United States ten-cent coin) (a.k.a. Roosevelt dime) The dime, in its current design since 1946, has a diameter of 17.91 mm (0.705 in) and an area of 62.250 nanoacres.

It's about one-seventieth as big as a Penny.
In other words, 1 nanoacres is 0.0142 times the size of a Penny, and the size of a Penny is 70.40 times that amount. (United States one-cent coin) (a.k.a. Lincoln penny, a.k.a. Cent, a.k.a. One cent piece)
The Lincoln penny, in its current design since 1909, has a diameter of 19.05 mm (0.75 in) and an area of 70.40 nanoacres.

It's about one-eighty-fifth as big as a Nickel.
In other words, 1 nanoacres is 0.01145 times the size of a Nickel, and the size of a Nickel is 87.340 times that amount. The nickel, in its current design since 1938, has a diameter of 21.21 mm (0.835 in) and an area of 87.30 nanoacres.

It's about one-one-hundredth as big as a Postage Stamp.
In other words, the size of a Postage Stamp is 100 times 1 nanoacres. A typical postage stamp measures an average of 100 nanoacres.

It's about one-one-hundredth as big as a Quarter.
In other words, 1 nanoacres is 0.008756 times the size of a Quarter, and the size of a Quarter is 114.20 times that amount. The George Washington quarter, in circulation in various forms since 1932, has a diameter of 24.26 mm (0.955 in) and an area of 114.20 nanoacres.

[stringsthings]

Nowz I'm going to go shoppin' for some seeds and grow lots and lots of nanoacres of baby corn.
The futures will be staggering!