Farnell and RS don't seem to sell anything suitable, which indicates they're a bit hard to find. I suspect they'd be quite big too.
LEDs are easy to find, so if they work well enough...
Cheers
-- Clive
ace.
Solar cells hard to find??? Did you look at eBay? You can find hundreds i n all sizes. I have some that are quite small. Heck, I have a little head swinging monkey toy from the dollar store that used a very small solar cel l to power it. If you can't find a solar cell, you aren't trying.
That's the IF that hasn't been determined yet. You still don't know how mu ch current you can pull before the voltage collapses. That's why I say the world is a strange place. I would find that information before I waste a bunch of time trying to optimize the color of the LED. I also wouldn't was te my time with a phosphor laden white LED.
-- Rick C. -- Get 1,000 miles of free Supercharging -- Tesla referral code - https://ts.la/richard11209
I have a friend who designs, among other things, street lights, and he gave me a few monster custom LEDs, about an inch square slab of aluminum with an array of chips and phosphor. At full power, they are literally dangerous to look at. They need a lot of heat sinking.
How far do you want to go between the chips? A light pipe might help. Maybe you could make your own mega-fiber, just a piece of soft plastic stuff.
Fill up a plastic tube with something else and make a giant step index fiber!
-- John Larkin Highland Technology, Inc lunatic fringe electronics
A photodiode is a current source in parallel with a diode, so if you put it across an LC tank, you get a bunch of p-p AC, if that's worth anything.
Lots of things would be possible except that you are probably working with microamps of photocurrent.
Aren't there photofets?
-- John Larkin Highland Technology, Inc lunatic fringe electronics
It would be part of a commercial product, so reliable, repeatable available etc apply. I could hack something, but it's sort of obvious that solar cells would work, so not worth trying unless they can be found small enough and through proper channels. It's just to turn a MOSFET on, so very little power.
The white LED was all I had to hand, others to arrive soon, though I since found a small SM green LED which will give 1.6V on the DVM (10M IIRC) and 1.5V with 1M across it, so already in the right ballpark for a few of these in series. This was illuminated with a green laser pointer, a red one does nothing and a UV torch does very little, so it looks like the wavelength can indeed be too short.
The white LED wasn't a waste of time - it showed me that I could get usable output despite it being non-optimal and that further investigation is warranted.
But my experiments are very, very ballpark at the moment, just enough to convince me that it can be done fairly easily. I should find out more next week.
Cheers
-- Clive
Yes Optomos and similar. I've used them elsewhere to provide isolated volt-free contacts.
Cheers
-- Clive
OK, here's a goofy idea.
Couple some surface-mount right-angle LEDs on a PCB, using a 2-mode lens (which I just invented in the shower)
This is a double-convex lens in one axis and a TIR internal reflection channel in another. The lens is just a plastic disk. It might even work with a clear washer that has a small bolt hole in the middle.
There's probably a standard polycarb or acrylic washer like this somewhere, for a penny maybe. Three blue LEDs would probably turn on a low-threshold mosfet. Maybe even two.
Don't solvent wash the plastic lens. Install the lens after the board is cleaned.
That may not be your geometry.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Photoelectric absorption has a sharp cliff in the light-below-threshold direction, and a slower falloff slope in the light-above-threshold direction, so you'd expect (due to random variation) green-driving-green to be less reliable than blue-driving-green or shortwave-UV-driving longwave-UV.
Cree Royal Blue (450 nm) would be a good driver for Cree Blue (465 nm) receiver, by that logic.
I did some crude testing with Red, Green, Blue and UV LEDs.
All 1210 SMD 'high brightness' types from the same supplier.
One of each in series at their rated 20mA, then holding another of each against each in turn, no air gap, switching in 1M as a load.
First thing to notice is how much brighter the green appears to be, yes I know that's where the eye is most sensitive, but it does stand out more than I expected. The red and blue appear about the same brightness as each other, the UV emits a little visible violet.
Volts OC/Volts 1M load (OC is DVM impedance)
Receive -> Red Green Blue UV Tx Red 1.53/1.50 3m/0 0/0 0/0 Green 1.59/1.58 2.03/1.15 200m/20m 3m/2m Blue 1.50/1.45 2.12/2.11 2.29/1.90 9m/2m UV 1.42/63m 2.11/2.09 2.34/2.29 2.61/2.47
So Blue --> Green looks favourite, no better than UV --> Green but easier to see.
If TX is longer wavelength than RX then nothing much, if TX is considerably shorter than RX then weak. Best seems when TX is a little shorter than RX as predicted by Martin Brown.
So five Green receivers gives me my 10V. Five Blue transmitters may work with a gap, may need focussing, could use more powerful types.
Cheers
-- Clive
I'm using one family of LEDs where at equal currents, green is dim, orange is nice, and blue is blinding. Had to change some resistors.
Good stuff. Thanks.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
Idea #1 - right-angle SMD LED array [ ] [ ] | )~~> ( | [ ] [ ]
[ ] [ ] | )~~> ( | [ ] [ ] [ ] [ ] | )~~> ( | [ ] [ ] [ ] [ ] | )~~> ( | [ ] [ ]Idea #2 - A search for "multi-chip series green LED" gets lots of results, e.g.
Power LEDs made with series die--
Idea #3 - If you've got lots of time, a simple 2-transistor relaxation boost converter can boost 2V to 10V, allowing you to use just one LED Tx-Rx pair instead of five.
Cheers, James Arthur
I suppose one could square-wave modulate the transmitter LEDs and somehowmumblemumble use the AC on the receive side. That would give a bit more peak voltage at the receiver, for a given average transmit drive.
-- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Only white or RGB as far as I can see?
Cheers
-- Clive
The modulated current seems to be low, so you'd need a lot of primary inductance to generate voltage being asked for.
Not an easy transformer to find or get wound.
-- Bill Sloman, Sydney > > > > -- > > John Larkin Highland Technology, Inc > picosecond timing precision measurement > > jlarkin att highlandtechnology dott com > http://www.highlandtechnology.com
Rats, you're right. I like Cree, but their website and datasheets are awful.
White-to-white wouldn't be ideally efficient, but white dual- and triple-die (6v & 9v) white LEDs are very common.
And Marktech will make whatever you want.
Cheers, James Arthur
There's an old circuit out there that uses an LED in PV mode to collect power, to blink that same LED. Rather than that, Clive could charge a cap up to a threshold, which fires a switch, that charges an inductor. When the switch opens, the inductor flies back.
It's almost easier to draw than describe, but I haven't the time to ASCII it just now. Too busy.
Make a photovoltaic C-W!
Cheers, James Arthur
Digikey has 12 and 24 volt blue LEDs
Look for "lighting" LEDs.
-- John Larkin Highland Technology, Inc lunatic fringe electronics
I looked at one of these and it's Parallel LEDs with a driver of some sort as far as I can tell.
Cheers
-- Clive
It is worth seeking out the LEDs with the highest lumen output and smallest beam divergence that you can for this sort of game.
The QE of some of the highly optimised blue LEDs has benefitted enormously from the white LED trade. The die surface luminosity is now slightly higher than the surface of the sun at that wavelength.
Cyan (505nm) driven by blue might get you slightly more voltage. Or if the Cyan LED is up to it Green driven by Cyan - the overlap is tighter for that pair but the output voltage marginally lower.
The ones with 15 or 20 degree beam collimation make the most sense more so if you can put them onto a clear Perspex rod light guide.
-- Regards, Martin Brown
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.