Post by thread:Re: White LEDs etc.

I recently wrote: >I visited the exhibition at Sandown (subject of recent message), looking >for white LEDs and other very bright devices. Very little in the LED line etc etc... Unfortunately I sent it to the wrong place. It was meant to go to a Folding Bicycle Discussion Group, or FBDG (OK, it's weird, but it's legal....). NOBODY on the PIClist noticed it had nothing to do with PICs! However, it generated some queries that are maybe worth answering: > (Leon) Maplin has white LEDS in stock. They are expensive: ~#5. What the FBDG folks want is a replacement for the filament headlamp bulb. The red LED has become very popular as a rear lamp because it has a near-infinite life. Headlamp bulbs don't last long, and dynamos give particular problems due to poor voltage regulation. If you pedal fast they blow. You can buy 2 zener diodes in a plastic case to connect across the dynamo to "clip" the peaks. It costs $10! Sometimes the (100mW) zener diodes also blow, because these bicycle folks aren't very smart. > (Mathias) Was the switch circuitry built in to the lamp, or did you make it yourself? If the latter, I would be very interested in the circuit design. It's a complete working lamp. Inside is an NPN TO92 plastic transistor marked S8050 and D518, 2 resistors, 2 capacitors, and a small transformer with 3 windings. 2 windings are arranged to provide feedback from collector to base, so as to make it oscillate strongly. The 3rd winding has a large number of turns, so a large voltage is produced across it; the tube is connected directly to this winding. It is very crude, but it runs very cool and may be quite efficient. As I said, it takes 0.4 amps at 6 volts; if I modify certain components to make it take 0.6 amps, I get - more light! This is the normal way of getting a high voltage from a low one. Take an inductor (a coil of wire) of value L (in henries), and connect it, using a switch (which can be a transistor) across a battery. The current will build up to some value I (amps). The energy stored in the inductor will be 0.5*L*I*I joules. Open the switch. The energy in the inductor generates a large voltage pulse, in my case large enough to "fire" my tube. All the energy in the coil flows into the tube, so we know where the energy went. We have to keep switching the coil on and off to get continuous light in the tube. If there's no tube across the coil, we may get a spark, so again we know where the energy went. It's then called an IGNITION COIL. Note that the coil should have as LOW a resistance as possible. High coil resistance means poor efficiency. Some of the energy is wasted; it just heats up the coil. > (Ram Krishnen) What was the largest size of white EL panel you saw ? I was wondering if it would be suitable for a compact slide viewer I was planning to build. What a good idea! Completely even white backlighting, instead of the yellowish uneven light with expensive, heavy condenser lenses, on which the dust is in sharp focus. That's what my slide viewer is like, anyway. To be commercially competitive, make the ON/OFF switch from tarnished brass that has lost its springiness, so that it flickers when you fiddle with it. My viewer has this feature. The largest EL panel was about right for the largest LCD screen you've seen; about 10 x 12 (inches), I should say. I don't know why they were set to such a dazzling level, except as an attention-getting gimmick. There were some small panels; you can see them in digital camera LCD viewfinders. It makes one think that your slide viewer has already been invented but it is called something else. See also the posting from William C Westfield for small available backlights. Also at the show were some 17-inch LCD colour displays, styled to replace normal monitors, but very slim of course. The quality was fine but they lacked brightness. They probably also lack cheapness, but I didn't check. So why didn't I ask the salesman, "Are these panels suitable for use as headlights on folding bicycles?" Because I would have been thrown out of the show, of course. > (Michael Coop) (Kingbright RGB LED - 5mm clear and milky versions tested) etc. Some very bright LEDs have, on close inspection of the data sheet, a viewing angle of about 8 degrees. Very hazardous for a rear light. The best one I have found (and indeed I am using it) is a Kingbright "ultrabright red", 12mm diameter, containing 6 LEDs, with a 60-degree viewing angle. At a total current of 120 mA, it's very bright. The "PIC related" bit (I'm really sorry about all this...) My folding bicycle has 16-inch wheels. I bought a cheap cycle computer but I find it can only be calibrated for 20, 22, 24, 26 and 27 inch wheels. Can I use a PIC-chip to adapt this computer to my bicycle? You need to know that there is a small magnet fixed to the wheel, which briefly causes a reed-relay contact to close as the magnet passes it. The contact grounds an input on the cycle computer; it has a 3 Megohm pull-up to the positive rail of the 3-volt Lithium computer battery. Reed-relays have a bit of contact bounce, but not much. There is no ON/OFF switch; a pulse fron the reed-relay powers up the computer. After 5 minutes with no pedal activity, it switches off. There's a challenge for someone. John Blackburn, South London UK.

On Mon, 24 Nov 1997 22:43:28 -0800 john blackburn <spam_OUTjohnbTakeThisOuTdircon.co.uk> writes: > >a near-infinite life. Headlamp bulbs don't last long, and dynamos give >particular problems due to poor voltage regulation. If you pedal fast >they blow. You can buy 2 zener diodes in a plastic case to connect >across the dynamo to "clip" the peaks. It costs $10! Sometimes the >(100mW) zener diodes also blow, because these bicycle folks aren't >very >smart. I think a bridge rectifier, then a "super zener" made of a low-wattage zener from the collector to base of a power transistor would do this job. Shouldn't cost very much. The shunt regulator is probably best since most simple series regulator designs would lose some voltage at low speeds. >The "PIC related" bit (I'm really sorry about all this...) >My folding bicycle has 16-inch wheels. I bought a cheap cycle computer >but I find it can only be calibrated for 20, 22, 24, 26 and 27 inch >wheels. Can I use a PIC-chip to adapt this computer to my bicycle? A 16-inch wheel should make 3 revolutions to cover the same ground that a 24-inch one does in two. So a circuit to discard every third pulse from the sensor but pass the other two on to the computer could be used. Then set the computer up for a 24 inch wheel. If the thickness of the tires makes the result not exact, add or drop a pulse every once in a while so as to remove long-term error. This kind of logic could be easily done with a PIC. Maybe put the PIC to sleep after every pulse and use the RB or GPIO wakeup to wake up. Of course just a RC oscillator would do, and keep count in RAM.