LED headlight again

[ducwiz]

Jim,

please, can you decribe your LED circuit more in detail, or post a simple sketch of it here?
In my opinion a voltage limiting regulator doesn't cure your problem of blowing the LEDs. The LEDs are current driven devices, they intrinsically limit the voltage across their terminals, independant of the current they carry. They behave exactly like zener diodes, the typical voltage for a single red LED is 1.6-1.9 Volt.
In my point of view you need a current limiting or constant current circuit. This also can be put together from cheap components - oh sorry, I forgot your limited free time ...
Afaik, there are such circuits available for home illumination by LEDs, but maybe for fixed larger currents only.

Hans

[DewCatTea-Bob]

[quote= ducwiz ...
" please, can you decribe your LED circuit more in detail, "

____ Until Jim has pleased our requests on THAT circuit,, in the meantime, you could further wet your appetite by checking-out his LED.taillight-story... viewtopic.php?f=3&t=1420#p9985



" In my opinion a voltage limiting regulator doesn't cure your problem of blowing the LEDs. "

____ I don't think Jim has started-out with bare LEDs, but rather LED.sets that were pre-circuited to properly work with a 12volt-DC.source. - (Right Jim ?)



" The LEDs are current driven devices, they intrinsically limit the voltage across their terminals, independant of the current they carry. "

____ I don't think that Jim had ever actually-meant to declare that it was specifically 'voltage' itself that actually burnt-out -("blew up") all the LEDs themselves. _ But rather, I believe that he actually had meant that the prefabricated 12volt-LED.circuits had become rendered inoperable, (and had merely-just stated "blew-up" to drive-across his point, right Jim ?).
__ So in such case as those '12v.LED-circuits' having been well overwhelmed with excessive 'power', then that occurrence either overheated the individual LEDs themselves or perhaps the LED.circuit's matched-resistor (with too-much 'current'),, thus ether-way, pretty-much killing-off the 'LED.circuit' (of each of his two LED.lights).
So rather than 'voltage' blowing-up his LEDs, it was actually excessive 'current' that either burnt-out or burnt-up his installed 'LED.circuits' (depending on which part of the LED.circuit that was actually heat-burnt).



" They behave exactly like zener diodes, "

____ I don't think I've ever been aware if that.
So then more certainly,, the matched-resistor (within each of his LED.sets), having been preset for properly handling under 24-volts,, once subjected-to further increased voltage, then allowed more current-induced HEAT than the LEDs (or-else their master-resistor) could possibly dissipate, and-so thus-then were overheated & burnt-out (or burnt-up).


____ Since Jim has left us in the dark (as to exactly what his headlight's actual LED.arrangement really is),, then assuming that his LED.sets were preset LED.circuits meant to be powered by a 12volt-DC.power-source, then regulating the power-source to keep it from developing higher-voltages (that would force too-much current through the LED.circuits), should naturally allow the LED.circuits to then perform as intended.
(So I don't see how voltage-regulation of the power-source may-not cure Jim's issue.)

____ BTW, all your input is much-welcomed Hans ! _ And I'm sure that Jim & myself are-not the only ones who'd like to learn more about what your thoughts were concerning the doubt that you've raised, (after Jim has provided adequate info).
__ (I wish MotoMike hadn't stopped providing his input !)

____ Hey Jim, how about telling some size & tech.details about your chosen LED.headlight ? _ And how you got it mounted. _ And-also, why-not mount a PAIR of those fairly-small (100mm ?) HL.bulbs, one atop each fork-leg ?


Enlightening-Cheers,
-Bob

[JimF]

This will be a quick response -

These are not conventional LEDs but rather are the high power LEDs.


see:

http://www.philipslumileds.com/products/luxeon-rebel

There is no Vf spec and current limiting resistor. They are driven by a charge pump type of circuit that monitors and adjusts the current flow as opposed to putting a bias voltage across the LED and setting the current limiting resistor based on expected supply voltage.

There are more than two dozen components driving three high-power LEDs - quite far removed from the convention 1 LED/1 resistor or several LEDs in series with 1 ressitor arrangement.

As for the details, let me get the thing to work on the bike before I post up the details... But for now let me tell you I modified a Hella 90mm headlight assembly and inside are three Luxeon Rebel white emitters.
The problem with the Hella 90mm as supplied was two-fold;

The Hella had a 12-volt halogen bulb of 65 watts.

The Hella had an overall lenght that was somewhere on the order of 8 inches. I wanted to hide a stubby headlight below the tach,

What the hella unit offers is mostly the appearance of legality. It has DOT or SAE markings that would pass for street legal if inspected. Never mind that the light is far brighter than the original and streel legal DOT approved headlight being replaced.

BA15S+HP White.jpg

BA15S+HP Lit.jpg

The LEDs are not the issue - I am looking to regulate the voltage that goes to the complex circuitry that drives these high-power LEDs. For our purposes the LEDs are not at all on the motorcyle's power source.

You probably would not want to plug your 12-volt Garmin power cord into a power source that could peak at 30 volts or higher - that's the same situation I am facing here. At some point the internal semiconductors will reach their breaking point.


Let's say I recitfy and regulate the ignition winding. How would the ignition system operate (that is to say the coil) if I switched the ignition to DC. I know the DC bikes received a different coil. What would happen if I used an AC coil with DC power?

To carry that one step further, since my lighting system will only consume 6 watts, if I put both the igntion on the DC and the lights too, what if anything do I need to do with the leftover pigtail from the un-used winding?



Jim

[DewCatTea-Bob]

[quote= JimF ...
" But for now let me tell you I modified a Hella 90mm headlight assembly and inside are three Luxeon Rebel white emitters.
Never mind that the light is far brighter than the original and streel legal DOT approved headlight being replaced. "

____ Are you really meaning to say that just three of those LEDs are actually brighter than the 65w.H4-bulb ?



" These are not conventional LEDs but rather are the high power LEDs.
There is no Vf spec and current limiting resistor. They are driven by a charge pump type of circuit that monitors and adjusts the current flow as opposed to putting a bias voltage across the LED and setting the current limiting resistor based on expected supply voltage. "

____ Your LED.setup is too modern for me to know anything about, so I can't be much help concerning them (and their-own circuitry).
(And it seems you're not somebody with the time to properly teach us.)



" There are more than two dozen components driving three high-power LEDs - quite far removed from the convention 1 LED/1 resistor or several LEDs in series with 1 ressitor arrangement.
The LEDs are not the issue - I am looking to regulate the voltage that goes to the complex circuitry that drives these high-power LEDs. For our purposes the LEDs are not at all on the motorcyle's power source. "

____ This is enlightening,, and it makes sense that the LEDs themselves are-not to be concerned with, as their-own rather complex prep.circuity eliminates any direct-concern for THEIR-requirements.



" You probably would not want to plug your 12-volt Garmin power cord into a power source that could peak at 30 volts or higher - that's the same situation I am facing here. At some point the internal semiconductors will reach their breaking point. "

____ So thus your only real-concern is-not directly-with the actual LEDs at all, but RATHER direct-concern with protecting their prep.circuitry from burn-up.
So it was fairly-MISLEADING to have stated that you had "blew up" your 'LEDs',, cuz with all their-own directly-involved circuitry, some part of THAT must've been what had really-actually gotten fried !
__ So you ought-to have, (if not actually need), a rather constant power-level most-like that of a battery,, so if you would consider such a sure-fix, a battery for a hand-drill could likely be easily set-up to be kept charged, while it reliably powers the LEDs.
And if you don't like THAT-fix, a small unnoticeable capacitor could likely 'fill in' !
Do you understand how a capacitor 'filters' p.DC into straight-DC (for such light loads) ?




" Let's say I recitfy and regulate the ignition winding. "

____ You have lot's of other options to consider than full-wave rectifying & regulating the ign.power-coil ! _ And if you realized the workings of them all, you quite-likely would realize that this-route is not YOUR best-choice !
__ But of-course it's my favorite-thing to discuss ALL the pros-&-cons details of each & every possible option ! _ So if you wish, I'm quite willing to exhaust this option (that you're seemingly stuck on).



" How would the ignition system operate (that is to say the coil) if I switched the ignition to DC. "

____ Your green-coil ign.system ALREADY works on DC !!! _ But just a single 'pulse' of it, (while the rest of the pos & neg DC.pulses are IGNORED & shorted to ground through the ign.points !) !
So if you regulated that required pulse of DC (along-with all the other DC.pulses [of the raw AC] ), then you'd no-longer obtain the full effect of the magneto-type ign.system and your ign.spark would be weakened & less reliable than either of the two established Duke-ign.system types ! _ As straight-DC has no flywheel-effect to jam-pack the current-juice through the ign.coil to shock it into QUICKLY building-up an induction-field.
__ If you insist on full-wave rectification & regulation,, then you will need to do the same as Nick did, and switch-over to the Monza/Mach-I type of ign.system.



" I know the DC bikes received a different coil. "

____ Yes, the 'RED-coil',, which rather works on the COLLAPSING-field process.



" What would happen if I used an AC coil with DC power? "

____ There's actually no such-thing as an AC.powered ign.coil,, your Mark-III's green-coil doesn't actually work off-from full-AC, but rather just ONE pulse of it (which is DC !). _ That DC-pulse is expected to have a greater punch of power-juice to shock the ign.coil-winding with, and-so the primary/secondary turns-ratio is not as great. _ So with regulated-DC,, you'd still be able to get a spark, but it would be weaker than it ought to be !
That your power-system would be 12-volts, means that your green-coil may work within the ball-park / well-enough, (but I'd look for a more suited ign.coil to take it's place later-on).



" since my lighting system will only consume 6 watts, if I put both the igntion on the DC and the lights too, what if anything do I need to do with the leftover pigtail from the un-used winding? "

____ Then just isolate/insulate it's terminal-end & ignore the wire-lead from the lighting-coil.
__ BUT however, even-though your lighting-system consumes such a small amount of power,, it would quite-likely have a very hard-time finding the 6-watts, since whenever your ign.coil is-not consuming it's high-rate of current-consumption, your M.III points-cam is then allowing the points to dump the REST of the available power-juice to GROUND, thus-then leaving your lighting-circuit to compete with a SHORT-CIRCUIT (until the cam opens the points [and the power is then (briefly !) bled-through the ign.coil]) !!
So therefore you'd need a 28-degree type AAU, to take the place of your stock points-cam.
And even with that AAU's well reduced dwell, your lights will likely flicker without a battery or a capacitor.


____ So Jim, why are you putting-off the simplest cure for your issue, and ignoring the AC.regulator-method that would limit AC.voltage to your rectified LED.circuits ??


Hopeful-Cheers,
-Bob