HID bicycle lights

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Haha, yeah if you only do 1.5 no need to fuss. That being said, actually, the bulb life gets extended the less wattage you push. So a lower watt bulb will last longer while a bulb pushed at higher watts lasts a shorter amount of time. The reason is the main determining factor in the life of an HID bulb is its output. Unlike halogens they typically don't just go out. They just get dimmer and dimmer over time. At 35W most bulbs are dimmer than a halogen at around 2000h of usage. Decreasing the wattage extends this time, increasing it decreases this time. So for example, running a 35W bulb at 55W (only do this on quality OEM bulbs) the lifetime drops to around 1000h or so. With cheap chinese bulbs, this lifetime is actually closer to around 300 or so hours.

here's where I got my info from:
Automotive Xenon Metal Halide HID Lamps

Metal halide lamps should not be overpowered, except where permissible for accelerated warmup and near-full light output during warmup. Overpowering one will shorten its life and increase the risk of the lamp exploding.
Underpowering a metal halide lamp is also bad. If the electrodes are not hot enough, they do not do a good job of conducting electrons into the arc and voltage drop in this process (known as the "cathode fall") is excessive. Excessive cathode fall causes positive ions in the arc to hit the electrode at excessive speed which "sputters" electrode material onto the inner surface of the arc tube. For more info on discharge lamp mechanics, look in my Discharge Lamp Mechanics File.
It is not recommended to experimentally operate metal halide lamps at reduced power. Besides the bad effects of high cathode fall on hot electrodes, an unusual temperature pattern can have the chemicals in the arc tube condense in locations that can block some of the light. And if the electrode cathode falls are excessive and unequally so, a DC electric field can result, which can cause destructive electrolysis effects on hot salts on hot quartz. This can cause the arc tube to crack.

Metal halide lamps should have power input within 10 percent of their rated wattage.
 
You are lucky then. Most of the chinese ballasts now have an INPUT power of 35W. This I have tested myself and they usually work out to an input power of right around 35W, some as low as 33 some as high as 36. Either way, the power at the bulb is well out side of the 10% range. I have tested DDM, VVME and about 5-6 other no name companies. Not ONE at the time when I tested them had an input power higher than 36W. It may be in the beginning things like this were an issue but from personal experience there aren't any issues with it in the past few years. AT ALL.
 
Last edited:
Luxeon headlight

continued from: http://www.diyaudio.com/forums/soli...lls-power-amplifier-book-167.html#post2554680

Hi Stein,

First, I'm glad you are back from your 'vacation'.
It took a long time to finish the mechanical part of the job, i.e. finding the right parts and the right guy with a lathe.

So I finally got the time to concentrate on the electronics. For those who didn't know yet, my objective is to power three Luxeon-III LEDs (in series) by means of two A123 LiFePO4 batteries (APR18650), which in turn are charged by a hub dynamo (SON).

The batteries, together with the charge controller, will be put in the steering tube. This means that there is not much room for a bunch of batteries, in fact only for two cells, that is 6.6V. As the three Luxeons need some 11.5V, the battery voltage has to increased by means of a boost converter. This part of the electronics will be put in the housing of the headlight. Again, not much room for big components.

Four functions has to be fulfilled by the electronics:
1. Current regulation of the LEDs (of course).
2. Over voltage protection in case the LEDs are disconnected.
3. Under voltage lockout in case the batteries are exhausted.
4. Over voltage lockout to prevent overcharging of the batteries.

Of course there are a lot of ICs out there that will do the job, but most of them are housed in a minuscule package. Very DIY unfriendly and at my age I can't handle these chips anyhow. So I had to limit my choice to old school DIP packages. Finally, I found them (for free :D)
The MAX1771 for function 1 and 2 and the MAX923 for function 3 and 4.
Now I'm waiting for a small but high DC current choke with a toroidal Ni-Fe core and distributed gap.
When all has been finished I will put the schematics, the test rig and some pictures on my website.

Cheers,
E.

PS: Don't put Li-Co batteries in the steering tube. If they explode you will get the debris right in your face! Also, don't use cheap LiFePO4 batteries, as they can't handle the high currents, at least the small ones.
 
Hi Wakibaki,

Thanks for the link.
As for my 'stomach', I know you can spend an insane amount of money on this stuff.
Happily, I got most parts for free, except the hub dynamo (SON28), which I bought at a discount of 33%. Nevertheless, it was still ridiculous expensive (120 euro), but I refuse to use a cheap Shimano hub that consumes even more mechanical power when the lights are off than when they are on. For lowest drag you better short circuit this kind of cr@p !!! Apparently, Shimano has never heard of eddy currents.

Cheers,
E.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.