Homemade LED Helmet Light Worked for 15 Seconds

[r97]

Hello,

Seeing as it is dark when I get home from school now, I decided I would make a helmet mounted light so I could ride my trials motorcycle a little when I get home. The bike doesn't have a very powerful lighting coil so I decided to go with LED lighting to get the most light from the power I have. With a bit of searching I found this article which I based my project off of, http://www.dirtbikeutah.net/light-1.html. I made a few changes and ended up with the following:

• My bike's regulated 12VAC running to a 2A inline fuse on the "hot" side, and a normally open rocker switch coming from the engine ground.
• A 3023-A-N-1000 BuckPuck, wired with the Vin(+) connected to my fuse, and the Vin(-) connected to the other side of my switch.
• A female 5.5mm center positive barrel connector wired to the BuckPuck's LED outputs.
• A male 5.5mm center positive barrel connector wired to a Cree XPG2 star 3-up cool-white LED in an aluminum housing with a narrow optic.

I wired this all together last night. All the connections were solid and between the correct wires. I started up the bike, plugged the barrel connectors together, flipped the switch and the light came one. At idle the light was a little dim and flickering, but once the engine sped up a bit the light became bright and steady. I set off on the bike to give the light a test and about 15 seconds after I had turned it on, the light went put with a faint popping noise coming from the wiring. Note that the BuckPuck has a built-in 15 second short circuit protection for the output; I don't think it's any coincidence that the light lasted about the same amount of time.

Today I started to try and find out what was wrong with the light, I determined the following:

• There is no continuity between the positive and negative wires of the LED.
• There are no shorts within the barrel connectors, the fuse holder, the switch, or their respective wiring.
• The fuse blew.
• There is not continuity between the Vin(+) and the Vin(-) leads on the BuckPuck.
• There is continuity between the LED(+) and the LED(-) leads on the BuckPuck.
• There is continuity between the red and yellow wires on the bikes voltage regulator (see links at bottom of post).
• There is continuity between the yellow wire going into the voltage regulator, and the engine ground.

I'm not sure what to make of these findings. My best guess is that the BuckPuck shorted somewhere due to "dirty" power from the bikes relatively unsophisticated electrical system, but the light worked fine for about 15 seconds thanks to the BuckPuck's short protection, after which the system shorted, blowing the fuse and possibly the BuckPuck.

Is there any reason why the components I chose shouldn't work together? Was the way I wired all of this correct, or am I missing something like a capacitor? Is the bike's likely "dirty" power the cause of my troubles? Am I on the right track as to what my problem is? Most importantly, what can I do to fix this?

Here are various links that can provide information that may be helpful in figuring out where my problem lies:

• The bike's wiring (my bike uses the late version of the fan circuit): http://www.gasgas.com/Pages/Technical/Electrics/wiring-ducati.html.
• The LED I used: http://www.ledsupply.com/leds/cree-xpg2-indus-star-3-up-cool-white-high-power-led, and its data sheet, http://www.ledsupply.com/content/pdf/leds-cree-xpg2_documentation.pdf.
• The driver I used: http://www.digikey.com/product-detail/en/3023-A-N-1000/788-1108-ND/3114443, and its data sheet, http://www.luxdrive.com/content/3021-BuckPuck.pdf.

I know this might be a very simple problem for most of you, so I really do appreciate someone taking the time to educate someone like me who doesn't really know what the're doing. Please let me know if there is anything that needs clarification.

Thanks in advance, Rory

 

[Illum]

I think harmonics killed it.

Somehow, after a dozen or so projects with the buckpucks, every project that involved installation in a vehicle failed in some comprehensible way. The household installations have remained largely intact and functional. With the DC models, adding an EMIFIL loop followed by a series of 2200uF capacitors seemed to improve its mortality rate on the car, can't say for the AC models.

 

[r97]

Okay thanks. Is my BuckPuck dead like I think it is? Is there a different driver that might work better for my application? Would it be worth getting a rectifier to convert the AC to DC power before it gets to the driver (in which case I would use a DC driver)? In that case the capacitor(s?) and/or EMI filter would still be necessary, right?

Thanks, Rory

 

[RTR882]

Consider powering your light from a battery which is charged from the bike's electrical system.

 

[mds82]

That driver is meant to only work at 60hz AC power. They do offer the BuckBlock which will work at other frequencies and probably will not die.

The other option is to use a rectifier to convert the AC to DC Power, then you can use a DC Puckpuck and use regular capacitors to smooth it all out. That would probalby be the best option http://www.gotwind.org/diy/images/Bridge_Rectifier.jpg

 

[r97]

RTR882, I would consider a battery if I had to, but my bike doesn't have a battery as it is and I would like to avoid adding one If I can.

mds82, the 60hz requiment would explain where my problem is coming from. I'll go the rectifier, capacitors, and DC driver route if I have to, but that seems like it is more complicated than it needs to be. The same goes for trying to run another AC BuckPuck with the EMIFIL and capacitors. I would like to use an AC driver that can simply plug into the bike and the LED without any extra components. Are you saying that the BuckBlock would do that? A quick look through its data sheet doesn't give any mention of that driver working with AC power. Maybe BuckBullets would work? If so could I wire two of the 700ma versions in parallel to power the led with 1400ma? The data sheet says it will run on AC power but I don't see where it mentions what frequency range it is meant to operate on.

BuckBlock: http://www.luxdrive.com/content/A009_BuckBlock_V1.pdf
BuckBullet: http://www.luxdrive.com/content/7006_Data_Sheet_v1.1.pdf

Thanks again, Rory

 

[r97]

RTR882, I would consider a battery if I had to, but my bike doesn't have a battery as it is and I would like to avoid adding one If I can.

mds82, the 60hz requiment would explain where my problem is coming from. I'll go the rectifier, capacitors, and DC driver route if I have to, but that seems like it is more complicated than it needs to be. I would like to use an AC driver that can simply plug into the bike and the LED without any extra components. Are you saying that the BuckBlock would do that? A quick look through its data sheet doesn't give any mention of that driver working with AC power. Maybe BuckBullets would work? If so could I wire two of the 700ma versions in parallel to power the led with 1400ma? The data sheet says it will run on AC power but I don't see where it mentions what frequency range it is meant to operate on.

BuckBlock: http://www.luxdrive.com/content/A009_BuckBlock_V1.pdf
BuckBullet: http://www.luxdrive.com/content/7006_Data_Sheet_v1.1.pdf

Thanks again, Rory