On 2018-05-03 19:36, bob prohaska wrote:
Sepp Ruf wrote:
David Scheidt wrote:
sms wrote:
:On 5/2/2018 7:49 PM, bob prohaska wrote:
: I finally caved and bought a Shimano dh-3n72 hub for use on my
: exercise bike. It works fine, but the LEDs flicker quite noticeably
: at all speeds. The circuit puts series trios of LEDs in inverse
: parallel across the dynamo, so each trio lights on alternate half-
: cycles. The flicker wasn't visible with the old Soubitez roller,
: because it was working at about 6x higher frequency. The circuit is at
: http://www.zefox.net/~bob/bicycle/schematic.gif
: The point of the design is to eliminate rectifier losses, which are
: substantial at low voltage.
:
: The new setup might be ok for fast riding. At low speeds and in traffic
: I'm less sure I'll like it and fairly certain bystanders won't like it
: at all, especially at night. As a DRL it certainly aids conspicuity 8-)
:
: Can anybody point me to a discussion of what outfits like B&M use in
: their dynamo-powerd LED headlights? I've searched intermittenly but
: never hit the right keywords.

The classic B&M circuits used rectification and current-doubling to fully
power one 1 amp led. Your 2x3S LED approach is quite different.

http://candlepowerforums.com/vb/showthread.php?430445
Good luck!


Note the three big 470uF capacitor in the schematic. That's the "flicker
reduction trick" :-)

Problem is, electrolytic capacitors do not live long in harsh climate
conditions such as the black light enclosure baking in the glistening
Arizona desert sun for hours every week or riding around the Klondike in
a harsh winter.

For me the only proper way of lighting a bicycle is with a rechargeable
battery. The LEDs get clean DC and the battery can be recharged from any
source, even a pulsating one. That way the lights will not dim to "stand
light" or go out when waiting at a traffic light. Why should bicycle
lighting be inferior to car lighting?

--
Regards, Joerg

http://www.analogconsultants.com/