On 3/28/2021 1:01 PM, Jeff Liebermann wrote:
On Sun, 28 Mar 2021 11:40:08 -0400, Frank Krygowski
wrote:

I tried the switch (and I do
dislike "soft" switches buried under rubber covers), I felt the wiring
connections, spun the wheel a few times, then it came back on. Rode ten
feet and it blinked off. Then back on and stayed on, mostly, but would
occasionally blink off then back on.

At one stop to listen to a chorus of frogs, the Eyc was running only its
standlight when it suddenly blinked off. That tells me it's not a wiring
connection or an internal dyno problem. It must be a problem inside the
headlamp itself. Also, it's not triggered by bumps AFAICT, and it
sometimes blinked off while riding a perfectly smooth surface.

Fortunately, it always came back to life and for the rest of the 15 mile
ride it blinked out only occasionally, and for never more than a second.
But its very worrisome.

I've seen similar symptoms with LiIon flashlights and battery powered
bicycle lighting. My guess(tm) is a loose connection somewhere in the
wiring system. As you mention, it could also be inside the light. The
erratic intermittent behavior suggests that it's NOT an electronics
component failure, which would be more cyclic.

Most commonly (in flashlights), I see this caused by the LED COB. The
LED is soldered onto the PCB (printed circuit board) forming the COB
(chip on board). The PCB will flex as the LED heats up eventually
breaking the solder connection. The solder doesn't really melt, but
instead cracks. I fix these using a hot air SMD soldering station:
https://www.google.com/search?q=hot+air+smd+rework+station&tbm=isch
Tear the light apart, and touch the LED or slightly bend the PCB while
the lamp is operating. If it flickers, you found the problem. It can
also be components other than the LED, which should be detectable by
bending the PCB or visual inspection. Using the SMD soldering station
takes practice. I suggest you find someone who knows how to use one.

Another problem I've seen are crappy wire connections. These come in
a variety of failures such as cold solder connection, bad crimp,
insulation displacement, jacket creep, and corrosion. The best way to
find these is by probing with a DVM (digital volt meter). The catch
is that you can only find a bad connection if the light is in failure
mode. If it seems to be working normally on the workbench, you're
probably better off taking the light and wiring apart and pulling on
each connection. If the connection moves, stretches, or falls apart,
you found the problem. Corrosion should be obvious. Also, look for
annual rings (usually black) around wires that run through holes in
the PCB. Such rings are cold solder joints, caused by the assembler
being in a hurry, and need to be restarted.

I've only seen one broken wire on a bicycle, which took me weeks to
isolate. I finally found it by pulling on each of the wire ends. One
wire seemed to stretch, indicating a break somewhere. I found where
the insulation necked down from the stretching and spliced the wire.
Problem solved. This wire was NOT under tension and showed no sign of
corrosion, so my guess(tm) is that the wire was defective from the
vendor.

However, my initial wild guess points to the Union bottle dynamo. I
haven't seen any failures with these because I don't use them.
However, if it's as old as you suggest, it might be a good place to
start. Try powering the light with a battery or power supply. Then,
disconnect one of the power wires. The B&M lights usually have a
stand light feature, which is basically a supercap to power the LED
while the generator is NOT turning. If you disconnect the power wire,
and the light goes into stand light mode, and slowly fades away, then
it's working normally. Therefore, the problem is likely NOT in the
dynamo or wiring that power the light. I would need to be after the
stand light circuitry inside the B&M light.

Good luck.

Drivel:
https://phobia.wikia.org/wiki/Dizoangphobia
https://phobia.wikia.org/wiki/Electrophobia
https://phobia.wikia.org/wiki/BMWphobia
https://phobia.wikia.org/wiki/Cyclophobia

Thanks for the information. I've emailed Peter White and will say what
he says before I dig in further.

But as I said, I suspect the problem is internal to the headlamp because
one of the time it shut down, I was stationary. The light was in
standlight mode, during which it gets no external current supply. The
standlight turned off just as suddenly as the main light was doing.

But again, thanks. (I'm hoping that on the bench, powering this with a
6VDC battery should be as valid as powering it with a dynamo.)

--
- Frank Krygowski