Roller dynamo slippage problem

On 4/4/2012 2:40 PM, Frank Krygowski wrote:
On Apr 4, 2:27 pm, wrote:
Frank Krygowski wrote:

Stephen Bauman wrote:

The B&M LED lights (head and tail) pack quite a bit of electronics and
optics in them. I'm an EE by profession.

I admire B&M's optics. They've done a good job. Their optical
efficiency is hard to duplicate using the tools in my basement!

Good thing you don't have to, and at a nominal 3W generator output
there are good reasons not to!

?

I've built a couple LED headlights. Although I'm not positive, I
suspect the total lumen output of my home brews is the same as that of
the IQ Cyo I own. But the Cyo will light up a stop sign nearly 1/4
mile away, and beautifully light the road at the same time. The home
brews don't do the same. If I could duplicate those optics, I
certainly would.

- Frank Krygowski

What was the beam angle of your home brews?

On 05/04/12 07:14, Peter Cole wrote:

What was the beam angle of your home brews?

With 4 CREE LEDs to focus, I used 3x 10 degree collimator lenses and 1x
25 degree lens.

--
JS.

On 03/04/12 21:56, Stephen Bauman wrote:
On Tue, 03 Apr 2012 08:29:07 +1000, James wrote:

I've never noticed my Sanyo Dynapower roller dynamo slip in wet weather.
AFAIK, the light is always on while the dynamo is on the tyre and the
wheels are turning at a little faster than walking pace.


snip

I experienced significant slippage in wet weather to the point that I
replaced it with the Soubitez equivalent. The Sanyo had a very light
spring whereas the Soubitez' was much stronger.

Sanyo's light spring also meant it would bounce off the tire whenever I
went over bumps. Sometimes the bounce was so great that it would snap
back and engage the stop. Like others, I cured this by using rubber bands
to get more pressure against the tire.

Sanyo has upgraded its classic Dynapower bottom bracket generator with
the NH-T10 model.

http://www.peterwhitecycles.com/sanyo.asp

I installed one, along with modern LED lights. The dynamo appears to
eliminate all the problems with the old model. There is a rubber covering
for the roller wheel that eliminates slippage and the spring is much
stronger. Also, the LED lights come to full brightness at only 1 watt so
there much less effort is required. This also reduces slippage.

I had another thought. Incandescent bulbs have low resistance until the
filament is hot. This would be like a short circuit on the dynamo
initially, and would cause much increased turning torque.

LED lights, OTOH, only start drawing current when the voltage has
increased to a level that causes them to turn on. This would allow the
dynamo to turn easier at slower speeds, and would lessen the tendency to
slip.

--
JS.

On Wed, 04 Apr 2012 12:14:23 -0400, Frank Krygowski wrote:



I admire B&M's optics. They've done a good job. Their optical
efficiency is hard to duplicate using the tools in my basement!

Regarding electronics: I'm an ME (retired) so not as competent as you
with electronics. But I'm curious about the specifics of the
electronics in those lights. My curiosity comes from the fact that the
critical parameter for an LED is current, and bike generators are
essentially constant current devices, 0.5 Amp. Obviously, there's need
for rectification, and for a capacitor if you want a standlight. But
what else are B&M doing in there?


I'm also a pensioner.

It's difficult to talk about voltage and current sources because there's
a Duality Principle in circuitry. There are Thevenin and Norton
equivalent circuits (one a voltage source with a series impedance the
other a current source with a parallel impedance) that are
indistinguishable from their terminal characteristics.

The dynamo model that I'm familiar with is a variable AC voltage source
with a large inductor in series. The voltage source varies in magnitude
and frequency with the dynamo's rotational speed. The rationale for the
high series inductance was to keep the voltage across the incandescent
relatively constant for large variations in dynamo output.

Electronics can provide much better regulation. There's a lot to play
with because the dynamo's open circuit (no load attached) is quite high.
I've measured a bottle generator putting out 30 vac and very low speed.
The homebrew circuits I've seen will put out higher voltage at low speed
and hold that voltage nearly constant over a wider speed than the simple
series inductor.


It seems many home brewers (like James) are doing fine just with
rectification, or just using multiple LEDs to handle the different
phases of the AC waveform. So what else is needed?

B&M uses a single LED, probably because it makes the optics easier. They
have also tailored their regulator to match the LED's performance. The
B&M light comes up to nearly peak brightness on just walking the bike. My
guess is that they are using a super cap to enhance low speed performance
and possibly another one for a standlight. There's also a design criteria
to convert every bit of pedal power into lumens.

Multiple LED designs that implement half wave rectifiers are not
conducting or emitting light 50% of the time. That's not an efficient use
of a fairly expensive component - the LED.

One other factor goes into B&M's design besides regulation and optics.
It's the heat sink for the LED. It influences not only the LED's life but
also it's light output. Given these three factors, plus a desired
production run of 1 unit, I decided it wasn't worth my time to try to
build a better mouse trap.

Stephen Bauman

On Thu, 05 Apr 2012 08:30:12 +1000, James wrote:



I had another thought. Incandescent bulbs have low resistance until the
filament is hot. This would be like a short circuit on the dynamo
initially, and would cause much increased turning torque.

LED lights, OTOH, only start drawing current when the voltage has
increased to a level that causes them to turn on. This would allow the
dynamo to turn easier at slower speeds, and would lessen the tendency to
slip.

Higher startup drag due to low filament resistance would not explain why
the original Dynapower dynamo would slip after it had been on for a
relatively long time.

We used rubber caps for the bottle dynamos when it rained to reduce
slipping. The problem with the bottom bracket dynamos is that they did
not have any such caps. The NH-T10 has the equivalent of a rubber cap. I
think that is a more important factor as to why it does not slip.

Stephen Bauman

On 05/04/12 11:41, Stephen Bauman wrote:

Multiple LED designs that implement half wave rectifiers are not
conducting or emitting light 50% of the time. That's not an efficient use
of a fairly expensive component - the LED.

Mine has a pair of LEDs conducting for the positive half cycle, and
another pair of LEDs conducting during the negative half cycle. That is
all as far as electronics is required.

Yes, multiple relatively expensive parts and the need for multiple
optical elements, but very simple, robust and efficient. In fact, can
you identify where power is wasted?

OTOH, most switch mode LED drivers can achieve 90% efficiency - see here
for example http://www.diodes.com/_files/design_...zetex/dn85.pdf

One relatively expensive LED and optics with one current regulator,
etc., is better for B&M. They can further complicate the design with
stand lights at relatively low additional cost.

One other factor goes into B&M's design besides regulation and optics.
It's the heat sink for the LED. It influences not only the LED's life but
also it's light output. Given these three factors, plus a desired
production run of 1 unit, I decided it wasn't worth my time to try to
build a better mouse trap.

I used a piece of off the shelf Al angle for a heat sink and mounting
bracket. Cost? I dunno, it was in my bin of offcuts.

All up cost was possibly $50.

Mine has no electromechanical switches or parts that must be protected
from moisture or dirt. In fact the LEDs have zero protection, aside
from the lens glued on to the front of them.

It has been through rain, dirty water, and down hills as fast as my
53x11 will allow. After 1 year of operation, it shows no signs of
weakness or degradation.

--
JS.

On 05/04/12 11:59, Stephen Bauman wrote:
On Thu, 05 Apr 2012 08:30:12 +1000, James wrote:



I had another thought. Incandescent bulbs have low resistance until the
filament is hot. This would be like a short circuit on the dynamo
initially, and would cause much increased turning torque.

LED lights, OTOH, only start drawing current when the voltage has
increased to a level that causes them to turn on. This would allow the
dynamo to turn easier at slower speeds, and would lessen the tendency to
slip.

Higher startup drag due to low filament resistance would not explain why
the original Dynapower dynamo would slip after it had been on for a
relatively long time.

However, you hit a bump and it starts to slip, and the effort to
reignite the candle increases.

We used rubber caps for the bottle dynamos when it rained to reduce
slipping. The problem with the bottom bracket dynamos is that they did
not have any such caps. The NH-T10 has the equivalent of a rubber cap. I
think that is a more important factor as to why it does not slip.

Is wet rubber on wet rubber any better than wet rubber on wet knurled metal?

Perhaps the grooves in the new dynapower rubber provide water channels
and additional contact pressure.

--
JS

James wrote:

:Is wet rubber on wet rubber any better than wet rubber on wet knurled metal?

By a factor of three or four, as I recall. That's why et railroad,
drains, and steel hole-covering plates are so dangerous o a bike.


--
sig 110

Peter Cole wrote:
On 4/4/2012 2:40 PM, Frank Krygowski wrote:


I've built a couple LED headlights. Although I'm not positive, I
suspect the total lumen output of my home brews is the same as that of
the IQ Cyo I own. But the Cyo will light up a stop sign nearly 1/4
mile away, and beautifully light the road at the same time. The home
brews don't do the same. If I could duplicate those optics, I
certainly would.

- Frank Krygowski

What was the beam angle of your home brews?

I'll have to work from memory here, and give vague estimates, because
I've moved LEDs from lamp to lamp. I had some high powered LEDs plus
optics given to me by a friend working in the field, and I also tried
some different optics with them.

One optic I tried was recommended by Chalo, an elliptical one from Deal
Extreme. I'm thinking perhaps 5 degrees vertical, perhaps 15 or 20
horizontal. Far too little throw for my purposes. That one barely made
it out of my basement.

I put more riding time on some sort of spot optic (maybe fifteen
degrees?), and spent some time experimenting with slight changes in the
relationship of the LED to the lens, hoping to shape the beam a bit
better for road use. It was somewhat more conspicuous than a standard
halogen headlight but it lit the road not nearly as well.

Part of my problem was that I was playing with donated LEDs, and some
donated lenses. Most of them were not a particularly good match either
to each other, or to the demands of the road.

About that time, my wife gave me the Cyo IQ for Christmas, which was ten
times better than what I'd come up with, so I stopped playing with them
for a while.

Then I built up a retro 3 speed with a retro Sturmey-Archer Dynohub and
the original matching headlamp, a metal thing nearly the size of a
softball. I reclaimed an LED and put it into that big lamp. On the
plus side, the huge size of the lamp gave more possibilities for playing
with beam shape. On the negative side, the SA headlamp lens was
essentially a small double-convex spot in the center of a plain lens,
sort of projecting an image of the original filament (or now, the LED)
down the road. The lamp also allows for fore-aft adjustment of the
filament or LED. Originally, IIRC, I got one bright spot in a dark
field, not very useful. My latest version of that has a fogged piece of
clear plastic hung in front of the LED, and a tiny aluminum shield as an
"eyebrow" above and in front of the LED, to direct light downward
instead of into the tree branches. That's working pretty well, although
it's used just for neighborhood cruising.

--
- Frank Krygowski

Stephen Bauman wrote:
On Wed, 04 Apr 2012 12:14:23 -0400, Frank Krygowski wrote:



I admire B&M's optics. They've done a good job. Their optical
efficiency is hard to duplicate using the tools in my basement!

Regarding electronics: I'm an ME (retired) so not as competent as you
with electronics. But I'm curious about the specifics of the
electronics in those lights. My curiosity comes from the fact that the
critical parameter for an LED is current, and bike generators are
essentially constant current devices, 0.5 Amp. Obviously, there's need
for rectification, and for a capacitor if you want a standlight. But
what else are B&M doing in there?


I'm also a pensioner.

It's difficult to talk about voltage and current sources because there's
a Duality Principle in circuitry. There are Thevenin and Norton
equivalent circuits (one a voltage source with a series impedance the
other a current source with a parallel impedance) that are
indistinguishable from their terminal characteristics.

The dynamo model that I'm familiar with is a variable AC voltage source
with a large inductor in series. The voltage source varies in magnitude
and frequency with the dynamo's rotational speed. The rationale for the
high series inductance was to keep the voltage across the incandescent
relatively constant for large variations in dynamo output.

Electronics can provide much better regulation. There's a lot to play
with because the dynamo's open circuit (no load attached) is quite high.
I've measured a bottle generator putting out 30 vac and very low speed.
The homebrew circuits I've seen will put out higher voltage at low speed
and hold that voltage nearly constant over a wider speed than the simple
series inductor.


It seems many home brewers (like James) are doing fine just with
rectification, or just using multiple LEDs to handle the different
phases of the AC waveform. So what else is needed?

B&M uses a single LED, probably because it makes the optics easier. They
have also tailored their regulator to match the LED's performance. The
B&M light comes up to nearly peak brightness on just walking the bike. My
guess is that they are using a super cap to enhance low speed performance
and possibly another one for a standlight. There's also a design criteria
to convert every bit of pedal power into lumens.

Multiple LED designs that implement half wave rectifiers are not
conducting or emitting light 50% of the time. That's not an efficient use
of a fairly expensive component - the LED.

One other factor goes into B&M's design besides regulation and optics.
It's the heat sink for the LED. It influences not only the LED's life but
also it's light output. Given these three factors, plus a desired
production run of 1 unit, I decided it wasn't worth my time to try to
build a better mouse trap.

Stephen Bauman

:-) "I'm also a pensioner" and "it wasn't worth my time" have different
values in my life! I find I can putter quite happily with things that
have very little value, beyond satisfying my curiosity!

But I understand. The acquisition of a Cyo did stop some of my experiments.

Anyway, it's true that one can measure 30 V out of a dynamo; but that's
open circuit. To me, the easiest way to understand these things is that
they put out a constant half Amp, or thereabouts. Give them a low
resistance load and the voltage necessary for that 0.5 A is very low.
Give them a high resistance load (or the extreme, an open circuit) and
they try their mightiest to pump half an Amp - that is, they produce
high voltage.

But it matters not to an LED. The LED has effectively low resistance,
and most high brightness ones seem designed for about 0.7 A. They work
nearly as well at 0.5 A, which is what the dynamo will deliver.

Perhaps, as you say, B&M is trying to get every last bit of efficiency.
And the possibility of using a capacitor to help very low speed output
is interesting. But I'd certainly like to see a circuit diagram. (And a
simple explanation, as a non-EE.)


--
- Frank Krygowski