Thinking Outside The Box

On 13/03/12 14:42, Ralph Barone wrote:
Peter wrote:
On 3/12/2012 5:50 AM, Phil W Lee wrote:
Jeff considered Sun, 11 Mar 2012
16:09:07 -0700 the perfect time to write:

If you're referring to my mention of PWM in bicycle lighting, that was
verified by me with several LED flashlights and one Bike Planet
bicycle headlight. At such power levels, one would need to be a
masochist to use a linear regulator (except in retrofit flashlight
bulbs). Would you like me to post an oscilloscope photo?

So based on your analysis of ONE bicycle light, you assert that all
use PWM?
Why on earth would they?
All you need is a current limiting device, which is typically a
resistor,

No, it's typically a switching regulator, which modulates duty cycle, AKA PWM.

The use of a switching regulator does not necessarily result in the LED
being switched ON and OFF at a high rate. I haven't analyzed any, so I
can't comment on how many bike light current regulators PWM the LED, versus
more sophisticated control methods.

You're on the money. It depends on the DC-DC converter topology, but
the current may be continuous and not cause the LED to turn off at all.

See here, page 6 shows the current through the inductor is continuous.

http://www.ti.com/lit/ds/symlink/lm3405.pdf

The same circuit _can_ be used to flash or dim the LED, by toggling the
enable/dim input pin.

--
JS.

On Mar 12, 7:35*pm, James wrote:
On 13/03/12 12:55, Dan O wrote:
snip

Mostly because it adds weight that I do not need to have, and it costs
me a minute to remove. *I would remove a battery powered light for the
same reason. *Our racing club now mandates rear red blinkies for winter
road races, so not all lights come off.

Mandates blinkies for road races? Why? A road race in daylight is
pretty unmistakable to traffic, particularly with all the brightly
colored riders, follow cars, lead cars, signs, corner marshals, etc.,
etc. Spring racing in Oregon is challenging because of all the rain,
but even with getting blasted in the face with wheel spray (and
wearing glasses), I never ran in to the guy in front of me. Not that
blinkie weight is a big deal. It just seems weird to require one.

-- Jay Beattie.

On Mar 12, 6:55*pm, Dan O wrote:
On Mar 12, 5:53 pm, James wrote:









On 13/03/12 09:43, Dan O wrote:

On Mar 12, 9:44 am, Frank
Second, there's the psychology. *It seems obvious to me that chain
squeaks cause far less energy loss than psychological loss. *The same
was true of the multiple squeaks in my ancient Dahon folding bike.
Maybe the old generators just _sounded_ draggy.

Psychology matters - especially when it's a continuous reminder of the
*real* drag behind it (and weight, and a bike goiter with a wire
hanging out of it). *I don't know about anybody else, but I revel in
relatively effortless speed (always liked to downhill, for instance),
and noticed it in very subtle ways.

Make no mistake, Dan, I like to ride fast and effortlessly too! *I still
don't feel impeded by my dynamo in operation, on a recreational ride
that is. *I take it off for a race...

Why?

..., but then I don't go as far as some,
who spend mega bucks to save a couple of pounds.

"Pounds"? *You don't feel the need to shave *pounds*? *Yet you remove
the dynamo?

I used a Sanyo Dynapower similar (probably not identical) to James's..
I'll confirm that the drag level (with halogen headlight) was just
barely noticeable. *That's from accidentally leaving it on for a
daylight ride to work, and noticing that I felt just a tiny bit slower
on the bike. *With my similar Soubitez, it (accidentally) measured out
to just a one mph difference at about 19 to 20 mph.

That's pretty significant.

I ride the same route 2-3 times a week. *The average speed goes up and
down by more than that depending on the elements and how I feel.

I don't measure my speed. *(Well, sometimes I kind of calculate on
estimates just out of curiosity.) *And absolutely, I know there are
many factors. *And I would like to have a (SON hub) dynamo someday on
my commuter (though I'll still use battery lights or no lights on most
bikes, and carry one with me as backup / supplemental / impromptu
spelunking gear, etc.) *But one of the first things I notice when I
get on a bike for the first time is how much subjective effort it
takes to produce speed.

OT -- riding home last night in a rain/wind storm, I had a negative
forward speed for a few seconds. I've ridden in that kind of wind in
the Gorge and on the coast in places, but downtown? Now I've got to
ride to work in the snow. What is going on here?

-- Jay Beattie.

On 3/13/2012 7:30 AM, Jay Beattie wrote:

Mandates blinkies for road races? Why? A road race in daylight is
pretty unmistakable to traffic, particularly with all the brightly
colored riders, follow cars, lead cars, signs, corner marshals, etc.,
etc. Spring racing in Oregon is challenging because of all the rain,
but even with getting blasted in the face with wheel spray (and
wearing glasses), I never ran in to the guy in front of me. Not that
blinkie weight is a big deal. It just seems weird to require one.

Perhaps the insurance company came up with that idea. Would not surprise
me considering what clubs go through with insurance companies mandating
safety equipment for club rides.

On 3/12/2012 12:07 PM, Peter Cole wrote:

Yes, perhaps, but that wasn't my point. I question the certainty that 10
lux is bright enough to see by.

The German laws stay in place not for anything to do with safety, but to
protect domestic manufacturers of dynamos and dynamo lights.

James wrote:
On 13/03/12 11:12, Phil W Lee wrote:
considered Sun, 11 Mar 2012 13:28:47
-0700 (PDT) the perfect time to write:

If the Germans have outlawed the use of a flashing light on a bicycle,
I think they're backward, and I don't care how many ex-spurts I have
just disagreed with.

Accident stats seem to indicate otherwise.
Maybe you are facing the wrong way?

Haha, you're funny.

I seem to notice flashing lights on bicycles from a long way away, that
I think would otherwise be lost in the noise.

Authorities and advocates here recommend flashing lights.

Ohhhh, "authorities" and "advocates" - and who's that, exactly? Where
were their methods and findings published?

http://www.bicyclensw.org.au/files/u...Queensland.pdf

"Flashing tail-lights are highly effective. Observers have
difficulty in seeing bicycle headlights, partly because they had
difficulty in identifying them as
bicycle lights."

So you copied from the summary, great. Where are the facts?

Where can we look up the demonstration of a necessity to identify
bicycle lights as pedal-bicycle lights, as opposed to being mis-
identified as motored-two wheeler lights?

Since that document was released, this has become part of the law, and
not just a recommendation.

Thanks a lot for the crappy user survey, here's a link:
http://www.austlii.edu.au/au/legis/v...9208/s259.html

"a flashing or steady white light (arguably then, the blue lights which
are becoming more popular as front lights offend this rule) that is
clearly visible for at least 200 metres from the front of the bicycle; and
a flashing or steady red light and reflector that is clearly visible for
at least 200 metres from the rear of the bicycle; and"

So they got around implementing vehicular-type norms, instead using
late-19th century categories such as "clearly visible for x metres."
Where is the Australian Plastic Toy Importers' and Manufacturers'
Association supposed to send the cheque to?

Is the flash cycle specified anywhere? Otherwise, bad luck if the
off-period of the flashing cycle happens to coincide with with a
driver's gaze duration.

On 3/12/2012 11:42 PM, Ralph Barone wrote:
Peter wrote:
On 3/12/2012 5:50 AM, Phil W Lee wrote:
Jeff considered Sun, 11 Mar 2012
16:09:07 -0700 the perfect time to write:

On Sun, 11 Mar 2012 15:26:01 -0700,
wrote:

On 3/11/2012 3:16 PM, Jeff Liebermann wrote:

Trivia: All high power (0.5w) LED lights use PWM (pulse width
modulation) for dimming, to control heating, and to improve
efficiency. Even in the full brightness mode, they flash. You can
see the PWM with a photo diode and oscilloscope, or just shine the
light on a rotating colored disk to see the strobe effect. I wonder
if Germany considers this as a flashing bicycle light.

I would view the Wikipedia article on bicycle lighting with a lot of
grains of salt. It's been hijacked by several individuals with a
specific agenda. A huge number of obvious errors and if you try to
correct them, someone undoes the correction.

I don't see any evidence of hijacking or an agenda in the history
since 2005:
http://en.wikipedia.org/w/index.php?title=Bicycle_lighting&offset=&limit=500 &action=history
Looks like mostly additions, spelling corrections, juggling
references, and discussions (talk). Which individuals are allegedly
responsible?

If you're referring to my mention of PWM in bicycle lighting, that was
verified by me with several LED flashlights and one Bike Planet
bicycle headlight. At such power levels, one would need to be a
masochist to use a linear regulator (except in retrofit flashlight
bulbs). Would you like me to post an oscilloscope photo?

So based on your analysis of ONE bicycle light, you assert that all
use PWM?
Why on earth would they?
All you need is a current limiting device, which is typically a
resistor,

No, it's typically a switching regulator, which modulates duty cycle, AKA PWM.

The use of a switching regulator does not necessarily result in the LED
being switched ON and OFF at a high rate.

I think it does, by definition.


I haven't analyzed any, so I
can't comment on how many bike light current regulators PWM the LED, versus
more sophisticated control methods.

As an EE, I can't think of any more "sophisticated" method, unless
you're modulating it to send data or something. FWIW, I've designed a
few PWM circuits for lamp (arc) control, a few switching power supply
regulators and a motor controller or 2.

On 3/12/2012 6:29 PM, James wrote:

So you've answered the question very nicely. Other manufacturers offer a
Germany only model without flash mode. Why shouldn't the German
manufacturers supply a rest of the world model that includes flash mode?

No market for them. The German law is intended to protect domestic
manufacturers' domestic sales. When was the last time you saw a German
bicycle light for sale at a shop in North America?

Authorities and advocates here recommend flashing lights.

http://www.bicyclensw.org.au/files/u...Queensland.pdf

Not just for bicycles of course either, flashing lights have been proven
to be more visible on emergency vehicles, road barricades, and
motorcycles. No one disputes that a flashing light is more visible.

On 3/13/2012 12:59 AM, James wrote:
On 13/03/12 14:42, Ralph Barone wrote:
Peter wrote:
On 3/12/2012 5:50 AM, Phil W Lee wrote:
Jeff considered Sun, 11 Mar 2012
16:09:07 -0700 the perfect time to write:

If you're referring to my mention of PWM in bicycle lighting, that was
verified by me with several LED flashlights and one Bike Planet
bicycle headlight. At such power levels, one would need to be a
masochist to use a linear regulator (except in retrofit flashlight
bulbs). Would you like me to post an oscilloscope photo?

So based on your analysis of ONE bicycle light, you assert that all
use PWM?
Why on earth would they?
All you need is a current limiting device, which is typically a
resistor,

No, it's typically a switching regulator, which modulates duty cycle,
AKA PWM.

The use of a switching regulator does not necessarily result in the LED
being switched ON and OFF at a high rate. I haven't analyzed any, so I
can't comment on how many bike light current regulators PWM the LED,
versus
more sophisticated control methods.

You're on the money. It depends on the DC-DC converter topology, but the
current may be continuous and not cause the LED to turn off at all.

See here, page 6 shows the current through the inductor is continuous.

http://www.ti.com/lit/ds/symlink/lm3405.pdf

The same circuit _can_ be used to flash or dim the LED, by toggling the
enable/dim input pin.


From the linked data sheet:

"The LM3405 is a PWM, current-mode control switching buck
regulator designed to provide a simple, high efficiency solution
for driving LEDs with a preset switching frequency of
1.6MHz."

Elementary circuit theory states that you can't instantaneously change
the current in an inductor (requires infinite voltage), or the voltage
on a capacitor (requires infinite current). In this example circuit, an
inductor is used, but that does not change the principle of regulation.
The switching FET is turned on & off, the goal being to spend 0 time in
the in-between state where it would be dissipating lots of power.

One reason to use an inductor is so you don't create a high frequency
radio transmitter. You will, to a certain extent anyway, since the
current is a sawtooth with a 1.6 MHz frequency, but a square wave at
that frequency will contain significant harmonics above that primary
frequency.

I have no idea why they want to switch at such a high frequency. They
estimate the power losses as having 4 components: 1) FET conduction (on
resistance) 2) FET gate current 3) FET switching losses 4) internal chip
consumption. #3 is over 50% of the total, and #1 is about 25%. #3 (and
#2) are proportional to switching frequency, so there is an efficiency
cost to running at such a high frequency. The 1W, 3W and 10W LED lights
I have used all seem to have switching rates low enough to see (when you
aim them at a moving fan), so it's seems like they're running in the low
audio range.

Running at high frequency allows you to use small inductors. Perhaps the
reason for using inductors is to limit the RMS (average) to peak
current, either to protect the LED, the internal FET, or perhaps to keep
the LED in its most efficient current range. It's quite possible that
the power losses incurred in switching at the higher rate are more than
compensated just by running the LED at greater efficiency.

Just thinking about this a bit more, I realize that I've been biased by
batteries (so to speak). Batteries often have enough series resistance
to effectively current limit, and often are sized to be close to the LED
operating voltage. When that's not true, you'd definitely need inductive
ballast, and the higher the frequency, the smaller the inductor. Of
course all this is moot with generators, since they have enough
inductance to be treated (effectively) as constant current sources.

Now that I think about it, the "strobe effect" I've seen on fan blades
was only at dimmed settings, so it seems that the high frequency
modulator is in turn modulated by a lower frequency signal, kind of like
AM radio.

Now that I think about it even more, I remember that the couple of times
I've disassembled 1W LED lights I think I saw small toroid ferrite
inductors. I assumed they were for RFI suppression. Now I guess I've
learned they were for ballast and even the cheap lights are being
regulated (PWM) at very high frequency. That also explains why the cited
regulation losses were higher than I expected.

All of this doesn't change the fact that even though the current through
the LED & inductor is a sawtooth, the voltage is modulated on & off, AKA
PWM.

sms88 wrote:
The German law is intended to protect domestic manufacturers'
domestic sales.

Good one 8)


Not just for bicycles of course either, flashing lights have been
proven to be more visible on emergency vehicles, road barricades,
and motorcycles. No one disputes that a flashing light is more
visible.

The only point being that 'visible' isn't the main thing in traffic,
the main thing is 'visible, easily predictable and not overly
disturbing'.

Try to navigate quickly through a bunch of flashing emergency
vehicles and you'll note that it is not a reasonable model for
standard traffic.


--
MfG/Best regards
helmut springer panta rhei