Why not generator lights?

[Frank Krygowski]

Marty Wallace wrote:

> "Frank Krygowski" <[email protected]> wrote in message
> news:[email protected]...
>
>>One problem is that bike generators put out AC, not DC, and LEDs are
>>inherently DC devices. You'd have to rectify the generator output,
>>which entails at least slight losses in efficiency. I don't know those
>>losses would have to be significant, though. But I also don't know that
>>the lumens/watt out of white LEDs are high enough (yet) to make up for
>>the focusing difficulties.
>
>
> The solution of course is to use two LEDs, since LEDs are DIODES you don't
> need to add a rectifier because they ARE rectifiers.
> Connect the two LEDs in parallel but with opposite polarisation and the
> current will flow with for both halves of the AC cycle. You might need to
> add some sort of current limiting device depending on the voltage and
> current characteristics of the generators.

I know.

There are complications, though. One is that the current capacity of
any LED you'd like to use for a rear taillight is almost certainly lower
than what the generator will be providing. Thus, you're dealing with
multiple LEDs in parallel each way, raising the general fussiness of
buiding the circuit. And I'm not really an electronics guy, but I think
you'd want a current limiting resistor on each LED, not just one
resistor, since you couldn't guarantee perfect matching (therefore
perfect current sharing). I saw one circuit for such a light that had,
as a rough guess, 20 solder connections or more. To me, that's fussy.

Another possibility would be to do a bridge rectifier out of LEDs, then
put any remaining LEDs in series.

One possible downside to an all-LED rear light is the narrow spectrum
emitted. It's been theorized that some color blind motorists may see
LEDs more dimly than they see incandescents. I don't know the latest
thinking on this, but I thought it might be wise to throw some yellow
LEDs into the mix - not enough to greatly change the net color output,
but just enough to help.

But since the yellow ones have somewhat different characteristics than
the red ones, it'll take a little time to design the entire array.

>
> Luxeon make a broad range of LEDs including special ones to retrofit
> Maglites.
>
> http://www.luxeonstar.com/sub_category.php?id=178&link_str=178

I'm not planning on using luxeons. I don't think there's any need for
that expense and power consumption in a taillight.


--
--------------------+
Frank Krygowski [To reply, remove rodent and vegetable dot com,
replace with cc.ysu dot edu]

 

[Robert Chung]

Frank Krygowski wrote:
> Matt O'Toole wrote:
>
>>>> "Frank Krygowski" <[email protected]> wrote in message
>>>> news:[email protected]...
>>>>
>>>>> FWIW, I did recently find a paper testing various night visibility
>>>>> schemes for cyclists and pedestrians. They found that an ancient
>>>>> leg light (probably 1.25 watt or less) was spotted over 1000 feet
>>>>> away, and was recognized as a cyclist several hundred feet away.
>>>>> The same was true for a pedestrian carrying an ordinary flashlight.
>>
>>
>> Where can I find this paper? Were these cyclists *recognized as such*
>> from 1000' away, or was the just the light visible?
>
> The paper is "Experimental Evaluation of Alternatie
> Conspicuity-Enhancement Techniques for Pedestrians and Bicyclists",
> Blomberg, R. et. al., Journal of Safety Research, Vol 17, pp. 1-12,
> 1986.
>
> The cyclist (and pedestrians) were recognized as _something_ to deal
> with from, actually, 1300 feet away. They were definitely recognized as
> cyclists 481 feet away, or in the case of pedestrians, 316 feet away.
> For the cyclists, even at a high closing speed of 40 mph, that's 22
> seconds telling to react to _something_, and 8 seconds to know it's
> definitely a cyclist. Obviously, in typical urban/suburban situations,
> typical closing speed isn't that high, so times would be even longer.

Thanks for giving the paper reference. I just looked it up. The test was
with "alerted" drivers on a closed course, so "1300 feet away" or "22
seconds" is the upper bound for people who were expecting and looking for
objects at the side of the road. The authors were testing different types
of visibility enhancement techniques, and they specifically warn that
their data are useful for determining relative, rather than absolute,
visibility. Compared to their baseline bicycle case (a cyclist wearing a
white T-shirt and blue jeans with CPSC reflectors on wheels and rear),
their main conclusions were:

1. Cyclists wearing dark clothing with no lights were basically invisible.
2. A small "active" light source (in this case, an old-fashioned leg
light) increased visibility over the baseline by roughly 50%.

 

[Robert Chung]

I wrote:
> Compared to their baseline bicycle case (a
> cyclist wearing a white T-shirt and blue jeans with CPSC reflectors on
> wheels and rear)

That should be, "reflectors on wheels, rear, and pedals."

 

[Steven M. Scharf]

Robert Chung wrote:

> Thanks for giving the paper reference. I just looked it up. The test was
> with "alerted" drivers on a closed course, so "1300 feet away" or "22
> seconds" is the upper bound for people who were expecting and looking for
> objects at the side of the road.

Ah, that explains a lot. It's like asking a friend to see if he can see
your bike as you ride by with your lights on, and then claiming that
your lights must be good because your friend said that he could see you
with no problem.

The movement of leg lights made them pretty visible as a "being seen"
light. Can you even still buy them? I haven't seen them for at least a
decade.

 

[Robert Chung]

Steven M. Scharf wrote:
> Robert Chung wrote:
>
>> Thanks for giving the paper reference. I just looked it up. The test
>> was with "alerted" drivers on a closed course, so "1300 feet away" or
>> "22 seconds" is the upper bound for people who were expecting and
>> looking for objects at the side of the road.
>
> Ah, that explains a lot. It's like asking a friend to see if he can see
> your bike as you ride by with your lights on, and then claiming that
> your lights must be good because your friend said that he could see you
> with no problem.

Not *quite* that bad. The researchers scattered different objects around a
roughly 10km long course: pedestrians wearing different kind of visibility
enhancements, bicycles with different kinds of visibility enhancements,
road signs, warning beacons, things of that sort. The drivers were
expecting to see stuff, but they weren't told exactly what or exactly
where. That's why the authors' conclusions emphasized relative visibility,
not absolute distances. The field work for the study was done in the Fall
of 1983, so they were using those old-fashioned leg lights.

 

[Frank Krygowski]

Getting back to the question I was asking about lights at night:

When observing night cyclists while driving, I've always been impressed
by the fact that cyclists are very visible when they use only what's
legally required - meaning, rear reflector and taillight, plus
headlight. They're even more visible when reflectors are added to
pedals, spokes or rims, and when the rear reflector is mounted low. And
ordinary LED blinkies have always been suitably conspicuous to me, as well.

As I've mentioned, I've done workshops with my bike club where we've
driven by others' bikes as they were being ridden. I've also had my
family rate my visibility this way, and I've had spontaneous compliments
from both friends and strangers who happened to pass me. The conclusion
was always unanimous, that "ordinary" equipment is fine for visibility.

So my question is, has anyone experienced an exception? That is, have
you, when driving at night, been surprised by a legally lit cyclist you
didn't see when you should have? If so, what equipment did he use, and
what were the circumstances?

--
--------------------+
Frank Krygowski [To reply, remove rodent and vegetable dot com,
replace with cc.ysu dot edu]

 

[Tim McNamara]

"Robert Chung" <[email protected]> writes:

> Thanks for giving the paper reference. I just looked it up.

Where? I couldn't turn it up via Google or the National Safety
Council's Web site (they are the publisher of the journal in
question).

 

[Tim McNamara]

"Steven M. Scharf" <[email protected]> writes:

> Robert Chung wrote:
>
>> Thanks for giving the paper reference. I just looked it up. The
>> test was with "alerted" drivers on a closed course, so "1300 feet
>> away" or "22 seconds" is the upper bound for people who were
>> expecting and looking for objects at the side of the road.
>
> Ah, that explains a lot. It's like asking a friend to see if he can
> see your bike as you ride by with your lights on, and then claiming
> that your lights must be good because your friend said that he could
> see you with no problem.

Not quite the same thing. Really, Steven, you should read more
carefully.

> The movement of leg lights made them pretty visible as a "being
> seen" light. Can you even still buy them? I haven't seen them for at
> least a decade.

A series of reflective dots- on the wrists, elbows, shoulders, hips,
knees and feet- would provide enough information for drivers to
definitively recognize a cyclist as such fairly easily, if seen from
the side. This isn't quite as true from the angle of overtaking, but
would still be an improvment as the movements attracts attention.
Easier of course in cooler weather than sticking reflective dots on
yor body during the summer.

 

[Robert Chung]

Tim McNamara wrote:
> "Robert Chung" <[email protected]> writes:
>
>> Thanks for giving the paper reference. I just looked it up.
>
> Where? I couldn't turn it up via Google or the National Safety
> Council's Web site (they are the publisher of the journal in
> question).

I have access to the text.

 

[[email protected]]

On Sun, 19 Dec 2004 06:06:45 GMT, "Steven M. Scharf"
<[email protected]> wrote:

>Robert Chung wrote:
>
>> Thanks for giving the paper reference. I just looked it up. The test was
>> with "alerted" drivers on a closed course, so "1300 feet away" or "22
>> seconds" is the upper bound for people who were expecting and looking for
>> objects at the side of the road.
>
>Ah, that explains a lot. It's like asking a friend to see if he can see
>your bike as you ride by with your lights on, and then claiming that
>your lights must be good because your friend said that he could see you
>with no problem.
>
>The movement of leg lights made them pretty visible as a "being seen"
>light. Can you even still buy them? I haven't seen them for at least a
>decade.

Dear Steven,

Lighted armbands at Performance:

http://www.performancebike.com/shop/profile.cfm?SKU=15324&subcategory_ID=4323

Strap-on light at Nashbar, warning that it may not fit
larger calves:

http://www.nashbar.com/profile.cfm?...rand=&sku=9979&storetype=&estoreid=&pagename=

And here:

http://www.defensedevices.com/lighted-armband.html

Carl Fogel

 

[Mark Janeba]

Frank Krygowski wrote:
> So my question is, has anyone experienced an exception? That is, have
> you, when driving at night, been surprised by a legally lit cyclist you
> didn't see when you should have? If so, what equipment did he use, and
> what were the circumstances?

Frank, I suspect you're asking the wrong group. Here's my reasoning: I
think cyclists are likely to "pick out" cyclists far more readily than
the general public. I could elaborate on this theoretical ability,
which I suspect is largely due to practice, but perhaps it's widely
accepted. At any rate, I know that while driving *I* will notice
cyclists when others with me don't. So to answer you question, my
answer is NO, but I strongly doubt I'm at all representative in this.

The opposite effect is when drivers have, whether consciously or not,
whether by neglect or design, classified cyclists as just part of the
visual "noise" of the driving experience.

Some formal tests of drivers' perception have been described in posts
here. I suspect the (driving) subjects were: 1) 100% sober, 2)
reasonably alert, not coming home after an exhausting day of work 3) not
talking on a cell phone during the test, and 4) reasonably responsible
drivers. In the case of tests with a bike club, the subjects are
actively looking out, and as I speculate above, better practiced at the
task at hand than most folk.

I'm not sure where I come down on the issue about the adequacy of
existing rear bike lights; the oft-quoted statistic is that only 10% of
car-bike accidents are rear-enders - but I still gladly tend toward
overkill in my lights. I commute year round, often in the dark, and
don't mind being silly if it gains me a fraction of perceptibility.

Still interested in the results you do find,

Mark "anybody remember the Kearney taillights?" Janeba

 

[(Pete Cresswell)]

RE/
>Frank, I suspect you're asking the wrong group.

Now that you've said it, I'd second that point.
I'd suggest misc.transport.trucking.
--
PeteCresswell

 

[Frank Krygowski]

(Pete Cresswell) wrote:

> RE/
>
>>Frank, I suspect you're asking the wrong group.
>
>
> Now that you've said it, I'd second that point.
> I'd suggest misc.transport.trucking.

They probably have no idea what bike lighting is required by law. Most
people don't, according to a survey Forester did. Specifically, most
people think reflectors (instead of lights) are sufficient.

Let's see what answers we get in this group before we go further afield.

--
--------------------+
Frank Krygowski [To reply, remove rodent and vegetable dot com,
replace with cc.ysu dot edu]

 

[Steven M. Scharf]

(Pete Cresswell) wrote:
> RE/
>
>>Frank, I suspect you're asking the wrong group.
>
>
> Now that you've said it, I'd second that point.

How is that question even answerable?

If you didn't see them, how do you know that they were there?

Often I see cyclists only when I catch them directly in my headlights,
and often these cyclists have a dim rear LED blinker. Technically, in
most cases, such a blinker doesn't qualify as a legal light, so they are
illegally lit, and hence they don't count as a legally lit cyclist.

There are actually very few legally lit cyclists, since most are using
flashing red lights.

Legal lighting does not always equal good lighting.

 

[davidd86]

I agree with the above; cyclists are much more likely to notice other cyclists at night because they have Feared the Reaper. Even more so, perhaps, are motorcyclists -- since we ride right in the road -- every motorcyclist knows that awful, sinking feeling of being tailgated or being near a squirrelly driver -- it just absolutely SUCKS and the pucker factor is way up there. So when a motorcyclist is driving in a car he's likely to be much, much, much more attentive to motorcyclists.

One dumb thing I've found that HUGELY increases your visibility on a bike at night near traffic is ... sorry guys I know these are uncool ... wheel reflectors. If you don't like the $.10 kind that come on bikes at Target, you can get some velcro ones that go on in basically the same place in the spokes. I put them on my commuter after observing people riding with them ... they're just so incredibly helpful. I got mine for the asking at a bike store, because no one wants 'em.

I'm also gonna try to find a pair of those Illuminite winter riding gloves. Anyone know where to source a pair?

(Also: I thought blinkers were legal in most jurisdictions as long as they aren't permanently attached to the bike; not so?)

(Pete Cresswell) wrote:
> RE/
>
>>Frank, I suspect you're asking the wrong group.
>
>
> Now that you've said it, I'd second that point.

How is that question even answerable?

If you didn't see them, how do you know that they were there?

Often I see cyclists only when I catch them directly in my headlights,
and often these cyclists have a dim rear LED blinker. Technically, in
most cases, such a blinker doesn't qualify as a legal light, so they are
illegally lit, and hence they don't count as a legally lit cyclist.

There are actually very few legally lit cyclists, since most are using
flashing red lights.

Legal lighting does not always equal good lighting.

 

[Peter]

Steven M. Scharf wrote:

> How is that question even answerable?
>
> If you didn't see them, how do you know that they were there?
>
> Often I see cyclists only when I catch them directly in my headlights,
> and often these cyclists have a dim rear LED blinker. Technically, in
> most cases, such a blinker doesn't qualify as a legal light, so they are
> illegally lit, and hence they don't count as a legally lit cyclist.
>
> There are actually very few legally lit cyclists, since most are using
> flashing red lights.

I presume you're mainly seeing these cyclists in California where you
live. In that case there's nothing illegal about the use of red
flashing lights on the back of a bicycle - only blue flashing lights
are prohibited. [Red flashing lights are prohibited on most "vehicles"
but bicycles are explicitly defined *not* to be vehicles by the CVC.]
It is required that the bicyclist have a front white light and a
red reflector on the back whether or not it also has any rear light.

However you are correct that few cyclists are technically legal in
California since they generally lack the pedal and side reflectors
that are also required in addition to the red rear reflector and a
front white light.

CVC 21201:
(d) Every bicycle operated upon any highway during darkness shall
be equipped (1) with a lamp emitting a white light which, while the
bicycle is in motion, illuminates the highway in front of the
bicyclist and is visible from a distance of 300 feet in front and
from the sides of the bicycle; (2) with a red reflector on the rear
which shall be visible from a distance of 500 feet to the rear when
directly in front of lawful upper beams of headlamps on a motor
vehicle; (3) with a white or yellow reflector on each pedal visible
from the front and rear of the bicycle from a distance of 200 feet;
and (4) with a white or yellow reflector on each side forward of the
center of the bicycle, and with a white or red reflector on each side
to the rear of the center of the bicycle, except that bicycles which
are equipped with reflectorized tires on the front and the rear need
not be equipped with these side reflectors. Such reflectors and
reflectorized tires shall be of a type meeting requirements
established by the department.
(e) A lamp or lamp combination, emitting a white light, attached
to the operator and visible from a distance of 300 feet in front and
from the sides of the bicycle, may be used in lieu of the lamp
required by clause (1) of subdivision (d).

 

[Frank Krygowski]

Steven M. Scharf wrote:

> How is that question even answerable?
>
> If you didn't see them, how do you know that they were there?

Wow. I hoped I wouldn't have to explain this in excruciating detail!

What I'm talking about is cyclists who had the sort of headlights,
reflectors, etc. that are sold in the greatest volumes. That would
include lights less than six watts, more or less standard reflectors,
and perhaps ordinary LED blinkies.

And I'm not talking just about "Gee, I personally don't think that guy's
sufficiently visible." I'm talking about "Damn! I almost hit him
because I didn't see his headlight, even though it's on!" Has that ever
happened to you while you were driving?

The laws vary state by state, of course. But IF you had a near-miss
experience with an apparently legal cyclist, why not describe it in some
detail so we can learn?

If nobody comes up with such an incident, or if we come up with very,
very few, we can learn from that too.

Again, if I can go first: I've _never_ had that experience. I've never
been surprised by any near-legal cyclist. And FWIW, nobody I know
(motorists or cyclists) have ever described such an incident to me. Am
I alone?

--
--------------------+
Frank Krygowski [To reply, remove rodent and vegetable dot com,
replace with cc.ysu dot edu]

 

[Robert Chung]

Frank Krygowski wrote:

> When observing night cyclists while driving, I've always been impressed
> by the fact that cyclists are very visible when they use only what's
> legally required - meaning, rear reflector and taillight, plus
> headlight.

and then in a subsequent post,

> They probably have no idea what bike lighting is required by law. Most
> people don't, according to a survey Forester did. Specifically, most
> people think reflectors (instead of lights) are sufficient.

From the Bicycle Transportation Institute's page on "Bicycles and Traffic
Law" (http://www.bicycledriving.com/trafficlaw.htm), which appears to have
been revised in 2004:

"Only five states require a rear lamp emitting a red light: Alaska,
Florida, New Jersey, New York, and Ohio. Of these five, Alaska, Florida,
and Ohio also require a rear reflector. All the remaining states require a
rear reflector, and specify that a rear lamp may be used in addition to,
or in some states, instead of, the rear reflector. All specify that the
reflector must be red, except South Dakota, which says, "red or yellow."
All but two states specify a minimum visibility distance for rear
reflectors ranging from 600 to 200 feet (see Table 1). In many states the
visibility requirement is less for reflectors (300 feet) than for the
optional rear light (500 feet). Three states also specify a minimum
reflector size (see also Table 1). Washington says that light emitting
diodes (LEDs) may be used instead of the optional rear lamp. Wisconsin
says that the optional rear light may be "red or flashing amber.""

 

[Steve Juniper]

Law notwithstanding, I think the easiest and most effective thing one
can simply do to enhance visibility at night is to wear white regardless
of what kind of lighting or reflectors you might have. An old white
shirt would stand out much more than the average flashing or nonflashing
bike tail light I see.

Steve J

Frank Krygowski wrote:
> Steven M. Scharf wrote:
>
>> How is that question even answerable?
>>
>> If you didn't see them, how do you know that they were there?
>
>
> Wow. I hoped I wouldn't have to explain this in excruciating detail!
>
> What I'm talking about is cyclists who had the sort of headlights,
> reflectors, etc. that are sold in the greatest volumes. That would
> include lights less than six watts, more or less standard reflectors,
> and perhaps ordinary LED blinkies.
>
> And I'm not talking just about "Gee, I personally don't think that guy's
> sufficiently visible." I'm talking about "Damn! I almost hit him
> because I didn't see his headlight, even though it's on!" Has that ever
> happened to you while you were driving?
>
> The laws vary state by state, of course. But IF you had a near-miss
> experience with an apparently legal cyclist, why not describe it in some
> detail so we can learn?
>
> If nobody comes up with such an incident, or if we come up with very,
> very few, we can learn from that too.
>
> Again, if I can go first: I've _never_ had that experience. I've never
> been surprised by any near-legal cyclist. And FWIW, nobody I know
> (motorists or cyclists) have ever described such an incident to me. Am
> I alone?
>

 

[Robert Chung]

Frank Krygowski wrote:
> If nobody comes up with such an incident, or if we come up with very,
> very few, we can learn from that too.

Actually, I'm not sure what we could learn from that. I'd think there'd be
considerable recall bias: for example, I can think of an event last week
where I made a left turn in front of a cyclist who was riding straight--I
didn't see him until we were both in the intersection, and since I didn't
see him until the last second I'm not really sure whether he was
legally-lit or not. I suspect if you'd asked this question a month from
now, I might not have recalled it. OTOH, I use more than the
legally-required amount of lighting, and I can recall similar "oh ****"
moments pretty damn well. It's not clear to me that either of those
recalled memories would make a good basis for estimating average risk.
Anecdotes aren't the same thing as controlled data.