New 1920s Technology... Today!

[AMuzi]

https://pedalcell.com/

--
Andrew Muzi
<www.yellowjersey.org/>
Open every day since 1 April, 1971

[Frank Krygowski]

On 4/8/2021 11:52 AM, AMuzi wrote:
> https://pedalcell.com/
>

I note that when touting their supercapacitor, they say:

"Raw Power
"Volatile, Unsafe"

Heavens! I wouldn't want to be unsafe! ;-)

--
- Frank Krygowski

[Mark J.]

On 4/8/2021 8:52 AM, AMuzi wrote:
> https://pedalcell.com/
>

At least they have the drive wheel run on the rim, not the tire. An
idea I think Frank has used and recommended. But yeah, kinda pathetic
overall.

Mark J.

[sms]

On 4/8/2021 8:52 AM, AMuzi wrote:
> https://pedalcell.com/

LOL, I especially like the "15-20W of Power Output." Uh yeah, from the
battery, but not from the dynamo. A large-screen smart phone running a
GPS navigation application with the LTE or 5G radio on can exceed the 3
watt output of the dynamo.

$299.99, wow! Before I clicked on the link I guessed $100 since it
should cost about $50. A 3W tire-driven dynamo costs about $6. The
electronics to convert 6-9VAC to 5VDC is about $1. A USB power bank is
about $5. I multiplied the retail cost of the components by 8 and
guessed $100.

Also love how their price comparison uses the most expensive, highest
quality, hub dynamo.

A much better solution is to buy a rechargeable bicycle light that also
functions as a USB power bank and charge that light from a dynamo. Buy
one of these <https://www.aliexpress.com/item/1005001507833824.html>
(actually it comes in a lot of 5 for $2.81), four 1N5817 diodes, and a
470uF capacitor, and hook it up to a dynamo.

[Tom Kunich]

Hey, my allergies are acting up and I just blew snot all over my keyboard!

[Jeff Liebermann]

On Thu, 8 Apr 2021 12:05:13 -0700, "Mark J." <MarkU...@comcast.net>
wrote:

Pathetic? The idea doesn't go far enough. That was the dynamo I was
thinking of when I suggested that it be used on disk wheels. The idea
was to move the dynamo along the radius of the disk wheel to adjust
the dynamo's rotational speed and thus obtain a constant output output
voltage at any speed.

Why bicycle stores sell so few dynamos:
<http://yehudamoon.com/comic/2017-03-15/>


--
Jeff Liebermann je...@cruzio.com
PO Box 272 http://www.LearnByDestroying.com
Ben Lomond CA 95005-0272
Skype: JeffLiebermann AE6KS 831-336-2558

[Frank Krygowski]

On 4/8/2021 8:43 PM, Jeff Liebermann wrote:
> On Thu, 8 Apr 2021 12:05:13 -0700, "Mark J." <MarkU...@comcast.net>
> wrote:
>
>> On 4/8/2021 8:52 AM, AMuzi wrote:
>>> https://pedalcell.com/
>
>> At least they have the drive wheel run on the rim, not the tire. An
>> idea I think Frank has used and recommended. But yeah, kinda pathetic
>> overall.
>> Mark J.
>
> Pathetic? The idea doesn't go far enough. That was the dynamo I was
> thinking of when I suggested that it be used on disk wheels. The idea
> was to move the dynamo along the radius of the disk wheel to adjust
> the dynamo's rotational speed and thus obtain a constant output output
> voltage at any speed.
>
> Why bicycle stores sell so few dynamos:
> <http://yehudamoon.com/comic/2017-03-15/>

I miss Yehuda. :-(


--
- Frank Krygowski

[sms]

On 4/8/2021 12:05 PM, Mark J. wrote:
> On 4/8/2021 8:52 AM, AMuzi wrote:
>> https://pedalcell.com/
>>
> At least they have the drive wheel run on the rim, not the tire.

<snip>

True. And those rim drive dynamos cost a lot more than the tire-driven
dynamos.

But you can buy a new 700c wheel with a 3W hub dynamo (Shimano) for
$125, less than the cost of a rim dynamo. For $275 you can buy a 700c
front wheel with a SP (Shutter Precision) hub dynamo. 20" SP equipped
("Joule") wheels are even less, $120 to $200. but of course you don't
get "Generator Drag Decoupling" (except with the SP dynamo hub that has
that feature).

I suspect that they've sold very few of the "Pedalcell" systems. The
people this is being marketed to probably didn't even pay $299.99 for
their entire bicycle.

The other devices that convert the output of a dynamo to 5VDC for USB
charging are being sold to people that already have a hub dynamo on
their bicycles, see
<https://www.cyclingabout.com/list-of-hub-dynamo-power-supplies-for-usb-devices/>;
some of the 27 devices they list are dynamo lights that have built-in
USB chargers.

As I stated before, if you really want to have a dynamo charge your
phone, or charge a battery pack, while also having a light that can be
operated from a dynamo, it makes a lot more sense to purchase a
battery-powered headlight with a USB output port, and then charge the
headlight from a dynamo. Many battery-powered headlights can operate
while being charged, though not all of them have this capability. For
under $3 you can buy a module that converts the AC output of the dynamo
to 5VDC.

[sms]

On 4/8/2021 5:43 PM, Jeff Liebermann wrote:

<snip>

> Pathetic? The idea doesn't go far enough. That was the dynamo I was
> thinking of when I suggested that it be used on disk wheels. The idea
> was to move the dynamo along the radius of the disk wheel to adjust
> the dynamo's rotational speed and thus obtain a constant output output
> voltage at any speed.

A CVT inside the dynamo would be a better design, whether a rim, tire,
or hub. But of course it's really unnecessary to have a constant voltage
out.

There are already dynamo powered lights with a USB port for charging
devices. If someone is really going to spend $300 then they'd be better
off buying a dynamo hub wheel and one of those lights, or a battery
powered light that can be charged and/or operated from a dynamo.

For the latter, I put together a Google Doc document that explains how
to do it
<https://docs.google.com/document/d/1UVk_yelr5UxrcKqDtGgsHTBVa3eWoUdpqDKhbZSHO_E>

[Jeff Liebermann]

On Fri, 9 Apr 2021 18:02:40 -0700, sms <scharf...@geemail.com>
wrote:

>On 4/8/2021 5:43 PM, Jeff Liebermann wrote:
>
><snip>
>
>> Pathetic? The idea doesn't go far enough. That was the dynamo I was
>> thinking of when I suggested that it be used on disk wheels. The idea
>> was to move the dynamo along the radius of the disk wheel to adjust
>> the dynamo's rotational speed and thus obtain a constant output output
>> voltage at any speed.

>A CVT inside the dynamo would be a better design, whether a rim, tire,
>or hub. But of course it's really unnecessary to have a constant voltage
>out.

I was looking for simplicity. A CVT (continuously variable
transmission) is not a simple device to build into a small dynamo
housing.

>There are already dynamo powered lights with a USB port for charging
>devices. If someone is really going to spend $300 then they'd be better
>off buying a dynamo hub wheel and one of those lights, or a battery
>powered light that can be charged and/or operated from a dynamo.
>
>For the latter, I put together a Google Doc document that explains how
>to do it
><https://docs.google.com/document/d/1UVk_yelr5UxrcKqDtGgsHTBVa3eWoUdpqDKhbZSHO_E>

Very nice but I screwed up.
The DC output voltage from a bridge is NOT:
DC_out = 1.414 * AC_Vrms_input
More correctly, the DC output is:
DC_out = 0.9 * AC_Vrms_input
For the calcs, see:
<https://www.electronics-tutorials.ws/diode/diode_6.html>
For such a low voltage dynamo, the Schottky diode voltage drop should
also be included in the calculations.

Another problem is you're ignoring the original problem(s) and solving
one that was not even mentioned (USB devices).

Jay arbitrarily decided that an 800 lumen light would be nice if it
could be powered by a dynamo. Without questioning the decision to
specify such a high power light, I went into what is necessary to
power an 800 lumen light solely on dynamo power. It was generally
assumed that building a battery operated 800 lumen light, where the
battery is charged by the dynamo, is possible. I later estimated the
percentage of power produced by the typical cyclists at 15 mph using a
commodity 3 watt dynamo light, and a higher power light suitable for
delivering 800 lumens. While there was an increase in drag, it was
within the abilities of most cyclists. All that was needed was a
dynamo that delivered both higher output power at higher efficiency.

I then arbitrarily decided that removing one of the key features of
most bicycle dynamos would offer better efficiency at high RPM. The
early dynamo power lights ran incandescent lamps. These lamps were
very sensitive to over-voltage. Even a little over their rated
maximum power, and the filament would fuse open. The resulting dynamo
included an inductor (saturable reactor) which would reduce the
voltage to the incandescent lamp at high RPM, where the voltage was
highest. This formed a crude and cheap voltage/current regulator and
is responsible for the "flat" voltage and power curve of the dynamo.
What I proposed was to remove the inductors and allow the dynamo to
produce more output power at high RPM. However, Frank offered a good
question. Does this also improve efficiency at high RPM. I assumed
that it does, but I'm not sure and won't know without some testing.

Another item that was mentioned was operating the headlight at low
bicycle speeds, where the dynamo is unable to deliver sufficient
output. This can be in various ways such as a CVT, PWM (pulse width
modulation) light dimmer to reduce the load, switch to fewer LED's,
etc. Of course, a battery/dynamo hybrid will solve the problem as
long as the battery can be recharged, but for now, batteries are not
part of the dynamo problem.

My best guess(tm) is that a higher power dynamo can only work if the
losses are reduced. That means, much higher voltages, no saturable
reactor regulators, very wide input range boost converter on the
input, high efficiency buck converter to drive the LEDs, and high RPM
for high operating frequencies (about 50KHz) so that ferrites can be
used instead of powdered iron. In other words, start from scratch and
design something that will drive a modern LED headlight, not an
adaptation of something designed to drive an ancient incandescent
light bulb.

[John B.]

On Fri, 09 Apr 2021 20:35:09 -0700, Jeff Liebermann <je...@cruzio.com>
wrote:

The solution? Just stick an internal combustion engine on that two
wheel device and you can have all the lights that you want :-)
--
Cheers,

John B.

[Roger Merriman]

Exposure sell such a light
<https://exposurelights.com/products/bike/dynamo-lights/revo-dynamo>

At a quick look do need to be traveling at a reasonable speed, ie 15mph,
it’s half that at 10mph according to the guide. Which certainly for my
uses, ie on my commute the faster sections are well lit, and it’s the
slower sections that if not need, certainly it’s useful to have a
reasonably powerful light.

Lot depends on your use case, if it’s urban or suburban on roads well to be
honest probably only need a light to be seen by.

My commute has a few woods and what not, which 600 Lumens in a fairly wide
flat beam is great, on gravel or otherwise roads it’s equally fine. Once
into MTB trails it’s shape more than anything let’s it down, as the beam
can’t cover the dips and angles, where as my MTB light which is fairly
modest as these go, *only* 1000 Lumens is much better able to light the
trail even on low settings.

You still can outpace the light but generally I take it easy on night rides
as certainly do so few MTB night rides, that stumping up for a 3000 or so
(which seems to be the sweet spot) Lumens MTB light for the 1/2 times I
might ride per year!
>
Snips
>
>
Roger Merriman

[Jeff Liebermann]

£240.00 or $331 not including the hub dynamo. Ouch. Thanks, but that
seems rather expensive. I like the idea, but there needs to be some
cost reduction for wider acceptance. Thanks for the link.

>At a quick look do need to be traveling at a reasonable speed, ie 15mph,
>it’s half that at 10mph according to the guide.

Presumably, adding a CVT (continuously variable transmission) to the
dynamo would provide more power at slow speeds. I haven't seen that
being done with hub or bottle dynamos. In effect, the dynamo would
produce a fairly constant output at speeds above some minimum, and
possibly eliminate the need for a voltage regulator in the light.

>Which certainly for my
>uses, ie on my commute the faster sections are well lit, and it’s the
>slower sections that if not need, certainly it’s useful to have a
>reasonably powerful light.
>
>Lot depends on your use case, if it’s urban or suburban on roads well to be
>honest probably only need a light to be seen by.

My use case is minimal and probably not representative of most
cyclists in this newsgroup. Whether a bicycle light can be properly
designed by someone who doesn't ride much or on different types of
machines and roads is an open question. For me, the main problem is
price and theft of the light, which haven't entered this discussion
yet. For lighting, the only times I've really needed a decent light
was when trying to read street signs and house numbers. I also had a
problem with being seen. My current headlight (Cygolite Streak 280)
allegedly delivers 280 lumens or about 175 lumens by my measurements.

>My commute has a few woods and what not, which 600 Lumens in a fairly wide
>flat beam is great, on gravel or otherwise roads it’s equally fine. Once
>into MTB trails it’s shape more than anything let’s it down, as the beam
>can’t cover the dips and angles, where as my MTB light which is fairly
>modest as these go, *only* 1000 Lumens is much better able to light the
>trail even on low settings.

Do you have a lux meter? Something like one of these?
<https://www.ebay.com/sch/i.html?_nkw=lux+meter>
If so, and your light an produce a roughly circular spot on a wall, I
have a way to measure lumens without an integrating sphere (or
integrating pipe). These are early versions (in desperate need of
editing) of how it might be done:
<https://groups.google.com/g/rec.bicycles.tech/c/UJdJQFTDgl8/m/NgOZUloVCwAJ>
<https://groups.google.com/g/rec.bicycles.tech/c/YQOILUCqIH8/m/StNB3J0_EAAJ>
<https://groups.google.com/forum/message/raw?msg=rec.bicycles.tech/UJdJQFTDgl8/NgOZUloVCwAJ>

>You still can outpace the light but generally I take it easy on night rides
>as certainly do so few MTB night rides, that stumping up for a 3000 or so
>(which seems to be the sweet spot) Lumens MTB light for the 1/2 times I
>might ride per year!

3000 lumens? Assuming 100 lumens/watt efficacy, such a light would
require 30 watts of power to drive. For the typical bottle dynamo,
that would require 10 dynamos. For the average riders at 15 mph, who
normally provides about 200 watts, that would produce approximately a
15% increase in drag to maintain the same speed. If you want battery
power, I can grind the numbers for you later. However, if you only
need that much light 1 or 2 times per year, a 2nd headlight, or some
kind of adaptation might be more effective:
<http://www.learnbydestroying.com/jeffl/pics/bicycles/bicycle-flashlight.jpg>

[Roger Merriman]

I don’t as such have problems been seen, I run my commute light at low, and
only flick it to high as and when I need it, ie generally going though the
big parks you get these parts or woods and what not.

>
>> My commute has a few woods and what not, which 600 Lumens in a fairly wide
>> flat beam is great, on gravel or otherwise roads it’s equally fine. Once
>> into MTB trails it’s shape more than anything let’s it down, as the beam
>> can’t cover the dips and angles, where as my MTB light which is fairly
>> modest as these go, *only* 1000 Lumens is much better able to light the
>> trail even on low settings.
>
> Do you have a lux meter? Something like one of these?
> <https://www.ebay.com/sch/i.html?_nkw=lux+meter>
> If so, and your light an produce a roughly circular spot on a wall, I
> have a way to measure lumens without an integrating sphere (or
> integrating pipe). These are early versions (in desperate need of
> editing) of how it might be done:
> <https://groups.google.com/g/rec.bicycles.tech/c/UJdJQFTDgl8/m/NgOZUloVCwAJ>
> <https://groups.google.com/g/rec.bicycles.tech/c/YQOILUCqIH8/m/StNB3J0_EAAJ>
> <https://groups.google.com/forum/message/raw?msg=rec.bicycles.tech/UJdJQFTDgl8/NgOZUloVCwAJ>
>

I used to years back ie light meter for various cameras. But no longer.

The MTB light is roughly circular though the commute light is shaped it’s a
exposure Strada so intended for road use, so has a lower flatter profile
even running on high, with both LEDs running.

To be honest I’m used to Lumens which gives a reasonable measure of how
much power the light has, plus beam shape/usage.

Not thus far been convinced that Lux as way of measuring is better, most
seem to be measured at 10m again you need to know the beam shape to make
any meaningful measure.

>> You still can outpace the light but generally I take it easy on night rides
>> as certainly do so few MTB night rides, that stumping up for a 3000 or so
>> (which seems to be the sweet spot) Lumens MTB light for the 1/2 times I
>> might ride per year!
>
> 3000 lumens? Assuming 100 lumens/watt efficacy, such a light would
> require 30 watts of power to drive.

Sounds about right my MTB light claims it kicks out 1000 Lumens/10watts

> For the typical bottle dynamo,
> that would require 10 dynamos. For the average riders at 15 mph, who
> normally provides about 200 watts, that would produce approximately a
> 15% increase in drag to maintain the same speed. If you want battery
> power, I can grind the numbers for you later. However, if you only
> need that much light 1 or 2 times per year, a 2nd headlight, or some
> kind of adaptation might be more effective:
> <http://www.learnbydestroying.com/jeffl/pics/bicycles/bicycle-flashlight.jpg>

I don’t need it I rarely take the MTB out for night rides, more likely to
be the gravel bike which I tend to stick to roads be they
tarmac/dirt/battered farm tracks but either way at that Strada at 600
Lumens is quite capable, what it can’t do is MTB trails which tend to twist
and dip and so on, and you end up riding into un lit sections, where as the
MTB light this is where it shines! Even at low settings it covers the
ground and low branches however the land might twist and drop off!

Low on the MTB light is approximately same as full on the Strada, in terms
of perceived brightness.

Roger Merriman.
>
>
>

[Sir Ridesalot]

Right. Use different lights for different needs/terrain.

What I like best about a good external powered battery light is the simplicity to move it from bike to bike. Plus in winter when it's extremely cold I can keep the battery tucked inside my jacket and thus keep the run time high.

Cheers

[Jeff Liebermann]

On Thu, 15 Apr 2021 22:41:43 -0000 (UTC), Roger Merriman

<ro...@sarlet.com> wrote:

>Jeff Liebermann <je...@cruzio.com> wrote:

>> Do you have a lux meter? Something like one of these?
>> <https://www.ebay.com/sch/i.html?_nkw=lux+meter>
>> If so, and your light an produce a roughly circular spot on a wall, I
>> have a way to measure lumens without an integrating sphere (or
>> integrating pipe). These are early versions (in desperate need of
>> editing) of how it might be done:
>> <https://groups.google.com/g/rec.bicycles.tech/c/UJdJQFTDgl8/m/NgOZUloVCwAJ>
>> <https://groups.google.com/g/rec.bicycles.tech/c/YQOILUCqIH8/m/StNB3J0_EAAJ>
>> <https://groups.google.com/forum/message/raw?msg=rec.bicycles.tech/UJdJQFTDgl8/NgOZUloVCwAJ>

>I used to years back ie light meter for various cameras. But no longer.
>
>The MTB light is roughly circular though the commute light is shaped it’s a
>exposure Strada so intended for road use, so has a lower flatter profile
>even running on high, with both LEDs running.
>
>To be honest I’m used to Lumens which gives a reasonable measure of how
>much power the light has, plus beam shape/usage.
>
>Not thus far been convinced that Lux as way of measuring is better, most
>seem to be measured at 10m again you need to know the beam shape to make
>any meaningful measure.

Sorry, I wasn't very clear. I'm not trying to switch from lumens to
lux. I'm trying to use a lux meter to measure lumens. The problem is
that to properly measure lumens, one needs an integrating sphere,
where you capture and measure ALL the light produced by the headlight
under test. Integrating spheres are not cheap or easy to use. In
general, they require a trained operator to obtain accurate results.

"Bike Light Lab Testing and Verification | Mountain and Road Bike
Lights | Integrating Sphere"
<https://www.youtube.com/watch?v=BEz34lxUL6I>

"Measuring light output with an integrating sphere"
<https://www.youtube.com/watch?v=mvyptpA-BmY>

The problem is that most of the lumens claimed by headlight
manufacturers and vendors are little better than science fiction. It
wouldn't be a problem if everyone had a test lab measure their lights,
but that isn't going to happen.

What my method does is provide an easier and cheaper method of
approximating lumens output by using the light measured over a small
area with a Lux meter, and extrapolating that area over a known spot
size. It makes a few marginal assumptions (circular spot pattern and
rapid fade to black at the spot margin) and assumes some comparison
factors that are not necessarily identical for different headlights
(operating temperature, color temperature, DC input power, lens and
reflector losses, bicycle air cooling, manufacturing variations, LED
brightness bin count, etc). Still, it provides a MUCH better
approximation of reality than most of the advertised lumens numbers.

>>> You still can outpace the light but generally I take it easy on night rides
>>> as certainly do so few MTB night rides, that stumping up for a 3000 or so
>>> (which seems to be the sweet spot) Lumens MTB light for the 1/2 times I
>>> might ride per year!
>>
>> 3000 lumens? Assuming 100 lumens/watt efficacy, such a light would
>> require 30 watts of power to drive.

>Sounds about right my MTB light claims it kicks out 1000 Lumens/10watts

Yep. If I know the LED type, LED temperature, type of optics, and
power consumed by the LED, I can make a fairly good guess(tm) for the
efficacy (lumens/watt) and estimate the lumens output. The 100
lumens/watt is typical for commodity LED's and optics.

>I don’t need it I rarely take the MTB out for night rides, more likely to
>be the gravel bike which I tend to stick to roads be they
>tarmac/dirt/battered farm tracks but either way at that Strada at 600
>Lumens is quite capable, what it can’t do is MTB trails which tend to twist
>and dip and so on, and you end up riding into un lit sections, where as the
>MTB light this is where it shines! Even at low settings it covers the
>ground and low branches however the land might twist and drop off!

Well, this is a tech newsgroups. Let's design a light that goes
around corners. Buy a small drone with a camera. Program it for a
"follow me" type mode to fly about 5 meters ahead of your bicycle. The
drone camera is aimed in the direction of travel by the position of
the handlebars. Video is transmitted from the drone back to a HUD
(heads up display), smartphone, or video display mounted on your
helmet or handbars. The drone has an LED light that illuminate the
field of view.

I can see plenty of potential problems that will need to be solved
before this is a workable solution for partially blind bicycle riding.
For example, if you're riding in a group, the riders ahead will
receive an unwanted hair cut from the drone. Maintaining a constant
altitude or track is difficult in areas with hills and canyons because
they block the view of the sky needed for GPS.

Soon everyone will be riding behind a drone "pace car" with lights and
perhaps a camera. For safety, of course.

>Low on the MTB light is approximately same as full on the Strada, in terms
>of perceived brightness.
>
>Roger Merriman.

[AMuzi]

I am indebted to Mr Scharf for the very useful term,
'chilumens'.

In our industry, management usually decrees something like,
'competitor says his gizmo is 950 lumens. Make our our
package say 1000 lumens at least.'

[Roger Merriman]

I found while that’s certainly the case with the eBay specials, if it’s
known manufacturers then it’s broadly in the same area, clearly measured
from the led, so no losses from lens etc, but it’s close enough to use a a
measure.

It’s why helmet lights started, as you can use a relatively low powered
spot to track the trail. Coupled with a bar mounted flood, I personally
don’t need such as I tend to stick to the tamer stuff at night! Another
reason that I tend to use the Gravel vs the MTB for such stuff.

some of the MTB brothers of the Strada have enough power to light up the
Forest but for my uses it’s not needed, though if you do ride manmade
trails at night doesn’t take much to reach the limit of my MTB light hence
why you get the such lights pushing out between 3000/5000 Lumens.

> I can see plenty of potential problems that will need to be solved
> before this is a workable solution for partially blind bicycle riding.
> For example, if you're riding in a group, the riders ahead will
> receive an unwanted hair cut from the drone. Maintaining a constant
> altitude or track is difficult in areas with hills and canyons because
> they block the view of the sky needed for GPS.
>
> Soon everyone will be riding behind a drone "pace car" with lights and
> perhaps a camera. For safety, of course.
>
>> Low on the MTB light is approximately same as full on the Strada, in terms
>> of perceived brightness.
>>
>> Roger Merriman.

Roger Merriman
>
>

[Roger Merriman]

Can’t say I’ve noticed much reduced run times, though the Strada is all in
one so would generate some heat, and London winters are never that cold -10
at absolute limit, mostly +3 or so.

Plus short winter days mean I’m running on high more, this said my old
light ran off AA batteries and that definitely did dip in winter!

The old battery for the MTB does dip a bit, it’s a external one you mount
to the bike frame, though new one I haven’t noticed.

I do like the Strada as it just clips on and off, no straps and whatever so
can just clip on to either the Commute or Gravel bike, and remove if I park
up somewhere etc.

Roger Merriman

[sms]

On 4/15/2021 5:52 PM, Jeff Liebermann wrote:

<snip>

> The problem is that most of the lumens claimed by headlight
> manufacturers and vendors are little better than science fiction.

That is true for most of the no-name lights from Amazon, eBay, or
Aliexpress. But there have been independent tests of some of the more
reputable manufacturers' lights that show pretty accurate lumen claims,
at least when the battery is fully charged. For example, the Magicshine
Monteer 8000s claims 8000 lumens and it was measured at 7800 lumens
<https://www.mtbr.com/threads/magicshine-monteer-8000s-bike-light-review.1173672/>.

Look at the graphic at
<https://www.mtbr.com/attachments/rn1200-ranger-1400-allty1500-jpg.1368263/>
where they test three lights at maximum power and lower power. The
Magicshine RN1200 actually exceeded its lumen rating for the first 80
minutes.

Some manufacturers take the rating of each LED, add them up, and claim a
theoretical number of lumens. If you had active nitrogen cooling you
might reach the theoretical number, but you'd never reach it in normal use.

Here's someone that tested some of the Chinese lights
<https://youtu.be/m1vFMbSEkdQ?t=957>. Not all of them are super-cheap,
some cost $100 or so. Two of the better ones he tests are the Towild
Professional 1800 ($65)
<https://www.wild-fires.com/bike-front-light-br1800/> and the Gaciron
V9D-1800 ($100, which I have)
<https://www.gaciron.com/V9D-1800bule.php>. The Towild is one of the few
LED lights I've seen that have two independent beams, spot and flood.
You used to see this long ago with MR11 and MR16 lights where they'd
have two separate beams for spot and flood.

[Oculus Lights]

From being on the inside of making and testing lights, one dirty secret is that the bike headlight "standard" only requires claimed lumen output for the first 30 seconds after being turned on.
The very small mass of heat bearing material in current small single LED lights saturates quickly. Light and Motion was a driving force behind that bike light headlight standard, and their lights are a noticeable offender.
The calibration, metering equipment, integrating sphere size, and mounting of a light makes a huge difference in measured lumens.
Another filthy secret is how MTBR and the related Francis Cebedo websites doctor their photography to make a sponsored product appear brighter in a photographs. Their reviews shouldn't be trusted as far as you can spit.
Before Francis was able to take them down, I downloaded pics he posted where the brighter looking Taz, had metadata showing that the picture was taking later than the picture of my Oculus (first generation 1500), and more exposure than the picture he took of the Oculus beam. But the picture of the Taz' beam looked abnormally brighter, with the blue night sky noticeably lighter blue than the beam picture from my Oculus which was taken earlier. He also aimed my beam downward more to the ground, resulting in smaller and uneven coverage than the beam picture of the Taz.
Still have the pictures with the metadata showing the time and exposure, if anyone wants to see how money and influence have as much effect on what a light gets tested and reviews like vs what a light's true output and beam pattern is.

[sms]

On 4/16/2021 3:27 PM, Oculus Lights wrote:

<snip>

> From being on the inside of making and testing lights, one dirty secret is that the bike headlight "standard" only requires claimed lumen output for the first 30 seconds after being turned on.

Yes, that's why the test results at
<https://www.mtbr.com/attachments/rn1200-ranger-1400-allty1500-jpg.1368263/>
are useful. They were done independently, and over a period of hours,
not seconds.