Let's design a road front light beam

[znomit]

Has anyone played with this reflector?

It seems to me that it may have a very fussy on focal point.

If it isn't too fussy, then more light from a broader source might be halfway decent. lovecpf

Unfortunately its the front window not the reflector he sells.

And its bulb specific, I don't think anyone wants to design around a 3w halogen nowadays.

 

[BrianMc]

It looked like my response to znomit did not load. If it did and this doubles up, the admin can delete it if I can't remove it in edit mode. 😗

Znomit made an excellent point that even IF (a big if) an optic from a 3-watt halogen would work, it has likely has little future production.

A previous thread on 'How Much is too Much' light for the road had those who supported a scorched earth approach: More is better. :thumbsup:

In my expereince: not always. 🙁

Using an HID on the road has hammered home the lumens versus beam issue. My neighbors complained, so most of the HID was aimed off-road to the right. There is little pedestrain traffic in the early morning or later at night, so that did not cause a second issue, unlike it might in say, NYC. I did not get the light where I wanted it.

One pickup driver was nearly blinded by the light (sounds like a hit song to me) when he ran a stop sign and turned left in front of me at around 6:00 AM. 500 lumens of HID shone full on his face through his driver's side window at maybe 10 feet away. I didn't feel very sorry for him with all that adenalin coursing through my system at the time. :devil:

To be fair, he probably had never met another vehicle at that time and place, let alone a bike, may have been half asleep :tired:, but obviously was not prepared to yield to cross traffic, and no Boy Scout. So a good bright light helps, but it's no safety talisman. 😱

The HID is gone and too expensive for its output to replace. I have a cobbled light giving me 80-100 lumens which is a bit more than a 'be seen' amount on dry pavement, and not enough with wet pavement.

So I have the need for more light, built some electronics kits, and lots of DIY experience, (mostly good), and a desire to experiment tempered by availability of funds.

A poster on the 'Too much' thread suggested there are multi LED solutions to the road high-low beam issue. I take it that these would be one or both, of either: a spot aimed low and right and a narrow beam aimed high and straight ahead (mostly) or a Hi-Lo current setting, as in the Taskled software. Any other solutions? Just reaim the light up/down?

So I searched CPF but did not find answers to:
1. Road experience with a current driven Taskled enabled Hi-Lo beam with one or various lights (beams shots yes, real world riding, no.

How well does this work in traffic? Any beam width issues?

2. Ditto for multi-light LED units. Although 10/20watt and 15/35 watt etc. dual light spot-narrow beam halogen systems have been around, their lack of lumens compared to new LED's makes those reviews irrelevant IMHO.

So how well do higher lumen dual/twin systems work as high/low beams?

(A lot of personal preferences are involved here, I know, but I think I can glean what I need.)

3. I am saving for a Schmidt wheel. Batteries can be incovenient. However, with only 3 watts, the lumens and distribution issues are critical. So using one, or at most two, LEDs ALL the time regardless of beam shape would be nice. Multi die LED's and focusing issues are an interesting problem within this larger low watts issue.

The CREE XP-G R5 will put 250-300 lumens on the road at 3.3 watts (1 A, 3.3 Ohms), depending on lens/reflector losses, how warm it gets, and losses in the rectifier and other circuits. The S2 should be 5-10% higher again.

I understand that 300 lumens was once thought adequate for night mountain biking. At 0.5 A output from a generator, the R5 is only 3.15 ohms, or 1.6 watts and puts out 140-200 lumens depending. So two at 500 are in the 280-400 lumen range. Moore's law says that in a couple of years we could see another 50% improvement on this.

So why do we want to mess around with hard-to-focus multi die LEDs? :shrug:

In the mean time, I need more light to ride, and to use as a benchmark for experimentation for my future generator system. I have about $100 or so, and a battery with chargers in hand.


I am thinking either 3 or 5 XP-G R5's to use a 12 volt 50 Watt-hr battery pack I have. Using the Bflex for three, or the Maxflex for five LEDs. (Four XP-G's puts me at a Vf of 12.3 at 300 mA to 13.2 Vf at 1 A and out of regulation with either driver). Running either system at lower amps will be more efficient, provide a lot of light, but permit much longer run times than the HID on the same battery. Runtime is important as the battery ages and loses capacity. Headroom to higher current will let me ignite the road, if I must. :devil:

Sorry this is long. I have read a lot of CPF and experience says to hone the design at least three times before ordering parts, measure twice and cut once.

Thoughts, anyone?

 

[BrianMc]

Two more ideas:

http://cgi.ebay.com/ebaymotors/Half...ptZMotorsQ5fCarQ5fTruckQ5fPartsQ5fAccessories

My oldest brother had a '53 Ford Mainline 2 door with the larger version of this sort of stainless steel cover for the larger dual headlight setup.

The 6 volt sealed beam pre-haolgens were very weak in low beam particularly with corrosion in the connectors. I seem to remember that these allowed using both beams even when meeting other cars. They also looked good. 😎

These practically scream 'simple'. And cheap.

If such a device reflected back to a reflector then out below the eyelid, a strong vertical cutoff should be possible with most of the light brought to bear on the low beam area.

It should be adaptable to either multi die or multi LED lights.

We haven't had the luxury of car headlight lumen levels on bikes for long so other than the German regs, and multi headlamp systems, the field looks open for experiment. Certainly the old classic incandescent headlight on a generator system needed no 'low' beam!

A flip down visor version would give hi-lo if mounted in an accessible spot like the handlebars or on the front of the stem.

This link looked like a nice brief overview of optical systems:

http://en.wikipedia.org/wiki/Headlamp#Optical_systems

Optics are optics, automotive or otherwise.

The shade method of the multi elipsoid projector lamp is an internal version of the eyelid concept, above. This would be a little harder to pull off in a bike light and remain water resistant, if not water proof. But it indicates that the removable shade concept has at least a little merit.

I have seen old Kodak slide projectors which have focusable lenses, reflectors, and used shades to cut all light between slides, at yard sales for less than a buck. A bit heavy for a bike but a cheap source of bits to experiment with. 😀

 

[clint357]

Hey, I know it's been a while, but could you do any outdoor beamshots with a set up like this? I'd be interested in seeing how the light is distributed in a real life setting...

The only trouble with this is that it does not condense the light horizontally, so you'll have a very wide beam. You might be able to put a reflector on the LEDs behind the rod though. Hopefully I'll have some pics posted soon of a P7 behind the DX euro aspherical lens.

 

[pepko]

this is my design ... ultrafire C2@SSC P7 DSVMI ...

 

[markus_i]

.....

3. I am saving for a Schmidt wheel. Batteries can be incovenient. However, with only 3 watts, the lumens and distribution issues are critical. So using one, or at most two, LEDs ALL the time regardless of beam shape would be nice. Multi die LED's and focusing issues are an interesting problem within this larger low watts issue.

The CREE XP-G R5 will put 250-300 lumens on the road at 3.3 watts (1 A, 3.3 Ohms), depending on lens/reflector losses, how warm it gets, and losses in the rectifier and other circuits. The S2 should be 5-10% higher again.

.....

Just one thought: the SON generator will produce 500 mA nominal, almost independent of the speed, and deliver ~0.6V/(kph) if mounted in the correct size wheel (~1V/kph if you use the 26" generator in a 20" wheel...). A simple and good solution is to use the SON as intended, driving a Schmidt Edelux. The doubler circuit in the Edelux drives the LED at ~1A and you get a decent aluminium housing, AR glass and near-perfect beam shape. Use a B&M Cyo and lose the aluminium case and AR glass.

If you still need more light at higher speeds (i.e. >20 kph), you might think about an additional high beam system: wire two LEDs with narrow beam optics of your choice in series, throw in a rectifier and a bypass switch that short-circuits the LEDs and then wire that stuff in series with the Edelux. With the switch closed, the additional LEDs are bypassed, if you open the switch, the current also goes through the additional LEDs and you can exploit the higher available voltage for road illumination.

Bye
Markus

 

[clint357]

this is my design ... ultrafire C2@SSC P7 DSVMI ...

Can you post a pic fo what the setup looks like with the full reflector? I have thought of doing something similar, but never got around to it.

 

[znomit]

this is my design ... ultrafire C2@SSC P7 DSVMI ...

Interesting but the spill is not the problem, its the shape of the hotspot we need to alter. The top of the reflector doesn't add to the top of the hotspot any more than any other part of the reflector.

 

[BrianMc]

Just one thought: the SON generator will produce 500 mA nominal, almost independent of the speed, and deliver ~0.6V/(kph) if mounted in the correct size wheel (~1V/kph if you use the 26" generator in a 20" wheel...). A simple and good solution is to use the SON as intended, driving a Schmidt Edelux. The doubler circuit in the Edelux drives the LED at ~1A and you get a decent aluminium housing, AR glass and near-perfect beam shape. Use a B&M Cyo and lose the aluminium case and AR glass.

If you still need more light at higher speeds (i.e. >20 kph), you might think about an additional high beam system: wire two LEDs with narrow beam optics of your choice in series, throw in a rectifier and a bypass switch that short-circuits the LEDs and then wire that stuff in series with the Edelux. With the switch closed, the additional LEDs are bypassed, if you open the switch, the current also goes through the additional LEDs and you can exploit the higher available voltage for road illumination.

Bye
Markus

Yes, those are resonably priced, and not easy to duplicate. Upgrading the LED over time should be feasible, even if the beam is slightly affected by LED dimensions.

OK. So I can concentrate for now and build skills to that DIY high beam unit.

The in-the-planning-stage battery system can stay on the errand bike.

So that returns me to an earlier question. Other than having an extra head unit (cost, weight, clutter issues , I know), what is wrong from a lighting up the road point of view with a wider beam aimed closer and not in the oncoming motorists line of sight, combined with a narrow that you can use as a high beam? You do gain some redundancy, and with two sets of batteries, full redundancy to get home, even if it means slower and/or a reaimed light.

 

[BrianMc]

These won't be light weight, but they aren't huge, about MR16 size and seem to have a tight road cutoff:

http://store.ijdmtoy.com/SUNWAY-Glass-Projector-Fog-Light-Lamps-HID-Ready-p/ot_projector_sku71.htm

http://store.ijdmtoy.com/Hyper-F-Glass-Projector-Fog-Light-Lamps-HID-Ready-p/ot_projector_sku75.htm

The first was a link suppied by Mboni on this thread:

http://www.candlepowerforums.com/vb/showthread.php?t=238400

Cheap enough to play with.

The glass lens would be a significant part of the weight, I'd guess.

Makes me think of a cirular version of Troutie's lucite bar lens. :twothumbs

It would be interesting to see how a bar lens like that one would perform with the traffic-side end "bull nosed" in the same radial profile as it has top to bottom, and polished to full translucence. Outdoor beam shots would be wonderful (as if there is nothing else to do). 😗

Galt asked in a previous post about a source for the lucite lens material. Was it anything special? Is it just a chunk of stock lucite rod with two flats cut off lengthwise for bottom and back?

Any tricks to cutting and polishing it that would spare an experimenter some of the learning curve? :candle:

 

[syc]

Any tricks to cutting and polishing it that would spare an experimenter some of the learning curve? :candle:

I was playing with some acrylic rods and a dremel the other night to make a safety light for the burley trailer I haul my toddler around in. The main thing I found that that if you are using a dremel, its too easy to have the rpm's too high and have your grinding stone melt the acrylic instead of grinding it. You end up with molten plastic caking the surface of your grinding stone.
A file seems to take off the acrylic pretty quickly and doesn't melt it.

I also filed down and polished one edge of an old luxeon optic that I had laying around, to see if it would be possible to alter the beam pattern to get a gradient. It didn't seem to have any such effect, though it seemed to have slightly chopped off the overall spill area so that the hotspot is kind of imbalanced. Maybe I'll try taking more off the optic, since it's toast now anyway.

 

[BrianMc]

A file seems to take off the acrylic pretty quickly and doesn't melt it.

Thanks. That's what I was looking for. :wave:

A previous poster in this thread said he was playing with these:

http://www.dealextreme.com/details.dx/sku.1919

The user comments on DX associated with these indicate use without a reflector or colimator lens.

I would guess they are optimized for light impinging the inside at a fairly narrow set of angles as one should get off a reflector for a fog or spot light. Using this lens for light from the wide angles of an LED seems less than optimal.


If a set of narrow lenses like these:

http://www.cutter.com.au/proddetail.php?prod=cut837

#10417 (Troutie provided beam shots of it)

or these:

http://www.cutter.com.au/proddetail.php?prod=cut875

or the narrow angle triple, quad, or 7-Up lenses were used with multiple XP-G's as the light engine, would sliding this lens in front of them change an array from a fill-throw into a fog-light-low beam?

Anyone played with them or something like them enough to hazard a guess, share experience, indicate whether this needs to be tested?

 

[panicmechanic]

Hi everybody :wave:

Joined cpf for this thread. Very interesting subject!

A previous poster in this thread said he was playing with these:
http://www.dealextreme.com/details.dx/sku.1919
The user comments on DX associated with these indicate use without a reflector or colimator lens.

From what I understand there, it may be collimators with Fresnel front structure. Thus limited to 1 emitter per lens. The beam shape to expect should look similar to this.

If a set of narrow lenses like these:
http://www.cutter.com.au/proddetail.php?prod=cut837
or these:
http://www.cutter.com.au/proddetail.php?prod=cut875
or the narrow angle triple, quad, or 7-Up lenses were used with multiple XP-G's as the light engine, would sliding this lens in front of them change an array from a fill-throw into a fog-light-low beam?
Anyone played with them or something like them enough to hazard a guess, share experience, indicate whether this needs to be tested?

Probably you're looking for a Fresnel lens only, like what you get, in mild form, in the front lens of most (halogen) bike front lights.
Placing multiple lenses behind a Fresnel will give a similar beam as with a single collimator, only vertically wider. No sharp cutoff 🙁

 

[BrianMc]

Hi everybody :wave:

Joined cpf for this thread. Very interesting subject!

Ditto. lovecpf

From what I understand there, it may be collimators with Fresnel front structure. Thus limited to 1 emitter per lens. The beam shape to expect should look similar to this.

Turn the lens 90 degrees, would you have a narrow down the road thrower?

Probably you're looking for a Fresnel lens only, like what you get, in mild form, in the front lens of most (halogen) bike front lights.
Placing multiple lenses behind a Fresnel will give a similar beam as with a single collimator, only vertically wider. No sharp cutoff 🙁

Thanks. That's one experiment I need not consider.

Can't afford to repeat failures or run dumb tests.

The Maxflex bike UI two-mode option should keep my neighbors semi-civil. This forum and particularly this thread has helped me hone my design. I am planning for lens swap outs as well as ungradeability so I can play with it some. 🙂

 

[panicmechanic]

Turn the lens 90 degrees, would you have a narrow down the road thrower?

This will never be a thrower, I'm afraid. You'll get a vertical stripe of light without real hotspot, yet still blinding to oncomers.

I found a constellation that might work. Still thinking about the concept though. This is as far from being a headlight as it can be, hasn't even been taken outside for test and beamshots.

 

[BrianMc]

Sorry this is long. Hang in there for some results from simple tests stimulated by other posts.

Such as this one:

I'm getting what I think are pretty good results with a simple and inexpensive system.

Low beam:

High beam:

Light with proto paper cutoff device: (low beam is 3+3, high adds the middle 4)

Light w/o paper (less blurry and overexposed):

A slightly more complex design could achieve an ECE-type right side to the beam that would light the non-traffic side of the road better. But that's for version 2.

The above system is similar in function to the two filament two focal point high-low beam headlights for cars but uses bills instead of lenses to control the low bean and turns on the highs do their thing.

A hooded DIY light out of testing and on the road:

http://forums.mtbr.com/showthread.php?t=295664 Post #31

I first thought this was to help night vision by reducing scattered light.

There was an earlier post in this thread (keep missing it, so no link) about the billed lights that mentioned a bike light, which may be this one.


My test results:
Summary: A bill had no effect on the cutoff from the fresnel lens/reflector system of my car. I was suprised at how much it helped on the bike light. Shades in front of a light don't work as cutoff devices.

Setup: This was done in a dark garage with a 100 lumen LED home made bike light in its best road position and some spill light hit walls, floor, and ceiling. I was interested in how much a cardboard bill wrapped around the top and/or left of the light would affect the edges of the beam and any spill. The light is a 10 degree spot and the cover lens is slightly frosted.

Results: Considerable stray light (of moonlight or night light intensity) lit the ceiling at almost 75-85 degrees from the horizontal that I only noticed after I blocked it with the cardboard 'bill'. The brighter main beam normally made this stray light 'invisible'. Billing the left side also provided a worthwhile amount of cutoff. Excessive extension of the bill was required to begin to approach a low beam profile. So much so, that side winds might aim the light for you or affect handling. The lenths of shade I needed to go to were made worse by the radial distance (MR16 bulb in larger housing) from the edge of the light source to the outside of the housing. In the thread above, the shades are close to the lenses and the lenses are not wide.

Discussion: Getting good cutoff with a bill or practical length needs a setup like the quote, above. Bills with shiny inside surfaces should reflect intercepted stray light down on the road right in front of the bike or to the right and down IF a light had much. I'd guess that it is more important to redirect that light for low output lights. However, the amount of benefit will be directly in proportion to how poorly the optics function in concentrating the light into the intended beam in the first place (or presence of a frosted cover?). (This result was from one light only so over extension of the results is very easy.)

So unless designed a s a system, this seems to be not so much a 'retrofit low beam cutoff method', as it is an 'easy to implement fix' for a light system that has significant stray light OR as part of a High Low beam system as in the quoted thread above.

I doubt that this weak part of the beam affects oncoming motorists much. Removing the bounce off the ceiling close to the bike made the main beam seem brighter so the scatter affected my eyes' dilation. That may be worse for the extremely near sighted, like me as that is another uncorrected confounding aspect of the test. 😱

Since ceilings are rare on the paths and roads outside a garage, this upwardly scattered light should only be an issue in rain, snow, or in paths and roads with overhanging branches or with tunnels, otherwise it should be 'lost' to the sky and not affect night vision acuity. :candle:

I will revisit this after I build my XP-G lights. I will have to go with what we have learned from dual light halogen systems. I can't wait for a perfect low-high light and this will be much better for my neighbors than the previous 500-lumen blinding HID.

Other tests:
Shades in front of the light do nothing unless you cover more than half the light, and then you only dim the light whether they are reflective back into the light or not, and they do not force a cutoff. Not the same as bending a reflector.

this is my design ... ultrafire C2@SSC P7 DSVMI ...

Other directions:
This leaves three avenues to explore all previously mentioned here:

1. the shade combined with on/off system of the first quote.
2. modding German lights to handle more LED light and heat, mimicking or
home made copys of the German lights' reflector systems
3. the bike version of the projector car headlight approach.

This last approach has an interesting development. Mercedes has upped the ante from a forward-rotated or 'slide into place' on-off hi-lo beam shade in its projector lights, to a profiled rotating one that slowly changes from high to low beam and the reverse. It can stay in intermediate positions. A rotating one would be more compact and better for a bike.

Such a sytem adapted to a bike light would be neither light nor simple, but with different currents applied to the LED's by virtue of one of George's TaskLed drivers, could be adjusted nicely from scorched-earth such as Troutie's 7-Up XP-G at 23 watts:

http://www.candlepowerforums.com/vb/showthread.php?t=251079

or dialed down to mild-mannered with driver-friendly cutoff.

Given how much lighter battery systems are than we had to put up with at one time, the weight issue for commuting or training bikes is not a major issue IMHO.

Any optics enthusiasts out there want to tackle a bike projector lamp?

 

[BrianMc]

From the Project "Hellmeet" (headlamp) thread:

http://www.candlepowerforums.com/vb/showthread.php?t=217415&postcount=17

Notice how the redirection spreads the hot spot whereas the folded over reflector quoted in post above simply lopped the top off the beam.

I don't think we need the extreme concentration of light we needed with 3 watt halogen generator lights. But the faceted reflectors of the German lights might not be that hard to emulate.

 

[pepko]

my new bike light with 10x Cree MCE 5C and hyperboost driver from taskled ... i have 5x MCE for dimmed lights and 5x MCE for main lights - all 10x MCE together ...

some pictures here:
http://translate.google.com/transla...tonmag.cz/index.php?showtopic=995&sl=sk&tl=en

 

[BrianMc]

Pepko: The translation link is acting weird for me: just trying to open repeatedly.
It might be my end. Anyone else have problems?

If it is mostly pictures, maybe the original site will do. Thanks.

But 10 MCE's! :devil:

Let's see: 370-430 lumens in the lab, about 30% in losses here and there, 250-300 lumens or so for 1250-1500 in low beam and double that in high at recommended max current. That's a lot of light on the road. About the same as a pair of halogen car headligts in high beam or one car HID unit.

I was concerned about keeping my neighbors as friends with 3 XP-G R5's low and 6 in high (circa 650-750 low and twice that high at 1 A). (Troutie has beam shots of his 6 x XPG R5's in a thread here, as well as the 7-up and those upper three levels are bright).

So I am thinking I'll be setting the max limit in the twin bflex's for road use at either 350 or 500 mA and running both high and low beam in duomode depending on how well my aiming and cutoff experiemnts turn out. Maybe I set the low at level 2 and the high level at 1 A for the idiots for a high beam for those drivers who won't dip their HID highs. :devil:

 

[BrianMc]

For shoots where they can't use shades to cut off unwanted light from a diffuse light source, photographers use eggcrates which essentially are a fabric grid of hoods:

http://www.lowel.com/edu/light_controls/egg_crates.html

They can reduce the angle of the diffuse source to 50, 40, or 30 degrees from nearly 180 but the drawback is that they reduce the amount of light delivered, as these grids absorb, rather than reflect light. Some light fixtures like fluorescents had similar lattice grids or in the case of circular lights, coaxial circular versions.

My searches have come up dry, so either no one has played with these in a projection light like a bike light, or I am looking under the wrong words.

Has anyone seen any beamshots beyond the upper hoods of lights like pepko's on any kind of light besides these diffuse photo shoot light sources? If so, where?