Exercises with road light beams

I've been playing about a bit to try and get a good road lighting solution, putting the light down on the road and not wasting it up in the trees. I had a bit of a discussion on here previously but unfortunately it was lost when CPF went down. Anyway, a quick recap:

One idea is to in effect use just half a flashlight, with an emitter firing down into a parabolic reflector as below:

This could work well, with the beam changing from focussed (part of the LED is at the focal point) through to wider flood (part of the LED behind, like a defocused Maglite). The next idea is to flip the design vertically and move the LED forward, which is very similar to the way a standard halogen headlight in a car works. There the dipped beam is effected by having the bulb element in front of the focal point, and a shield preventing light emission down into the bottom part of the reflector. The design would be like this:

Either method requires cutting a reflector in half of course, but will give very good heatsinking as the LED star can be attached directly to a flat heatsink on the case.

Michael

Firstly the test kit:


An XM-L T6 driven by a Shiningbeam 2.8A driver. The reflectors, from rearmost clockwise, are:

• Standard Mag C/D reflector - 52mm smooth plastic
• Underwater Kinetiks 4*AA - 23mm smooth plastic
• P60 incandescent dropin - 26mm orange peel aluminium
• Standard Minimag - 20mm plastic
• Litho123 reflector for Mag C/D - 52mm smooth aluminium

Unfortunately the Underwater Kinetiks reflector couldn't cope with the heat of an XM-L, and the Minimag was just too small to usefully focus the light from the XM-L. So the beamshots here are from the two aluminium reflectors, the excellent Litho123 Mag reflector and the P60 orange peel drop-in.

Here's an animation of the Litho123 reflector, moving from having the LED behind the focal point to in front. The beam starts off below the cut-off when the LED is behind the focal point, as in the first method in the initial post (I'll call it method A).

Both methods could work. With method A most of the light is directed under the cutoff, including the spill. Method B would have to be inverted of course, and the spill is then upwards so outside the semicircular cut-off close to the front wheel may be a problem.

What's interesting is that the beam can be tightened up by moving the LED in the vertical plane as well. For method A, moving the LED further up and away from the reflector, and for method B moving the LED further up into the reflector has the effect of tightening up the beam as follows (Method A first):

And Method B:

Finally, here's a couple of shots with the P60 reflector. It is such a shame I can only find orange peel ones as I really need a smooth one! Method A:

And Method B:

Interesting, but aren't you loosing a lot of light since it seems there is a hole in the back of the reflector.
I think method A would provide a better beam on the road. It is really difficult to judge the usefulness of a bicycle light beam on a flat wall, but basically you want your beam to be brighter at the top (further on the road) than at the bottom.

Calina said:

Interesting, but aren't you loosing a lot of light since it seems there is a hole in the back of the reflector.
I think method A would provide a better beam on the road. It is really difficult to judge the usefulness of a bicycle light beam on a flat wall, but basically you want your beam to be brighter at the top (further on the road) than at the bottom.

Click to expand...

Yes, there is a hole in the back of the reflectors (unfortunately) as they are all designed for flashlights and not custom made. I did investigate the custom route but it would be a bit expensive. Method B loses less light, and with the Litho123 reflector Method A doesn't lose very much light out of the back as the hole is only 4mm high. The dome of the XM-L is nearly that height, so any light that would be reflected back would only hit the LED dome anyway.

Both methods provide a light that is brighter at the top for throw, just underneath the cut-off (of course Method B would be mounted inverted from what you see above). However from preliminary outdoor tests Method a seems to work a bit better, with marginally more throw and less light close to the front wheel. Method B looks impressive nearby with a big swathe of light, but doesn't throw quite as well I think. I need to create mountings for the reflectors to do some outdoor beamshots to make sure.