Things are never quite what they seem, especially if they appear straight forward.
I started this project wanting some cheap lights so that I could ride with the C1XV on week nights, rather than being restricted to weekends. The plan was to use a (rather tired) Sealed Lead Acid (SLA) battery that I had left from my aeromodelling days. Although this worked fine, the wretched thing weighed the same as my full susser's frame and shock, and was therefore 'not suitable'. Cue the start of this quest in earnest.
The key part of getting any light set to an acceptable weight is the battery, and for the best power * weight vs cost ratio you need Nickel Manganese Hydride (NiMH) cells. These require a specific charger, so instead of being able to use my old one, I had to buy that too, although it is at least a 'fast' charger. This contributed the main cost of the system, but if low weight is important then it's unavoidable. At least 2 hours run time was needed, so I chose 4.5Ah cells, which should give 2 ¼ to 2 ½ hours with a 20W bulb. If you feel the need for a big burn time and have deep pockets then NiMH cells are also available in 7Ah, 9Ah and 13Ah capacities, priced at around £1 per Ah. I decided not to over-volt the lamps (with a 13.2V pack) because it reduces run time.
The other area that became 'upgraded' from the original plan was the bulbs. I already had some 20W MR11 (35mm) floods, bought for 50p each in Aldi last year. With the SLA these gave about 2 ½ hours burn time. Although the light was strong, with a 38 degree spread the beam was diffuse and poorly patterned, so I sought a narrower beam. Eventually I selected 24 degree and 8 degree 'low energy' bulbs, to provide a main riding light plus a (very) long range light for fast downhills. Although rated at a 20W current draw, they are reckoned to have similar output to a standard 35W Dichroic unit. Not bad, since the Lumi 10W HID is equivalent to a 40W bulb.
Construction
Battery
Battery ingredients
A bottle pack was chosen for ease of swapping between bikes. I started by soldering the tagged cells together to form 2 X 4 cell packs and a 2 cell pack, to which the fuse was attached. These were encased in rubber sleeves made from an old inner tube and slipped into the bottle. Short sections of power cable were soldered to join the cells in appropriate sequence.
After drilling the top of the bottle and feeding the power cable through, one strand of the cable was connected to the fuse and the other to the final cell at the other end of the pack. Joints were covered with heat shrink insulation.
The fuse was pushed down between the cells and spaces between packed out with rubber from the old tube.
Finally, the cable entry hole was sealed with silicone sealant.
Lamp holders
Making the holders
12mm holes were drilled approx 1cm off-centre in the back of the blanking caps, and a slot cut in the skirt to allow the off-centre mounting of the switches for clearance of the lamp connector.
4mm holes were drilled in the body of the couplers near one threaded end and a length of power cable passed through (for one lamp there were 2 sets of holes for the second cable to the other lamp).
One end of the cable was soldered to a switch tag, the other soldered to a wire from the lamp connector. The second lamp connector wire was soldered the free switch tag. Joints were covered with heat shrink insulator.
Both lamps holders were wired this way and in addition, in one lamp holder a second cable was routed out to the other holder, and was soldered alongside the first, in parallel. Pipe clips were screwed to the 'bottom' of the holders so that they were a tight fit, and the ends of the screws inside the coupler were superglued to stop the screws coming undone.
The switches were mounted in the end caps before fitting to the holders, wires tucked out of the way and the bulbs connected.
Note: the rubber compression rings were left out of the bulb end of the holders when everything was assembled.
Finally, the wires from the first lamp holder and the wires from the battery were soldered to barrel connectors. This was done such that it would only be possible to connect one way round, preventing the charger being connected in reverse. A short length of cable was also soldered with connectors to allow the charger to be attached with crocodile clips.
Evaluation
The light result
The 24 degree Osram IRC bulb produces a very even and controlled circle of light with excellent projection. The Philips masterline spot suffers slightly uneven focussing, but produces a small spot of very intense light. A quick 'Saturday night sortie' at the guitar playing badger's house matched them up against some popular budget lights: Sigma 5w/20w and Cateye RC230 10w/10w.
In terms of light intensity and evenness of beam the 24 degree Osram lamp was a long way ahead, completely swamping the Cateye unit, even with both lamps on. The Sigma unit could produce the odd hot spot of light with both bulbs on, but suffered poor focussing and unevenness. Switching on the Philips spot put that lamp in the shade, providing an intensely bright centre to the beam. Battery weight was also noticeably lower than either of these units, despite having a longer run time.
The light result
In practice the 2 lamps balance each other very well. The Osram lamp was good for general riding, producing a decent spread of light at sufficient distance to avoid normal trail obstacles and supply information about the terrain ahead. For most riding it would be quite sufficient. On a fast-ish local descent I switched to spot only. This was fine, provided the terrain was read well in advance and the trail not too twisty. With both lights on everything is well lit, but it doesn't seem to provide quite so much detail as with a single light. There is also rather a lot of light spill from the housings, and this reduces the effectiveness. It may be necessary to paint the inside black to reduce this. Brown-out arrived after about 1 hr 10 mins with both lamps being used together about 25% of the time. I expect this to increase as the cells are cycled
Both lamps ran quite cool. At the end of the ride the spot glass was cool enough to touch with bare skin, although it was quite a cold night. The mounting system worked OK, with the lamps remaining as placed, despite riding across a VERY bumpy field.
When time permits I'll try a comparison with some Lumi or cateye HIDs. From memory, this system is less powerful than Doc Dolittle's Stadium, but not so far behind to be ashamed. Total system weight - around 900g.
Conclusions
The performance seems to be everything I'd hoped for, especially with the superb focussing on the Osram bulb. The Philips spot definitely has it's place too, although it isn't absolutely necessary.
Building the battery unit was surprisingly easy, however the lamp unit was actually a little tricky, due to the lack of space inside. Also the cost was significantly higher than expected, although like me, many will already have the consumables and some parts. About £10 (and 130g) could have been saved by using only one lamp, which would be quite sufficient. Tamiya or even phono type connectors would also have been a little cheaper and quite adequate, although less classy. Ditto the heavy power cable. The battery could be made lighter by packing out with expanded polystyrene or building foam, rather than rubber from the inner tube.
Overall then, I reckon similar performance to an upper mid range lighting system for around half to two thirds of the cost.
For a full set of photos taken during the building process, see tertl.fotospot.net
Detailed written instructions will also be available shortly for download.
David Lister
