David W. Knoble, Ph.D., PE (Tupelo, MS)

How much light do the dimming multi-wattage bicycle lights give? This article is a simple-scientific evaluation of the NiteRider Digital Pro 12 bicycle light, both (easy) theory and measurements.

In December, 1996 I was given a NiteRider Nitehawk (now named the Digital Pro 12) bicycle light. This light consists of dual reflectors and bulbs — a moderately wide angle (nominal 17 degree) 20W beam, and a wattage-controllable narrow angle (nominal 10 degree) 12W beam. The 12W bulb is step-controllable to 6W, 9W, and 12W levels. The advantage of the reduced wattages is increased battery usage times. Having used this light for several midnight centuries, I enjoyed the extended usage times, but I questioned the actual amount of light produced at the reduced wattages. Indeed, the 6W setting seemed to give lighting similar to my Cateye 2.4W HL-500. Why?

To affirm that this was not an illusion of the light-free roadway, I measured the light output at 12W, 9W, and 6W multiple times with a calibrated Minolta model T1 lightmeter.  The results confirmed the road observation. The 9W setting used 75% wattage (unmeasured) and gave 43.6% of the light (or 5.2W worth). The 6W setting used 50% wattage (unmeasured) and gave 17.6% of the light(or 2.1W worth). That is, the 6W setting gave about as much light as my Cateye HL-500. Of course, the HL-500 has a somewhat different beam pattern, so this is not a direct comparison.

How can this be so? Well, in first grade physics (you did take first grade physics, didn't you?) you learned about the Stefan-Boltzmann Law. Energy is radiated proportional to the fourth power of temperature. The production of visible light from incandescent filaments is strongly dependent on the temperature of the filament. The temperature, in turn, is dependent on the power expended by the filament. If you assume a reasonable 3100° K temperature for the 12W filament, then at 9W of power into the same filament, you’d expect the temperature to be slightly less than 2900K. At 6W of power into this filament, you'd expect the temperature to be lowered to about 2600K. At these lower temperatures, the efficiency of the filament in producing light (called efficacy) is much less. Thus, at 9W, you calculate the light equal to 6.5W (54%), and at 6W, you calculate light equal to about 2.6W(21%) of the same filament operating at 12W. These numbers don’t exactly match the measurements because the filament temperature at 12W is not exactly known, and because tungsten filaments are not perfect at radiating light. But it doesn't matter how you dim it, a cooler incandescent filament is always less efficient than a hotter one. Thus, it would have been more efficient to have had a 5W bulb in one light, and the auxiliary 10W in the other. The 5W would have given me twice the light as the 6W setting, for slightly less wattage consumed.

How did the Pro 12 work in practice? This light was used while riding night centuries (no moon) on the Mississippi Natchez Trace Federal Parkway. Of course, an additional, redundant, headlight was taken for the 5-6 hour ride — a helmet-mounted 10W light (alkaline cell). Because of the lack of street lighting, and since most house lighting was either hidden or quite far away, the 6W setting (giving the light of a 2W lamp) was quite adequate. I was able to see the road and striping for a considerable distance, and to read signs well before I approached them. The few oncoming cars dimmed their headlights while well ahead of me, indicating good visibility. On down-hills the additional 10W (or a higher setting) was added for safety at higher speeds. Later, on a portion of rough road surface, the combination of the 10W helmet mounted light, and the Nitehawk 6-9W light was more than adequate to pick my way between the holes. The 9W setting was used sparingly. The NiCad battery lasted the entire 5:30 ride, with some energy to spare. Subsequent night rides indicate that the 6W setting is quite adequate as long as traffic is very slight. Significant on-coming traffic causes enough night-blindness that I prefer the 12 or 20W setting. Again, the lower settings are principally for use on long rides when battery use times are critical. Short night rides (e.g., a short commute) permit the luxury of the higher settings, even including the 32 W setting. By the way, since the reduced wattage settings are obtained by switching the filament on and off rapidly, there is enough radio frequency interference that my Cateye heart rate monitor would not read. This is no surprise, since an HRM works on very low levels of radio power. The HRM works fine when the Digital Pro 12 is on the 12W or 20W setting.

Bottom line – dimming incandescent filaments makes them considerably less efficient. It is more energy efficient to have the bulb wattage fit your intended need. However, if you need extended battery times occasionally, or can't find a bulb that's appropriate for you, dimming headlights are a useful alternative. You would want to take a redundant lighting system with you on a long night ride anyhow, in case one light fails. (If you’re walking in Mississippi, you’re mosquito bait. If it’s really dark, you probably would sit till morning twilight, and be a mosquito-fest.) You can also consider two systems either of which would permit you to limp home (ride slowly) if the other failed. But each system should be adequate for the length of the ride, including a tire change, etc.