BISY headlamp patterns

The following series of pictures attempts to fairly compare the illumination patterns of BISY-FL lamps, one pair with standard halogen bulbs and the other pair with LEDs. Each of the conventional BISY lamps uses a new Philips HPR-64 halogen 6V 3W bulb over-volted to 6.8V in an attempt to reasonably duplicate performance with a dynamo and no taillight. At 6.8V each bulb uses about 530mA, approximately 3.6W. Each of the LED-converted BISY lamps uses a Luxeon I type LXHL-DW01 sideways emitter, color temperature 8000K, driven at 350mA by a TaskLED nFlex constant-current regulator. The two lamps are driven in series. This results in approximately 1.3W delivered to each LED, slightly more than 1/3 the power used by each halogen bulb.

The pictures are as they came from my digital camera, except that each was reduced to half size using Paint Shop Pro 5. Colors, contrast, brightness etc are unchanged.

The characteristics of the light from the halogen lamp and the LED are markedly different. The halogen lamp produces light which to my eyes looks white, while the LED light looks decidedly purplish (rather like the HID lamps used on some automobiles). The difference in the light makes the images from the camera highly influenced by the white-balance setting, so I've specified that setting for the pictures which follow.

 

I'll give the comparison game away by saying up front that (1) the BISY light pattern is better controlled with the halogen lamp it was designed for, no surprise; (2) the halogen lamp provides better close-in light and peripheral light; and (3) two LED-equipped BISY lamps (using a total of 2.6W) appear in most respects more than a match for one over-volted halogen-equipped BISY lamp using 3.6W. Overall I'm glad I took the trouble to convert these lamps but I think it's fair to say that in light output the over-volted halogen may still have the edge in lumens per watt.

 

My digital camera has five white-balance settings: Daylight, Cloudy, Tungsten, Fluorescent and Fluorescent-H. These make the same illuminated scene appear remarkably different. In the photos below, the white-balance settings are specified. The images produced by the "Fluorescent" setting are the closest to what my eyes perceive.

The first set of six pictures is of two BISY lamps aimed at a cream-colored wall approximately 2 feet away, powered as noted above. The halogen lamp is on the right, the LED-converted lamp on the left.

The pictures above are with white balance = Daylight and exposure = 1/125 second. The halogen light looks excessively yellow/orange and the LED light doesn't look blue enough. Note also that the halogen pattern is more closely controlled than the LED pattern.

This is the same picture as the first one, still with white balance = Daylight but with exposure time doubled to 1/60 second. Now you can see some scattering of light from the halogen pattern but scarcely any from the LED pattern. When you see this with your eyes the scattered light is much more obvious, because your eyes have more dynamic range than the camera's image sensor. This scattered light produces the close-in and peripheral illumination we'll see in later pictures.

The pictures above are with white balance = Cloudy and exposure = 1/125 second. The halogen light looks excessively yellow/orange and the LED light looks white. Quite unrepresentative of reality in my opinion.

The pictures above are with white balance = Tungsten and exposure = 1/125 second. The halogen light now looks too white and the LED light looks too blue.

The pictures above are with white balance = Fluorescent-H and exposure = 1/125 second. The halogen light looks excessively yellow/orange and the LED light doesn't look blue enough. Similar to the Daylight results.

The pictures above are with white balance = Fluorescent and exposure = 1/125 second. The colors of both patterns look very similar to the way my eyes perceive them, but the LED pattern seems a bit dimmer compared with the halogen than my eyes perceive. Note the tighter beam pattern with the halogen lamp. Both patterns are brighter at the top as you'd desire for higher-speed riding, but the halogen lamp has a greater variation in brightness from top to bottom. As we'll see in later photos, the halogen lamp interacts with the optics to provide a lot more close-in and peripheral light. You can't see the scattered light in this photo, the exposure time isn't long enough.

 

Next is a series of pictures taken in my back yard. I'd have preferred a parking lot with striped parking spaces, but couldn't find one large enough and dark enough anywhere nearby. So I set up a line of small orange "traffic cones" at 10-foot intervals, with double cones at 50 feet and 100 feet. I also put a line of cones at the same spacing, 5 feet to either side of the center line of cones, to simulate a 10-foot-wide lane of the road. The cones farthest away are at 120 feet, since the central cone at 130 feet would have been just behind the tree at the center. 120 feet seemed far enough; may as well make a virtue of necessity! It had been cold, note the remnant of a prior snowfall at right foreground.

This photo of the test corridor was taken about 30 minutes before dusk, preceding a clear and moonless night. The two cones in the foreground are at the "0-foot" point and I'm standing perhaps 8-10 feet closer to the viewer. The stepladder I used for mounting the lamps was placed directly over these "0-foot" cones together with the tripod for the camera, and the lamps were on a step about 40 inches above ground level. Thus the test setup reasonably replicated handlebar mounting. The lamps were spaced about 8 inches apart horizontally, a requirement imposed by the C-clamps I used to hold them. On a bicycle I would mount them either vertically spaced, or much closer together horizontally.

It was pretty dark that evening. Here's an 8-second exposure with white balance = Daylight. You can't see much except a little skyglow and tree shadows in the upper RH corner!

Here are the two halogen-equipped BISY lamps mounted as noted above, using white balance = Fluorescent and an 8-second exposure. The first cone you can see is 10 feet from the lamps. The lamps are aimed so that one illuminates mostly the left-hand side of the lane and the other, the right-hand side, with light extending approximately 3 feet off the ground (as well as I could set it up) at the far end of the range. Note the clearly defined illumination pattern, the close-in light, some light scattered on the surrounding trees and bushes in the foreground, and the clearly visible tree trunk at center rear about 130 feet away.

Here is the pattern from the LH halogen lamp, with white balance = Daylight and an 8-second exposure. Note the well-controlled but narrow pattern. Unfortunately I didn't take any single-lamp pictures using white balance = Fluorescent. The lights at the left are due to the neighbor behind our property turning on his outside garage lights.

And here's the pattern from the RH halogen lamp, again with white balance = Daylight and and 8-second exposure.

 

Now let's look at the patterns from the LED-converted BISY lamps. First, with both lamps on.

Here are the two LED-equipped BISY lamps mounted as noted above, also using white balance = Fluorescent and an 8-second exposure. The first cone you can see is 10 feet from the lamps. The lamps are aimed so that one illuminates mostly the left-hand side of the lane and the other, the right-hand side, with light extending approximately 3 feet off the ground (as well as I could set it up) at the far end of the range. Same setup as for the halogen lamps. In comparison with the halogen patterns above, note that the illumination corridor is less well defined and narrower, the illumination up close and scattered into the nearby trees and bushes is significantly less, the tree at the 130-foot mark is less well lighted, and in general there seems less light available.

I must add however, that while the halogen bulbs clearly had the advantage my eyes didn't perceive the difference in illumination to be as significant as it appears in these pictures. Also bear in mind that we're comparing 7.2W of halogen illumination with 2.6W of LED illumination.

It would have been interesting to have used Luxeon III emitters instead of Luxeon I, the heat sink approach certainly appears adequate. But I wanted power consumption which would allow me to run both lamps at all times from a 3W dynamo, and have enough power left over to charge the battery pack which will also see duty as a source of power for camp lighting etc. I was also aiming for all-night run-time with both lamps on a series string of four LiIon 18650 cells and no dynamo. See the "Related Headlamp Topics" page for some details.

Here is the pattern from the LH LEDlamp, with white balance = Daylight and an 8-second exposure. Unfortunately I didn't take any single-lamp pictures using white balance = Fluorescent. The Daylight white-balance setting makes the LED light seem dimmer.

And here is the pattern from the RH LEDlamp, again with white balance = Daylight and and 8-second exposure.

Return to Converting a BISY headlamp to LED