Why I was interested in LED headlight bulbs:

My 3.0 engine with custom CF intake prompted moving to the smallest alternator I could find, for clearance. It is a 70 amp used in some older Corolla and Geo cars. While it should be fine for the meager needs of a street 818, I still want to minimize the electrical system power demands.

The 818's "new" style front end uses Hella 90mm Premium projectors, outfitted with H9 Halogen lamps. These cast a yellow tinted light of course. They meet all vision requirements of course. They use the most power of the three types I tested, of course.

I had an H9 HID lamp and driver already here, from another project. But what LED lamp to try? There are way too many different LED bulbs on the market. Hyper marketing claims have blown away any hope of deciding what is the right bulb just from reading specs. Some of the WEB reviews and discussions seem to say projector style lamps cannot work right with the LED bulbs, because the LED's are arranged on sides, and don't create a 360 output point source inside the housing. Hella sells an LED version of these projectors now, for a rather large sum of money - but is the projector itself redesigned or is the LED bulb made just for the projector?

From multiple WEB reviews these bulbs seemed to have prospect: DiodeDynamics SL1, Beamtech, Lightning Dark, Supernova V4. And I know Frank818 lists a bulb he is happy with. The DiodeDynamics is not cheap at about $80 each, but is a US brick and mortar, they have a no questions return policy and decent warranty, so I decided to test their SL1.

H9 bulbs tested:
----------------
Halogen bulb: GE 65W, 11.98 volts @ 5.00 amperes
- gray card reflected Lux = 87
- night view: dim appearing, strong yellow tint, consistent pattern, sharp low beam cut-off, soft high beam shape
- comment: dim is subjective, but seems to be about half to two-thirds the intensity of the other bulbs
----------------
HID bulb: Xentec 35W 6500K , 12.27 volts @ 3.38 amperes
- gray card reflected Lux = 88
- night view: good white faint bluish, minor uneven patterns in beam shape, sharp low beam cutoff, tight high beam spot
- comment: takes about 10 to 15 seconds to brighten from dim startup, not huge power savings
----------------
LED bulb: Diode Dynamics SL1 6000K (tested two bulbs), 13.00 volts @ 1.60, 1.52 amperes
- gray card reflected Lux = 107, 64
- night view: good white no blue, matches HID brightness, instant on, consistent pattern, sharp low beam cut-off,
- comment: second bulb down on Lux but on night test brightness seemed almost as good, low current
----------------

I'm going with the SL1 LED bulbs. I discussed the lower brightness bulb with Diode Dynamics sales and they will replace the bulb if I wish. The bulbs can be rotated in the socket somewhat to adjust their beam and I'm to try that next. For the moment, I'm not sure it was enough different in actual appearance at night to ask.

-----------------------------

Some background about testing lamps if you are still interested -

I did not take pictures, because I don't believe they convey sufficient information for anyone to compare and decide from.

I've done a ton of reading and mulling over the test process, but it comes down to how these look to my eyes more than anything else. The question to answer: "is the power savings worth whatever differences are observed for each lamp type?". There is also a question of cost, which is a consideration in selecting cheap vs expensive HID & LED devices - will low cost bulbs perform, last and match up, or is the high price item really no more than a "prestige buy"? I didn't test a low cost LED bulb so I'm in the dark there (!).

In each measuring test the light meter was sensing reflected light after it was projected from 12 feet onto a neutral gray photography standard card. The meter was 6" from the card, aimed at the card center. The card was angled 45 degrees to the lamp and the meter was at 90 degrees to the original beam. The beam was adjusted until the central and highest reading part of the (low) beam was hitting the center of the card. In each case the portion of the beam used was roughly the same area, just below high beam cut-off line and roughly the center of the left-right "sweep" of the overall beam. The high beam's "hot spot", when uncovered was always about 10 inches above and to the right of the low beam "center".

For the night distant viewing test, the beam was projected onto my white garage door in near pitch dark from 100 feet. The "levelness and pattern consistency" of the cut-off beam was viewed, as was the high beam mode's "pattern". This subjective test is really the one that is of most value, it's a little closer to the practical application.

I also did direct measurements of the beam as it comes from the lamp, but these are very sensitive to position, distance, etc. and are harder to replicate and compare on a one-to-one basis, even though such tests are how a real laboratory would conduct them. I think those measurements with my meter were misleading and difficult to replicate (due to its limitations or design) and so I omit them here.

Technical materials I read stated Lux is not a proper measure of source intensity, rather candela (cd) is the desired parameter. My Lux meter provides a conversion to get cd based on a specific distance, etc., but since it reads directly in Lux and is all I have on hand, I just tested and listed the values it gives using the reflected card method, which is as simple as it can be and is easy to repeat with acceptable accuracy. I did find that the reflected light intensity test is easier to setup and replicate with consistent results. I've also used it in testing existing OEM brake and tail lights while designing my custom tail-lights, more on that forthcoming.