See the following for some information from Daniel Stern regarding Hella, IPF, and Cibie H4 conversions:
On Tue, 2 Nov 2004, Parcell, Jason wrote:
> I have a 2001 Jeep Cherokee, which uses the 6054 type sealed beam
> headlights. I previously had Sylvania Xtravision sealed beam
> replacement headlights, and upgraded to Hella E-code H4 headlamps. I
> upgraded based on some advice that I would have much improved nightime
> vision. After installing the E-Codes, I noticed they didn't put out
> nearly as much light as the sealed beams. The lighting pattern on low
> beam was nice, but still didn't put out much light, and the high beams
> put out a lot of light, but it was so focused, that I couldn't see the
> sides of the road. Could the lack of light output be a bulb-related
> problem, or is this just the nature of the E-Code lighting pattern?
It's the nature of *those particular* E-code lights that you installed. The Hella units are widely sold but -- in the 200mm x 142mm rectangular format you have as well as the 7" round format -- poor performers. The problem is worse than you state -- the low beam has a sharp cutoff, which is good, but that's all that can recommend it. The beam is too narrow, the hot spot too far over to the right, there's not much light within the beam pattern, and there's a *vexing* problem with those lamps related to beam
focus: The high beam hot spot is separated from the low beam hot spot by an excessive vertical amount. This means: Aim the low beams properly, and the high beams are focused up in the trees. Pull the high beams down where they can actually do you some good, and the low beams end 30 feet in front of the car. I exaggerate for illustration, but the effect is very real and not solvable; it's in the optics.
At this point you may feel as though you've been "burned" and not want to try and fix it, but if you're up for trying again with some good lamps, I do have the ones that were factory installed on Jeeps originally destined for export. They are Cibies, and they exhibit none of the problems your Hellas have.
Take a look at
http://www.torque.net/~dastern/Photometry/isocomparo.html .
These are isocandela diagrams for four different 7" round H4 headlamp units. I haven't got the 200mm units scanned yet, but they are very similar.
If you're not familiar with isocandela diagrams, these will look like random squiggles and lines. Think of it as a topographic or "contour" map of the correctly-aimed beam pattern. Each differently-colored line represents the threshold of a particular intensity level, with the color legend located to the right of the isocandela diagram. The diagram is plotted on a chart calibrated in degrees. Straight ahead is represented by (0,0), that is, zero degrees up-down and zero degrees left-right.
To get a mental approximation of the units and amounts under discussion
here:
Parking lamp: About 60 to 100 candela
Front turn signal: About 500 candela
Glaring high-beam daytime running lamps (e.g. Saturn): 8000 candela
The parameters to pay attention to are the luminous flux (total amount of light within the beam), the maximum intensity and its location within the beam relative to the axial point (H,V) -- the less downward/rightward offset, the longer the seeing distance -- stray light outside the beam pattern and effective beam width (contained within the dark-turquoise 500 candela contour)
The two lamps at the top of the page are no longer produced, which is sad
-- they're number one and number two in performance in this comparison. Focus on the bottom two diagrams, Cibie vs. Hella.
Things to notice about these two diagrams:
(1) The Cibie produces a much wider beam pattern than the Hella. The 1000 candela line of the Cibie's beam pattern extends from 25 degrees Left to 25 degrees right, while the 1000 candela line of the Hella extends from 18 degrees Left to 20 degrees Right. At a distance of 50 feet from the car, this means the 1000 candela-and-brighter portion of the Hella's beam is 10.5 feet narrower than that of the Cibie. The 300 cd contour of the Cibie's pattern is *far* wider, extending from 43 degrees Left to 50 degrees Right, compared to 26 Left to 25 Right for the Hella. This means the overall useful width of the beam pattern at 25 feet from the car, as perceived by the driver, will be 40.7 feet for the Cibie and 22.3 feet for the Hella.
2) The total luminous flux (overall amount of light) within the beam pattern is 695 lumens for the Cibie, 463 lumens for the Hella - the Cibie is 50.1% more efficient. (the TLF data is listed as "Luminous Flux" in the readings up above the isocandela diagram)
The high beams for these two lamps (isocandela diagrams not yet scanned
in) are somewhat similar in overall performance and amount of light -- the critical difference is that the Cibie's high beam hot spot is located closer to (0,0) and closer to its low beam hot spot. The Hella's high beam and low beam hot spots are separated by a fairly large vertical amount, giving the problem described above.
So...whaddya say? Try again with good lamps?
DS
--------------------------------------
On Tue, 2 Nov 2004, Parcell, Jason wrote:
> You described exactly what I was seeing! I had my lows set perfectly,
> even though they weren't that great, and my highs barely even touched
> the road surface!
...and yet the trees were nice and bright.
> If I can get the vendor to let me return the Hellas, I might consider
> the Cibies. Where are Cibie lights made?
Belgium.
> Do you have any opinions of the IPF E-Codes?
Yes:
(1) They're not E-codes; there is no E-mark, which means they've passed no photometry or environmental-resistance tests, which means all we have to go on is IPF's breathless hype.
(2) Contrary to IPF's breathless hype, the lenses are *NOT* 3/8" thick and are *NOT* tempered glass. What else are they lying about?
(3) The performance of the "old" IPF lamps, with lens optics, is fairly decent. Not as bad as the Hellas, not as good as the Bosches or Cibies. The new "IPF" clear-lens headlamps are garbage. Massive upward stray light, poorly-formed beam patterns -- they're made in China by a company called NS-Sirius.
(4) IPF's line contains some good stuff. Their 840 fog lamp is a terrific lamp given its size and shape, and I'd pick it in a moment...if it were imported to North America. But the line appears to be moving towards hype and nonsense used to promote BS products. Their latest gimmick/scam product, the "X51 Fatboy" bulb, which is a 60/80w bulb (sixty over eighty, 60w high beam and 80w low beam). They've revved up their hype machine with all kinds of meaningless gobbledygook about this bulb, e.g. "190w optical effect on low beam, 150w optical effect on high beam". There is no such thing as "optical effect", and a bulb's wattage is only a measure of its electric power consumption, NOT its light output. Also, the large size of the glass on the "Fatboy" bulb is visually impressive, and I'm sure it'll trick lots of people into shelling out $70 for a set of these bulbs, but here's something for potential buyers to remember: The *smaller* the glass tube of a halogen lamp, the better its performance!
So yeah, I'd buy 840s, but I'd do it grudgingly because I don't have a lot of respect for companies that feel the need to lie to sell their products.
> What kind of bulbs would you recommend with the Cibies?
Osram ultra high efficiency items, $17/ea.
> If I stay with stock wattage, should I upgrade my wiring?
It always helps, but it's not necessary. First priority ought to be getting good lamps with good bulbs in there.
------------------------------------------------
On Tue, 2 Nov 2004, Parcell, Jason wrote:
> The Osram ultra high efficiency bulbs you spoke of, are these
> Silverstars
Osram Silverstars, yes. NOT Sylvania Silverstars. The difference is critical. Sylvania Silverstar is short-lived junk with blue glass. Osram Silverstar is the highest-performing road-legal stock-wattage H4 currently extant on the planet.
Info here:
http://www.candlepowerinc.com/pdfs/SBL_Headlamp.pdf
Even more info:
AutoExpress finally released the results of their new H4 (=9003, =HB2) bulb tests.
Nothing that shocked me. In a nutshell, OSRAM (not Sylvania) Silverstar or Philips VisionPlus is the best standard-wattage bulb you can buy, the best blue-coated bulb's performance doesn't match a decent standard bulb, PIAA is garbage, and the range of performance in bulbs on the market is huge.
Viz:
Standard and blue bulbs:
http://www.autoexpress.co.uk/product_test/product_test_story.php?id=39920
"Plus 30" high efficiency bulbs:
http://www.autoexpress.co.uk/product_test/product_test_story.php?id=39919
"Plus 50" ultra high efficiency bulbs:
http://www.autoexpress.co.uk/product_test/product_test_story.php?id=39917
Annd surprise of surprises, people who come to me wanting blue crapola or demanding to know where I get off trashing PIAA's "good name" are always stunned at how well they can see when I send them away with...with...well, by golly, with the bulbs this test, done an ocean away from me, says work best!
Stern's Lighting Law Number 1: There is no such thing as a bulb that turns a bad headlamp into a good one.
COROLLARY: There are lots of bulbs that turn good headlamps into
bad ones.
DS
Who is beginning to tend towards cynicism but hates to say "Toldjya"
----------------------------------------------
On Tue, 2 Nov 2004, Parcell, Jason wrote:
> Very nice information. I'm just about sold!!! Any way you could get
> the isocandela diagrams for the 6054's low/high to me? That would
> definitely seal the deal!
Not quite, but can do the next best thing.
Isoscans for H5006s are at
http://www.torque.net/~dastern/Photometry/575.html (top two diagrams). Performance closely mirrors that of an H6054 in terms of beam formation and strength.
And here's something I wrote a few months ago regarding some dissective analysis of various H6054s:
--
Spent some time on Thursday and Friday disassembling and experimenting with 200mm x 142mm rectangular sealed beams.
GE just released these brandy-new super-duper sealed beams ("Night Hawk") that're supposed to be the hot ticket for vehicles that use sealed beam headlights. Their beam formation is certainly better in several respects than other sealed beams: Focus is obviously much better, the hot spot is very well shaped and placed, and the beam is considerably wider.
Foreground light is still lacking, and upward stray light is still excessive. They didn't use a mask during aluminization, so the optically useless floor and ceiling of the lamp are covered with shiny stuff, same as the reflector. Therefore, light from the filaments hits the floor and ceiling and is emitted from the headlamp as optically uncontrolled upward stray light, which causes glare and backdazzle in bad weather.
I wanted to get in there and put matte black paint on the floor, ceiling and other optically-useless areas to see how much I could knock down the glare and backdazzle. It's tough to disassemble sealed beams, because they're...well, they're sealed. Time was, the glass lens was "welded" to the glass reflector; the lens-reflector junction was selectively heated so the two components fused together. Nobody does that any more because it's expensive and difficult; they mostly use adhesive of one sort or another. GE uses what appears to be some kind of a grey epoxy.
Drawing on experience gained from oven-assisted disassembly of other "permanently" assembled (but unsealed) lens-reflector assemblies, I experimented by putting a cheap GE 6052 in the kitchen oven. After just 30 minutes at 325 F, there was a "Popf!" from the oven, and I was able to use the edge of a butter knife to pry the lens away from the reflector.
GE stopped making nonhalogen sealed beam headlamps some time back; everything uses a halogen burner now, even the ones in "standard tungsten" boxes, which use twin-transverse-filament burners with the wattage and flux adjusted to approximate the characteristics of the previous nonhalogen assemblies. They suffer from all the typical problems: Very poor beam focus, low peak intensity, excessive uplight and upward stray light.
Once I removed that lens, it was easy to see why. The unlensed beam is a very poorly focused diagonal(!) oblong of "blotchy" light (gradients/streaks within the oblong). There's just no way to get anything even remotely approaching a well-defined beam pattern out of this. The best one can do is use the lens optics to shift enough light away from the upward-leftward area to comply minimally with the glare maxima. Heel distortion in the reflector was evident and can't have been helping any, either. Then add-in the aforementioned floor and ceiling reflections
*plus* light escaping the burner and travelling directly through the lens, and the resultant beam pattern is a sick joke.
OK, now I've proven it's possible to open a glued sealed beam without getting killed or injured -- onward to one of my only two examples of these H6054NH "Night Hawk" units. These proved considerably more, er, mayhem-resistant. It took three sessions of scraping adhesive out of the lens-reflector junction gap, and ramping the oven to 440°F, but shortly thereafter I was rewarded with a loud "POPFK!" from the oven. I'd placed the headlamp lens-down on the oven rack; eventually this heat level caused the reflector to separate forcefully from the lens.
After the reflector cooled, I applied power and examined the unlensed beam. These lamps use a burner with C8/C8 twin axial filaments -- essentially an HB5 burner of 65/55w, emitting an estimated 1400/1100 lumens high/low beam. The unlensed beam itself was very impressive -- a tight, centrally-focused round spot of light with moderate "fingers" upwards and downwards. Not so impressive was the amount of stray light coming primarily from the reflectorized floor (directly and via the reflectorized ceiling), also directly from the filaments.
I used some high-heat-resistant matte black paint on the lamp's floor, ceiling, distorted reflector heel directly behind the burner, burner "legs" and grommets, and the distorted areas of the reflector directly in front of the locating lugs. The resultant unlensed beam pattern exhibited considerably less stray light outside the beam pattern, while the beam pattern itself was scarcely changed (going by eye here -- no photogonio range available).
Next, I took a look at a Sylvania Xtravision H6054XV headlamp. This lamp uses a twin-transverse C6/C6 burner -- essentially an HB1 burner but with a 65/55w high/low filament pair emitting an estimated 1300/1000 lumens. The beam pattern from the assembly is *very* poorly formed and *very* poorly focused. The low beam hot spot is large and of low peak intensity, there is a large and intense vertical "spike" of light, absolutely zero cutoff and *very* high levels of upward stray and "flare" light outside the beam pattern. High beam throws light practically everywhere except straight ahead.
Therefore, I was surprised when I separated the lens from the reflector to find that the unlensed beam is very tight and well focused -- a properly oriented (horizontal and square with the vertical and horizontal) "bowtie" of light as one might expect to see from a transverse filament placed on the focus of a parabolic reflector. Stray light levels remained fairly high owing to near-heel distortions in the reflector and ghost images (reflections of the transverse filaments in the burner envelope glass, subsequently picked up and distributed by the reflector).
I used matte black paint on the floor and ceiling, the distorted reflectorized area of the heel, and the burner "legs". This reduced reflected upward stray light considerably, but a considerable amount still remained since the levels were initially very high and ghost images from transverse filaments cannot be really effectively be dealt with -- the only way I know of to mitigate or eliminate them is to use a burner with spherical rather than tubular glass, and no such burners exist. Of course, as was the case with the GE NightHawk, the addition of black masking material to reflective but optically-useless areas of the reflector/housing did nothing to attenuate the stray light coming directly from the filaments. A halfway job of containing this light was effected by masking the inside of the lens rim.
The really interesting part was the results when I experimented with different lenses in front of different reflectors. Using the *lens* from the standard nothing-special GE H6052 headlamp with the *reflector-burner* from the Sylvania H6054XV Xtravision headlamp resulted in a much better-focused, better formed beam pattern than from the Xtravision lens. Both of these lenses were designed for use with transverse filaments. Using the GE NightHawk lens in front of the Xtravision burner-reflector yielded a less well-formed beam (this lens was designed for axial
filaments) but it was still markedly better focused than with the Xtravision lens.
Using the Xtravision lens in front of the transverse-filament GE H6052 burner-reflector resulted in a poorly-focused, poorly-formed beam. Using the Xtravision lens in front of the axial-filament Night Hawk burner-reflector resulted in an extremely poorly-focused, poorly-formed beam. That Sylvania Xtravision lens optic is junk!
-DS
