308/328 Headlight Shootout - The Results Are In!

Over the past several years I've been working to find good headlights for the 308/328. From the very beginning with my stock 308, the 7" sealed beams were inadequate--I just couldn't see.

I'm now on my third iteration of 308/328 headlights, and have finally found something that I think is perfect. It is by far the best of the five headlights setups that I have personal experience with.

First, some FerrariChat history. The thread that started it all--by Mike Charness:

Headlamp Upgrade for 308/328 (long)
http://70.85.40.84/~ferrari/discus/messages/256120/98730.html


Basically, from the very beginning, we had the right idea. The basics are:

1. The sealed beam headlights in the 308s and 328s absolutely suck.

2. We need to replace these with aftermarket 7" replaceable-bulb headlamps, like Cibies (discussed below). Hella has historically been another common choice.

3. When we replace the headlamps (aka the housings, or lenses, or assemblies), we then get the option of running our choice of H4 bulbs. The wattage and brand we choose is important.

4. Headlights are high-current items. As such, running them through the stock fusebox is not the greatest idea in the world. The alternative is to use a special harness, which is a fused set of wires and relays which runs the headlights from current direct from the battery. Essentially With the harness, the power that used to run the whole headlights, going through the switch through the fusebox and through the relays and up to the headlights (say, 10-20 amps) now is only switching power for the relays (less than 0.25 amps). This is a Good Thing.

5. Running a headlight harness as above, without changing anything else, will result in brighter lights, because there is a voltage drop associated with going through the fusebox, relays, wiring, and headlight switch. That is, the voltage of the battery--say, 13V, is not on the stock setup the voltage that the headlights see (speculating, 1 volt less). The result is a direct, visible, and measurable reduction in brightness of the headlights.

6. As a fact to consider as we proceed here, if you upgrade the headlights with bulbs that use more power than stock 55W bulbs (greater current draw), you MUST use a harness because the stock wiring is NOT designed to carry the current drawn from higher wattage headlights. Fire and/or smoldering wire and/or ruined switches and/or blown fuses will result.


So this was a pioneering work by Mike Charness but, several years later, I have come to find out some flaws in this initial setup.

1. Using a headlight wiring harness--with relays, fuses, and heavy wire, direct from the battery--was not stressed. This is critical.

2. The choice of "Xenon" bulbs--not true HID (high intensity discharge), but bulbs which appear more blueish/whitish in color--is actually NOT recommended by experts. This color temperature of bulb actually results in less brightness overall given a fixed bulb wattage, because the blue tint/filter use to obtain the color temperature filters out some of the actual light.

3. The lenses ("headlamps") recommended are cheap crap. I ran these in my 308, and they did OK. But they're made of plastic on the back, are not made by any reputable headlight/optics manufacturer, and are basically in line with all the other Chinese/Korean imitation look-alike crap that is flooding among other industries the automotive headlight industry. It's just not the best way to do it. Another important fact is that in general, running overwattage (>55W, i.e. 90W/100W, 130W/90W) bulbs with plastic headlamps is a big no-no. Finally, for what it's worth, the vendor where we got these lamps from (KyotoUSA) is now apparently out of business.


All that said, Mike Charness was right on the money though with the idea of upgrading these things.


At some point, FerrariChatter Dr. Tommy Cosgrove followed up on Mike Charness's thread with some very useful suggestions:

OWNERS WITH FACTORY DIM HEADLIGHTS; PROBLEM SOLVED FOREVER CHECK THIS OUT!!
http://70.85.40.84/~ferrari/discus/messages/256120/217706.html


Here, we have the idea of the headlight harness introduced. Also introduced is the use of overwattage bulbs, or bulbs that are rated greater than 55/60W.

So all the pieces are in place. Problems? Yes.

1. The APC wiring harness is junk. I know. I've had two of them. Crappy ghetto Chinese relays (you want BOSCH), no fuses (you have to splice in your own), wire quality leaves something to be desired, and it is a bit TOO SHORT for our application (308/328). I had to fabricate extenders.

2. There is talk of Hella headlights. This is a good step forward, in that Hella is a very well respected automotive lighting manufacturer. But it's not the best (you want Cibie--more on this to come).


What about HID? Well, I've spent HOURS investigating this. You'll see posts from Modified348ts / Modman about HID. I was really into trying to get HID into our cars. I've spent hours upon hours researching options and possibilities.

My conclusion re: HID in 308/328? Not possible to do right. Doing it the right way is to install projectors, and they just won't fit inside our headlight buckets. Everything else is crap.

Here's a thread that really summarizes my quest with HID.

HID Headlights for 308/328: Sylvania Xenarc X6024
http://www.ferrarichat.com/forum/showthread.php?t=17003

I have since come to find out that the whole design process of the Sylvania Xenarc 6024s was a joke, and of course they're now discontinued.

After a fair amount of research, I've since come to understand that HID is NOT the "end all be all". HID done properly in our cars--projectors--would be pretty could. $1000 to $1500 for this, if you can ever find a system to fit.


Here's the NEW thinking in 308/328 headlight upgrades.

My suggestion is to get yourself to the Daniel Stern Lighting website located at http://www.danielsternlighting.com/. I have no personal affiliation of this company, but I have spoken with Mr. Stern, the proprietor, at length and have exchanged tens of emails as he's helped guide me towards a positive lighting result for my 328. So I'm a customer, but that's it.

My suggestion is that you read everything you possibly can at that site. It will take about an hour. Read about how HID is not the end all be all; and how blue bulbs (i.e., the blueish-whitish bulbs that I was running before) are NOT what you want. You will also find Mr. Stern posting in the USENET forums (i.e., Google Groups) as well as on the Philips forums and others. He approaches automotive lighting with a disciplined, cut through the hype, common sense, and scientific approach. As a scientist myself I can vouch for the way that he presents his thoughts on automotive lighting. I consider him an expert.

Which is fortunate, because if you read the above threads, with this lighting stuff, we've really been bouncing around with what "Seems good" with no basis or expert opinion on how or what we're ACTUALLY doing.

With Mr. Stern's input, and my motivation for better lighting, along with the basis of past FerrariChat threads on the subject, I've finally been able to converge on what will be my final iteration with upgraded 308/328 lighting.

For those of you who can't wait, here is the setup.

1. Headlamps: Cibie 7" E-code (European ECE Spec), replaceable beam, H4
2. Bulbs: German-made Narva overwattage 90W lowbeam, 100W high beam (90/100W) H4 halogen bulbs
3. Harness: Self-fabricated harness from Relay Kit furnished from Daniel Stern Lighting

All of these were purchased from Daniel Stern Lighting (http://www.danielsternlighting.com/products/products.html). The best part? The price. It's really in line. The ultimate in lighting upgrade, for $199, LIST PRICE. That's two headlamp assemblies, two bulbs, and the relay kit. Figure another $20 or so for heavy gauge wiring you'll need to buy to complete the headlight harness.

Two data points to contrast with.

Old Bad Lighting Option #1: KyotoUSA headlights with H4 bulbs supplied ($60), APC ghetto harness ($65), + inline fuse holders ($5). Total: $130. Result: better than stock, but still crap. Not a bad value, but not what you want. This is what I ran on the 308 for about a year.

Old Bad Lighting Option #2. Adjure / Crystal Eyes headlights (http://www.coolights.com/index.asp?PageAction=VIEWPROD&ProdID=363) ($154), Raybrig Whitesonic H4 bulbs (http://www.coolbulbs.com/raybrig.asp) ($79), APC ghetto harness ($65), inline fuse holders ($5). Total: $303. Result: better than stock, but still crap. Very bad value. This is what I've had on the 328 for the past year.


Contrast these costs with the $200 ultimate setup above, and we have a really good value. Beats $1000 - $1500 for decent HID, which isn't "better" or worse than the above halogen solution.

So that's the money. On a side note, it took me two hours to fabricate the relay harness, and about two hours for the headlight to install. I could do it again in half the time, of course...
So I wanted to contribute some actual PHOTOS of all of these. And so I bit the bullet and had a headlight "shootout" (what the hell else am I going to do on a Saturday night?) so you could see the difference in the beam patterns yourself. Pictures are included below.

OK. So we've established that (a) you DON'T want blue bulbs, (b) you WANT a relay harness, and (c) if you have a harness, you can run higher wattage bulbs. So that's a given. You pick the bulb wattage for the brightness you want. The light color temperature is good, and your wiring is good. The only variable now is the headlamps. If you're running higher wattage bulbs, you need a glass/metal headlamp. So those are the requirements.

What is all comes down to is BEAM pattern. That's what these photos show.

The contenders are:
1. 308 sealed beams (I still have these from my old 308--I sold it with the KyotoUSA headlights still installed), 55/60W, TS manufacture, 6014
2. 328 sealed beams, 55/60W, GE manufacture, 6024
3. Adjure / Crystal Eyes Diamond Cut Headlights with Raybrig Whitesonic 55/60W H4 blueish-tint (looks white) bulbs [Japanese Manufacture on the bulbs], as noted above in Old Bad Lighting Option #2.
4. Cibie headlights with Narva [German Manufacture] 90/100W bulbs (my final iteration in the 308/328 lighting saga)

Test conditions:
All photos taken in a 20 minute session of rapid headlight swapping, with the car warmed up and idling, Interstate battery, steady voltage of approx 13 volts measured at the battery, using Daniel Stern Lighting self-fabricated relay kit. Relay kit tested with a 0 volt voltage drop, that is, voltage at battery = voltage at headlights. Temp = 70 degrees, some humidity, 9:00pm test time (full darkness). Distance from headlight plane to garage wall was 17.5 ft +/- 0.5 ft. Headlights were not necessarily perfectly aimed.

Attached to this post is a picture of all the headlights.

Legend: Clockwise from upper left: Adjure Diamond Cut, 308 Sealed Beams, 328 Sealed Beams, Cibies.

Still to come: Results from the headlight shootout.

Quick note about the photos below: The lighting pictures are in black in white. This is done on purpose, so that only the beam pattern of the headlight is considered (that's what we're showing here). Lighting color should not be confused with "brightness" or headlamp performance, and so we take that variable out here with the black and white.

 

First up: 308 Sealed Beams.

Pic #1: Low Beam.
Pic #2: High Beam.

Conclusions: These suck. Notice rough "blob" of light as the focus area. Not enough spread. Beam pattern has light going all over the place, which translates into glare for other drivers. DOT spec. 6014, from way back in 1978. One of these is "TS" for TungSol, since bought up by a couple of companies (Westinghouse etc.). The other is a replacement of different brand.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Next up: 328 Sealed Beams. GE Manufacture, 6024 (slightly more modern), slightly improved optics. DOT-compliant.

Pic #1: Low Beam.
Pic #2: High Beam.

NOTE: The headlight aiming is a bit off. This is not the headlight, just how I had it mounted in the headlight bucket (was doing this pretty quickly).

Conclusions: Slightly better, same objections as above with 308 sealed beams.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Next up: Adjure / Crystal Eyes Diamond Cut replaceable H4-bulb headlamps with Raybrig Whitesonic H4 bulbs, 55/60W.

Pic #1: Low Beam.
Pic #2: High Beam.

Conclusions: LAUGHABLE! I spent $300 and what I got was a freaking set of bright foglights. Notice FLAT, WIDE HORIZONTAL BAR of light with NO focus areas or hotspots (you want these--lights need to be brighter in focused areas so you can see way down the road). This explains why I really couldn't see distance with the damn things on lowbeams. These are not what you want. Raybrig bulbs are respectable (Japanese--better build quality than Korean bulbs, not as good as German bulbs--remember, majority of major car manufacturers use german-made bulbs as OEM--Philips, Osram, Narva). Raybrig bulbs make a very white, perhaps slightly blueish light, I think around 5000K color temperature. Reasonable, not obnoxiously blue, but this does nothing for you (although you may subjectively like this white light better than white light with slight yellow to it).
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Finally: Cibie 7" E-code (Euro spec) replaceable-bulb H4 headlights, with Narva 90W lowbeam, 100W high beam (90/100W) halogen bulbs, standard color temperature.

Pic #1: Low Beam.
Pic #2: High Beam.

Conclusions: Ahh. This is what you want. Very well crafted beam pattern with hot spots on the light focus. Good spread for side lighting. Beam pattern goes up and to the right, to put more light on the road without distracting oncoming drivers. Sharp beam cutoff--no glare for oncoming drivers. Properly aimed, you can run overwattage bulbs without offending oncoming traffic / blinding other drivers. "Kink" or upward bend in the beam pattern is by design, and also facilitates visual horizontal aiming for the headlights.

Subjective experience with these is VERY good. I can see without the high beams. This is what I've been looking for.
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login
Image Unavailable, Please Login

 

Well, I hope this has been useful. After five iterations of headlights, I've finally found the ones I'm sticking with. Only other possible modification would be to upgrade to even higher wattage bulbs (100/130W) ($38), but right now I see no need and am happy with my 90/100W Narvas.

On a side note, I also upgraded my foglights (stanrard 55W H3s, special selective-yellow light).

I wanted to post the results for all to see to make up their own minds. I can only point you in the right direction, so feel free to draw your own conclusions!

I've invited Mr. Stern to register here and post response to any questions you have. He's very quick to respond to email though and I'm sure can answer any of your questions personally himself via email. [email protected].

For $200, this upgrade represents a tremendous value. I only wish I had done it the right way instead of throwing away $400 in previous lighting attempts, but those had value to as learning experiences and for the community.

Finally, if you do decide to do this, I can recommend Daniel Stern Lighting as a supplier--my transactions with him have been flawless and I have not a single complaint with any product I've received from him.

http://www.danielsternlighting.com/

Again, I have no affiliation except as a customer, but he has certainly earned my business and will not hesitate to do further business with him in the future. I would expect any fellow Fchatters' experiences to be equally positive.

Feel free to post back here with any other comments or questions that I can help answer, or PM me if you want.


Yours in awesome lighting (FINALLY!),

--Mike

 

Additional notes.

The EXACT same concepts apply to 308 GT4s (Cibie also makes a tremendous 5.75" high beam and low beam lamp), as well as with most Mondials (also 5.75" lamps, except for Mondial T). Not sure about the headlight setup on other cars.

 

Also, wanted to post the foglights!

Notice flat, wide, bar-like beam pattern with sharp vertical cutoff line (as high beams should). These lamps need vertical aiming (which I've since done)--you can see the left lamp pointing to the bottom! The main beam that you see is actually the spread of the RIGHT foglamp alone. I think the glare above it is the reflection of the left foglamp from the ground.

I would suggest reading up at Daniel Stern's website (http://www.danielsternlighting.com/), to understand more about fog lamps. They are NOT auxilliary "driving" lamps. Direct link to info is http://www.danielsternlighting.com/tech/lights/fog_lamps/fog_lamps.html.
Image Unavailable, Please Login

 

Mike,

Awesome. A huge thank you. Wonderful work.

Another item on my to do list!

regards,
Pete

 

Mike,
Great research. I used the Kyoto lights for a few years, but I think I will switch. However, with 22 hours of sunlight up here during the driving season, I don't think I will use the lights much.

Thanks for the hard work, it is very helpful.

 

My thanks too. Great information

 

This is brilliant (so to speak) work!!
Many thanks!
I'm changin lights!

 

...and just checked the Testarossa--they appear to take the 5.75" sealed beams, too, so the above applies to them (you just need to get four of the 5.75" housings--since one light on each side is a dedicated high beam).

 

Mike,
Excellent write up. I would only add that the Hella brand lights are of similar if not equal quality to the Cibie (I have used both brands). The key is to get the "euro code" and not the "sealed beam" DOT compliant variants.

Hellas are more widely available, and more steeply discounted.

http://www.rallylights.com/hella/7in.asp You want the ECE ones, NOT the "vision plus" which are the sealed beam/DOT compliant ones.

Daniel Stern's website is a wealth of information, and for that reason alone, worth buying from and being loyal to. But if you are a cheap bastard, Cibie's are cheaper here: http://www.cibieusa.com/cibie_main_frameset.htm

Note also that Cibies typically have a round or convex face, versus the Hella's flat face. Primarily a cosmetic issue, but some people prefer the more traditional look of the convex face. The flat face is superior when applying a 3M plastic film to protect the expensive lights though.

I have never used this company's products, but their wiring harness's are pre-made: http://www.suvlights.com/ I will add that if you are using stock wattage bulbs (55w/60) the stock wiring will work just fine. That is what I did in my previously owned 308 and later 328. A wiring harness could improve the output slightly, but the big advantage is giving you the opportunity to use a higher wattage bulb. But the stock bulbs with the improved Hella or Cibie lenses is a 200 to 300% improvement alone, and was enough for me.

As always, compare closely before purchase, and consider that Daniel Stern's products and recommendations are very good.

Mike, thanks again for all your time, and research!!!

Dave

 

Another item to add to the project list for when I return.

Great write up, this is the stuff that makes Fchat THE place for Ferrari info. Thanks.

 

Thanks Mike for taking the time to do the writeups, photos, and tests. I feel OK with my lights for the momemt, but I'm sure that if I ever perform this upgrade, it will be night and day(pardon the pun). And then I would wonder how in the world did I ever drive around with my old set of lights... Ahhhh another thing on my wish list/things to do.

 

Hey Dave! These are excellent thoughts--thank you for chiming in.

Yes, Hella is an alternative, but I do believe the Cibies are better. You are correct that Euro code here is important, over the DOT sealed beams. However, Mr. Stern is very clear to point out that there are good and bad Euro code headlights, and good and bad DOT headlights.

Now, consider the two isometry charts at the following link:

http://dastern.torque.net/Photometry/Isocomparo.html


In it, Cibie 7" lamps are compared with Hella 7" lamps. This chart is basically a "brightness" topagraphy graph--it shows you where in 3D space light output from the headlight is concentrated. (If Mr. Stern gives me permission, I'll post actual images--these are his I believe.) The image reference is the roadway.

I personally am not qualified to really draw conclusions from these. They are very good data points, however, and there are measured differences between the two lamps (Hella vs Cibie). They are NOT the same. Perhaps Mr. Stern (Daniel) can comment further on the attributes of the Cibie lamps, according to the Isometery data he has.

Correct, that site has the Cibies for $49 each, Daniel Stern Lighting has them for $59 each. I am confident that to earn your business, Daniel will be able to come close to that price, perhaps meeting it.

This is a great point! I was not aware that Hellas were flat. My opinion is that flat vs. convex DOES change the optics of the bulb, and somethings that are possible with one may not be with the other--so it would naturally follow that beam patterns will be different (Daniel can confirm). I can definitely see that flat face is easier to apply 3M plastic film. (Is this a common practice? On cars, with exposed headlights all the time, I can see it (especially if tracked).

Dave is right--and I should underscore Dave's emphasis that stock wiring will be just fine with stock wattage bulbs inside Cibies or Hellas. But I will also underscore that there IS a voltage drop associated with stock wiring, and this WILL affect the brightness of your headlights, no matter what you do. Once you "go harness", and understand the new path of the current (from the battery through the relays to the headlights--all of two or three feet), bypassing the headlight switch, fusebox, and inner-cabin trip altogether, it begins to make a whole lot of sense. On a side note, I believe that I saw in my 328 parts manual room for relays in the battery area for European spec cars--perhaps a similar setup?

Finally, I would caution against SUVlights.com. I've spent about two hours on the phone total with Jonathan there, and it wasn't the most pleasant of conversations. He sells the Sylvania Xenarc 6024 7" HID Setup (that at least two F-chatters use, but the design of which, unfortunately, is crap)--he bought up all the remaining stock when they were discontinued. He was ready to sell me any and all sort of ghetto HID retrofit. He also apparently had a hand in designing the APC ghetto harness, of which I have used two, and that he still sells. He DOES have a new harness apparently, which requires only one power lead to the battery. Given the $50 and two hours is took me to custom make a perfectly-fit harness for the 328, I wouldn't go near it. Daniel Stern's kit includes Bosch relays (you WANT these--NOT ghetto relays--this is IMPORTANT), and good Hella brand components (relay sockets and H4 adapter sockets). Bottom line is any of these will do. Re: SUVlights, I personally think it's shady, but that is just my opinion. Remember, at two points in the past, I was close to buying from them. It probably would have worked out fine.

Right on, Dave! Awesome to be able to pool our knowledge like this.


--Mike

 

How come you didn't try out the Sylvania SilverStar? I used them to replace my H4 in my MR2 and I must admit the difference was night and day (no pun intended). Direct replacement sealed beam headlamp.

I know you considered the Sylvania Xenarc (but a joke) but, why did you then overlook the SilverStars? They aren't a blue light, nice clear and crisp. Bang for the buck I don't think you can beat them!

 

Mike -- GREAT research and write-up! You've taken what I and Tommy Cosgrove did another great step forward.

 

Eeeeyargh.

Same as PIAA's Stupidwhites and everybody else's blue-glass bulbs: just another short-lived, lower-output, heavily-hyped blue bulb.

First, let's look at the Silver Star bulbs.

Here's manufacturer data, from an internal engineering database, for output and lifespan at 13.2v for all the Osram/Sylvania H1 bulbs. Lifespan is given as Tc, the hour figure at which 63.2 percent of the bulbs have failed.

Osram or Sylvania H1 (regular normal):
1550 lumens, 650 hours

Osram or Sylvania H1 long life:
1460 lumens, 1200 hours

Osram H1 Super (if Sylvania Xtravision line included H1, this'd be it):
1700 lumens, 350 hours

Osram H1 Silver Star (NOT Sylvania Silver Star!):
1750 lumens, 350 hours

Osram H1 CoolBlue or Sylvania H1 Silver Star:
1380 lumens, 250 hours

Now, looking over these results, which one would you rather:

(a) Buy?
(b) Sell?

The answer to (a) depends on how well you want to see versus how often to change the bulb. If you want the best possible seeing, you pick the Osram
Silver Star. If you don't care as long as it works and you don't want to hassle with it, you pick the long life or Daytime Running. The answer to (b) is determined by how rich your company's shareholders want you to be, and is obvious: You want to sell the bulb with the shortest lifespan and highest price. That'd be the Sylvania Silver Star.

Note that my selection of H1 to illustrate the point was arbitrary. The absolute values differ from bulb format to bulb format, of course, but the comparative patterns remain alike.

The fundamental problem with Sylvania Silver Star from a technical standpoint is that it's based on a lie ("whiter" light = better seeing).

More? Sure:

AutoExpress finally released the results of their new H4 (=9003, =HB2) bulb tests.

Standard and blue bulbs ("Osram CoolBlue" is what is sold in North America as "Sylvania Silverstar"):

"Plus 30" high efficiency bulbs ("Osram Super" is what is sold in North America as "Sylvania Xtravision". Others of this type are Philips "Premium" (Europe) and "High Visibility" (North America), "Wagner BriteLite", "Candlepower Bright Light" and "Narva Rangepower"):

"Plus 50" ultra high efficiency bulbs (Osram Silverstar is available in North America as Candlepower Super Bright Light, GE Night Hawk, Narva Rangepower+50, or —like the Philips VisionPlus—can be ordered from one of the overseas websites that ships worldwide):

The AutoExpress website will let you view up to two articles before it wants you to "register"; throwing phony info at it will make it shut up and let you see more stories.

A bulb with colorless clear glass is ALWAYS better than one with tinted
glass. Always.

Now, what about those Sylvania Silver Star sealed beams? Also junk: a garbage bulb in a garbage optic. If someone really wants to do a cheap headlamp upgrade on one of these cars (...?!), then the sealed beams to pick (the ONLY sealed beams to pick!) are the GE Night Hawks.

GE just released these. 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 is quite intense, and the beam is considerably wider. The big news is that GE did not go with a trendy blue-glass burner for so-called "whiter" light, but instead maximized the actual light output with a high-efficacy burner with colourless clear glass.

Foreground light is still rather lacking, due to antiquated DOT beam regulations that apply to even new sealed-beam designs, but the big "dammit" is that GE 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 removed the lenses from the reflectors of two GE (standard, Night Hawk) and three Sylvania (standard, Xtravision, Silver Star) sealed beams and evaluated them on the test bench. Doing so permits the separation of beam factors caused by the burner-and-reflector assembly on the one hand, and beam factors caused by the lens optics on the other hand.

The unlensed beam of the GE standard sealed 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 (this is when the actual reflector shape isn't what the design specified, because the stamper pulls the molten glass, taffylike, when it is lifted, and the resultant deformation distorts the beam. This is also why more and more sealed beams have larger and larger nonreflective areas in the heel area of the reflector...the tooling was old 10 years ago, it's even older now, and nobody's retooling), 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 bit of a sick joke.

Next up, GE Night Hawk. 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, reflector heel directly behind the burner, burner "legs" and grommets, and the small 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 focused. The 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.

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 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.

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 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 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.

I looked at Sylvania Silver Star and Sylvania standard 7" round and 200mm rectangular sealed beams. What I found was surprising even given the hype-heavy nature of the Silver Star product: the unlensed beam of the Silver Star was much less well defined than that of the Xtravision. The lensed beam (pre-disassembly) of the Silverstar sealed beams is a worse mess than that of the Xtravision, which probably is a result of using the type of burner they used in combination with the "Silver Star" lens, which is identical to the standard Sylvania sealed beam lens (i.e., 1979-spec optics that are long overdue for retool and weren't all that well conceived in the first place!).

Conclusions:

• GE did a fairly good improvement job with new lens optic tooling, given that the project probably was not very well funded; there isn't much money in sealed beams nowtimes.
  
• Sylvania's line of sealed beams is junk. A shame, since it was once a very
  respectable product line. About the only thing left in the line worth buying are their Xtravision replacement bulbs (not sealed beams), their SilverStar signal bulbs (not headlamp bulbs or sealed beams) and their signalling bulbs.

I will back up and explain why the "whiter" light scam is a scam, if that's needed.

 

Similar/equal quality, definitely. Both companies use quality materials and careful assembly techniques; definitely much better than the 3rd-worldy junk (Adjure and Kyoto and Maxtel and Eaglite and others from China, Autopal from India, Zelmot from Poland...).

Equal performance? Well...no.

It looks like Mike already pointed to the comparative isocandela diagrams for Cibie vs. Hella European-code 7" round H4 headlamps, plotted on the same photogoniometer in the same light tunnel with the same bulb powered by the same power supply set to the same voltage.

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. Points other than straight ahead are referred to as Left ("L") or Right ("R") of the vertical axis V-V of the beam, and as Up ("U") or Down ("D") of the horizontal axis H-H. So, a point referred to as 0.6D, 1.3R means 0.6 degrees below H-H, and 1.3 degrees to the right of V-V.

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)

Things to notice about these two diagrams:

• 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.
  
  
• 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 very 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, such that setting the lows where they belong results in most of the high beam light going up in the trees, but pulling the high beams down so they send light straight ahead puts the low beams 10 feet in front of the car. This works OK on a vehicle with high-mount headlamps (truck, van, SUV), but is a damn nuisance on a passenger car. It effectively means you have to pick between properly-directed low beam intensity or properly-directed high beam intensity.

If you're enough of a geek to be capable of crying over dead headlamps, pick this link instead of the one above. This one contains the Cibie and Hella isoscans, as the third and fourth diagrams on the page. The first and second are scans of 7" round H4 headlamps no longer produced (sob!).

Well...yes and no. Getting a "Euro code" (E-code, ECE, European-code) headlamp does not guarantee that it is a good one. Just like the DOT headlamp specifications contained in FMVSS 108, the ECE headlamp specifications contained in R112 permit a very wide range of headlamp performance. There are plenty of cruddy European-code and cruddy DOT headlamps, just as there are high-performing, well-designed headlamps conforming to each regulation.

Definitely true. Pays money, takes choices, etc.

The Hella "Vision Plus" is not a sealed beam; it's a replaceable-bulb headlamp that takes an H4 bulb (called "HB2" in North American regulations; go here to read up on the reasons and differences). The beam pattern is DOT-compliant, and the low beam isn't all that bad, as DOT headlamps go. Definitely not my first choice of low beam light distribution, but not my last choice, either. The high beam is very weak in the middle (where it should be strong) and these lamps have vent/drain problems that tend to make them into fishbowls. So, yeah, avoid this one.

A few years ago, Hella released a "Super" version of their 7" round ECE H4 headlamp. Its performance isn't significantly different on low beam compared to the standard 70476 Hella ECE H4 unit, which is a damn shame; I really wanted to like this lamp. At least they went a long way towards fixing the vertical beam separation between low and high beam hot spots I carp about above.

I appreciate that. Thank you. I spend a great deal of time and effort (witness these posts!) trying to get complete and correct info out there, regardless of whether any particular individual buys from me or not. I'm sure this results in a significant number of people who ask me all their questions, then go give their money to someone else. Nevertheless, it's surely nice to see this kind of sentiment, which reassures me that not everyone is doing that.

Few things here:

• 3M does not make any headlamp protective films. Yeah, I know, the brown-and-green waxpaper you pulled off your "Støngärd" film has "3M 3M 3M 3M 3M" all over it. 3M made the adhesive and the backing paper, and they would really rather they hadn't. I was talking with a 3M Adhesives engineer a couple years ago and he told me 3M gets an endless stream of complaints about how that film is not optically clear as claimed in the ads (true, it's not), about how it clouds up and yellows (true, it does) and how it doesn't stop headlamps being broken by rocks (true, it doesn't). They refer such complaints to the maker of the film.
  
  
• No headlamp film will ever protect a lamp from a rock that has your headlamp's name on it. A good film will definitely eliminate sandblasting, and will shrug off the smaller road missiles that might otherwise put a bullseye in your lens, but if there's something big enough to cause a crack flying towards the lamp, it'll do so with or without film. If you must have a film, use the surface-hardened stuff from e.g. www.xpel.com .
  
  
• Films can increase the risk of lenses thermally stress cracking. This is mostly a concern with smaller lamps running hotter bulbs, but it is something to be aware of in those situations.
  
  
• If you have an additional 1/2" or so clearance in front of your headlamp lens surface (some hidden-lamp setups do not; many do), then there are much more effective, less deleterious ways to protect your lenses. And no, I am not talking about those silly mesh guards or the yucky hard-plastic shell covers.

Don't be so sure! If by "work just fine" you mean "not set the car on fire", you're completely correct. But you might be surprised how much voltage drop is commonly found in the headlamp circuits of even brand-new, quality cars (and let's face it: Ferraris are several dozen different kinds of wonderful, but the ones we're discussing in this particular thread are neither brand new, nor are their electrical systems of the highest quality.)

Take a look for yourself. Spend some quality time with your voltmeter.

ALL headlamps connected - you may have to backprobe the sockets, but removing the socket from any headlamp invalidates the test.

Engine off, circuit to be tested (low or high beam) energized

First, connect your voltmeter across the battery + and - terminals. Record the voltage reading, this is the battery voltage.

Connect your voltmeter positive lead to the battery (+) and the voltmeter negative lead to the + terminal of whichever headlamp beam you're testing—use the bulb farthest away from the battery. With the lamps on, your voltmeter will give a direct reading of the voltage drop. Write it down.

Then connect the positive voltmeter lead to the ground terminal of the headlamp bulb, and the negative voltmeter lead to the (-) terminal of the
battery. With the lamps on, your voltmeter will again give a direct reading of the voltage drop. Write it down.

Add the two voltage drop figures obtained, and this is the total circuit drop.

Subtract the total circuit drop from the battery voltage to find the voltage at which your bulbs are operating.

Remember, light output drops exponentially, not linearly, with voltage drop. Small voltage drops = large light losses. The formula for determining the change in light output with a change in voltage is:

lumens @ old volts x [(new volts /old volts) ^3.4] = lumens @ new volts

So for simplicity's sake, let's take a 9006 low beam bulb rated 1000
lumens at 12.8 Volts and plug in different voltages:

10.5V : 510 lumens
11.0V : 597 lumens
11.5V : 695 lumens
12.0V : 803 lumens
12.5V : 923 lumens
12.8V : 1000 lumens <--Rated output voltage
13.0V : 1054 lumens
13.5V : 1198 lumens
14.0V : 1356 lumens <--Rated life voltage
14.5V : 1528 lumens

Drop 0.8v across the circuit, lose 200 lumens = 1/5 of the output.

Again I thank you for this.

 

It doesn't quite work exactly like that. Yes, you're correct that there are optical differences between flat and convex lenses, but there's no optical-superiority reason to pick a flat-lens lamp over a convex-lens unit or vice versa (and there have even been some con<b>cave</b>-lens round headlamps over the years, some of them excellent and some merely adequate...)

Another 7" round H4 lamp of which I mourn the discontinuation is the unit Carello used to make (the car unit, not the motorcycle one!). It had a convex lens, an effective bulb shield, and a very nice beam pattern.

Yep, see my previous post. What a lot of people don't realise is that the blue "extra white" bulbs simply mask the brownish appearance caused by starved bulbs. Feeding the bulbs (any bulbs) properly "whitens" the light considerably but gives you more light, rather than less.

 

scheinwerfermann: Who ARE you? Your profile isn't filled out... do you work in the lighting industry?

Thanks for sharing the great information!

EDIT: Ah, I now see that it's Daniel Stern! Thanks for joining us and contributing.

 

thanks MIKE328 and to Daniel for joining us here. this is a great thread !!!

 

You do know that you lose points at an FCA concours if you use a non-OEM brand/model headlamp?