Odd HID (D1S) failures..

[deilenberger]

Recently I'd noticed the output from my HID factory headlights didn't seem to be what it used to be. SWMBO's Lexus RX350 had better lighting, and that just isn't right. After refinishing the front surface of the headlight (by polishing it out then coating it with a Dupli-Color clear coat made for headlights) - there wasn't much improvement.

 

I had one of the small parking lights crap out - so I had to remove one of the headlights and open it up. Curiosity prompted me to pull out the HID "burner" to see if I could find any reason for the lower output. I rather expected at 82,000 miles that I might see some darkening of the glass envelope.. but to my surprise, this is what I found: 

 

 

Both bulbs had these deposits on them.  I've never had any condensation problems with the headlights, and it appears that the '06 headlights have an air filter of some sort on the bottom of the housing, and the rubber tube on the rear cover for the housing (which isn't really an "open" tube - it has a sort of nozzle on the end so stuff can go out, and not go in.)

 

 

Both bulbs were the same, both had this coating on it. The coating didn't come off with gentle fingernail picking, so I ordered two new Philips bulbs from Amazon (about $116 for the pair - and they were genuine Philips - authenticated on Philips webside with the codes on the packaging.)  Obviously the new bulbs didn't have this coating/film on them.

 

Things are now back to the expected brightness.  As to what caused the crap on the burners - I'm puzzled. I know at one point with the early P!G there were condensation problems, and Porsche addressed this by putting packets of desiccant (silica gel) inside the headlight housing.  I couldn't find any packets of desiccant - but I wonder if Porsche put the desiccant into the filter looking thing in the bottom of the housing..  If a desiccant salt was used I could see it being absorbed by moisture inside the housing, then deposited out at the hot spot on the HID burner.  The burners were the original ones installed by Porsche at the factory (Osram Xenarc.)

 

Puzzle..  anyone else seen this? Anyone else want to look at theirs?  I may try a bit of acid on a Q-Tip to see if I can dissolve the crap without damaging the quartz envelope/bulb.

 

More on silica gel: http://science.howstuffworks.com/innovation/science-questions/question206.htm

 

Update: I can update this - after driving for an hour tonight after dark it really became evident what was missing.. light. The lights are now as good as or superior to the Lexus HID servo-controlled projectors..

Edited August 3, 2014 by deilenberger

[RFM]

This is a normal phenomenon, after a certain time of use there are deposits on the spot where the arc flash is standing. This is applicable to all xenon lamps after time.

[bigbuzuki]

Xenon lamps have a finite service life with a reduced output over time.

They require replacement every few years.

[deilenberger]

I would agree to that - except the hazy coating is on the outside of the envelope.

 

A darkening of the inside of the envelope (actually - the inside of the sphere where the two electrodes are) might be expected as the electrodes burn down a bit, but this is different, the coating is on the outside of the quartz tube that covers the quartz sphere where the actual arc/discharge takes place. 

 

If you look closely at the pics - you can see the sphere where the two electrodes are and where the discharge takes place. That's sealed, then there is an additional sealed quartz envelope that the sphere is contained in.

 

While HID lamps/burners do have a finite life - the usual cause of failure is the gap getting too wide and the discharge failing to start. As the gap gets wider there may be some deposited electrode material on the inside of the quartz sphere.

 

Photo (from AdvancedAutoParts) of an Osram Xenarc bulb/burner:

 

The inner sphere and outer envelope are fairly easy to see in this photo..

 

[wrinkledpants]

I've never had an HID actually fail to ignite.  They always end up with reduced lighting output and that haze you're talking about.  I seem to get about 2-3 years or 20-30K out of my HID bulbs, and that spans various cars.  

[hahnmgh63]

I have had the D1S bulbs like ours fail to ignite and quite a few failures here and on Rennlist. Probably the ignitor as I've never had a D2S fail. Deilenberger is right, never seen deposits on the outside, only on the inside. I believe the deposits on the inside are the Metallic salts added that become a plasma gas once ignited, and after multiple firings they deposit a bit on the inside of the glass.

[deilenberger]

Also posted on Rennlist..

 

Resolution..

I tried acetic acid (happened to be what I had kicking around in the darkroom..) It had no effect on the coating. So - I decided mechanical means were needed.

 

Used some 3M Perfect-It Rubbing Compound (actually a very fine polishing compound) rubbed on with a paper towel, then careful rinsing with plain water.

 

It took a few tries - but it cleaned off the coating.

 

A before:

 

An after:

 

Side by side - left one polished, right one untouched:

 

 

Good as new? Probably not since there is still a bit of hazing/deposits on the inside of the sphere where the arc takes place.. but bet it's about double the light output of what the bulb provided before, and much less diffuse (ie - better beam quality.) If you're hesitant about spending $150-200 for HID replacement bulbs (that's what good ones that aren't counterfeit cost) this would seem a viable option.

 

Before use - the bulb has to be carefully cleaned of any finger-oil that has gotten on it (the oil will char and burn into the quartz, possibly causing it to fracture.) I usually use pure ethanol (200 proof stuff) to clean bulbs before installing them. Probably some MAF cleaner would be equally as effective (and won't leave deposits.)

 

The question of what causes this - since it seems unique to Porsche HID headlights is still up for debate, but my WAG is - if you look at the photos where the headlight is disassembled - there is what appears to be a vent in the bottom of the headlight housing, with some sort of filter material. I'm betting that the filter material has a layer of silica-gel to keep moisture out of the headlights. I'd bet this eventually gets saturated and outgases fumes that then deposit on the hot spot on the HID bulb.

 

You can see the filter in Navaro911's picture from his thread (on Rennlist) on disassembling and painting headlight housings:

It's the white thing in the right corner of the housing bottom. On the other side it sticks out of the housing about 1/2"..

 

At least I now have a useable spare bulb to put on the shelf instead of just throwing them away.

 

Note - on Rennlist several other people reported seeing the deposits that I had - and one even provided photos of them, and for comparison an HID bulb from an Audi, which had no deposits. It's something specific to the 955/Cayenne housing for sure (and perhaps other models of Porsche) - and I'd bet money on the desiccant used to keep the headlights from fogging up.

[hahnmgh63]

Just FYI, Amazon still has the Phillips D1S oem color bulbs for $50 each. They are in Phillips boxes and stamped Phillips on the Ignitors so I would definitely say they are Phillips.

[deilenberger]

Just FYI, Amazon still has the Phillips D1S oem color bulbs for $50 each. They are in Phillips boxes and stamped Phillips on the Ignitors so I would definitely say they are Phillips.

 

Actually there are a range of prices for the Philips D1S bulbs on Amazon. Some are undoubtedly fakes - even with what appear to be Philips boxes and the correct printing on the igniters. Two things I looked for - (1) the bulbs I bought were being sold directly by Amazon, meaning if they were fakes or I had any problems with them - Amazon stands behind them, and Amazon is very good about standing behind what they sell. Many of the bulbs are sold by secondary companies in the Amazon "marketplace" many of the secondary companies may be selling fakes. (2) Reading the reviews is usually a clue to this being a problem. There are lots of reports of bulb failures on bulbs from the secondary companies.

 

It's a big enough problem that Philips has a holographic numeric tag on each of their genuine bulbs boxes - and a scannable code that lets you go to a Philips website where you can enter the holographic number - where it's matched up with the printed barcode and authenticated (or rejected.)

 

Apparently Chinese rip-off "reproductions" are quite common, and of variable quality. I think it's especially easy to spot them when they are listed as 10,000 kelvin.. I doubt if  Philips would ever make a bulb with purple light coming out of it.  What's interesting is visiting the Philips USA website - they only list one D1S bulb, the 4200K, 3350 Lumens bulb, which makes me suspect of any bulb labeled Philips that has a different output color (I see a lot of 6,000K ones listed on aftermarket sites.) - From the Philps USA website: http://www.usa.philips.com/c-p/85415C1/xenon-standard-automotive-headlight/specifications

 

Interesting discussion on the fakes: http://www.hidbulbsrus.com/identifying-counterfeits/ - and a PDF link of info to help ID them. Note that the newer boxes from Philips are no longer like the ones shown in the PDF - the new ones have the holographic label on them with the unique ID number.