Edit (2017): If you found this via Google, the information in this thread is vastly out of date. Please see this thread or skbowe.com for current information on BOW and WMM failures. The process below still applies to the physical conversion, but you do need to consider the electrical ramifications regardless of using OEM or aftermarket ballasts
I'm helping out with a project that is, IMHO, a major step forward for the P1 community. Until now AFAIK there have been only two categories of HID conversions for P1 cars with halogen low beams: (1) Traditional retrofits, where you shove a H11 re-based bulb into the halogen projector, and (2) projector retrofits, where you find some other european car with the correct size/shape projector and bolt it into your existing housing. Neither of these routes are optimal since both require extensive modification to the housings.
So someone got the idea to take the *entire* bi-xenon housing and completely replace the halogen housing. If they had asked me first I would have said, "It won't work because you need a new engine harness and have to reprogram the ECM to talk to the GDL modules over LIN, plus install all the auto-leveling sensors and that's at least $2500 right there." But fortunately, I wasn't asked, and it turns out that I am (completely) wrong.
In fact, contrary to the wiring diagram that's been on my website for the last few years, the OEM bi-xenon housings are almost completely plug-and-play, with the exception of exactly one +12v wire which isn't in the halogen harness EDIT: AND THE PWM Voltage Regulation in the Halogen CEM I've assembled a corrected wiring diagram here. Granted, the auto-leveling and road-following won't work (that's a project for another day
) but everything else, including the OEM ballast, does.
The headlight assembly that will be used in this writeup was removed from wreck, so is pretty well banged up, including a broken ball joint and a cracked lens. So the first step will be to fix that stuff.
Started by stripping everything out of the assembly so I could actually get to where the ball joint failed. This required baking the housing at 250 degrees (F) for 15 minutes to loosen the butyl rubber weather sealant around the lens. Next the reflectors were removed one by one, and finally the optics tray was extracted through the front. The resulting shell is rather disparaging. Excuse the iPhone pics:
![Image]()
</img>
You can see (sort of) where the ball fell off. If you can't spot it, look for the laser:
![Image]()
</img>
Here's the ball. It was supported by a piece of plastic that I had to cut away, and of course the resulting knurled shaft doesn't fit in the hole:
![Image]()
</img>
There is a lot of load on this piece, so I had to come up with something that had actual structural integrity. How about a screw?
![Image]()
</img>
Now how to get the screw to support the ball joint that was previously cast in plastic? Drill and tap anyone:
![Image]()
</img>
Hopefully you can see where this is going
![Image]()
</img>
Of course, never one to leave well enough alone I found some alloy tubing with an ID appropriate to press fit over the ball joint shaft and the screw.
![Image]()
</img>
The outside of the alloy tube got roughed up a bit, then pre-heated and melted into the housing. Before the ball is screwed in...
![Image]()
</img>
Industrial size JB weld. There's enough JB weld in there to put the Titanic back together, and still have enough left over for this project. While the JB weld is setting, I thought I'd test out the electronics.
![Image]()
</img>
Note that the bulb is a D1S, which is the same base as an D2S but includes an integrated ignitor. This means that an aftermarket kit with D2S bulbs will work just fine. Seems like it works to me:
![Image]()
</img>
An interesting note is that while the ballast is labeled 35W, it draws around 4A (which is around 45W or 50W) - much closer to 55 than I expected. It's possible that the bi-xenon cars use the same bulb-failure circuitry in the CEM.
Stay tuned folks, this is gonna get interesting.
I'm helping out with a project that is, IMHO, a major step forward for the P1 community. Until now AFAIK there have been only two categories of HID conversions for P1 cars with halogen low beams: (1) Traditional retrofits, where you shove a H11 re-based bulb into the halogen projector, and (2) projector retrofits, where you find some other european car with the correct size/shape projector and bolt it into your existing housing. Neither of these routes are optimal since both require extensive modification to the housings.
So someone got the idea to take the *entire* bi-xenon housing and completely replace the halogen housing. If they had asked me first I would have said, "It won't work because you need a new engine harness and have to reprogram the ECM to talk to the GDL modules over LIN, plus install all the auto-leveling sensors and that's at least $2500 right there." But fortunately, I wasn't asked, and it turns out that I am (completely) wrong.
In fact, contrary to the wiring diagram that's been on my website for the last few years, the OEM bi-xenon housings are almost completely plug-and-play, with the exception of exactly one +12v wire which isn't in the halogen harness EDIT: AND THE PWM Voltage Regulation in the Halogen CEM I've assembled a corrected wiring diagram here. Granted, the auto-leveling and road-following won't work (that's a project for another day
) but everything else, including the OEM ballast, does.The headlight assembly that will be used in this writeup was removed from wreck, so is pretty well banged up, including a broken ball joint and a cracked lens. So the first step will be to fix that stuff.
Started by stripping everything out of the assembly so I could actually get to where the ball joint failed. This required baking the housing at 250 degrees (F) for 15 minutes to loosen the butyl rubber weather sealant around the lens. Next the reflectors were removed one by one, and finally the optics tray was extracted through the front. The resulting shell is rather disparaging. Excuse the iPhone pics:
You can see (sort of) where the ball fell off. If you can't spot it, look for the laser:
Here's the ball. It was supported by a piece of plastic that I had to cut away, and of course the resulting knurled shaft doesn't fit in the hole:
There is a lot of load on this piece, so I had to come up with something that had actual structural integrity. How about a screw?
Now how to get the screw to support the ball joint that was previously cast in plastic? Drill and tap anyone:
Hopefully you can see where this is going
Of course, never one to leave well enough alone I found some alloy tubing with an ID appropriate to press fit over the ball joint shaft and the screw.
The outside of the alloy tube got roughed up a bit, then pre-heated and melted into the housing. Before the ball is screwed in...
Industrial size JB weld. There's enough JB weld in there to put the Titanic back together, and still have enough left over for this project. While the JB weld is setting, I thought I'd test out the electronics.
Note that the bulb is a D1S, which is the same base as an D2S but includes an integrated ignitor. This means that an aftermarket kit with D2S bulbs will work just fine. Seems like it works to me:
An interesting note is that while the ballast is labeled 35W, it draws around 4A (which is around 45W or 50W) - much closer to 55 than I expected. It's possible that the bi-xenon cars use the same bulb-failure circuitry in the CEM.
Stay tuned folks, this is gonna get interesting.
