HIDs dont ignite, intermittently

[FlashingBlade]

I installed Porsche's bi-xenon headlight assembly in my 2002 996 C2 about 4 years ago (http://www.renntech.org/forums/index.php?showtopic=9851&hl=)

They have worked fine until recently (colder?), they don't ignite on the first turn, but require multiple turns of the headlight switch before they come on. Sometimes, they don't come on at all.

Several others on this forum have complained about the problem, but I have not seen any definitive solution. The choices so far are:

• Headlight switch going bad
  
• Electronic modules going bad (at the same time?)
  
• Lamps going bad (at the same time?)
  
• Not enough initial current to the ignitors (colder temps, 15A fuses)
  

My general approach to problems is to start with the least expensive.

The last item is the most intriguing to me. One poster claims that the OEM Xenon equipped cars have 30A fuses in slots A9 /A10 while the non OEM equipped cars have 15A fuses (which is indeed my case). There may be some merit to that theory because fuses work by generating heat with a bit of resistance, lower amped fuses have higher resistance, so less current to the ignitors. Combined with colder weather -> lower voltages may make some sense.

Unfortunately, I don't have the circuit diagram for my car and so I cannot verify this information (can anyone send it? or verify that claim).

In the mean time, I cleaned the lamp housing contacts and now of course all is well. I cannot further diagnose but would like to solve the problem so as not have surprises while on a trip.

Your help as usual will be highly appreciated.

[Loren]

If both headlights start slowly or are intermittent then I would suspect the headlight switch.

A1 & A2 (High beam) for each side are 15 A regardless of headlight used.

As are A9 & A10 (Low beam) for both headlight styles.

[wwest]

I installed Porsche's bi-xenon headlight assembly in my 2002 996 C2 about 4 years ago (http://www.renntech.org/forums/index.php?showtopic=9851&hl=)

They have worked fine until recently (colder?), they don't ignite on the first turn, but require multiple turns of the headlight switch before they come on. Sometimes, they don't come on at all.

Several others on this forum have complained about the problem, but I have not seen any definitive solution. The choices so far are:

• Headlight switch going bad
  
• Electronic modules going bad (at the same time?)
  
• Lamps going bad (at the same time?)
  
• Not enough initial current to the ignitors (colder temps, 15A fuses)
  

My general approach to problems is to start with the least expensive.

The last item is the most intriguing to me. One poster claims that the OEM Xenon equipped cars have 30A fuses in slots A9 /A10 while the non OEM equipped cars have 15A fuses (which is indeed my case). There may be some merit to that theory because fuses work by generating heat with a bit of resistance, lower amped fuses have higher resistance, so less current to the ignitors. Combined with colder weather -> lower voltages may make some sense.

Unfortunately, I don't have the circuit diagram for my car and so I cannot verify this information (can anyone send it? or verify that claim).

In the mean time, I cleaned the lamp housing contacts and now of course all is well. I cannot further diagnose but would like to solve the problem so as not have surprises while on a trip.

Your help as usual will be highly appreciated.

Install the largest, within reason, electrolytic capacitors (>15 volt rating, polarised) as close as posible directly across the two leads supplying power to the HID ballast/ignitor. Also, make certain sure that the battery/alternator system is operating at proper capacity and charge level.

HID initial ignition requires >25,000 volts, a serious level of DC-to-DC upconverting from the battery's 12 volts. The gap voltage drops to ~70 VDC once the arc is established.

Edited January 19, 2010 by wwest

[FlashingBlade]

Thanks for all the suggestions (and some PMs).

I have had the problem demonstrate itself in several ways. The first time, only one light went on. After that it's been fairly consistent that nothing happens when I turn the lights on and I have to cycle the switch a bunch of times before it works. It's hard to tell whether the lights initially flicker or not, but my wife claim they do not.

PMs reported that they fixed their problems with either a new switch or a new ballast, which apparently is the same as those used on MBs (available on ebay for ~120US$).

After cleaning the contacts with the lamp housing, the problem seems to have gone away, at least for the moment. Perhaps the small amount of resistance that built up over time lowered the voltage sufficiently that it wont power the ignitors sufficiently.

At this point, I cannot trace this problem any further. I am however keeping a Volt meter in the trunk in case it goes out again so I can see measure the voltage at the contacts.

[wwest]

Thanks for all the suggestions (and some PMs).

I have had the problem demonstrate itself in several ways. The first time, only one light went on. After that it's been fairly consistent that nothing happens when I turn the lights on and I have to cycle the switch a bunch of times before it works. It's hard to tell whether the lights initially flicker or not, but my wife claim they do not.

PMs reported that they fixed their problems with either a new switch or a new ballast, which apparently is the same as those used on MBs (available on ebay for ~120US$).

After cleaning the contacts with the lamp housing, the problem seems to have gone away, at least for the moment. Perhaps the small amount of resistance that built up over time lowered the voltage sufficiently that it wont power the ignitors sufficiently.

At this point, I cannot trace this problem any further. I am however keeping a Volt meter in the trunk in case it goes out again so I can see measure the voltage at the contacts.

A voltmeter will be of no help.

You need a digital storage oscilloscope or logic analizer.

The ballast inrush current flow for initially igniting the HID arc lasts less than a 10th of a second, maybe even as little as 1/100th of a second, not enough time for most voltmeters to react.

Once the arc starts the current load on the wiring is even less than what you would have with halogens, 35 watts HID vs 55 watts Halogen

Edited January 20, 2010 by wwest

[FlashingBlade]

Thanks for your suggestions about the oscilloscope. Before I dust mine off, I want to exhaust all the simple causes first.

Today I got lucky, of sorts. The lights did not go on, to the immense amusement of the people at my work, and I popped out one of the lights to measure the voltage across connector pin outs 7 (gnd) and 8 (Lo beam): 5 Volts. WOW!

I popped out the other light: 14V (i.e., no load), popped the other one back in: 5V. That basically means that somewhere in the circuit I have a huge build up of resistance.

Back home, I retrieved the original Halogen lamp housings from the attic, removed the Xenons, popped one of them in: no light and 0.2 V at the other connector.

A back of the envelope computation indicates that I have a build up of an additional ~275Ohms in that circuit (i.e., from the battery through the switch to the lamp). Not good. Ground seems fine.

So like Loren originally suggested (of course), it's probably the headlight switch gone bad.

Before I order a replacement, one more question: Are there any relays or other components in this circuit that could be the culprit. I only have a 2000 circuit diagram (not a 2002) and I only see a fog light relay (why), but none for the low beams. So unless I have a fried wire, only the switch is left.

Thanks

[perryinva]

None anywhere near as likely as the switch. This is so common as to be funny. (not) When you replace it, be sure to get out that one screw that's on a diagonal from the side. It's pretty hard to see, unless you know it's there. It's a cheap (for Porsche) and easy (for Porsche) fix. ANd buy it online at Sunset or Suncoast. The difference from the dealer was like $40 here.

Edited January 22, 2010 by perryinva

[wwest]

Thanks for your suggestions about the oscilloscope. Before I dust mine off, I want to exhaust all the simple causes first.

Today I got lucky, of sorts. The lights did not go on, to the immense amusement of the people at my work, and I popped out one of the lights to measure the voltage across connector pin outs 7 (gnd) and 8 (Lo beam): 5 Volts. WOW!

I popped out the other light: 14V (i.e., no load), popped the other one back in: 5V. That basically means that somewhere in the circuit I have a huge build up of resistance.

Back home, I retrieved the original Halogen lamp housings from the attic, removed the Xenons, popped one of them in: no light and 0.2 V at the other connector.

A back of the envelope computation indicates that I have a build up of an additional ~275Ohms in that circuit (i.e., from the battery through the switch to the lamp). Not good. Ground seems fine.

So like Loren originally suggested (of course), it's probably the headlight switch gone bad.

Before I order a replacement, one more question: Are there any relays or other components in this circuit that could be the culprit. I only have a 2000 circuit diagram (not a 2002) and I only see a fog light relay (why), but none for the low beams. So unless I have a fried wire, only the switch is left.

Thanks

I seem to remember that a number of years ago I accidently discovered that if one of my headlights was disconnected then the other one would not work. That could have been on my Canadian origin '99 w/DRLs or the '01 C4.

Work is..??

Our office is in Redmond.

Edited January 23, 2010 by wwest

[FlashingBlade]

It was the switch!

I just replaced it, and all is well. It was a lot easier than I thought, give or take figuring out where to pull on things. If you do a search in this forum you'll see a variety of explanations and diagrams, but they don't mention some of the gotchas.

• To remove the housing the light switch is in requires pulling hard enough to overcome the force of the two clips (top and bottom). This is after you remove the screws of course. I used some combination of plastic "Trim and Molding Tool Set" from HarborFreight tools.
  
• To remove the switch you actually need to pull the illuminated ring from the front of the panel forward a bit. It's not obvious but that ring serves as a lock for the switch and just slides out (forward). After you slide it forward, you can twist out the switch.
  

My diagnostic contribution is that if your headlights go out, pull one of your lights out (takes 2 seconds with the tool in your kit), and measure the voltage across pins 7 and 8 (upper left most two in my case when looking at the connector from the front of the car). Then take the other light out, and see what you get then. If the difference is large, it's probably the switch.

BTW, the best price for the switch was at Sunset Porsche in Beaverton. Got the part in a day (80US$ + 9US$ shipping), which was well below Pelican (120US$) and the local P dealer (160US$ + tax). I guess BillG doesn't care how much things cost.

Hope this helps.

Edited January 24, 2010 by FlashingBlade

[perryinva]

Well, was I right, or was I right? FWIW, I tried to clean my old one, and it was a waste of time on mine. I should have mentioned the collar, too, sorry.

It was the switch!

I just replaced it, and all is well. It was a lot easier than I thought, give or take figuring out where to pull on things. If you do a search in this forum you'll see a variety of explanations and diagrams, but they don't mention some of the gotchas.

• To remove the housing the light switch is in requires pulling hard enough to overcome the force of the two clips (top and bottom). This is after you remove the screws of course. I used some combination of plastic "Trim and Molding Tool Set" from HarborFreight tools.
  
• To remove the switch you actually need to pull the illuminated ring from the front of the panel forward a bit. It's not obvious but that ring serves as a lock for the switch and just slides out (forward). After you slide it forward, you can twist out the switch.
  

My diagnostic contribution is that if your headlights go out, pull one of your lights out (takes 2 seconds with the tool in your kit), and measure the voltage across pins 7 and 8 (upper left most two in my case when looking at the connector from the front of the car). Then take the other light out, and see what you get then. If the difference is large, it's probably the switch.

BTW, the best price for the switch was at Sunset Porsche in Beaverton. Got the part in a day (80US$ + 9US$ shipping), which was well below Pelican (120US$) and the local P dealer (160US$ + tax). I guess BillG doesn't care how much things cost.

Hope this helps.