[melz_belz]

Hi everyone,
I've a 2009 petrol model and my first globe blew around the 6 month mark, and it seems one's gone every 3 to 6 months after that. Not the same globe though, it likes to alternate from offside to nearside depending on it's mood. It's ridiculous and making me mad. My previous car I had for seven years and only changed one globe right around it's 17th birthday, and it was also the easiest thing in the world to do unlike with this car!
Sorry to be lazy and not read through all the comments but has anyone been successful with solving this problem yet? I want to avoid dealerships if possible, they've been nothing less than painful from the get go.

[neoto]

This morning a left low headlamp blew with a flash just after a very rough start. The car (2011 diesel) has been sitting for a week in a garage and when I started it I though that it barely runs. After a few seconds the rpms became steadier and then a flash... Damn it  Anyway, I did manage to replace it on my own.
My statistics: 18 months, 60k km, 3 blown headlamps (2 left, 1 right). We are obligated to run 24/7 with the headlights on and I start my car with the headlights off.

A question for those members in countries that require headlights on during daylight: do the LED DRLs qualify for that rule or must it be headlights?

[neoto]

As far as I know, DRL are qualified.

As far as I know, DRL are qualified.

Thanks neoto, I should have asked even though your answer covers it impliciltly, but just in case, what about tail lights? I'm assuming these do not light along with DRLs and are not required in daylight.

In which case I'd persue a method to reduce the voltage slightly to the headlights in all conditions other than darkness. My thinking is simply that reduced voltage will lower stress on the filaments to extend their life.

[Phil №❶]

Here is an idea for people to consider.

A long time ago I purchased a device similar to this one.

http://www.ebay.com/itm/DC-DC-Converter-Regulator-24V-Step-down-to-12V-10A-120W-wide-I-P-14-5-40V-/250885209834?pt=LH_DefaultDomain_0&hash=item3a69ec7eea

Unfortunately, I have not tested it yet so I can't help anyone here.

The required input voltage minimum is 14.5 V so I'm not sure if it will work at all. But if it does, it just may save many blown bulbs.

My cars do not blow bulbs, so I can't rig it up to give any testimony to it's suitability.

Perhaps someone who is having a particularly rough time with this problem could rig up a test and see if

a. It works at all.
b. Bulbs stop blowing.

@120w it would only be working @ approx 50% of it's capacity for low beam (1 side) only of course.

It's not that expensive to purchase. Do we have any volunteers 

[Asterix]

A question for those members in countries that require headlights on during daylight: do the LED DRLs qualify for that rule or must it be headlights?

LED DRLs  do qualify for that purpose.

Since start 2012 an EU rule have made it illegal to have the taillights on with the DRL's, in the name of anti pollution. 

[Phil №❶]

That would make a very interesting calculation actually.

No of Eu cars on road x 2 lights @ 5 watts = lots of watts . Divide 746 (I think) for 1 electric HP, now how much fuel into 1 Hp? = fuel saved   

Here is an idea for people to consider.

A long time ago I purchased a device similar to this one.

http://www.ebay.com/itm/DC-DC-Converter-Regulator-24V-Step-down-to-12V-10A-120W-wide-I-P-14-5-40V-/250885209834?pt=LH_DefaultDomain_0&hash=item3a69ec7eea

Unfortunately, I have not tested it yet so I can't help anyone here.

The required input voltage minimum is 14.5 V so I'm not sure if it will work at all. But if it does, it just may save many blown bulbs.

My cars do not blow bulbs, so I can't rig it up to give any testimony to it's suitability.

Perhaps someone who is having a particularly rough time with this problem could rig up a test and see if

a. It works at all.
b. Bulbs stop blowing.

@120w it would only be working @ approx 50% of it's capacity for low beam (1 side) only of course.

It's not that expensive to purchase. Do we have any volunteers 

I don't think it will work.

The minimum of 14.5 volts will be to allow at least 2.5 volts across the regulator. Even if it did work the output is regulated to only 12 volts where we probably need 13-14 volts as running voltage else our dim lights will be even dimmer.

I gave cursory thought to something like this when I posted which I thought would probably require some sort of switching supply to create the higher voltage then regulate that to say 14 volts. It all sounded too hard... and like you I've not blown any globes yet either in four years so no incentive.

I still think my idea of pre-heating the headlight filaments by a bleed resistor fed from the parking lights might help if the blowing is related to the cold inrush current. This would only work though if the headlights were turned on manually so that the headlights get a few milliseconds of low current as the switch is turned from off through park to headlights. If the headlights are left permanently on then an alternative would be to have some logic to sense initial switch-on with the slow turn on activated from that.

As I said before all this is speculation as we have not had a diagnosis of the cause yet.

[Phil №❶]

Here is a slightly unrelated article, but take note of the quoted standard light circuit design, food for thought  Comments anyone.

http://faqlight.shopower.com/headlamp-harness.html

and another good article, read about inrush current.

http://sound.westhost.com/articles/inrush.htm

Quite interesting and has caused me to start calculating. Ohm's law lives...

[Asterix]

That would make a very interesting calculation actually.

No of Eu cars on road x 2 lights @ 21 watts = lots of watts . Divide 746 (I think) for 1 electric HP, now how much fuel into 1 Hp? = fuel saved   

Remember, only from Jan 2012 onwards, so not all EU cars..

Can't quite follow you regarding the 2x21 watt.. ...maybe I misunderstand you, but taillights are 5 watt. (unless this have been changed in the past few years)

This law have actually caused a lot of trouble, as people forget to turn on the normal lights when driving at dusk and dawn. They can see their headlights are on, but forget that their taillights are not. They have been used to that for 22 years...

IMO, if they should make a law regarding DRL's that would make sense, they should demand LED in the front instead of the low beams 2x55 Watt.

The way it is right now, only 10 watts are spared. I will at any time use more wattage when using heated seats, A/C, etc.

[Phil №❶]

Quite correct Asterix, the brakes are 21 w aren't they   

[Asterix]

Quite correct Asterix, the brakes are 21 w aren't they   

Yep.   

I calculate that a cold headlight will be approx 0.173 ohms with an inrush peak current of 69 amps. If the headlight was pre-heated with a 2.7 ohm resistor it would need to dissipate 20 watts. Not a small resistor that and probably a heatsink type. This would reduce the switch on inrush to 5.04 amps which is near to the normal current of 5.58 amps when heated. These figures assume 14.5 volts where the 55 watt lights will dissipate 80 watts by my calculation.

The headlamp incidently would use 19.5 watts while the park light was on. That's about one third normal so the brightness would be quite dull and yellow.

Just for interest, without the current limiting resistor the instantaneous power is 832 watts. Of course this lasts probably less than a millisecond but does explain why bulbs blow when you switch then on.

[Phil №❶]

That's great Pip, thanks I appreciate your efforts.

Does the resistance vary if the temp is say -20 Deg or more. Parked outside Mishas house in Russia, this could certainly happen. If it was my car, I would not be concerned if it took even 5 seconds to come up to full because at startup in cold conditions, I always wait for oil and engine to run properly before moving off. Unfortunately, this wouldn't help if you use the passing (flash) function of the bulbs would it.

Also in the texts, there is reference to "thin spots" on the filament which is where the bulb will eventually fail. I guess, if there were modifications made, new bulbs at the time would be a good idea.

[csirac2]

Disclaimer: I have an EE background but these days I mostly do software. I also have very limited experience with auto electrics. So it's possible my suggestions are entirely bogus :-)

BUT if it were me I'd try to rule out the charging system. I would try to measure the voltage as close to the running headlamp as possible (aside: not all cheap DMMs are calibrated properly!), with engine running at different RPM, and with different equipment turned on/off. Even a few % over-voltage can decrease the life expectancy by half.

I realize everybody knows to be careful to void fingerprints on halogen bulbs generally, but I never bothered to discover exactly why... http://www.frankshospitalworkshop.com/electronics/training_course_halogen_bulbs.html says it's because the resulting hotspots deform the quartz which causes the halogen to leak. It also says:

A reduction from 12V to 11.4V (5%) will double the life expectancy. The loss of brightness is hardly visible.

I assume that means for a bulb with a design-voltage of 12V. I imagine automotive bulbs are designed for a slightly higher voltage. In this case, you could get a reasonable voltage drop with a diode, I wonder if something like http://ie.farnell.com/vishay-formerly-i-r/16f20/diode-standard-16a-200v/dp/365439 would work - still a waste of heat though.

It also mentions some "expensive" (non-automotive) halogen bulbs have a soft-start facility, so the cold in-rush current must be a non-trivial problem.

If it's any consolation, my '99 Ford Falcon only gets about 12 months/40,000km between bulb replacements. Improper installation (where the retaining clip isn't installed properly) seems to accelerate the failure substantially, I assume due to additional vibration of the hot filament or perhaps not enough heat being conducted back into the headlight housing.

[Phil №❶]

@csirac2,

Thanks for your input, discussion will eventually bring a solution to this problem I feel and the best way is to get thought from as many members as possible.

It appears that the problem is simply current, way too much because the filament is initially (very) cold and it so happens that as the engine starts there is back emf from the starter motor, the initial output from high power alternators and the alternator voltage regulator circuit adjusting to whatever conditions it finds. One of the last reports of bulb failure mentioned that the car had not been used for 1 week. Now, even when switched off, our cars are not so there is always constant battery drain as in clock and auto unlock waiting for a signal. Therefor for the car mentioned I expect the alternator went berserk charging the battery and to load cold filaments with the current was fatal for 1 bulb.

Our other member Pip, has kindly done some sums to work out what would be required to soften the blow on the bulbs but what remains is exactly how to achieve this. I think his idea is to power the filaments at a reduced rate all the time, however, I would prefer filaments off and then power up over 5 seconds when the light switch is activated.

I do not have any EE background so apologies to those who know better, but I'm interested in this issue. We've never blown a bulb in our cars,,,, yet 

That's great Pip, thanks I appreciate your efforts.

Thanks.

Does the resistance vary if the temp is say -20 Deg or more

Yes it does but the difference between -20 and 20 compared to the operating temperature of "white hot" is insignificant. Cold temperature can be considered to be any place you are likely to take your car. The resistance difference between cold and hot is about 15:1 from my reading.

Our other member Pip... I think his idea is to power the filaments at a reduced rate all the time

Not quite all the time. Yes all the time you might have the park lights switched on but otherwise only activating the preheater as you pass though park to headlights on. I'm assuming here that the headlamps will preheat in the few milliseconds that the park circuit is energised before you get to the next click of the switch. It is just a cludge really.

And there are 2 theories at play here. One is that the filaments just pop at switch on due to inrush current and the other is that there is high voltage hash spikes of some sort from the starter motor or the alternator.

The inrush current is a known problem for (all) incandescent lamps due to their positive temperature coefficient.

[Phil №❶]

@pip

Is it necessary to treat voltage spikes differently to inrush current. It seem one is current and the other is voltage 

[beerman]

I recently got some repco globes, so will see how they go.

I also note Hyundai have a 12 month/20000kl warranty on spare parts. So next time one goes, I might get them to throw the new one in.

With globes costing $15 bucks a pop, even with Hyundai's stupid pricing it might prove a wise investment. If we all did it, the cost of globes may well cause them to come up with a fix.

[Dazzler]

I recently got some repco globes, so will see how they go.

I also note Hyundai have a 12 month/20000kl warranty on spare parts. So next time one goes, I might get them to throw the new one in.

With globes costing $15 bucks a pop, even with Hyundai's stupid pricing it might prove a wise investment. If we all did it, the cost of globes may well cause them to come up with a fix.

Good thinking 99 AKA Beerman 

[beerman]

Mind you, they are probably $122.06 each.

[Surferdude]

Mind you, they are probably $122.06 each.

Plus fitting.
And bulbs are excluded from warranties.

@pip

Is it necessary to treat voltage spikes differently to inrush current. It seem one is current and the other is voltage 

If they come from different sources then different treatment (probably) applies. Ultimately it is the temperature of the filament exceeding its ability to remain solid that "burns" it out. It literally melts. Actually, burn out is incorrect as burning implies take up of oxygen of which there is none inside a globe. The temperature reached is directly related to the power dissipated and in turn power is derived from the product of applied voltage and the current that flows as result of the voltage.

With a water analogy voltage can be thought of as a header tank full of water and current is that water which flows through the pipe from the tank. Resistance is simply determined from the bore of the pipe. One can see I think that the higher the tank (voltage) the more water (current) flows for a given resistance (bore) and the other interrelationships can be seen I think.

To the original question, excess power can be caused by too high a voltage or too low a resistance. The first is the suggested spikes while the second is the low cold resistance of the globe. Similar result but different cause.

If the spikes came during cranking from the starter motor then simply turning the lights of while starting should fix that. If the spikes came from the alternator randomly then this is more problematic and having a slow turn-on current would not have any effect as they occur at different times.

Hope that helps basic understanding and apologies if presumptuous.

[Phil №❶]

The water analogy is a good one, it helps understanding. No apologies necessary. 

Hmm, lots to consider