Headlights Revisited

[JohnEd]

Lin,

I understand completely and for a very good reason, I had all those thoughts and tried all those fixes.  Your posture was mine and while I was told to just install the new relay and be done with it the irresistible challenge was in the prophecy that I would fail.  I did and I am a pretty resourceful guy, thank you very much.  I think you will end up where I am after many tries if you continue to pursue the lighting issue.  I hope you don't think I am being pushy with this advice and not giving you space.  Farthest thing from my mind and I want you to have your own adventure to what ever extent individualism takes you.  Me?  All I ever was concerned with was results. As a fellow Knut and a friend I am obligated to share whatever info I think you might find useful.  Friends don't let friends drive drunk sorta thing.  I can, however be educated.

If the insides of the dimmer switch are compromised then they will present a voltage drop and that is a easy test.  After I replaced my dimmer switch I found that the new one had pretty much the same voltage drop.  In my adventure I finally figured out that all those little crimp connections are not sealed and the wire corrodes slowly and deteriorates the "quality" of the electrical connection while the mechanical connection remains good.  There are a ton of those crimp connections and the more current you ask them to carry the faster they deteriorate and one of the reasons is they get warm or hot where they aren't making a perfect connection and the heat expedites the erosion of the quality of the connection.  I determined that the wire alone was undersized and the circuit would still be unsat if all the components were perfect.  Disappointing!

On some cars I started out with 2 volts lost.  When I ran a +12 volt wire to the filament and just touched the bulb contact while the bulb was on the bulb literally exploded with new light.  With that firmly fixed in my mind I was hell bent on "getting the bugs out".  I used a digital voltmeter even way back then and I could see when I picked up a fraction of a volt.  When I was done soldering connectors and replacing connectors and replacing the L relay and cleaning fuse contacts till I was blue I figured it out.  Designed to be inadequate.  Crap!

Measure across every wire run and component and determine what you will gain from resolving their contribution to the issue of final voltage drop across the filament.  Where the rubber meets the road.  It would benefit all if you documented all the "bad" components and gave a before and after.  Please do that for everyone.

I replaced my headlight foot dimmer cause it was getting difficult to push and on occasion it would stick in.  Was probably 25 years old and very tired and dry and corroded.  The new one, now ten years old is still working fine.  had the same drop when I put it in as did the old one.

I am not now, nor have I ever been a perfectionist.  I shoot for 110% but I am happy with 80% from a reasonable effort.  There is a problem with adequate and that is "good enuf is the enemy of better".  Still, I get your point.

Every time I satisfy myself with a compromise it comes back to bite.  Any more, I can't do very many things as well as I know they should be done so I ain't preaching to you....just say'n.

I have concluded that all headlight circuits have at least a volt of drop designed in and that's almost 10% loss.  For this reason and others I  have never considered using the stock system.  BUT....the stock system is only deficient in terms of current handling capability.  Using the stock system as a "signal" voltage of a small 30 amp relay to drive the lights....perfect and entirely usable.  Not saying that there aren't some out there that are suitable but only in my sheltered life I ain't found any.


Good luck on your project Lin.  Wish I was there to hold the flashlight for ya.  Don't forget...we want to know every detail.  Please.

John

[Lin]

John,

I have no doubt that everything you say is correct.  Changing the system to include a nice, new, stout wire run through a relay appears to be the best way to go.  If when I test the new lights on the road I find them inadequate, the relay would have to be the next step.  However, if they light up the road enough, I might be prone to put further improvement way down the list.

[Tim Strommen]

On-topic, but not in a bus...

I recently (last month or two) had the experience of helping a friend with headlights on his 2003 Subaru Impreza WRX.  He had aftermarket lights that came with a new wiring harness that attached directly to the battery on the car.  He should have had every expectation that the included wiring harness was designed for the loads the lamps would create.

What I found after a few minutes was terrifying...

He had 55Watt H7 low-beam bulbs, and 65Watt H9 high-beam bulbs.  Total for one side (both low and high beams on) is 120Watts.  Each fixture (passenger and driver side) had a relay that allowed for the switching of the battery power directly to the fixture, and a second DPDT relay on the passenger side for enabling the high-beam.  This was all powered from a single 18Ga wire to the battery!!!!  When I measured the voltage drop at the bulbs (actually pulling the fixture out and putting the probes on the bulb leads), I found that the WIRES were dropping about 30% of the power to the headlights (about 8Volts from a 12Volt supply got to the bulbs... I'm still amazed that the setup didn't catch fire).  It also was powered from a single 15 amp fuse ( [65 + 55 + 65 + 55 = 240Watts] 240Watts/12Volts = 20Amps!!!)

I ended up building him a completely new harness with 10Ga wire from the battery and 12Ga wire to each independent bulb, and each side got its own 20Amp fuse (BTW it wasn't "cheap", I spent at least $100 on his setup).  Voltage drop now is less than 1%, and the comment is that it's like driving with a completely different set of headlights.  The maximum rule of thumb design drop for non-critical circuits (dome light, fan, etc... convenience features?) is 5%.  The maximum rule of thumb design drop for engine-compartment circuits and critical loads (like headlights, engine computers, wipers, etc... things that keep you safely operating the vehicle?) is 1%.  Non-critical loads run through the engine compartement are tighter tolerance because they are exposed to higher temperatures and the rating of the insulation becomes an issue with high-draw self-heating circuits.

Moral of this story, take a look at the voltage first, then take a look at the causes/problems - you may be suprised at what you find.  Don't just assume that the circuit is build right (even from the factory).

• Relay contacts have voltage drops (they are not perfect conductors, good manufacturers publish this info).
• Fuses have voltage drops (they aren't perfect either, that's why they can "blow")
• When possible, it is preferrable to use a return ground to battery, not the chassis - the chassis is an unpredictable conductor
• OEM builds for cheap/shippable, and they're insured for the build quality.  If you are going to mess with the wiring, you are directly responsible for the performance afterwards.  Overbuild it by %20, make the bulb the weak point so nothing starves it (them).  For a 65Watt bulb design for 1% drop with a ~80Watt load...

-Tim

BTW, the aftermarket system was from France - and he told me a joke once I showed him the math: "In Heaven - the Germans make the cars, the French do the cooking, and the English are the Police.  In Hell - the French make the cars, the English do the cooking, and the Germans are the Police."  First-hand I agree with that -T

[belfert]

My batteries are way back just in front of the engine.  If I run wires from the batteries to a relay at the front for the headlights will I have voltage drop from running the wires a good 30 feet or so?

Would a continuous duty solonoid be better than a relay for this?

[Lin]

I wanted relays when installing my Jakes.  The only 24v continuous duty relays I could find around were solenoid relays.  Considering their size and power capacity, they could be considered overkill, but that's what was available.  Any type of relay that can handle the DC volage, the amperage, and is designed for continuous use will work.

[JohnEd]

My batteries are way back just in front of the engine.  If I run wires from the batteries to a relay at the front for the headlights will I have voltage drop from running the wires a good 30 feet or so?

Would a continuous duty solonoid be better than a relay for this?

Belfert,

You have lights, heating/defrost blowers, fog lights and whatever else you have up front.  You should have a light "buss" run all the way to the front.  You have to figure your total current draw and decide if #4 or 0 or 00 is needed.  These wire sizes have a spec of X ohms per 1,000 feet.  You need to break that number down to 40 feet and figure the voltage drop using ohm's law....simple division and multiplication.  OR WE COULD ASK SEAN.  I can help with this if need be and be happy to do it.  Let me know.    It isn't as big a deal as it might seem at this point.  You first need to figure the "load" in amps by adding all the loads.  amps = watts divided by the volts.  To start you off.  There should be a fuse panel up there and that panel should have a feed from the bats.  If you put an ammeter in series with that wire and turn "everything" on that will be the amps for the full load.....unless your voltage drops a lot.  Then we can add the amps your new lights draw, add that and have a total load.
Like I said, Sean is your best bet.

John

[JohnEd]

On-topic, but not in a bus...

I recently (last month or two) had the experience of helping a friend with headlights on his 2003 Subaru Impreza WRX.  He had aftermarket lights that came with a new wiring harness that attached directly to the battery on the car.  He should have had every expectation that the included wiring harness was designed for the loads the lamps would create.

What I found after a few minutes was terrifying...

He had 55Watt H7 low-beam bulbs, and 65Watt H9 high-beam bulbs.  Total for one side (both low and high beams on) is 120Watts.  Each fixture (passenger and driver side) had a relay that allowed for the switching of the battery power directly to the fixture, and a second DPDT relay on the passenger side for enabling the high-beam.  This was all powered from a single 18Ga wire to the battery!!!!  When I measured the voltage drop at the bulbs (actually pulling the fixture out and putting the probes on the bulb leads), I found that the WIRES were dropping about 30% of the power to the headlights (about 8Volts from a 12Volt supply got to the bulbs... I'm still amazed that the setup didn't catch fire).  It also was powered from a single 15 amp fuse ( [65 + 55 + 65 + 55 = 240Watts] 240Watts/12Volts = 20Amps!!!)

I ended up building him a completely new harness with 10Ga wire from the battery and 12Ga wire to each independent bulb, and each side got its own 20Amp fuse (BTW it wasn't "cheap", I spent at least $100 on his setup).  Voltage drop now is less than 1%, and the comment is that it's like driving with a completely different set of headlights.  The maximum rule of thumb design drop for non-critical circuits (dome light, fan, etc... convenience features?) is 5%.  The maximum rule of thumb design drop for engine-compartment circuits and critical loads (like headlights, engine computers, wipers, etc... things that keep you safely operating the vehicle?) is 1%.  Non-critical loads run through the engine compartement are tighter tolerance because they are exposed to higher temperatures and the rating of the insulation becomes an issue with high-draw self-heating circuits.

Moral of this story, take a look at the voltage first, then take a look at the causes/problems - you may be suprised at what you find.  Don't just assume that the circuit is build right (even from the factory).

• Relay contacts have voltage drops (they are not perfect conductors, good manufacturers publish this info).
• Fuses have voltage drops (they aren't perfect either, that's why they can "blow")
• When possible, it is preferrable to use a return ground to battery, not the chassis - the chassis is an unpredictable conductor
• OEM builds for cheap/shippable, and they're insured for the build quality.  If you are going to mess with the wiring, you are directly responsible for the performance afterwards.  Overbuild it by %20, make the bulb the weak point so nothing starves it (them).  For a 65Watt bulb design for 1% drop with a ~80Watt load...

-Tim

BTW, the aftermarket system was from France - and he told me a joke once I showed him the math: "In Heaven - the Germans make the cars, the French do the cooking, and the English are the Police.  In Hell - the French make the cars, the English do the cooking, and the Germans are the Police."  First-hand I agree with that -T

That is my exact experience after many years.  I never found a stock system that wasn't dropping 10 to 20 %.  And that included my 70 Datsun Z car with the "gold plated connectors" and big wires. 

Thank you for posting in detail.  Good read.

John

[Tim Strommen]

...My batteries are way back just in front of the engine.  If I run wires from the batteries to a relay at the front for the headlights will I have voltage drop from running the wires a good 30 feet or so?...

Short answer, "yes".  The point you want to watch for though is not "if" but "how much".

Some popular Ohms/Foot for copper pulled from Basic Car Audio Electronics:

0000 (4/0) = 0.00004961
000 (3/0) = 0.00006250
00 (2/0) = 0.00007875
0 (1/0) = 0.00009921
2 = 0.00015749
4 = 0.00025000
8 = 0.00062996
10 = 0.00100000
12 = 0.00158740
14 = 0.00251983
16 = 0.00399998
18 = 0.00634956


An example of how to use this:

If you have a 100Watt CRITICAL load you want to power with a 12Volt system, and your load is 50 feet from the battery, here's how you'd work it out.

100Watts / 12Volts = 8.3Amps

Remember Critical load: 1% voltage drop max!!, Non-Critical load: 5% voltage drop max.

12Volt-system * 1% (0.01) = 0.12Volts max voltage drop

Maximum resistance the wire can have:

0.12Volts / 8.3Amps = 0.0144578313253012Ohms <-- this is the value where you would SUBTRACT the resistance of the fuse and the relay/solenoid contacts.

Add some design margin (20%):

0.0144578313253012Ohms * 0.8 = 0.011566265060241Ohms

Divide the allowable resitance by the length (I'll multiply the length by "two" first to account for a ground return wire):

0.011566265060241Ohms / ( 50Ft * 2 ) = 0.0001156626506024096Ohms-per-foot maximum

Now you look at the table above to see which wire has this value or less and that's your "okay" mark - thicker is okay, but more expensive and takes more space/weight, and may be harder to handle and terminate (connect to).

I see that a 0Ga (1/0) wire would be suitable for this load at <1% voltage drop from the wires (this example DOES NOT account for switching via relay, or a fuse!!).

0Ga = 0.00009921Ohms-per-foot
Max = 0.0001156626506024096Ohms-Per-Foot MAX

0Ga resistance is lower than Max resistance = OK


As always, if you feel this is out of your league, don't do it and if possible seek help from an experienced hand.

Best,


-T

[Tim Strommen]

...That is what I am planning to use on my baby, except with xenon bulbs and ballasts.  Xenon aftermarket kits have come down drastically in price over the last few years and provide a much closer color to actual daylight.  Not the blue and purple lights the rice cars are installing!!  I purchased a kit for my truck about 3 years ago and the performance has been flawless!  Animal eyes are much more visible from a much greater distance!  I would recommend a color spectrum around 5000K-6500K...

First it should be said that putting "retrofit HIDs" in a halogen fixture is definitely not legal.  The fixture housing is marked with the bulb type that the fixture was certified with (for example "H4").  If you install a bulb that is not "H4" in that fixture, the certification and beam compliance is invalid.  I don't advocate that this be done.

That said, I have done the same thing that bioVenture has done, and have installed H4-biXenon retrofits in my Cibie's on my pickup, but with extreme caution, research, and tweaking.  I disagree with the "color temperature" suggestion that bioventure gave - I don't recommend 5000K-6500K light for night-time headlights, but I agree that the super-blue and purple HIDs are stupid.  At night when the average light availability is lower, your eyes naturally shift the white-point to a greener shade, and become more sensitive to blue light (more of which can cause crippling glare in older drivers).  This is why OEM factory HID fixtures use a 4100Kelvin bulb, and why putting out 6500Kelvin light like we get during the day is not as good an idea as the marketing people may advertise.  Incidentally, the warmer color-temperature bulbs also put out more light.  I also went through the trouble of simulating the bulb and fixture which showed the HID retrofit needed some modifications, then I had the ACTUAL output independently verified by an optical lab to ensure the HID lamp's distribution matched the requirements for E-code HIDs.  A lot of extra work, and the combination is still not "legal" per-say, but a nice CYA if someone starts blaming me for something... (I should be able to cut down the "he did this with wreckless abandon!!" comments from the other lawyer).

-T

[JohnEd]

Tim,

I found that the HID bulbs did not focus like a halogen.  My Hellas that I am in love with were terrible and would have ruined the oncoming traffic.  I took them out after the first night of testing.  I have heard the same tale from others.  If you can get a fixture for HID E speck you will be amazed at the light.  Anybody want to buy some almost never used except one night HIDs?  Cheap.

John

[Tim Strommen]

...I found that the HID bulbs did not focus like a halogen.  My Hellas that I am in love with were terrible and would have ruined the oncoming traffic... ...I have heard the same tale from others...

This is slightly off-point...

Yes, I found that the kit manufacturer had incorrectly lined up the electrodes to the filament line on the H4 kit I have, which required re-seating the lamp in the base (the arc should be lined up with the filament as it curves upwards from the electrodes).  The next problem was the assembly itself - the bulb is designed to slide from the forward-most "shielded" position (low-beam) to an unrestricted-output position in the rear-most filament position (high-beam) using a solenoid.  I found that the shield was not correctly clocked (they had both sides level, the passenger-side edge should be higher by 15degrees!!), and the bulb wasn't hard restrained so it would shimmer if the assembly was vibrated (like in normal use on the road).  I had to replace the metal shielding with a cast ceramic shield with a sliding end-piece attached to the tip of the bulb to keep it aligned.  Then I had to dip the tip of the bulb in a black paint to obtain the front of bulb light blockage.

Once I did all of that, I still needed to sprinkle the glossy black ceramic shield with flat black bulb dip to knock down the light output to ECE standards.  Then I had is tested (passed). Note that the pattern was what I had tested - ECE requires HIDs to be used with an automatic headlight leveling system, which I do not have.

I have seen and felt (ouch) poor HID conversions many with the too-cool >=6000K bulbs that are just, well just awful...  It's even noticable that the light output isn't anywhere near what it should be (regular Halogens are often actually brighter).  Before I settled on the Cibies, I had run the Hella DOT/ECE fixtures - my opinion was that they were well intentioned but an optical disater.  I think they were a Marketing ploy to get DOT compliant fixtures with something related to ECE on them to sell them as ECE equivalent... (they were not).

By far the worst HID conversions I've seen are the ones on Fords, Dodges, and Chevy's with 9007 or equivalent bulbs, which instead of relying on a metalic light shield to control light just relies on the 3mm difference from low to high beam as well as a minor output wattage (45W low, 65W high).

HIDs in these fixtures basically output both high and low at the same time, or on the "biXenon" versions they put out LOW+HIGH or just HIGH-out-of-focus and not pointing on the road at all.

-T

[belfert]

I did some research on Hella vs Cibie.  Most of the forum postings out on the Internet seem to recommend Cibie over Hella.  They also all said that Hella would be better than DOT headlights.

I think I might spend twice as much and get the Cibie lights instead of the Hella lights I was planning to get.

[Tim Strommen]

I agree with the results of your research having done the following progression on headlights over the years on my truck (6"x8" rectangle):

Osram Sealed-Beam --> Sylvania Halogen-Capsule-Sealed-Beam --> H4 Hella DOT/ECE --> H4 Cibie --> HID Cibie

I personally recommend that if people can afford it, the upgrade path should skip the middle two steps between Sealed-Beam and H4 Cibie, and if you can do 24Volts to the bulbs (better for voltage drop!!), use Osram "TruckStar" 24V-H4 bulbs not the Hella 24V-H4 bulbs.  And again for clarity, I don't recommend the HID path...

-Tim

[JohnEd]

Tim,

Yours are some informative and valuable posts.  I have made copies for my RV  elect folder.  Thank you very much for your effort and sharing.

I finally got around to looking into the Cibie vs Hella issue "again" after all these many years.  Well things have changed and I should have known.  The 7 inch Cibie that does the trick is model number HCR and it is E code.  It has a lens pattern similar to the legendary, in my circle of old, Marshals.  That's the lens that has the converging long vert lines in the center of the lens that seems "cat eyes like" and has a razor horizontal cut-off on low beam.  I have never used the Cibies but I can imagine they are stellar.

The Hella lights for sale on Amazon and such for $34 ARE NOT the high performing lights I was referring to.  Those on the amazon are a clear lens with the reflector having the focusing element engraved in wide flat panels of the reflector.  That might be an improvement but it won't come close to the Cibies, I strongly suspect.  The "real" Hellas are number WO 133 1615 927.  In the catalog these lights are labeled "FOR OFF_ROAD USE ONLY" just as mine were way back when.  The catalog then goes on to list that model for all the Porsche s, Z cars, etc, high performance sports cars of the day.  That model sells for something like $68 each and not the $34  mentioned.  I think the Hellas don't have the great price advantage we thought they did.  Also, they are not street legal but they are over looked by law enforcement.  Cops can tell at a glance or the smart ones can and they have done so with mine but only smiled when mentioning it.

The advantages of the Hella in performance are that they produce a wide and flat low beam but at the edge of the right side of the pavement there is a "up-swept" beam that illuminates people and game very well and very early.  I bought them way back when for that exact feature.  They are also sold in singles.  They are marginally cheaper.  The Cibie is the storied performer but I am starting to have second thoughts about their performance edge over the off-road Hella.  Tim seems to point out that the stock Hella bulb is not the hot performer and I suspect that the standard issue Cibie comes with a better performing bulb.  I suspect that neither comes with the highest performing Osram item. (thank you, Tim).  My Hellas didn't come with a bulb and I have purchased 6 of the critters and never looked back.  I think the Cibies include a high performance bulb and that is maybe as much as $40 worth buried in the price....if not?

You guys have an enviable choice in that both are winners as far as I can see.  Double "antandrey" intended.

Thanks again Tim,

John

[artvonne]

  My Hellas didn't come with a bulb and I have purchased 6 of the critters and never looked back.  I think the Cibies include a high performance bulb and that is maybe as much as $40 worth buried in the price....if not?

John

  Interesting observation. IIRC the ones we purchased "many moons ago" came with bulbs, and if the Hellas do not, that could create quite a cost differential favoring the Cibie, as long as thats the particular bulb you need or want. In our case we wanted to blind everyone and put the 55/60's in the box for spares and bought 70/100's. Eventually you fall back to the 55/60 as those lamps with 100 watts on high are simply far to bright and I can see why they are illegal. IIRC, 100 watts is even illegal in Europe for road use. Seriously, road signs 1/2 a mile away are almost to bright to look at from the reflected light.

  Also, its interesting you note the difference in Hella lamps. If your going to put in good lights, you want the illegal ones. As long as they are adjusted properly (down) and have "normal" wattage bulbs, and you are a couteous and conscientious driver, no one will ever question them. Fail to drop them at a cop at night and he may start looking. Blind him 100 watts on high and he might just bust them out with his club.