Piston Slap: Droppin' Knowledge on Headlight Wiring Voltage Drop

Sajeev Mehta
by Sajeev Mehta

Randy writes:

A few weeks ago I installed a pair of 9003 Philips Vision Plus 30 bulbs in an attempt to improve my wife’s Tucson headlight output. The light down the road is pathetic yet the headlight lenses are clear and the reflectors are still mirror like, at least to the extent I can see from the bulb mounting opening. And the light patterns are well defined, both low & high beam, which also suggests the reflectors are okay. But no joy whatsoever from the new bulbs.

I then did some voltage tracing down the headlight schematic. There is ~14.4 V coming from the alternator & battery into the SJB (smart junction box) that the Tucson uses to switch power to the bulbs. But only 12.5 V on the pins out to the bulbs. And then another couple of 1/10th V loss down the 0.5 mm2 (~20 AWG) wire to the headlamp connector netting 12.2 or 12.3 V at the bulb. The lead electrical tech at a local Hyundai service department says this is 2V headlight wiring voltage drop normal, even as much as 2.6V.

BUT, by my calculations, this means that although DOT specs for 9003 bulbs indicate 910 lumens (+/- 10%) at 12.8V, my wife’s car will be generating a little less than 800 lumens at the voltage actually supplied to the bulb. No wonder night driving is such a frightening experience.

I’ve toyed with the idea of trying some off road 90/100 9003 bulbs under the notion that at this lower voltage the output would be only 30 percent or so above the DOT standard and not likely to be blinding to oncoming traffic. And one hopes not so hot as to melt anything in the headlight assy. But I’d still wind up putting ~7 amps down a pretty thin wire which at 14 A/mm2 is double the recommended maximum current density I’ve seen for automotive wiring. I don’t wish to trade better light for burnt wiring or perhaps a fried SJB @ several hundred $.

My best idea at the moment is to wire the headlights via some fused relays using the car’s high/low circuits to switch the relays. I’d happily trade shorter bulb life for safe night vision.

Can anyone suggest an alternative solution? And yes, I’m aware of HID and LED “conversions” but am not willing to go that route, trading more lumens for rogue beam patterns.

BTW, is it common nowadays for headlight bulb voltage to be so low? Or is this unique to Hyundai? I realize that this improves bulb life … but at such a cost to safe night vision.

Sajeev writes:

I was about to recommend installing headlight relays to ensure maximum voltage to your headlight bulbs. Then I remembered my experience: blowing out headlight bulbs (when flashing high beams) because said headlight relays were juiced directly from the alternator’s output wire. Apparently the correct source (this is on a Fox-body Mercury Cougar) was the starter solenoid since everything else starts there.

The point of this ramble? I have no earthly idea how to answer your question, but I know a guy…

Daniel Stern answers:

You’ve made some good observations, and you’re sniffing in a productive direction asking about bulbs and wiring and voltage.

First, let’s sharpen the picture of the voltage drop in your headlamp circuit. Voltage drop — that is, lower voltage at the demand end of the wires (the bulbs) than at the supply end (battery or alternator) — is caused by resistance in a circuit’s wires, switches, and connections. More resistance in the circuit means more voltage drop, and more current through a given circuit will mean more voltage drop than you’ll get with less current flowing through that same circuit. This means unplugging the headlights and probing the sockets won’t answer the question you’re asking, because that’s measuring the circuit with virtually no load on it: minimal current flow, so minimal resistance, so minimal voltage drop.

To get an accurate reading of the voltage drop, you have to measure the circuit with its normal current flow. That means all the bulbs have to be hooked up and working, and you’ll have to find a way to probe the socket without unplugging it. Some headlight sockets provide easy rear access to the metal terminals inside; if not, you may have to use a bit of cleverness: carefully insert a straight pin or unbent paperclip alongside each wire and into the socket until it makes contact, then touch your voltmeter probes to the exposed parts of the paperclips or pins. Or another technique is to remove the seal boot and probe the bulb terminals ahead of the connected socket. Avoid puncturing wire insulation unless there’s no other option and you’re prepared to effectively patch the hole immediately afterward.

On the HB2 bulbs in your car (other designations: 9003, H4), the terminal arrangement is as follows:

Left vertical terminal: Common


Top horizontal terminal: Low beam filament


Right vertical terminal: High beam filament

If your car uses a switched-feed headlamp circuit, the common terminal is the ground for both filaments, and the individual filament terminals are fed depending on the position of the headlamp on/off and high/low beam switches. If your car has a switched-ground circuit, the common terminal is the feed for both filaments, and the individual filament terminals are grounded depending on the position of the headlamp on/off and high/low beam switches.

You can put your voltmeter probes across the positive and negative wires and directly read the voltage at the bulb. To figure out where in the circuit the voltage drop is happening, put one probe at one end of the circuit (or the portion of the circuit you’re examining) and the other probe at the other end. For example: put the voltmeter’s positive lead on the battery positive terminal, and the voltmeter’s negative probe on the feed (+) wire or terminal of whichever headlamp beam (low or high) you’re testing at the moment. Use the bulb furthest away from the battery. With the lamps on, your voltmeter will show the voltage drop in the feed portion of the circuit, upstream of the bulb. Then connect the positive voltmeter probe on the ground terminal of the bulb, and the negative voltmeter probe on the (-) terminal of the battery. With the lamps on, your voltmeter will display the voltage drop in the ground portion of the circuit, downstream of the bulb. Adding the two voltage drop figures will give you the total circuit drop.

You raised the idea of using overwattage bulbs (100/90w), which you figure will still be brighter than stock despite voltage drop, but that’s actually not correct. A high-wattage bulb draws a lot more current than a standard-power bulb — about 65 percent more, in the case of a 100/90w bulb in place of a 60/55w bulb. That’s not just more current, it’s an overload, a whole lot more current than headlamp circuits are built to tolerate. Remember, more current makes more voltage drop and more resistance makes more voltage drop, so right off the bat you’re going to be much more significantly underfeeding the overwattage bulbs. The overload is going to heat up every part of the circuit and gradually cook stuff to death. Before that happens, though, your headlight performance is getting worse, not better: resistance causes heat (which is why light bulb filaments glow), and heat increases resistance, so depending on your luck you’ll have a vicious cycle of slow-roasted or quick-fried headlamp circuitry; neither is a tasty dish.

There are other good reasons to avoid overwattage bulbs, too; we’ll get to those in a minute. For now, let’s quantify the light loss we can expect as a result of voltage drop. Light output from a filament drops exponentially to the power 3.4 with voltage drop, so the formula to find the change in light output with a change in voltage input is: lumens @ spec volts x [(operating volts ÷ spec volts)^3.4] = lumens @ operating volts

910 lumens on low beam is the nominal spec for an HB2 bulb at 12.8v. There’s a tolerance allowed of plus-or-minus 10%, so an HB2 bulb producing anywhere from 819 to 1001 lumens on low beam is equally legal, but for these calculations let’s assume the bulb is right smack on spec and plug some numbers into our formula.

If the bulbs are getting 12.2 volts:

910 x [(12.2 ÷ 12.8)^ 3.4] = lumens we’re getting

910 x [(0.953125)^3.4] = lumens we’re getting

910 x 0.8494 = lumens we’re getting

773 = lumens we’re getting. So slightly less than a 5 percent voltage drop took a 15 percent bite out of our intensity.

If the bulbs are getting 11.8 volts:

910 x [(11.8 ÷ 12.8)^3.4] = 690 lumens (a little under 8 percent volt drop, 24 percent light loss)

It works the other direction, too; suppose we feed the bulbs 13.5v:

910 x [(13.5 ÷ 12.8)^3.4] = 1091 lumens (a little over 5 percent above spec voltage, a little under 20 percent more light than spec).

Or suppose we connect the bulbs via fat wires directly to full-boogie 14.4v off the alternator:

910 x [(14.4 ÷ 12.8)^3.4] = 1358 lumens (12.5 percent above spec voltage, 49 percent more light than spec).

So yes: feeding the bulbs a complete breakfast makes them put out more light, and starving them makes them feeble. But we might not want a zero-loss power path between the alternator and the bulbs. Bulb lifespan is also exponential with input voltage, but the exponent is much larger at -13 (negative 13), so the formula for bulb life looks like this:

hours @ spec volts x [(operating volts ÷ spec volts)^-13] = hours @operating volts

Let’s take a long-life HB2 bulb with a 1000-hour-at-12.8v low beam filament and run those same voltage calculations:

12.2v: 1867 hours (87 percent longer than spec)


11.8v: 2879 hours (almost 3x spec)


13.5v: 500 hours (half of spec)


14.4v: 216 hours (less than a quarter of spec)

This tells us why an automaker would prefer a little less rather than a little more voltage reaches the bulbs; it cuts down dramatically on bulb replacement costs under the vehicle warranty. Add the cost savings from using lighter-duty wire and switches, and multiply by the number of cars built, and, well, in the auto industry a fraction of a penny shaved from the automaker’s cost of a car earns someone a fat bonus and a promotion.

Warranty cost reduction is also a major reason why automakers tend to specify long-life bulbs as original equipment. But the filament changes required to make a long-life bulb tend to reduce luminance and beam focus, which means shorter seeing distance and a browner light color. The image shown here is from the R&D journal of Hella, one of the world’s foremost vehicle lighting manufacturers. It’s a comparison of the low beam seeing distance from a headlamp using a long-life bulb (3000 hours) versus that same headlamp using a bulb with higher luminance and better beam focus, but shorter lifespan (500 hours). This should go a fair ways toward answering why picking the longest-life bulb you can find and feeding it as high a voltage as you can would be a no-gain, zero-sum deal.

Now come those other reasons I mentioned for avoiding overwattage bulbs. For one, their lifespan is very short even at carefully-controlled voltage. But let’s say you really don’t care about bulb life; you’re happy to carry around a case of spare bulbs and change them every 10 days as long as you can see. But overwattage bulbs won’t help you see better — not really, it’ll only feel that way. High-watt filaments are bigger than lower-watt filaments, and a bigger filament has a strongly negative effect on beam focus, so effective seeing distance plummets. At the same time, the big increase in foreground light destroys your distance vision; the overlit foreground causes your pupils to constrict and drags your gaze downward, so you can clearly see all the things close to the car (at realistic road speeds you’re going to hit ’em no matter how brightly lit they are), but you can’t see the important stuff further down the road. Double-whammy, too; the bright foreground creates the false feeling that you have excellent lighting — we humans are very poor judges of our visual performance; our vision just doesn’t work the way it feels like it does, so it’s very easy to create situations where we think/feel we can see much better (or much worse) than we actually can, and while foreground light is the top correlate with subjective impressions of headlight performance, it’s really very far down the list of safety performance factors in a headlamp’s light distribution. Down-the-road distance light is much more important.

And on top of all that, you’re in a Tucson, which is an SUV with headlamps mounted higher than they are in a passenger car. Put overpower bulbs in, and you’re producing unsafe and illegal amounts of blinding glare for other road users (some of whom can write you tickets for it), unless you lower the lamp aim — and even a seemingly minor drop in aim angle severely limits your seeing distance (while increasing foreground light even more). Lose/lose.

But wait, there’s more! Since overwattage bulbs aren’t legal in any country developed enough to have traffic lights, there’s no reason for automakers to specify (or pay for) headlamps that can tolerate the much greater heat output of high-power bulbs. You might get away with it for awhile, but even if your headlamps don’t turn into flaming rivers of molten goo right before your very eyes, they’re suffering. The reflector is shiny by dint of a super-thin layer of vapor-deposited aluminum with a super-thin protective clear topcoat. The heat from a high-power bulb degrades the topcoat, then the aluminum oxidizes and the reflector’s shot to hell, even if it still looks nice. Optical degradation of the reflector is severely advanced well before you can see it with the naked eye; by the time it’s progressed far enough to be described as “pretty good, just a little imperfect” the lamp is a zombie, long past dead. In optics the important stuff happens on a scale we can’t see with our eyes: a headlamp reflector and the bumper of a vintage show car look equally shiny to us, but an as-new headlamp reflector is over 97 percent reflective, while even the most costly, beautiful show chrome is only about 67 percent reflective — not nearly adequate for optical purposes.

So let’s look at better bulb options. Bulb selection matters a lot to how well you can (or can’t) see at night. It pays to be picky, because there’s a mountain of junk on the bulb market, a lot of it is overpriced, and of course it’s all promoted as an “upgrade”. You were headed in the right direction when you picked up those +30 bulbs, but you’ll want to go further down that road. The +30 bulbs you got were state of the art about 20 years ago; since then we’ve progressed: +50, +80, +90, +100, +120, etc.

What do those plus-numbers mean, though? It’s not the amount of light coming from the bulb, because again, that’s pretty tightly controlled: plus or minus just ten percent on the HB2 low beam. In fact, the plus-numbers refer to intensity within the headlight beam. We already talked about the performance-degrading effects of optimizing a filament for long life; the opposite filament changes are made to create the “plus” bulbs, with the opposite effects: lifespan is reduced, but luminance and beam focus are improved so you get to see better and longer, and the light isn’t as brownish. The important thing to understand is that even if all the filaments put out exactly the same amount of light — the same lumens from a long life bulb, a standard bulb, +30, a +80, etc. — the headlamp performance and appearance with the long-life bulb would still be inferior compared to the same headlamp performance and appearance with a “plus” bulb.

That’s enough theory to understand the bulb test results posted here by a far-flung colleague of mine. Take a look at the giant differences in beam performance between a standard 60/55w bulb and a “+80” 60/55w bulb! Also note how the only advantage the 100/90w bulb gives is in amount of light on high beam. Seeing distance on low beam is inferior to that from the “plus” 60/55w bulb.

So okeh, to see as well as possible you’ll want to skip the overwattage bulbs and the weak standard and long-life bulbs; now let’s eliminate some more bad options in another fell swoop: any bulbs claiming to produce “extra white” light is best avoided (or super white, hyper white, platinum white, metal white, xenon white, etc). It doesn’t matter whose name is on the bulb — Sylvania Silver Star or ZxE, Philips BlueVision or CrystalVision, Wagner TruView, anything from PIAA or Hoen,, Nokya, Polarg, etc. — all the same bad deal. They have blue-tinted glass which changes the light color a little but blocks light that would reach the road if the glass weren’t tinted, so they give you less light than ordinary bulbs (not more). It’s purely for fashion; these bulbs (poorly) imitate the color, but not the performance of higher-technology HID and LED headlamps. The filament has to be driven very hard to get legal-minimum levels of light through the blue glass, so these kinds of bulbs have a very short lifespan, and there’s nothing about the tinted light that improves your ability to see — the opposite is true (less light = less seeing, no matter about the tint). A couple of years ago Sylvania got spanked pretty hard (30 million dollars’ worth!) for false and misleading “upgrade” claims on Silver Star bulbs, and theirs are among the least-bad ones, so the math kind of does itself: you’ll want to reject the whole category.

My current favorite 60/55w H4/HB2/9003 bulb is made by Tungsram (GE of Europe); it is a +120 item that is a few developmental iterations improved over the +80/+90 bulbs in the linked comparison tests. I bring them over from Europe in wholesale quantity, but they’re also available (usually at a price I have a hard time seeing as cost-effective) in American GE packaging as seen here — replete with the usual clue-deprived “reviews”. Another solid pick is the Philips Xtreme Vision +100, more reasonably priced as seen here. Occasionally there’ll be a screamin’ deal price drop on one or another of these premium bulbs; right this minute here’s a super price on the Philips Xtreme Vision +130, though it could be gone by the time you read this, and Canadian prices and deals are going to be different.

Narrowing in even further, what’s to pick among these +100, +120, +130 types of bulbs? Rather than just shooting for the highest plus-number, you’ll want to pick ones with the least amount of blue-tinted area on the bulb glass. The Philips and GE +130, the Sylvania Silver Star Ultra, and the various Osram Night Breakers have an uncolored ring surrounding the low beam filament, but the whole rest of the bulb, including the area the high beam filament shines through, is blue-tinted. That means you’re losing some performance on high beam for useless fashion points. The Philips +100 and the GE-Tungsram +120 have only a thin blue ring right below the tip of the bulb, which doesn’t affect performance at all. The marketers say it’s is there to make a fashion statement with your headlamps, etc, and yes, from various off-axis angles as you observe the headlamp, you do see some blue glint, but it’s really there to sneak the bulbs past their type-approval tests. The regs strictly limit bulb output to not more than 10 percent above the nominal spec, remember? That’s tested in a device called an integrating sphere, which measures the bulb’s total output in all directions. The blue ring filters a part of the bulb that has nothing to do with beam formation because it’s not in the light path between the filament and the reflector. So that blue ring gives fashion-fixated kids of all ages a blue flash they can point to from certain irrelevant angles without coloring the beam or filtering out any usable light. It gives the marketers something to babble about so the bulbs fly off the shelves. Meanwhile, the filament is pumping mad lumens through the uncolored glass where the reflector is looking — a clever trick that really works.

If you want to wring maximum possible light out of the headlamps, no matter how short the bulb life might be, that’ll take the optimal bulbs and a relay install done thoughtfully and properly (i.e., make sure you’re putting in components dependable enough to entrust your life to, because that’s what you’re doing); hints are here. But there’s one other reason why you might not want to install a headlamp relay harness: if your vehicle uses the headlamps as DRLs (daytime running lights), then in many cases you can’t put in a relay harness without making problems: most headlight-based DRL systems operate the low and/or high beams at reduced voltage compared to what they get when you turn on the headlamps to drive at night. High beam DRLs operate at about 50 percent of normal voltage; low beam DRLs at between 75 and 85 percent. Full-intensity low beam DRLs are only allowed if all the lights normally lit with the headlamps — parking lights, tails, side markers, license plate, etc — either come on full time with the DRLs, or come on automatically when it gets below a certain darkness outside. Full-intensity high beam DRLs aren’t allowed at all. But the lights will run at full intensity if you install a headlight relay harness on such a vehicle, meaning the DRLs will be operating in an unsafe and illegal manner until the relays burn out, which they will. If you’re bound and determined, you can disable the headlight-based DRLs (easier on some vehicles than others) and move the DRL function to the front turn signals with a module like this, or install a reputable-brand (Hella, Philips, Osram/Sylvania) set of LED DRLs.

Now we’ve finally sorted out what to do about bulbs and wiring: lamp aim is by far the main thing that determines how well you can (or can’t) see at night with any given set of lamps, so this is crucial: you will need to see to it that the lamps are aimed carefully and correctly with an optical aiming machine as described here. It can be difficult to find a shop that has (and uses) an optical aiming machine; keep calling around until you get the right answer — and “we shine ’em on a wall/on a screen” is the wrong answer. To get an idea of what a proper lamp aim job looks like, see the VW document presented here.

One final point: you’re absolutely right and very smart to steer clear of so-called “HID kits” and “LED bulbs”, every last one of which is illegal and unsafe. Halogen lamps can only work effectively, safely, and legally with halogen bulbs.

[Image: Shutterstock user Fahroni]

Daniel Stern has been a freelance vehicle lighting consultant for decades. He’s on several of the world’s technical standards development boards for vehicle lighting, and is Chief Editor of DrivingVisionNews, the global vehicle lighting and driver assistance industry’s journal of record.

Send your queries to sajeev@thetruthaboutcars.com. Spare no details and ask for a speedy resolution if you’re in a hurry…but be realistic, and use your make/model specific forums instead of TTAC for more timely advice.


Sajeev Mehta
Sajeev Mehta

More by Sajeev Mehta

Comments
Join the conversation
4 of 39 comments
  • Speedlaw Speedlaw on Jun 16, 2019

    Thank you Daniel. Back in the day, we'd toss the DOT spec lights and insert H4 or H1 bulbs, depending. I had a Golf and a Jeep, each had the 7 inch H4 spec with normal 55/60 bulbs, then as an additional switched circuit, relayed off the high beam side, 100 watt H1 bulbs in separate driving lights. Provided you aimed the H4 correctly, you were in good shape. The current "blu bulbs" or the guy who runs around with an LED strip on, are just a public nuisance. Now, I get it...I've had bad lights on cars. The nonsense 9004 bulb comes to mind, but even that was good in the Volvo 240, which had a huge lamp to aim it....although most interations were horrid. Luckily, that bulb appears to have died out, finally. Lights can be wonderful. I have an LED setup on my C class that is bright, white, aims itself, and does not glare at all, while giving me the best light I've ever had . My MDX, with HID low and halogen highs, also does an excellent job. I've not felt the need to adjust or augment either car. One of the few places the IIHS and my issues align is in headlights and crash tests. I've been followed and glared by too many bogus LED lights. I'm not a violent person but I have a fantasy about carrying around a ball peen hammer.....

    • See 1 previous
    • HotPotato HotPotato on Jun 17, 2019

      I had H4s in my 914: amazing. I imported a Euro grill with factory monoblock aero H4 lamps for my gen-1 Jetta: amazing. However, my mechanic wired in a second set of wires & relay with a switch under the dash. Flick the switch off, and they were as sickly-weak and yellow-brown as the lights on any gen-1 water-cooled VW over 5 years of age. This was handy when the local cops were looking to bust people for aftermarket lights.

  • Jb2wheels Jb2wheels on Jun 18, 2019

    Excellent read. I finally subscribed after lurking for a few years just to comment. Nice to understand why so many oncoming cars with headlight "upgrades" annoy me to no end.

  • Corey Lewis Think how dated this 80s design was by 1995!
  • Tassos Jong-iL Communist America Rises!
  • Merc190 A CB7 Accord with the 5 cylinder
  • MRF 95 T-Bird Daihatsu Copen- A fun Kei sized roadster. Equipped with a 660cc three, a five speed manual and a retractable roof it’s all you need. Subaru Levorg wagon-because not everyone needs a lifted Outback.
  • Merc190 I test drive one of these back in the day with an automatic, just to drive an Alfa, with a Busso no less. Didn't care for the dash design, would be a fun adventure to find some scrapped Lancia Themas or Saab 900's and do some swapping to make car even sweeter. But definitely lose the ground effects.
Next