Headlights...

I just dug through the old invoices, David. I bought a NAPA P/N 819-5562. 

Thanks for the quick reply, Stan. Does your lift strut attach just South of the hinge and then go forward thru the engine firewall through the vacant space and then attach to the rear of the rear seat bulkhead ? I didn't look up there when under the car last. 

Glad you had a nice time in Mexico   ~  one of my Boys just came back from a wedding there and had a great time. 

I don't have the engine strut like you guys... they now use ... the SS metal wand you see below.  It goes into a small dimple that is simply drilled in the FGlass and is on a hinge point. If you look on the edge there is a plastic clip that the rod gets pushed into to secure it so that it does not flop around.  Works well 

That’s a better setup. 

David- the front of the strut goes to a dedicated bracket on a frame member in the cavity between the “bathtub” and the bulkhead.

And Mexico, yeah. Mexico is my happy-place. 

Where do you like to go to Mexico?  We have gone a few times to Puerto Vallarta and one other place Cancun I think. 

I know about Terminations!!  (not so much about Mexico, but stayed on Oaxaca once in 1967!)

All computer assemblies use cable harnesses very similar to automotive/airplane ones and all of those (including cars and planes) have crimped terminations.  What makes the difference between good (what you and I aspire to) and Mil Spec (what the airplane guys want to have to keep their butts from falling from the sky) are better quality terminals themselves and better quality crimping tools that provide consistent crimps.  So let's look at those two:

1.  The terminations can be made from cheap, semi-malleable steel (or other metals, at a price premium) or better grades of metals that can be deformed (while making the crimp) and stay that way.  

The cheaper material (think Chinese terminals from Pep Boys) can be crimped, but the metal then often has a lot of stress cracks which lead to a weakened connection and/or corrosion of the joint.  This is especially true if the joint is over-crimped.  Pep Boys and others carry everything from poor quality parts to relatively higher quality parts.  I get my electrical parts from NAPA or Fleet Electric Supply (a local shop) so can't really comment on the other sources.

With higher quality terminations, the metal accepts the distortion of the crimp without stress fractures and then holds that shape with integrity, providing a solid connection between the wire and termination that doesn't degrade over time.

There are also a number of different terminator materials used, depending on application and cost point, from steel (Pep Boys stuff and many others) to copper to bronze to gold (Augat, AMP or Amphenol, sometimes Molex) in ascending cost points, so the designer has to decide between the application's current/resistance/termination integrity and environmental considerations (will it be exposed to elements or protected).   The terminator used depends on wire size, application and whether the terminator is exposed to the elements and needs to have a hood/cover or water-proof enclosure.  The stuff from the automotive parts stores are color coded by wire size from #18 (Red) to #12 (Blue) to #10 (yellow).

2.  Crimp tools - can be either hand-operated or air/pneumatic operated, but either one should provide an even crimp (evenly, all the way around the wire) of consistent force.  All computer harness makers use air crimpers for operator comfort (try making a couple thousand crimps per day and see how long YOU last) and crimp force consistency.  They are usually tested/adjusted before every shift and there are QA tests for crimp quality (only one of several is wire pull force allowed in pounds after crimp).  

The hand ratcheting crimp tools can be adjusted for varying crimp force, and you apply hand pressure as it ratchets until it no longer "clicks" and then stop and the handles open automatically, providing a consistent-force crimp, crimp after crimp - You cannot under-crimp (the tool won't open until the full crimp force is applied), you cannot over-crimp.  Over-crimping fractures or distorts the terminator barrel (where you insert the wire) so it loses structural integrity and weakens the connection (so you can pull the wire out after time passes).  The variables are terminator placement in the tool, placement (depth) of the wire in the terminator barrel and quality of the terminator material.  Non-ratcheting crimpers cannot provide a consistent crimp because the pressure applied will be different every time.  Might not be a big deal for the 2 - 3 crimp per year user, but the inconsistencies remain.  I have been known to use water-pump pliers to do a good crimp in a pinch (Gasp!), but it would not be my first choice tool.

So what is a good, ratcheting hand crimper?  There are a bunch between $20 bucks and $400 bucks, but for our use, I would recommend this one, with a very long URL, from Amazon:

https://www.amazon.com/Foxnovo...1574373977&psc=1

It does everything mine does (plus) for 1/3 the cost.  If you're building or vastly renovating a car and doing a lot of crimps, get one of these or one like it.

Does it eliminate soldering connections?   Well, if the connection will be exposed to the elements, no.  For those, you can either solder it, put some shrink tube over it and forget it or use the proper environmental terminations with a solid crimp and even then I would put a dab of silicon caulk on the wire end of the terminator, just to seal it.

I have, from time to time, made a good electrical connection (twisted wires) then soldered it (which provides more environmental protection than an electrical connection), then put shrink tube over it, filled the shrink tube with silicon caulk and THEN shrunk the tube to make an environmentally secure connection.  That's pretty extreme, but it works really well (it's very similar to under-ground electrical connections for lawn irrigation systems which use a dielectric grease in place of the silicon caulk).  

Remember that solder is only a mechanical connection (although it provides some environmental protection, too - it is not to be relied upon as an electrical connection and can introduce high resistance if not done properly (look up "cold solder joint").

Long rambling, and more than you wanted to now about this stuff, I know.  Sorry.

gn

Gordon,

Will this tool crimp an aluminium stop, for say accelerator cable or do I need a swaging tool?

Art

It should.  You would have to have an un-crimped electrical terminal (to compare sizes) and then use the jaw set that matches, but, yes....It should crimp an aluminum barrel as well as anything else.  The difference in force needed is less, with aluminum, but for that application it should be fine.

Bear in mind that some tools crimp a round barrel while others do a hexagonally shaped (six-sided) crimp - Coax cable ends come to mind - If you look at the crimp jaws the hole is 6 or 8-sided.  While those work, for a while, a full perimeter crimp is better, which is provided by that tool up above.

Thanks Gordon!

Battery cable lug ends for starter and main chassis ground :  I no longer live  10 minutes from a "everything" battery store that would do the ends for the cost of the parts so I now do my own. Strip the cable end back 3/8 " - 7/16 " I set the copper lug horizontal into the vise jaws with a 1/8 machine screw or 10 penny nail parallel  against the side of the open lug closing the vise jaws against the lug it dimples the lug and making a secure factory looking clamp.

Thanks, Gordon. I still do all mine with a plier crimp and I suppose I ought to up my game. I also usually solder and heat shrink but it's pretty time and labor intensive.

fwiw my crimping career began about age 11 when my uncle Pete handed me a role of coated steel wire and a tiny bag of barrel crimps and instructed me to make a half dozen live bunker rigs.

Bunkers, aha Menhaden, are an oily fish used for bait. They usually weigh one or two pounds. We'd fish them live, with one hook just below the dorsal fin and one near the tail, and when the blues were getting big and ferocious near the end of the season—mid or late October, say—you better have at least 60-pound-test wire leaders on every hook. 

Absent a liberal drag setting on the reel, a 14-15-pound blue could just about pull an 11-year-old boy overboard. 

So to make the rig you cut off about three feet of wire, loop one end  about two inches and crimp it with a barrel crimp, using a channel lock. Slide another barrel and a hook on the other end and make a loop about six-eight inches up from the end with the hook in it, then finally one more barrel and a hook with a loop at the bottom.

I made these for decades and never saw one break or even slide no matter what kind of monster fish grabbed it.

Which is why loose electrical connections—and the specialty tools we need to make them tight—mystify me. 

Geez, I take my wife out in the Speedster for a nice lunch in the wine country on a warm, sunny, January afternoon, and am overwhelmed on my return.

More posts than you can shake a stick at about cable grades, crimp connectors, soldering technique, hydraulic deck lifts, and Mexican vacation spots. Just a typical day here, I guess.

For the record, I soldered all the connections under the car and crimped the ones at the relays inside the battery box (which I figured are a lot better protected from the elements and from things that go bump in the night).

Gordon is right. There is a knack to proper soldering. You don't want to just drip melted solder over a joint and call it done. You get the joint hot enough to draw solder in and make sure it flows into both things you're trying to join. A good flow has a nice, shiny look when done. If it has a matte, bubbly look to it, start again.

My trusty 230-watt Weller had barely enough power to heat up two ends of 12-gauge multistrand copper wire. One of those little butane torches probably would have been better. Don't even think about using one of those 25-watt pencil irons, intended for lightwight electronics work.

And I see there are more professional-grade crimpers than I knew existed. Air-powered and 400 bucks? I'm going to omit here the part where I explain what kind of crimper I use.

I will say only that I have been crimping away under the dash of this car for about five years now and, so far, everything is still crimped, and all the stuff that's supposed to light up, buzz, beep, and whir still does.

One significant discovery in my wiring install was the state of grounding connections on my Vintage Speedster.

First, I finally found the grounding point on the front beam where the heavy cable from the battery is connected. You have to remove the wheel and get way in there to see it. You can't see it from under the car. The battery cable ends in a ring crimp connector that's bolted to the beam. It's one of the most vulnerable spots under the car, exposed to splash from the wheels and dirt and debris from the road. I think this is probably the best place to start looking for trouble if there are electrical problems of almost any kind in the car. I connected my new ground wire for the headlights directly to that point, BTW. But I'll be checking up on it periodically now that I know where it is.

Second, there are five wires going to the headlight bucket (high beam, low beam, running light, turn signal and ground wire). The ground wire is shared by all four circuits and is the skinniest wire of them all - maybe 16 gauge!

And it doesn't go straight to the battery negative post. It disappears into the wiring harness and probably makes a few complete tours of the car before connecting to anything that's a real ground. That right there could be the main reason why the headlights on these cars start out so freakin' dim.

Alan Merklin posted:

Battery cable lug ends for starter and main chassis ground :  I no longer live  10 minutes from a "everything" battery store that would do the ends for the cost of the parts so I now do my own. Strip the cable end back 3/8 " - 7/16 " I set the copper lug horizontal into the vise jaws with a 1/8 machine screw or 10 penny nail parallel  against the side of the open lug closing the vise jaws against the lug it dimples the lug and making a secure factory looking clamp.

In a very critical batt cable connection, I've been known to finish by drilling a 1/8" hole thru it and finish with an AN bolt and nut. I know.....

David Stroud posted:

Aircraft guys swear by connectors sold by www.steinair.com  and those sold by Aircraft Spruce too. This stuff is aircraft rated but I don't know what qualities of the product make them better than the hardware store variety though. 

I'm betting they're the same connectors.

The aviation stuff is probably backed up by a higher grade of liability lawyers.

Ryan in NorCal posted:

...I suspect we can measure voltage drop before switching out the wiring to test, right? Can you share what you're getting to your relays? If you have 13v on the battery, do you see 13v at the relays now?

@Ryan (formerly) in NorCal , I'm too lazy to pull the spare and all the stuff that's on top of it out of the boot to get to the relay but...

After sitting a week in the cold garage without a start, the battery was at 12.7 volts this morning. The relays are only a 12" length of 12-gauge wire from the battery plus terminal, so they must have been at virtually the same voltage.

Voltage to my lights before the conversion was about 9.5 volts!

At one point, I had one light on the new wiring and the other still on the old, and shining the lights at the garage door, you could see an obvious difference.

Sacto Mitch posted:

 

Geez, I take my wife out in the Speedster for a nice lunch in the wine country on a warm, sunny, January afternoon, and am overwhelmed on my return.

More posts than you can shake a stick at about cable grades, crimp connectors, soldering technique, hydraulic deck lifts, and Mexican vacation spots. Just a typical day here, I guess.

For the record, I soldered all the connections under the car and crimped the ones at the relays inside the battery box (which I figured are a lot better protected from the elements and from things that go bump in the night).

Gordon is right. There is a knack to proper soldering. You don't want to just drip melted solder over a joint and call it done. You get the joint hot enough to draw solder in and make sure it flows into both things you're trying to join. A good flow has a nice, shiny look when done. If it has a matte, bubbly look to it, start again.

My trusty 230-watt Weller had barely enough power to heat up two ends of 12-gauge multistrand copper wire. One of those little butane torches probably would have been better. Don't even think about using one of those 25-watt pencil irons, intended for lightwight electronics work.

And I see there are more professional-grade crimpers than I knew existed. Air-powered and 400 bucks? I'm going to omit here the part where I explain what kind of crimper I use.

I will say only that I have been crimping away under the dash of this car for about five years now and, so far, everything is still crimped, and all the stuff that's supposed to light up, buzz, beep, and whir still does.

One significant discovery in my wiring install was the state of grounding connections on my Vintage Speedster.

First, I finally found the grounding point on the front beam where the heavy cable from the battery is connected. You have to remove the wheel and get way in there to see it. You can't see it from under the car. The battery cable ends in a ring crimp connector that's bolted to the beam. It's one of the most vulnerable spots under the car, exposed to splash from the wheels and dirt and debris from the road. I think this is probably the best place to start looking for trouble if there are electrical problems of almost any kind in the car. I connected my new ground wire for the headlights directly to that point, BTW. But I'll be checking up on it periodically now that I know where it is.

Second, there are five wires going to the headlight bucket (high beam, low beam, running light, turn signal and ground wire). The ground wire is shared by all four circuits and is the skinniest wire of them all - maybe 16 gauge!

And it doesn't go straight to the battery negative post. It disappears into the wiring harness and probably makes a few complete tours of the car before connecting to anything that's a real ground. That right there could be the main reason why the headlights on these cars start out so freakin' dim.

 

Be glad your not using these.... Back then At least, 1910 you probably could not outrun your headlights with the roads as they were.

Stan Galat posted:

Mitch,

...There is actually a wire-type for "extension cord" wire (stranded and jacketed) in the electrical wiring world. The version with 250v-rated insulation is "SJ" cord, and can be purchased in various gauges and number of individual wires (14-3, 12-2, etc)...

...an electrical supply house (or probably somebody on eBay can sell you what you need by the foot...

@Stan Galat , fwiw, the Amazon cord was rated SJTOW 12-3. I've learned the 'T' is for the type of jacket (thermoplastic), the 'O' is oil resistant, and the 'W' weather resistant.

I figured the cable was available by the foot somewhere, but for the short length I needed probably not worth the hassle or any potential cost savings. The Amazon cord was $20 and on my doorstep in 48 hours.

Since I wanted color-coded wires, three short spools of 12-gauge from NAPA would have been more than that.

Besides, I only used about eight feet out of the middle, so I can solder the unused ends back together and still have a perfectly good seven-foot extension cord!

Ray, those look about as bright as mine did before the upgrades.

And if you forget to turn them off, you don't kill your battery:

"Back then At least, 1910 you probably could not outrun your headlights with the roads as they were."     I'm sitting here thinking  when they hit something back in 1914 did the wonder if they hit a deer or, a neighbor ?   

Sacto Mitch posted:

 

David Stroud posted:

Aircraft guys swear by connectors sold by www.steinair.com  and those sold by Aircraft Spruce too. This stuff is aircraft rated but I don't know what qualities of the product make them better than the hardware store variety though. 

 

I'm betting they're the same connectors.

The aviation stuff is probably backed up by a higher grade of liability lawyers.

 

No, I think they won't be the same connectors just like they're not the same type of nuts and bolts either. I'm no expert but I do know that the AN ( Army Navy ) specified variety of bolts for instance are of a composition that will allow a certain range of flexibility and strength that is suited for general aircraft use and then there are others intended for less common but specific aircraft use....high strength, close tolerance etc. Add to that, these nice pieces of hardware require a high level of mfg. inspection for stated compliance during manufacture which adds to the cost.  

I'd guess that the "aircraft grade" connectors are made to some high end specification too with the metal and plastic and would follow a similar inspection cost for conformity too. But, yes....for sure they'rd be a sling of high grade lawyers on standby not far away. 

David, was sort of joking.

Don't know about the specifics here, but I'm sure hardware made to aviation standards is often better than garden variety. But I think a large part of the price difference is often due to certification and liability costs.

A pilot friend often bemoans the cost of parts for his plane's engine - an air-cooled Continental six cylinder based on what's now ancient technology.

The costs of getting newer designs certified and liability insurance for manufacturers has almost eliminated the chance of more modern options becoming available at prices the average guy can afford.

Cheers.

Sacto Mitch posted:

 

David, was sort of joking.

Don't know about the specifics here, but I'm sure hardware made to aviation standards is often better than garden variety. But I think a large part of the price difference is often due to certification and liability costs.

A pilot friend often bemoans the cost of parts for his plane's engine - an air-cooled Continental six cylinder based on what's now ancient technology.

The costs of getting newer designs certified and liability insurance for manufacturers has almost eliminated the chance of more modern options becoming available at prices the average guy can afford.

Cheers.

Agree completely. Gonna put another log on the fire.... -26 C tonight. Out, Amigo. 

Earlier in my career I used to get directly involved with that rarified atmosphere that exists between the design engineer and the manufacturing engineer.  THAT environment  was usually in the office of someone in Purchasing and it was always an adventure in product development.  

The designer would pour through his part-spec catalogs (I'm dating myself, they were actual books back then) and come up with something that was a truly elegant part for his design but vast overkill for what was needed and specify that part because the spec looked really good (never having actually touched one of the parts).  

The Manufacturing guy, who was looking at ease of assembly and overall product cost, would look at the spec and say "That'll cost a friggin fortune and kill the assembly cost, let's find somethin' cheap" but that would fluster the esteemed and elegant designer and we'd all end up at an impasse in the Purchasing guy's cubicle. Ultimately, a 3-way argument would ensue and we would end up calling "Ron, the guy from AMP".

Ahhh........ "Ron from AMP" (RfA).  AMP is a HUGE suppler of all sorts of connectors, terminators, lugs, you name it.  BIG name in the electronic, automotive and aircraft industries.  Ron's territory was the East Coast and he could handle that big an area because, back in the 1980's Ron had a small, dedicated staff at AMP corporate AND flew his own twin turbine Citation jet.  Yes, he sold so much stuff for AMP that they gave him a company jet.  In the 1980's.  Think about that.

So RfA would arrive (there were always at least several different issues we needed his help on), we would offer some of our better donuts and gourmet coffee, we'd show him what our design looked like, he would ask the designer a couple of questions, then ask the Manufacturing guy a couple of questions, then ask the Purchasing guy "Can y'all tell me how many, in what timeframe and what's yer sliding cost point?"  Then, invariably, he would open one of the Amp component books, turn quickly to a specific page, point to a part and spec line and say "That baby, rhat 'cher, that should do it for the volume y'all be buildin'."

Now, if you or I had done that, the designer and Manuf. guy would still be arguing to this day, but because RfA came up with it, it was the best part available in the world for that particular application.  Period.  

Moments of elation would erupt in the conference room, gourmet donuts (not really) would be passed around, much high-five-ing and hand-shaking and then the next small trio (usually including the same purchasing guy) would walk in with their specific problem to be solved, by the legendary "Ron from AMP."  

What a guy...

Gordon, I thought of you this morning when I saw this headline. 

A Tiny Screw Shows Why iPhones Won't Be 'Assembled in the U.S.A.'

NUT: "Tests of new versions of the computer were hamstrung because a 20-employee machine shop that Apple’s manufacturing contractor was relying on could produce at most 1,000 screws a day."

They needed 20 guys in a machine shop to make this one special frickin' screw?

All I could think then was, why would Apple not just alter the design spec to use one of the bazillions of commercially-available screws?

So when did Ron from AMP retire?

I would guess that RfA retired sometime around 1990 or so.  All he really needed, at his sales volume, was ten years and could retire (with his jet) anywhere he wanted, with more money than God.  My guess would be a place in the lakes region of New Hampshire, a second place near Lane at Kiawah Island, and a third place on Grand Cayman (for the banks).  

I saw that Apple screw article - look at the back cover screws for any MacBook laptop and voilá!  A slightly larger “tiny screw”.  While most screws are made by rolling a piece of round dowel stock between two dies, THAT one has a machined head and in relatively low volumes has to be finished singularly, either on a special CNC machine or by an operator.  That might explain the 20 operators and relatively low volume.  Real high volumes on that part would be made in a totally automated environment cranking out many thousands per day.

In the 1990’s, my company was growing by over 300% per year and we were ramping the volume of our products VERY quickly.  One thing we had to really watch was burning out our pre-production sources, like machine shops, as we went to full production.  Some shops could adapt and grow, others could not.  It was not unusual to visit a shop and write a contract that would have them making a pre-production part starting at, say, 50 per week and in six months we would be needing 1,000 per week and still ramping.  Some places wanted to stay small so we would start with them and they would work with a subsequent high volume vendor to transfer the part to them for production volumes.  We would then give the little guy the next pre-prod thing coming along and everyone was happy.  A few would take on the pre-prod part and do OK, but would choke on the higher volume so part quality or timely deliveries would suffer til we pulled the plug and went elsewhere.  Some of those marginal places still survived, some went away.  Others grew with us and did very, very well.

Anyway, any decent designer will check all available sources for an off-the-shelf part to satisfy his design and go to something custom as a last resort or suffer the wrath of his manager or the Manufacturing guys (or Program Manager, like me).  If it is a costly part it had better be REALLY needed to justify the inevitable increase in overall product cost. And then, of course, you need to find someone to make the thing and on and on...

—all of which speaks to having the kinds of design, product development and management chops that I guess I always assumed Apple had? 

I mean, just the idea of it—your flagship computer is chudded by capacity issues at a single Texas machine shop? 

The story just doesn't make sense, particularly the part where they say "well Americans don't have the skills to make special screws at high volume or to manage production schedules blah blah blah." 

Bullshit. That's what training, and industrial process engineering is for. It's management's job.

There was a guy around here a few years ago, runs a wire basket company, got a lot of mileage in the national business press out of saying some similar nonsense about not being able to get good machine operators. "Skills gap" and all that rot. 

Turns out he was offering like $16 an hour and the job required mastery of CNC machines, plus some old mills, PLUS—in order to hose down all the customers—full familiarity with various flavors of publishing software and the ability to write about as well as me (minus the snark, of course). Also: nights and weekends for answering their emails.

It's like Henry Ford never happened out there.

But back to Tim Cook, billionaire.

I mean, dude. You need a screw. Either have one of your minions redesign the back of the screen to accept an oval head (?!) or spec your absolutely crucial magic screw design out to, I dunno, other machine shops!

^ I regret that I have one "like" to give this comment.

"Jobs Americans don't want to do" is just shorthand for "Nobody wants to be my slave".

Here is what annoys me about automotive industrial design.  If someone had just allowed the little itty bitty part that you need to fu#k with to cost, oh say maybe, another 5 cents, so it could actually be handled by a human after the whole damn car was assembled, maybe I would not be so cynical.  and electric connectors: is there a law that says every one of these ever designed has to be impossible to take apart after its put together, and why does every one of the 100 or so that might be in any given car have to all be different, with no obvious method to take any of them apart?  Not to mention using materials that, again for just one more nickle, don't disintegrate after a few years under the hood. and yes I know that one more nickle times a million cars can amount to real money, but --- just no.  I would gladly pay the extra nickle, ,just sayin'

Sorry, off topic I guess . . .

I recently replaced a battery in an Apple iPhone - took special screwdriver called a Pentalobe screw driver.

Congrats, Greg. You could have left it apart and Tim Cook would've traded you a real '55 Speedster for that screw.

El Frazoo posted:

Here is what annoys me about automotive industrial design.  If someone had just allowed the little itty bitty part that you need to fu#k with to cost, oh say maybe, another 5 cents, so it could actually be handled by a human after the whole damn car was assembled, maybe I would not be so cynical.  and electric connectors: is there a law that says every one of these ever designed has to be impossible to take apart after its put together, and why does every one of the 100 or so that might be in any given car have to all be different, with no obvious method to take any of them apart?  Not to mention using materials that, again for just one more nickle, don't disintegrate after a few years under the hood. and yes I know that one more nickle times a million cars can amount to real money, but --- just no.  I would gladly pay the extra nickle, ,just sayin'

Sorry, off topic I guess . . .

$50,000 for that 1 million cars shouldn't keep them from using better parts but it does.

Ed, I totally agree about the Apple thing - it was the responsibility of the product introduction team to make sure the product intro didn't stall because of the want of a nit part.  Doesn't make any difference if it was for a phone, a computer, a toilet or a car, you have to be on top of all that stuff and make sure it ALL happens when it's supposed to or you're screwed (along with the product intro, as the article pointed out).  

On the fading away of machine shops (and many other skilled jobs in America), that'll take a decent pub and a beer or two to discuss.  I've seen both sides of the equation, and can support both points of view, but most on here might be bored reading through it.  Suffice to say it is a problem we (America) can fix, but a lot of people will have to learn new skill sets along that path.

Kelly, too true.  And how about having to remove four different big plastic shields in the nose of my wife's Outback (all held in with those little expanding mushroom plugs) just to replace a damn headlight bulb?  These things fail.  We all know it.  The silly designer(s) knew it.  So why did they make it a 45 minute job to replace a headlight bulb where the only access is through the wheel well?  And it's always the driver's side, too, so there is more going on there than meets the eye.  

Friggin designers - don't get me going....

IMHO, the passenger side bulb goes out anywhere from a week to a couple months after the driver's side dies.......every. single. time.

Not on this car.  Now running on our third driver side bulb and never yet replaced the passenger side.  Got me baffled, but I'm getting better at swapping bulbs.   

Although it was not at all obvious, at least not to me at first, but bulbs on the '03 Mazda 6 --may she rest in peace -- can in fact be R&R without moving or taking apart anything, but you really need some contortion skills to do it -- and small hands fingers.  there is nothing anywhere I could find that gave what you REALLY needed to know to make this happen.  And one contra-corollary to "If it aint broke don't fix it" might be: If you have to replace a headlight bulb, go ahead and replace the other one too at the same time. Because, well, like Danny said . . .

Sorry Gordon, that car is messing with you!

DannyP posted:

Sorry Gordon, that car is messing with you!

Maybe not.

A failed filament is pretty much just metal fatigue, no? And what causes metal fatigue?

Hot/cold cycling or vibration mostly?

There are a million things under the hood that could expose one lamp to more heat or vibration than the other. Is one closer to an exhaust manifold, radiator hose, alternator, power steering pump, A/C compressor, etc. etc?

Look around the lamp that fails with an engineer's eye and I'll bet you'll find the culprit, Gordon.

Possibly the new bulb isn't getting totally seated and is vibrating, just a little bit? There's your culprit!