There are many ways to fail

Whew! Glad that was an easy fix, El Frazoo! It's always the simple things that get ya. 

Glad you have a mechanic who didn't take advantage and turn it into bad battery, alternator, wiring, etc.

Glad you have it "fixed.

However, I am not a person to trust a butt splice in a cable that carries 300 amps on startup.  It might work just fine in warm weather, but will give you issues when the weather turns cold and the starter calls for twice as much current to turn your cold, dead motor over.  I butt spliced the "+" cable in Pearl and it simply wouldn't turn the starter over when everything was cold.

I would replace the entire length of ground cable with a new one.  NAPA sells them with the proper end connectors for your car and the right length (and no corrosion).  Get one of those installed (you can do this, yourself) before cold weather, eh?

On any car, and modern unibodies especially, I run 2 or 3 ground straps. A good idea people.

I just had the negative cable in my speedster disintegrate inside the terminal. Looked fine on the outside. 

El Frazoo posted:

...my car-guy intuition tells me that this corrosion thing ought to be a gradual process, with starting becoming harder and harder, and obvious low voltage to the starter motor...

 I'm guessing the corrosion itself was a gradual process - the car's 14 years old, right? But you might very well not see any difference in starting until there's no cable left.

If it broke while cranking, the rocking of the engine probably did the final deed.

I'm with Gordon, just replace the whole deal. If it's corroded in one place, there's probably more corrosion lurking, just waiting to bite you on a dark and stormy night.

And one cable every 14 years amortizes out to a pretty reasonable annual cost.

I bought new cables when I rebuilt/built it again, whatever I did....

New cables after 10 years is probably a good thing. On another note the impact fried my voltage regulator??? Which is under the shroud, in the alternator. No part of the engine was touched by the accident in any way. Alternator itself is fine, but had to replace the regulator.

I have thought about inspecting the nature of the splice they did, currently covered in loose electrical tape.  They said they used a "cable kit", not sure what that means, but assumed to be specific to battery cables.  It was something they had to order from their parts jobber. They did pay attention to the right gauge of the wire.  Further, they inquired with Mazda about replacing the cable, and their reply was that the negative ground cable from the battery merges with other wires into a mutli conductor harness of some description, and a new full-up harness is . . . . please everyone, sit down . . .  OK everyone safely seated??  . . . eleven hundred dollars, +/-.  Hence the splice kit.  If it were me doing it, I would solder that splice, and have a good heat shrink cover on it.  The guy mentioned soldering before we got started, but not sure how they ended up.  Will have to undo the tape and have a look, then decide.  If it is just a clamp, then I may redo it.

I did not use a "splice kit".  I visited the biggest commercial electrical contractor supply house in my area and used a commercial butt splice for 8ga cable (about 3/8" diameter cable core) on what's called "welding cable" for my battery +12V cable. That means that the cable itself is quite flexible and made up of many, many small strands.  The butt splice was a machined, extruded part with four stainless set screws to tighten against the cable strands on each cable end.  No soldering was involved (I don't have anything with enough heat for a cable that size, anyway).  Total length of the cable was about 8 feet.  

Anyway, I saw a significant current drop across the butt splice, such that, under the right circumstances, like a cold start at 40°F, the engine turned over a LOT slower than "normal".  That was evident almost from day one.  The ultimate cure for me was to go to a totally new +12V 8ga cable with no splice which I bought at a local truck fleet maintenance supplier at about 2 bucks per foot.

My only advice would be to drive it - It's probably gonna be fine.  If you DO have starting issues with it, especially in much colder weather, rather than buying a new, OEM harness I would get a new ground cable (be careful of any additional grounds ganged at the battery connector/clamp and get something similar) that will reach from the battery post to where-ever your current ground attaches, connect it at both ends and dress it off with suitable tie-wraps (or zip-ties, whatever) to make it looks nice.  Liberally spray the frame connection with paint or battery connector shield (looks like a paint crossed with parrafin) to protect it from future corrosion.

In the meantime, if it starts, drive it!

My curiosity will have me peeling the tape just to see what's what.  Meanwhile, the shop warranties their work for quite a while, and I may sell car before then. On the other hand, car drives well despite its ~115K mi, and you can't get the V6 anymore. Nice torquey mill there, seems to run as well as ever. New models are "Skyactive" whatever that means.  Part of what that means is they have a turbo 2.0 l 4 banger in there that is supposed to feel like the V6.  Well it doesn't, quite. Six spd tranny vs 5; even more plastic and thin metal, so is a bit lighter; and probably some tweaks to the EFI too.  Trying to get the fleet mileage up.  About two years ago, they said they were 6 months out to introducing their race-proven diesel engine to the US.  Been selling in Europe for a while now.  Every few months they would press release about how they were almost there, just a few more tweaks and then it would be ready for the US market.  Well, actually no, not so.  Maybe the VW diesel fiasco put a damper on all of that. Is it time for a Bimmer??

2017 Mazda 6 comes only with a 2.5L at this time - no turbo (yet).  Only model I'm aware of with turbo is the new (introduced late last year) CX-9 - their biggest SUV.  The 6 is one of only sedans (outside of BMW and VW ?) that you can still get with a 6 speed manual or the 6 spd auto with manual shift feature.

I have a 2016 MX-5 with the 2.0L (no turbo available) that produces 148 hp.  The 2.5L in the 6 and several other models is 184 hp but a whole lot more weight than the MX-5.  The CX-9's turbo is a nice 250 hp but 4300 lbs.  Despite Mazda's efforts at weight reduction, that's 2x weight of MX-5 (but it seats 7 and not 2).

Hopefully, the MazdaSpeed models will be back soon!  I like the new CX-5 mid SUV with the 2.5L non-turbo and only 3,500 lbs to pull around.

Zoom-Zoom!

Possibly I did not hear right, or the guy said a fib.  Anyway, I swear he said the engine had a turbo.  But it seems that what they have is 2.5 l, 13:1 and direct inject regular gas.  That said, I found it a little lacking in zip, although he swore it performed "just like the six".  Well, I don't think so, since in fact I do have a six.  I'll check, but think it runs out at 230 hp or thereabouts. Time for a Bimmer??

Buy a BMW, and be prepared for love/hate. 

... tell me about the hate part.  Lane seems to love his unequivocally.

I've had two, and both have turned out to be extremely reliable as well as a helluva lot of fun to drive.  I traded the first after ~136k absolutely reliable miles.  I have ~76k on this one, including 6-8 track days.  I did have one potentially expensive failure (VANOS system, the variable valve timing control), but the extended warranty we purchased covered it.  Rock solid otherwise.

Of course I tend to buy low-end models without a lot of fancy-shmancy options.

German universities apparently produce much better MEs than EEs. The electronics in my 540i did not age particularly well-- and by, "not particularly well", I mean, "the entire reason I sold the car".

Mechanically, dynamically, and ergonomically the best vehicle I will ever own. That's the "love" part.

I had an ABS module that could not be fixed for a love or money. I shelled a water-cooled alternator ($1500. No kidding).  Every last digital component on the car was ridiculously expensive, laughably fragile, and horribly unreliable. That was the "hate" part.

Stan, My man, what year production are you talking about?  and how much does the extended warranty cost, and how far does it go these days??

A water-cooled alternator, eh? 

My 540i was an E39 (2000, I think), which was a "holy grail" car (kind've like a 993 in "P-car" circles), but this was 10 years ago, when it wasn't so out-of-date.

Lane's report of a potentially very expensive controller failing within the warranty period kind've reenforces my opinion of the lack of ruggedness inherent with German electronics, at least as it pertains to the Bavarian Motor Werks. Yes, the expensive gizmo was covered, but I guess I expect stuff to work past the warranty period, and given my past track record-- I don't believe it would.

The really surreal thing is that my wife drives a Chrysler Town and Country minivan. I was terrified of the sheer number of gizmos on this thing enough to buy an extended warranty. Aside from a couple of very minor issues, the car has never been in the shop in 5+ years and 85K mi. It's kind've the anti-BMW: it aspires to nothing and has exceeded every expectation for reliability, as opposed to aspiring to greatness and being let down by stuff Asians (and North Americans) seem to be able to handle as a matter of course.

Don't get me wrong-- I LOVED my BMW. I just always got the feeling it was waiting to let me down.  

Should have bought the BMW last year when it still had the 4 year/50k B to B maintenance free warranty.  Only 3 years and 36 months now and excludes once included brake pads, brake rotors, wiper blades, engine belts or manual transmission clutches. And then there are the run flat tires

" I LOVED my BMW. I just always got the feeling it was waiting to let me down.  "

You know, that kinda describes how I view my Speedster.  And it has not disappointed me; every time I get it sorta right and going well, by damn, it surely does disappoint me.  Pretty much all has been documented here at one time or another.  Biggest single issue: F-ing Webers. In fairness, many of my issues turned out to have root causes best described as "User Error", but still . . .

So maybe here is another data point, German auto engineering at its not so best, Audi this time not BMW.  Sparing you the details, I had two four door sedans (I forget the designation of these models) , and they were acquired used, not new, and that may have been most of the problem.  My experience with those two cars ('80s vintage, I think) made me swear that I would never again put one dime toward an Audi.  Ever.

Our Mini Cooper (engineered by BMW) has some maddening quirks thanks to Bavarian cyber geeks who had way too much time on their hands. But maybe not enough time to get the details right.

The electric windows - which are controlled by microprocessors - go up when you push the 'up' button - about half of the time. The other half of the time, they go down. They will also sometimes go up when you push the 'down' button. The 'fix' is to have the car's entire computer system reset by the dealer (Big Bucks). Or, I guess you could just buy a Toyota.

The 'hill assist' - which, ironically, is needed only by those who can't drive a stick shift - can't be turned off. So, if you're in gear on a slope and maybe want to roll back a few feet on purpose, you can't.

The headlights are not turned on by the headlight switch, but by a microprocessor that receives a command from the headlight switch - after the microprocessor processes your request to turn on the headlights and thinks it over for a few seconds.

But the most maddening cyber overindulgence of all is what you have to go through to spring forward or fall back the clock twice a year. Our Toyota clock has two labeled buttons - hours and minutes. A chimpanzee could reset the Toyota clock in two seconds (not that I've actually witnessed this, but I'm pretty sure.)

Not so, the Mini Cooper. The clock setting is controlled by the car's central computer system - along with about seven hundred other functions, all of which are controlled by the same unlabeled button. That's right - seven hundred different settings, one button. You navigate through all of the seven hundred settings in 16 different menus and 42 sub-menus by pressing or holding down that one damned button in the right sequence of clicks and pauses. Get down to, say, the third sub-menu and hold the button down when you should have only clicked it and you've got to start all over again.

I know now. Every fall and spring, when it's that dreaded Saturday before the dreaded Sunday, I set aside at least a half hour for resetting the clock in the Mini.

Exactly.

Well fu#* that. No Minis for me either.  That would drive me insane.

... what?? a water cooled alternator?? Oh my.  Sounds kinda like a good idea, technically, but how executed?  And Mucho dinero, no doubt.  Better (and more complex) being the enemy of the good (simpler) here too?

El Frazoo posted:

... what?? a water cooled alternator?? Oh my.  Sounds kinda like a good idea, technically, but how executed?  And Mucho dinero, no doubt.  Better (and more complex) being the enemy of the good (simpler) here too?

Yes. If it's electric, the boyz in Munich cannot embrace the KISS principle, for it would be beneath them. There's never been a function too insignificant or simple not to solve by use of some ridiculously complicated, Rube Goldberg arrangement of year and model specific logic board. One module shall rule them all.

In my (limited) experience:

The Asians: build electronics that will keep reliably computing long after the rest of their vehicles have oxidized and returned to base elements (10 years or so in a salty environment). The ECU will still be computing away when roaches are the only other life left on this ball.

The Americans: build vastly underrated, robust vehicles, with dead-nuts reliable electronics and engines and suspensions overbuilt by a factor of 2x. The engine will run 200k mi-- but the cup-holders will fall off and the headliner will fall down about the same time the paint on the hood oxidizes into caulk (or, roughly when an Asian car begins to rust in half). The transmission will either go for 500k mi or will begin to slip and hunt as you pull out of the showroom (depending on whether or not your name is Jim Ignacio).

The Germans: build cars that want to be driven 12 hours a day, at 120 mph. They eat miles like fat kids eat donuts, and have available roughly 3x the torque required for any given situation. They chafe at your plebeian commute, and will punish you by throwing a code for the airbags or antilock system every 20k mi or so. They laugh at your petty problems. If you can't afford a $3000 electronic window raising module, then perhaps you should just not touch the buttons.

Then there's us. We'd just like to keep the hubcaps on the cars we love.

Actually the VANOS failure was mechanical or hydraulic - I forget which - and it occurred between 60-65k miles.  Still, that doesn't excuse it.  On the other hand, everyone I know, has had some sort of potentially expensive failure in a recent model car, regardless of manufacturer.  My previous car was an '04 Acura (Honda) TSX, and it suffered a catastrophic failure of the A/C compressor at 70k miles, rendering the car undriveable because the the destroyed serpentine belt.  The entire A/C system had to be replaced at a cost of ~$4k.  This was a known problem at roughly the same mileage on every Honda product that used that compressor.  I knew three people with Honda CRVs that had the same issue.  Despite all of the bad press online Honda refused to do a recall.  I copied a bunch of that stuff to them and mentioned that I was a repeat Acura buyer, after which they paid for all but $1200 of the repair.  My friends with CRVs weren't as lucky as Honda refused to help them out.  I'm not a believer in the "Japanese reliability is the best" hype.  The most troublesome new car I've ever owned was an '83 Celica GTS.  I don't have the time or energy to detail all the issues that car had.

One more thing...  The most reliable car I know of is a friend's '06 911 Carrera.  He bought it used and has done a few mods to it.  After many track days and one highway accident the only reliability issue that I know of at nearly 100k miles was a failure of the shift cables.  It was not an overly expensive fix.

Just about all cars built today are worlds more reliable than the ones of two decades ago. But I think they may have peaked around 2005 or 2010. There seems to be a sweet spot after which more electronics do not improve reliability. I believe that spot is just before the point of universal drive-by-wire throttle controls. 

I have nothing to back that up but a queasy feeling.

The urge to digitize all functions does not seem to be based on utility or reliability. It appears to be an indication of power within the organization. When MEs had the power, ME solutions were dominant. When EEs and CS post-docs hold sway, you get, well...you get one button with elevendy-billion submenus. 

It is The March Of Progress.

The sad part is that while electro-mechanical components can be repaired/replaced relatively cheaply, digital systems are almost universally unrepairable, requiring replacement at high cost.  EM systems are likely to be repairable or replaceable for years.  Digital components are unlikely to produced by the aftermarket and will eventually become unobtainable.

Drive by wire: the '03 Mazda 6 has this and I hate it.  Very spongy, really no such thing as a throttle blip.  On an automatic, may be OK, but I have a 5 spd and enjoy some sporty moments now and then.  You get used to it, but for the first few weeks/months, it is annoying as all get out, and you are constantly stalling out on take off.  Salesman even warned me during my third stall during a test drive: yeah, everyone does that.  And when you really get on it, and need to throw a quick shift after winding it out toward the red line, it does not back off with your foot, so it wants to over rev during that quick shift.  You know what you are doing, but the throttle does not get the message for a while, and makes the whole thing just not work very well. Compared to the Speester, is night vs day.  I hear you can get a mod for the ECU that allows for a tighter feedback loop and quicker response.  Pricey though.

Hmmm...  I have DBW on the BMW and it's quite responsive, almost to much so in sport mode.

Makes a simple throttle cable seem both quaint and desirable, doesn't it?

"The sad part is that while electro-mechanical components can be repaired/replaced relatively cheaply, digital systems are almost universally unrepairable, requiring replacement at high cost."

Well, yes and no.

When I was working, we designed and built full-blown computer systems, roughly the size of a kitchen refrigerator, that each contained 24, 18" square computer boards, 164 disk drives, two completely redundant power systems with battery backup, along with operator consoles and other ancillary stuff.  ANY part of the system, digital or otherwise, could be removed and replaced in less than 5 minutes, while it was running, with no interruption of operation in any way.  I described that ability in car racing terms for customers:  "It's like you're driving a Formula 1 race car on the race track at full speed - 200 miles per hour - and a service truck pulls up alongside, pops off the driver's helmet and gives him a brain transplant, then replaces his helmet, all before reaching the next turn".  

If something is designed to be serviced, however unlikely the occurrence, then is is not costly to do so.  Unfortunately, very few people in the car manufacturing world seem to get this.

"get this"??  I will say none do. its all about saving two motions on the ass'y line and making it easier/faster on production, service be damned.

Lane,

I'm 100% in agreement regarding overblown hype regarding Japanese cars' reliability. Mechanically, they are generally spindly and under-built.

The electronics, however are ridiculously robust. 

Stan Galat, '05 IM, 2276, Nowhere, USA posted:

Lane,

I'm 100% in agreement regarding overblown hype regarding Japanese cars' reliability. Mechanically, they are generally spindly and under-built.

The electronics, however are ridiculously robust. 

BUDDY had a BIG FAT LEXUS...car would NOT start, NOT turnover even with JUMP. Was STUCK and COULD NOT be put into neutral. He had to manually  REMOVE the computer to get in neutral and 5 guys had to push it out of his garage and thru 100 feet of high grass to get to where the AAA flatbed could load it! My 1964 Volvo Amazon had only 3 wires, made 300,000 miles and NEVER gave an ounce of trouble and cost LESS than the faulty LEXUS computer chip!

Gordon,

I used to do quite a bit of work on the Liebert units that cooled those bad boys. Even the AC units were overbuilt and redundant.

No more. The smart phone I'm composing this on (using voice recognition technology, over a cellular network) probably has more computing power than that refrigerator sized machine. This one cost me $200 because it's three generations old.

The difference between what we've got now, and those old mainframes is more than serviceability. It's reliability: nobody anywhere is thinking about a service life beyond about five years out. I just bought a new computer, and would like to run software from 2010. Talking to the Tech Support guys, you would think I was asking them how to tan the hides of animals I killed with a spear using my own urine.

I'm fully aware I'm retrogrouch, but when it comes to microprocessors "more" is very often not better, it's just "more". I'm not sure where that line is, but I'm pretty sure we are well past it when a microprocessor controls how the windows go up and down.

edsnova posted:

 

...The urge to digitize all functions does not seem to be based on utility or reliability. It appears to be an indication of power within the organization. When MEs had the power, ME solutions were dominant. When EEs and CS post-docs hold sway, you get, well...you get one button with elevendy-billion submenus...

Methinks neither ME's nor EE's have the power that bean counters do today.

I think what's driving this move to computerizing everything is cost. Copper wire, decent quality switchgear, and mechanical linkages all cost a lot more than what's replacing them.

I'm no expert here, but apparently the 'bus architecture' that a digital control system uses can run more devices with fewer wires, and the long runs can be lighter gauge wire. The switches for sending digital signals can be rated at a hundredth the current capacity required for old school circuits. And fly-by-wire requires no heavy cables, bellcranks, or any constraints in the physical design of where controls are routed.

Doesn't have the 'feel' of the old system? Well, shoot, our focus groups tell us 92.3 per cent of the target demographic won't notice the difference. You want to double-clutch downshift? What the heck is that?

You old curmudgeons will just have to change with the times.