By on June 18, 2012

Third in a series, the format of which I described in the introductory remarks of Part One, this article will focus on specific modern technologies—not at all egregious in themselves—that have, in my opinion, been applied and utilized in a most egregious fashion in the modern automobile. One is a direct continuation from Part Two, although covering a new area. The other one is a component technology within one of the technological areas criticized in Part Two. Tally Ho—to THE BIN we go!!

The Plastic Exhaust Manifold

Don’t think for a minute that it hasn’t been considered, attempted… or had the proverbial towel thrown in on the idea, yet.

As with the overuse of the technologies already covered, the use of plastics in the construction of the automobile is not exempt from criticism.

Plastic, of course, has played a major role in component composition since back when Henry Ford got the ball rolling with the Soy-based “Bakelite”—developed as a light, durable, versatile and attractive alternative to other materials used up until that time.

Plastic has been used to excellent—sometimes surprisingly so—effect in ignition systems, fuel systems, exterior body panels and trim, and a wide variety of interior fitments.

While manufacturers have continued to push the development envelope of plastics, which is not unreasonable, they have also pushed the production envelope too. This has led to some noteworthy disasters in the past, and it continues to create problems for support networks, customers, and independent repair facilities.

A few of my faves:

  • Plastic FUEL INJECTION DISTRIBUTION RAILS, that leaked and caused fires (The same company, many years earlier, put plastic-bodied CARBURETORS into production. NOT surprisingly, they had durability issues, as well.).
  • Plastic-bodied STEERING BALL JOINTS. Do I need to say that I never saw a replacement part for these joints with plastic bodies?
  • Plastic composite INTAKE MANIFOLD GASKETS. I’m referring here to engine designs wherein the gasket has to seal major coolant passages.

From my experience, these gaskets last anywhere from 30 to 50,000 miles, whereupon the plastic media where the urethane seal is inserted, deteriorates to the point of causing internal and external coolant leaks. The gasket design these gaskets replaced would usually last at least as long as the entire engine!

Receiving dishonorable mentions:

Plastic Intake Manifolds, Valve Covers, Radiators (and other Cooling System components), Air Cleaner Housings—pretty much anything made of plastic and used in an environment where high heat and pressure are present.

It’s not that these parts don’t work for a while, just not nearly as long as the designs they replace; which doesn’t sound like progress in the right direction to me.

What’s a little more disconcerting is when it appears that plastic is the RIGHT choice for a particular component, but the WRONG compound is used, thereby fast-tracking the whole matter into the realm of the egregious.

A good example of this is the plastic used in underhood MULTI-PIN ELECTRICAL CONNECTORS. I have experienced that, after only a few short years of normal operation, these connectors just don’t want to separate when impressed to do so. No doubt, there are a few contributing factors to this phenomenon, but the ultimate “deal-killer” for a clean separation without damage are the prematurely aged and brittle plastic connector housings.

The “Self-Fusing” Electrical Connector: PLASTIC—NOT SO FANTASTIC!

The Flash Memory Computer Program

I’m not here referring to the way computerized systems are being programmed, although I do have not a few bones to pick in this area (think Traction Control operation on modern All-Wheel Drive vehicles with Throttle-By-Wire, and other such offenses). What I’m speaking of is the use of “Flashable Memory” Program Modules—Electronic Control Units that require the uploading of program information after assembly (and often after installation), as opposed to being assembled with Program “Read-Only” Modules, or PROM’s.

My gripes are thus: 1) There seem to be a lot more actual program failures with these units, 2) Diagnosis of this type of failure is more difficult for the independent repair shop, 3) requiring considerably more expensive equipment to successfully “Reflash” the program and effect a complete repair, and 4) is more costly, in terms of both time and money—especially if it’s a “Dealer Only” type of repair, and there are no dealers in the immediate area. That’s enough to qualify for “Egregious”, in my book!

In the days of the PROM, I don’t recall seeing very many computer failures actually involving a “corrupted” PROM. Sure, I saw failures involving over-voltage issues, which led to the destruction of other internal components within the ECU. I saw bad soldered connections, degradation of circuit boards, tweaked or otherwise damaged pins in the ECU’s multi-pin connector, or even water damage (which was often repairable by cleaning and drying the circuit board after removing the covers).

Every now and then, I’d get a vehicle with some annoying, if not somewhat transient driveability problem that I could, after checking for PROM updates, solve by replacing the PROM with the updated unit. Just remove the ECU, remove an access panel on the ECU, and carefully unplug the old PROM and plug in the new one! But never a really serious issue caused by a “Failed” PROM.

With the Flash Memory Program Module, it has been a different story.

A classic example involved an ’02 RAM (its funny how even the name has taken on a “Computer-Age” meaning) Pickup.

All of my diagnostic information told me the problem was in the Park-Neutral Switch—now installed deep within the bowels of the transmission, for technician and customer “convenience”. Impressing my trusted transmission shop experts for a second opinion, my findings could not be confirmed. Apparently, their findings varied enough that they could not be certain of WHAT the actual problem was.

We both concluded that it was a programming issue, which neither of us had the capability to correct. We sent the customer to the dealer, where the Engine/Transmission management ECU was reflashed with an updated program, and VOILA: everything back to normal! What a profit-less hassle!

We’ve all had to come to terms with program updates for our personal computerized devices. While keeping them updated and running can be a hassle, the manufacturers have been able to keep the cost-to-benefit ratio within somewhat acceptable tolerances. The fact that the individual owner can affect most of this process while online in their own home or office, without additional gear or expense, is one of the main factors keeping it all tolerable.

Wish I could say the same for our modern Automobiles.

Stay tuned for Part Four…

Phil Coconis is the owner of a West Coast independent auto repair shop.

Get the latest TTAC e-Newsletter!

36 Comments on “Memoirs Of An Independent Auto Repair Shop Owner: World’s Most Notorious Automotive Technologies, Bin Three...”


  • avatar
    Robert.Walter

    Flashable memory is necessary because it is not unusual for the initially released software to have errors, or for there to be later changes made to coding to increase fuel Econ, reduce emissions, or improve drivability.

    A NSS installed inside the trans case may be a decent solution to protect the switch from a large variety of external environmental factors.

    • 0 avatar
      Athos Nobile

      The funny part is that he doesn’t complain about having to remove the ECU to install a new chip in it, which does essentially the same as reflashing it.

      Although I like GM’s old MEMCAL system, it is more convenient to have something that can be relatively easily updated with either a laptop or a Tec2 or Rotunda or whatever name the service tool has. And modern cars have so many computers that the PROM system just won’t cut it anymore.

      Sadly, the independents are left outside.

    • 0 avatar
      Phil Coconis

      As I said in article, I never personally encountered a PROM with errors; and by the time the monitored system could benefit from an updated PROM, they were available.
      I also said that the problem with the trans op in the RAM pickup was NOT the P-R-N switch, merely the flashed computer program that operated the trans.

  • avatar
    Robert Gordon

    “…pretty much anything made of plastic and used in an environment where high heat and pressure are present.”

    So you take issue with headlamp reflectors then? Strange, they seem very robust to me.

    • 0 avatar
      mikedt

      I’m guessing a headlight reflector doesn’t reach the temp of an engine block, nor does it have to deal with any pressure above atmospheric.

      • 0 avatar
        Robert Gordon

        Reflector temperatures are significantly higher than average block temperatures and experience significant pressurisation when in use.

        Small driving lamps are even more extreme and quite a feat of engineering.

      • 0 avatar
        Kevin Kluttz

        mikedt +1. And there is only atmospheric pressure inside a modern headlamp. Otherwise, fogged lenses would be a constant problem. Every headlamp unit I have had on every car I’ve had has a vent to equalize the pressure. Off track there, Robert. And they also do not have to deal with combustion gasses. (Gases?)

      • 0 avatar
        Robert Gordon

        “Every headlamp unit I have had on every car I’ve had has a vent to equalize the pressure.”

        You assume they are there to equalise pressure, they are not. A typical headlamp has two or more vents which are placed at locations to create a specific convection pattern within the lamp with the aim of moving moisture from the area adjacent to the reflector to less critical areas of the lamps. In order to do this the lamp must be at higher internal pressure than ambient -a feat that is effected with heat from the bulb. One is never trying to equalise pressure in the lamp, merely reduce it slightly locally. Pressure per-se is never a problem in lamps. As a demonstration of this one of the tests of a headlamp to test the integrity of the glue track is to pressurise it to 25 psi with all vents blocked. A maximum leak rate then determines a pass or fail.

        “And they also do not have to deal with combustion gasses. (Gases?)”

        Have you ever studied the temperature gradient on cylinder wall? Thought not. It’s not that dissimilar to a headlamp. remember a filament is heated to 2500-3000 deg C

    • 0 avatar
      LeadHead

      Most headlight housings I’ve seen are vented. I also have not seen one headlight housing that didn’t start UV degrading and turn cloudy.

      • 0 avatar
        Robert Gordon

        Headlamp housings are vented primarily as an anti-moisture measure – many employing a one-way breathable scrim and a cap to further aid this.

        Headlamp housings are not usually exposed to UV however they are usually constructed from a UV stabilised grade of polypropelene.

        However I presume you were talking about the lens with respect to cloudiness. Your statement “I also have not seen one headlight housing that didn’t start UV degrading and turn cloudy.” is grossly hyperbolic. Polycarbonate lenses employ a resin hard-coat that prevents the UV deterioration. In early iterations of PC lenses the hard-coat process was less durable than it might have been. Basically it used to be dripped on like treacle which led to uneven coverage on the lens. Most headlamp manufacturers now use a spray system which provides a much more even coverage.

        If the hard-coat is mechanically removed for example with abrasive polish then it will allow the lens to yellow. This is why it is important never to clean a Polycarbonate lens with anything but mild detergent and water.

    • 0 avatar
      Phil Coconis

      Are you SERIOUS, Robert?!! I have seen MANY a plastic headlamp housing with a degraded reflectorized coating! Not talking about the usual and customary “fogged” lens area, either. More egregiousness!!

      • 0 avatar
        Robert Gordon

        “Are you SERIOUS, Robert?!! I have seen MANY a plastic headlamp housing with a degraded reflectorized coating! Not talking about the usual and customary “fogged” lens area, either. More egregiousness!!”

        How many have you seen and what percentage failure does that represent? Did you perform a rigourous root cause analysis on them or merely assume the failure mode?

        For what it is worth, headlamp housings do not receive a coating in any case the material used to make them, polypropelene does not coat well. It might help you to understand a bit of headlamp anatomy. Modern headlamps are made of four main moulded components as well as numerous assorted adjuster mechanisms, caps, springs bulbs and connectors indicator lenses etc. The four moulded parts are 1) Lens made from PolyCarbonate (PC) with hard coat. 2) Reflector – Made from Polyester Dough-moulding Compound. Unlike the other polymer components this is a thermoset and extremely heat resistant. Once moulded into the reflector shape, a coating may or may not be applied (depending on the manufacturer) to the reflector surface to improve the surface quality of it. The reflector is then metallised with an aluminium coating in a process known as vacuum deposition. 3)Housing – made from polypropelene it is the part that bolts into the car. The reflector nests insode the housing and the lens is glued into the glue track around the periphery of the housing. 4) Bezel – this fits over the reflector and form both a decorative function and sometimes a photometric function in preventing light bled at incorrect angles. This may be finished in a number of ways including metallising depending on the designer’s whim. A metallised surface is usually avoided where possible as it makes part handling more critical.

  • avatar
    rporter

    I first heard of the new use for plastic when I got a code at 30000 miles. Dealer after couple hundred dollars said part of the problem was a plastic intake manifold. Ok for headlights perhaps not engine.

  • avatar
    Athos Nobile

    I’ve already seen 2 part welded stamped steel exhaust manifolds.

    Regarding the connectors, been there, it sucks.

    I also hate those plastic radiator tanks, once you remove them to change the gaskets because they’re leaking, they can’t be resealed properly… so in comes a new part.

  • avatar
    cdotson

    The “flash” memory type required for read/write programming to the ECUs has a finite write cycle life. I’d bet that a lot of the memory sectors are frequently written to with new variables during operation and when that sector dies (goes bad from too many write cycles) the whole program quits working properly. The ultimate failure mode would be random as in you won’t be able to tell what will go wrong, but once it goes wrong it would be a fairly consistent occurrence. Rewriting the entire code to the unit would randomize the allocation of the memory sectors, and the installation routine may be able to identify and avoid “bad” sectors, allowing the program to run properly again.

    It kinda stinks, but it’s one of the trade-offs made to allow the electronic read/write. Ultimately it’s usually cheaper and easier in the long run to buy the equipment needed to do this than to have to continue paying to manufacture physical chips to update or change programs.

    • 0 avatar
      BigOldChryslers

      The flash memory that the OP is complaining about contains the program that the computer runs, not operating variables. These are usually stored in a RAM which is maintained by the vehicle’s battery. If power is disconnected then the computer has to re-learn the settings which are stored in RAM.

      There are undoubtedly some settings that are programmed at the factory and stored in a section of the flash. These may need to be recalibrated occasionally, such as if your transmission is overhauled or a sensor replaced. Flash memories are partitioned into separate areas for code and data storage. Re-flashing the ECU will not “randomize” the memory usage as you say.

      Even if the ECU does alter some variables that are stored in flash instead of RAM memory, flash memory today typically has 100,000 write cycles or more before failure. If the entire flash was rewritten 10x per day, it would take almost 30 years before a bit died. A robust memory manager would use checksums to detect defects in the data memory and mark the bad sections so they would not be used, extending the life of the flash further.

  • avatar
    Crabspirits

    GM 3800 V6:
    If it wasn’t for the stupid plastic coolant “L” and composite gaskets, this would have been the most reliable engine ever made.

    • 0 avatar
      DubTee1480

      Haha, I keep spares in my glovebox and will leave them there for the next owner when I sell the car.

    • 0 avatar
      ajla

      The L36 lived off the reputation of the LN3, LG3, and L27.

      Honestly, I consider the Series II to be the least durable of the Buick V6 family, even taking into account the poorly oiled RWD 231s.

      It’s still better than anything Chevrolet ever came up with though.

  • avatar
    Educator(of teachers)Dan

    Plastic components? Blame the push for profit and weight savings for fuel efficency.

    Flash programs? Blame the increased computerisation of cars. More things that could go wrong because of computer programs, ergo those computer programs need to be easier to fix.

    Yeah I’m starting to agree with “Right To Repair” laws.

    • 0 avatar
      psarhjinian

      It’s funny, though, that with increased computerization, while the theoretical number of things that can go wrong has increased, the actual likelihood of any one problem occurring has gone down.

      It’s also (potentially) much easier to get information on why things have gone awry: you can more sensors, and more powerful tools for much less money to monitor those same sensors.

      There is something a little funny in embedded computing circles vis a vis openness, though. Imagine if you had something like SNMP (for reading back data) and TFTP (for pushing updates) in a car? These standards are not new, nor is it particularly difficult to implement them, but commoditization and openness are really uncomfortable concepts for embedded systems designers, and they really ought not to be, not in an era where silicon is so cheap.

      • 0 avatar
        Educator(of teachers)Dan

        I agree cars are more reliable than they’ve ever been, even with lowest price, cheaper than cheap, components. But now that less things go wrong it is harder to diagnose those things that do go wrong, unless you are a dealership mechanic.

        Recently the AC in my 2004 F150 Heritage stopped working (suddenly, not a gradual decrease in performance that would indicate a leak) and I realized that when I switch on the AC I can’t here the compressor engaging. Back in the pre-computer days it would be something obvious like a fuse or the electromagnetic clutch that engages the compressor would be out or something like that. Todays cars however there is a complex system of relays under the hood and of course in many modern systems computer programs that tell certain vents to open and close or when the MAX/AC can engage… ect. Visiting the stealership should be wonderful. (eye roll)

      • 0 avatar
        DubTee1480

        @ Dan: Replace your AC switch first, should be no more than $20 and worth a shot before taking it in. Should be on the AC line with a two wire pigtail going to it. When the switch goes bad, it sees it as low coolant pressure and will not allow the clutch to engage. Did you by chance notice the clutch engaging and disengaging frequently before it went out?

      • 0 avatar
        Educator(of teachers)Dan

        Actually I did not notice that, but the truck is also a standard cab which means the AC never had to work very hard with a small square footage to cool. Thanks for the tip. I had already checked the fuses (in cab) because a few years back the fuse blew (never repeated) that controlled the AC. I attempted to open the relay box under the hood but it seemed to require the sort of force that would break the lid, which of course needs to be water-tight for obvious reasons. There is a relay in that box that is connected to nothing but the AC. (FYI the AC compressor is the only thing not working. Ventilation is just fine. Thankfully I live in a low humidity state.)

      • 0 avatar
        theirishscion

        “There is something a little funny in embedded computing circles vis a vis openness, though. Imagine if you had something like SNMP (for reading back data) and TFTP (for pushing updates) in a car?”

        Having spent quite a bit of time around embedded systems designers in the past, I’ve felt much the same thing about car systems. They do tend towards a very closed-system mentality, and there are reasonable, bordering on good, reasons for that (SNMP and TFTP servers may be relatively trivial to implement on a non-realtime system with effectively unconstrained memory, storage and CPU cycles available, but the same is simply not true of many truly embedded systems. That’s changing, but quite slowly. There simply isn’t the Moore’s Law-like drive for bigger-better-faster-more like there is in the rest of the IT industry) but I suspect that as time goes by, that closed nature is going to become more company policy/sop to the dealer network and less recalcitrant developers. Silicone may be cheap but manpower and testing are most certainly not, and the liability issues around automotive safety are nothing to be trifled with.

        Put more simply, I dream of the day that I can SSH into my car and read the live data and log files in a reasonably well documented form, but I’m not holding my breath until some serious legislative action takes place. The ability for me to SSH in opens the possibility for someone else to do so as well, what’s stopping the kid at the quick-e lube leaving a little malicious code in my car that allows his dumb little buddies to easily defeat the anti-theft system later in the month? What’s stopping that badly written code from backfiring and stalling my car as I try to make a left turn across traffic? What’s stopping me from deciding to sabotage my ECU just before my extended warranty runs out so I can get a new engine? It’s a Pandora’s Box alright. One I still want desperately to open, but I can see why it’s still being kept locked up.

  • avatar
    noxioux

    Although it was just a small mention in the article, I’d like to reinforce the author’s dislike of self-fusing connectors. Fusible links are the work of SATAN.

  • avatar
    schmitt trigger

    Fusible links are there to appease insurance companies…. with so much wiring, the opportunity for an electrical fire increased so much that automakers installed them.
    But of course, when one fails, it is pure hell.

  • avatar
    Les

    The complaint about Flash memory reminds me of moaning and kvetching I’d read on some gaming forums when the original X-Box was in the pipeline and rumors were floating that it would have a hard-drive.

    “It’ll be brilliant.”

    “NOOOOOOO!”

    “What? But we won’t have to fiddle with memory-cards anymore.”

    “NOOOOOOO!”

    “We can install game-discs so we won’t have Loads and Loads of Loading.”

    “NOOOOOOO!”

    “Okay smartarse, what’s so bad about a game console with a hard-drive?”

    “Don’t you see you damned fool? If the system has a hard-drive they’ll start releasing games before they’ve properly finished their beta-testing cycle, it’ll become acceptable to release games with glaring programming errors and game-breaking bugs because those will now all be fixable with Patches.. just like they do with PC games.”

    And damned if the curmudgeons weren’t right. :P

  • avatar
    Landcrusher

    PROMs work. Software can usually work or be updated.

    All sorts of firmware and programmable crap in between is to be avoided if possible. It would be cheaper to run a Linux PC on cars today, but they don’t. The reasons cannot be both rational and customer friendly. They just can’t.

    • 0 avatar
      Les

      Linux is open-source, and you gotta keep all the bits and bytes of code running all the electronics inside the car proprietary to maintain competitive advantage…

      ..you just Gotta. ;p

  • avatar
    robc123

    Why can’t anyone in the entire world make an aluminum bodied simple but classically styled car?

    One with roll up windows, old school heat to cool slider, and a simple efficient engine with no plastic and a cardboard glove box for a decent price.

    little 4 or 2 seater, 2 or 4 doors, doesn’t have to 0-60 in under 6 sec. like a simple fiat 500 with a couple of air bags.

  • avatar
    pistachiojoe

    Wow! It’s like you are writing about my 2001 Malibu while protecting the guilty. I had to get rid of it after the 4th intake manifold gasket was beginning to leak 2,000 miles in. The BCM was equally buggy with the passlock “security” leaving you on the side of the road randomly. Great article.

  • avatar
    Racecar spelled backwards is racecaR

    Underhood plastic not spec’ed to the conditions: My 2000 Audi 1.8t has destroyed 2 oil dipsticks. The dipstick itself is metal, but has an orange plastic handle, and is guided into the block via an orange plastic tube. The plastic becomes brittle with heat (and oil?) and breaks into little gravel-sized pills. Oh, and the soft-feel coating on interior plastic parts (door panels and shifter surround) peels off (VW, too). The damped ashtray linkage has snapped (was handy for holding change) and the passenger side headrest can’t be adusted. The plastic adjuster button broke off inside. Oh, and the “airbag” warning light has been lit for many years because the plastic electrical connecter underneath the passenger seat traps water inside and the contacts grow a layer of electrolytic fuzz which interrupts the warning circuit. Sigh.

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • AnalogMan: It’s still only available with an automatic transmission, which is disgraceful for a Corvette. Even...
  • golden2husky: Yeah, all the shortages are clearly Biden’s fault…./s. There’s plenty of things to...
  • dukeisduke: OTOH, it probably doesn’t take 50,000 gallons of water, and several hours, to extinguish.
  • dukeisduke: At first, I thought, “What is a heater intake grid relay?” Turns out it’s an...
  • golden2husky: I had the luck of driving a friend’s new C8 – 1200 miles on it. First time driving a dual...

New Car Research

Get a Free Dealer Quote

Who We Are

  • Adam Tonge
  • Bozi Tatarevic
  • Corey Lewis
  • Mark Baruth
  • Ronnie Schreiber