A Mechanic's Life - Tales From Under the Hood

Mr_Shiftright

Ok, so GM has a problem with multiple failures, not just one. Swell.

thecardoc3

Close. The cars can develop any number of problems that have the potential to result in the system disabling. That's why "The Problem" isn't a single answer that anyone can solve with a single fix. Every event has to be dealt with as it's own entity, and the practice of trying to dumb it down is counter productive to doing that efficiently.

stever

Seems like the issue is not identifying the problem in the first place (by the systems, not the techs). You may as well have the one idiot light that says "it's broken".

Cars are pretty primitive when you get right down to it.

I'm on a tear about garage door openers too. Torsion springs that rust and break and kill people. A noisy chain that runs to the middle of the garage (at least you can get somewhat quieter belts). Just stupid old tech.

Be a lot better to have two quiet little electric motors at each edge of the door that raise or lower the door to any position you like, and hold it there.

Of course, by the time they figure that out, car ownership will be passé and we'll be using that extra space for it's intended purpose - junk storage.

thecardoc3

stever said:

Seems like the issue is not identifying the problem in the first place (by the systems, not the techs). You may as well have the one idiot light that says "it's broken".

Cars are pretty primitive when you get right down to it.

Some of them is very basic, and there are some of them that took 21st century tech to make available.

stever said:

Be a lot better to have two quiet little electric motors at each edge of the door that raise or lower the door to any position you like, and hold it there.

Of course, by the time they figure that out, car ownership will be passé and we'll be using that extra space for it's intended purpose - junk storage.

You mean one of these....http://www.amarr.com/residential/product_detail/8500

stever

Better, but it still uses cables and a torsion bar. This is closer but still has compromises.

Mr_Shiftright

That's right. GM should just install a "Something's Wrong" light.

thecardoc3

Between this thread and the Stabilitrak thread there is a contradiction that keeps being played.

One statement wants the computerized systems to be even better at detecting faults, while the opposite is also being pushed in the idea that the system is flawed when it does detect a fault in any of the sub systems. If a system is made better at detecting faults, then that is what it will do, detect more faults. Are you sure the consumer wants that?

Fault tolerance and FMEM (Failure Management Effects Mode) otherwise known as "Scripted Failure Mode" are operational strategies that the systems employs when a failure is detected. When you see the system like Stabilitrak shut down with multiple warning lights FMEM is actually what is going on. The codes set in what ever module that is generating them are clues to the source of the issue for that given failure, but that is all they are, clues.

Lets take this to a single module perspective keeping in mind that the level of fault detection can change between different modules in a given system. A module runs tests on its inputs, outputs, and communication circuits. It knows what the voltages (sometimes current) should be at its connector pins when certain operations are being carried out. It also knows what the sensor voltage limits are and when certain failures occur that force those voltages outside of the operating range it responds by generating a trouble code. The same goes for the output circuits. The computer does not know why a given voltage falls outside of a given range it only knows that it does. The computer cannot test beyond its connector. If the software allows, the computer will usually substitute an expected value based on other inputs and attempt to continue to operate the system. If the software is written to not allow that for what-ever reason then the system shuts down or goes into a limp-in mode.

One of the reasons (there are more of course) you wont get computers to diagnose this is that the software isn't written to display signals that fall outside of reasonable operating ranges. On top of that, noise filtering tries to allow for minor defects in the vehicle without the system crashing. If you have an O2 sensor circuit that gets shorted to power, the scan tool only displays what ever the maximum input voltage that the software was written to display. If the normal operating range for an O2 sensor is 0v-1v, the software will usually not display above 1.2-1.5v on the scan tool. Meanwhile a voltmeter or oscilloscope connected to the circuit displays the full system voltage indicating the short to power revealing that the PCM and scan tool are effectively lying to the technician. The tech only has the code for the O2 sensor voltage too high. It's up to the technician to decide how to proceed and all of the noise on the outside about the cost to do diagnostics serves to discourage them from taking a disciplined approach to do just that. When you grasp this that goes a long way to explain why there is always enough time to diagnose something when it comes back a second (or more) time, but not enough the first time.

The Stabilitrak "System" relies on the PCM, TCM, ABS, SRS, TAC, BCM, and AWD, modules and each of them has its own way of being responsible for acquiring and communicating data and producing outputs. This complexity is repeated throughout the car with virtually anything that the driver may want or need to have operate. The tech has to sort through the obvious failures as well as the not so obvious ones and yes it takes decades to learn how to do that. But it never happens if that isn't allowed to happen or if the tech leaves the trade.

stever

Display signals that fall outside of reasonable operating ranges would show up as a fault - it'd be just another PO455xya (always wondered why so many OBDII codes start out PO or P0 -- sometime about off.... ).

The computer cannot test beyond its connector - unless you design with a back door to monitor downstream and upstream. Say you have a dead coffeepot at home. The issue is a tripped GFI caused by a funky plug. You find that out at the office because the wifi in the coffeepot sent the error code to your smartphone, using the power in a capacitor somewhere. You know it's the outlet by the sink that tripped because your house texted that info to you.

Right now you have one Stabilitrak idiot light in your car. You need a readout on the dash that says the AWD isn't working right because of xyz. Or even better, the ABS is off-line because the sensor in the right front tire has a discrete component failure in the little board in the RF sensor and the ETA for the self repair is six minutes. You may need park the car in a hotspot and approve the d/l of the updated repair data.

We already have a micro-electromechanical system that processes a bunch of info from sensors very fast and decides whether to inflate your airbag. Or steady the digicam on your windshield. You don't program these chips telling them how to respond to every possible scenario. You tell the chip the results needed and let it figure out the rest, based on what the sensors are saying. This isn't rocket science and the tech is already out there for self-healing systems that diagnose and repair these same chips and circuits.

Mr_Shiftright

So I was going to do pads and rotors on my Mini Cooper but I knew right off I was going to have trouble getting those rotors off. My car spent a few years in Michigan and I have struggled every time with rusted bolts on this car. The rotors are held on by 50mm torx and they are presently just laughing at my rust-buster and my impact hammers (electric and hand-held).

I'm tempted to just install the pads for now (rotors are smooth and clean) until I can gear up for a new assault. I'd presume the rotors are somewhat out of spec, but the car stops great.

I wonder how risky this actually is? I'm really not keen on drilling these out right now, nor am I keen on paying a technician $160/hour to sit there and go whirr-whirr for an hour.

BTW, the check engine light is off for good. New gas cap was the problem all right.

thecardoc3

When they are seized and the Torx bit fails to remove them, I take those out with a large center punch. Do the same with the Philips screws that hold some rotors.

xwesx

You can check them with a caliper and compare that to specifications if you're uncertain about leaving them in place. If they aren't scored, worn, or warped, why bother replacing what isn't broken?

explorerx4

@mrshiftright,
I don't have your mechanical skills, but I would weigh the cost of new pads only, plus upgrading your tools and something could break vs having it done.

roadburner

Mr_Shiftright said:

So I was going to do pads and rotors on my Mini Cooper but I knew right off I was going to have trouble getting those rotors off. My car spent a few years in Michigan and I have struggled every time with rusted bolts on this car. The rotors are held on by 50mm torx and they are presently just laughing at my rust-buster and my impact hammers (electric and hand-held).

I'm tempted to just install the pads for now (rotors are smooth and clean) until I can gear up for a new assault. I'd presume the rotors are somewhat out of spec, but the car stops great.

I wonder how risky this actually is? I'm really not keen on drilling these out right now, nor am I keen on paying a technician $160/hour to sit there and go whirr-whirr for an hour.

BTW, the check engine light is off for good. New gas cap was the problem all right.

Just drill out the head of the set screw. In the alternative, whacking the back of the rotor with a big enough hammer will usually snap the screw head right off- then you can either use locking pliers to remove what is left of just forego replacing the screw altogether.

roadburner

Apparently some vehicle owners aren't using oil with properly sized molecules; this from another board that I frequent:

So I am about to get my car back this week after a spun bearing at 22000 miles. The dealership and Subaru of America deemed, after 4 months of delegation, that the cause was oil starvation. I disagree. But they are fixing it at no cost to me. The tech told me that Mobil 1 was bad for this specific car. Said the molecules were too small.

thecardoc3

Mobile 1 would easily meet the specs for Subaru if they used the proper formulation. Exactly what car does he have? The majority of the Subaru models require the North American/Asian spec products and they are easy to identify with the regular API SN and ILSAC GF5 approvals.

The quote above appears to be a distortion of the information that may have been shared, and that could have occurred between anybody in the communication chain.

The 5W30, 5W20, 0W20 that are dexos approved are thinner versions of the API SN and ILSAC GF5. They still meet that spec. they just have additional requirements to meet the dexos specification. GM and Ford require thinner oil because of their tighter bearing clearances. They both state that at very cold engine starts the shearing forces between the crank and bearings can tear the oil molecules apart damaging the oil. If that has been confused in reference to the size of the molecule by someone then that would at least start to explain the misstatement. Lots of people struggle with the idea that a 5W30 that is dexos approved (ACEA A1/B1) is thinner than an API 5W20, while a 5W30 that is LL-01 approved (ACEA A3/B4) is thicker than a 10W40.

The statement that a bearing failed because of oil starvation, on what exactly does he disagree? What is the rest of the story?

Mr_Shiftright

I hope they inspect that block very carefully and also measure oil pump pickup clearance to the pan.

Mr_Shiftright

Thanks for the tips on the rotor. I do in fact have the appropriate hammer. I've never buy another car from the rust belt. You should have seen what it took to get the supercharger pulley nut off! (answer: a hydraulic press).

thecardoc3

I hope you considered that trying to break off the head of the bolt in such a fashion could easily result in causing run-out in the wheel hub flange.

Mr_Shiftright

Well I don't intend to murder it, but thanks, I will be careful. I hope to drill or grind the head of the bolt first. I'm generally not meat-handed with cars, I take my time. Maybe I'll soak those bolts overnight. Any recommendations for a good rust-buster?

roadburner

PB Blaster and Kroil are good.

isellhondas

Are you using a hand held impact screwdriver? Maybe an overnight soak would work.

thecardoc3

Mr_Shiftright said:

Any recommendations for a good rust-buster?

With as much of that stuff gets sold and used, did anyone ever stop for a moment and consider that bleeder screws, brake lines, etc. use a straight metal to metal contact and keep a fluid that is under pressure from passing in-between them? That being the case, what chance does a fluid that is not under pressure have of forcing its way past? "Penetrating oil" has its uses, but it is more often misused and wasted.

If you want to do something that would release the screw from the rotor and hub, heat will work, but you have to be precise on exactly what you heat and just how much is applied. Another trick is to heat the screw and rotor and then shock cool them with water. Each method has its strengths and appropriate usage as well as times when they are ill-advised. A favorite for those with sufficient skill is to weld a nut to the screw head. This works really well with a broken exhaust stud in a manifold or cylinder head. One routine for that is to weld a washer to the bolt/screw first, and then the nut to the washer especially when dealing with aluminum.

Mr_Shiftright

I always thought shock-cooling warps a rotor---like when you come down a long mountain road and then hit a puddle at the bottom.

thecardoc3

There is a difference between heat and temperature. There is a quantity factor that has to be understood. You can add heat to a small portion of a brake rotor and increase the temperature of a very small area to the melting point and not be anywhere near the quantity of the heat generated in the entire rotor during heavy braking.

The same goes for dealing with the broken bolts in a cylinder head. Welding a nut to a broken stud isn't difficult at all even when the stud is in an aluminum head. The head barely gets warm because while the weld is quite hot, the quantity of the heat used is small.

Mr_Shiftright

Got it, thanks.

thecardoc3

This is interesting. http://www.payscale.com/research/US/Certification=ASE_Master_Technician/Hourly_Rate

Less than one year exp. 1% of the workforce. One to four years 10%. Five to nine years 15%...... Over twenty years, 42%.....

Mr_Shiftright

Well at $30 an hour, that's $60K a year plus benefits in some cases. Not bad for some parts of the country, real bad if it's San Francisco Bay Area and you have a family.

thecardoc3

Take note of how the younger the group the fewer that there are in the trade. It's something to see it broken out by age and its no surprise that those under ten years experience only make up 25% of the workforce. Now factor in that the trade won't keep most of them for the long haul career wise and things will really be interesting in the future. While that $30/hour is out there the scale starts somewhere in the $12-$14/hour and the top wages are earned by the 20+ years in the trade segment.

Mr_Shiftright

The way car tech is going it's possible that not as many technicians will be needed. There's going to be more and more "not serviceable" in cars, especially if we see more and more EVs and hybrids. And of course there will always be work for commercial diesel, which is where I'd go if I were young and ready to start my tech education.

Michaell

Mr_Shiftright said:

The way car tech is going it's possible that not as many technicians will be needed. There's going to be more and more "not serviceable" in cars, especially if we see more and more EVs and hybrids. And of course there will always be work for commercial diesel, which is where I'd go if I were young and ready to start my tech education.

My son's best friend went to school to be a diesel mechanic ... is working for a local road striping company taking care of the fleet. I think he's making somewhere between $60-70K here in Colorado at age 29.

thecardoc3

Michaell said:

My son's best friend went to school to be a diesel mechanic ... is working for a local road striping company taking care of the fleet. I think he's making somewhere between $60-70K here in Colorado at age 29.

65K works out to 42 hours a week at $30/hour. You should see what he would get to earn if he chose to try and work for a retail chain store.

Michaell

thecardoc3 said:

Michaell said:

My son's best friend went to school to be a diesel mechanic ... is working for a local road striping company taking care of the fleet. I think he's making somewhere between $60-70K here in Colorado at age 29.

65K works out to 42 hours a week at $30/hour. You should see what he would get to earn if he chose to try and work for a retail chain store.

I know he switched from hourly to salary earlier this year, but the company also provided him with a company truck for his commute (about 40 miles one way).

I don't think he'd want any part of a retail chain or a dealership.

Mr_Shiftright

nor would I if I were a young man in the trade. I'd avoid both of those like the plague. Either go into a specialty indie shop, or go commercial fleet work. In a chain store or dealership, there's no "up".

thecardoc3

Here is another example of what one can expect if they want to become a technician. This scan tool is being presented as having O.E. level coverage. http://matcotools.com/catalog/product/MDMAXME/MATCO-MAXME-TABLET/

The first thing to notice is the return policy.

The tool's price is a big red flag, while it might seem like a lot for a DIY'er tool it is way below the cost of most aftermarket professional level tools. For example it costs more per year to have a full subscription for just GM's tools and software. https://www.acdelcotds.com/acdelco/action/subscribehome For a shop/tech to try and support for additional manufacturers (Ford, Chrysler, Toyota, Honda etc.) and the costs quickly get to be unmanageable.

Tools like the one linked above have very minimal capability. It can be enough of course for the easiest repairs, but the gaps in coverage reveal themselves very quickly. FWIW. It was gaps in older tools that had people believing that they were locked out of certain information and started the whole R2R issue.

Mr_Shiftright

The reviews for the product (some of them at any rate) make it pretty clear that this tool isn't for people who need it to make a living.

thecardoc3

They say a picture is worth a 1000 words so here is one to test that idiom.

1997 Jeep Wrangler 2.5l manual. The customer reported a loss of power (under powered) during acceleration especially at higher RPMS. There are no codes setting.

This is a modified vehicle that has had a turbo-charger added to it. The engine and its wiring harness have been replaced along with a long list of other repairs in attempting to repair the reported issue. Here is a capture from my PICO scope that shows the failure that was occurring.

All of the traces were taken at the PCM's connector.
The green trace at the top is the crankshaft position sensor signal.
The red trace is the camshaft position sensor.
The blue one is cylinder #1 injector.
The yellow one is the ignition coil primary.

These measurements were taken because during a hard acceleration the computer would report that it lost synchronization between the CKP (crank sensor) and CMP (camshaft sensor).

Any idea what was wrong with the car? The multicolored overlay shows each cylinder and what stroke that cylinder would be on during that capture between the cursors.

Mr_Shiftright

Well not sure but if ckp and cmp synchronization are being lost, one might suspect a mechanical malfunction rather than an electrical one (the latter also being possible of course, perhaps in the PCM harness?).

thecardoc3

It's one thing to suspect something, and then there is proving what is wrong. For the trained eye, the failure is obvious in that scope capture.

thecardoc3

Here is a capture without the overlay on it.

Mr_Shiftright

Oh did we lose an injector pulse there?

thecardoc3

That's part of it. BTW. That is the result of the failure, not the cause.

xwesx

Mine is the untrained eye. There's definitely a "skip" on the crank position sensor voltage where it does not return to zero (e.g., stays at ~5v) at ~70ms - during the ignition stroke on #2. But, across the chart, that's the only spot where that occurred.

If that then caused ignition failure on the stroke, I can see how one might notice it for a moment. What would cause it to hold voltage like that?

thecardoc3

Each rise in the CKP signal is caused by a window in the flywheel moving past the sensor. Each group should be four pulses. The one that is a single normal pulse and then a large one is the result of damage to the tone ring attached to the flywheel. The resulting signal causes the computer to lose the crankshaft speed and position and it not only fails to trigger the injector, it mistimes the spark command and that is what causes the back-firing and loss of power.

There is always a chance that a crankshaft sensor being loose, failing or a wiring or PCM issue could cause a similar failure but in this case it only happens when the CMP sensor signal is going down which means it is specific to that one point of the flywheel. If the failure was any of the other possibilities it would 't be specifically timed like this one is. This was proven by collecting a few dozen captures and comparing the position of the anomaly.

Mr_Shiftright

How exactly does a tone ring on a flywheel get damaged? If it somehow got loose, then I was right--it was a mechanical failure that caused loss of synchronization.

xwesx

thecardoc3 said:

If the failure was any of the other possibilities it would 't be specifically timed like this one is. This was proven by collecting a few dozen captures and comparing the position of the anomaly.

Wow; that's pretty impressive. The truly interesting part is that the explanation makes perfect intuitive sense, yet is a complete mystery (even if the cause is quite simple) if one does not know exactly how everything interacts.

Mr_Shiftright

And can you imagine the future? There are some high end new cars that have over 90 interacting computers.

PF_Flyer

Mr_Shiftright said:

And can you imagine the future? There are some high end new cars that have over 90 interacting computers.

It's why I chuckle when I see someone stopped on the side of the road, hood up, hands on the fender, staring into the engine compartment as if they can actually DO something with a tool they forgot they had in the glovebox.

thecardoc3

The future is already here. The vehicle related to that capture will soon be twenty years old. We have been using the kinds of tools and training that scope capture demonstrated to solve problems like that since the late eighties.
That scope and scan tool usage is what it really takes, on top of the training and decades of experience to work on the high tech systems in today's cars. Nobody can go to a school for two, or even four years and walk out the door and be truly competent as a technician. There should be an apprenticeship program that runs at least two years for the four year graduate, and probably a five year one for a two year graduate. Then they would need another ten years to really polish their skills so that they can cope with not only the things that have been on the road for a while, but be ready to handle anything new the first time, and maybe only time in their lives that they encounter a specific failure.

BTW, the idea of just using Google (or whatever) to find an answer to a given failure only works to thwart the technicians intellectual growth. Every individual job ticket is a test, that Mustang with the offset wheels is a perfect example. For techs there is no finish line when it comes to education and experience, there is always going to be something new that you have never seen before.

People really shouldn't be surprised that it is difficult to find qualified techs today and that is only going to get worse. When you know what to look for, you see all of the reasons why almost no one who is really capable of doing the work even considers the trade as a genuine career choice.

Think about the brake pad (and now pads and rotors ) post. Imagine making thirty dollars an hour, so that job puts eighteen dollars before taxes into your pocket, no matter how long it takes you to do it. Then we have the Mustang with the offset wheel incident. Look at the responses where all they care about is using it as leverage to further deepen the wound for the business. They don't grasp that given enough chances they would easily make that same mistake, (or worse) . They think that they are too smart for them to make such a mistake. They aren't.

Mr_Shiftright

Are you saying that the technician makes only $18 on the $425 front brake job on my Mini? I find that hard to believe.

stever

Nope, you don't need better techs, you need better programmers. Those 90 computers should be talking to each other so that the readout on your phone would tell you exactly what's happened whenever something isn't optimal.

Those rims? It would be trivial to match the rims with the "correct" position on the correct axle so that the car would send you a message whenever a wheel was removed from the car and if replaced, the car would require you to train the wheel to the car. The training would fail because it would have been the incorrect size rim, requiring a manual override to proceed.

You guys focusing on techs are trying to reinvent the wheel.