Be an auto tech for a day · 10:40pm Jul 2nd, 2017
What follows is because of a debate that the shop manager and I had regarding the diagnosis of a Buick last week.
Source (YouTube Link)
This isn't actually the Buick I was diagnosing, but a 2006 LaCrosse is a pretty boring car, so I wanted to put up a better image.
Those of you who have taken ASE tests (and in some cases, state mechanic tests) may be familiar with technician A and technician B. A lot of questions are worded like that, and here I'm going to give you a simple, made-up example. Regarding ponies, because I figure anyone reading this blog would know the answer.
Technician A says that Rarity is a unicorn. Technician B says that Applejack is a pegasus. Who is right?
a: Technician A
b: Technician B
C: Both
D: Neither
See, it's pretty simple, usually. Of course, on the real tests you often have to read the statements more carefully. They don't normally try to trick you on the tests; however, badly-worded questions do come around sometimes, and like all multiple choice tests you have to be careful not to outsmart yourself.
Hold on to that thought.
Once upon a time, cars were pretty simple. Really, up into the 70s or thereabouts. A lot of automakers saved money by re-using parts across multiple vehicle lines, so you wound up with things like GM B-body cars having the exact same brakes as the 1/2 ton trucks, or the same alternator (my Oldsmobile constantly got new alternators, because I kept stealing them for my pickups). Back then, it wasn't too much to expect a technician to be fairly familiar with most cars, to the point where he generally didn't have to look in a book to know what the rotor thickness was supposed to be, or how fast the engine should idle hot, or any number of other things.
These days, of course, most of the functionality of the car is controlled by computer, and every year they add more things in. Back in the 70s, your headlights were controlled by an on/off switch (basically, a fancy toggle switch); now all that happens when you turn the headlight knob is that you tell the computer you want the headlights to be on and then the car decides if you're going to get them or not.
Of course, that means that you really need to look up the diagnostic information and follow it, rather than rely on your memory. Which isn't so bad . . . in the 70s, you'd have to look in a book; now you can just put the trouble code into the computer and nearly instantly get the test.
This, of course, leads some people to think that all we have to do is hook 'the machine' up to the car and it will tell us what's wrong . . . and it does, to a point*, but there's still diagnosing to do. And, unfortunately, sometimes the diagnostic instructions are more of a hinderance than a help. In fact, I remember many years ago diagnosing a mass airflow sensor on a Nissan that not only had the wires reversed on the connector diagram, but if you followed the instructions exactly, you'd be replacing the PCM every single time. You could not possibly come to any other conclusion . . . because they missed a step in the diagnostic instructions.
The automaker, I assume, generally writes pretty good diagnostic procedures. They designed and built the car; they know where every wire goes and what the computer is supposed to do. And if you work for a dealer, you have access to that information. If you don't . . .
Well, if you don't, you'll be subscribing to some aftermarket source of information. Mitchell, Identifix, and Alldata are three of the popular ones. Many times what they do is take the information that the dealer provided for them, and then they re-write it into an easier-to-use format--or what they think will be easier to use.
Sometimes that does make things easier. Each automaker seems to have a different name for some of their parts: the domestic automakers have solenoid valves, while Toyota has VSVs. GM stubbornly calls alternators 'generators.' German automakers have their own special way of laying out service manuals which are completely incomprehensible, and Mercedes puts different model numbers on the car than they do in the book (I just worked on an older C320 which according to Mercedes is a model 209 . . . if it says that somewhere on the actual car, I couldn't find it).
Other times, they manage to mangle the instructions beyond all recognition, as you're about to see.
__________________________________
*The easiest way to think of how engine diagnosis works is that the car's computer constantly runs tests, and gives an error code when a test fails. The technician then has to figure out what component failed. So, to give a common example, a car might set a P0171 System Too Lean code, but that doesn't tell me what part is bad, just that the car thinks that it's getting too lean of a mixture. It could be an intake leak, a faulty MAF sensor, a faulty HO2S, low fuel pressure, clogged fuel injectors, E85 in a non flex fuel car, and probably some other things I'm not thinking about right now.
I used to work at a GM dealer, so I know how GM writes their diagnostic instructions for their technicians. They're what's called a decision tree or a diagnostic flowchart (or as I liked to call it, 'choose your own adventure').
Here's a picture of one I found on the internet. It's a partial chart, but it will give you the general idea.
Source
Basically, as you look at it, the first column is the step number, the second column is what you do in that step, the third (value) is what value you should expect if there is a value to expect (you find those mostly on circuit tests), and the last two columns are where you go if it passes the test (yes) or fails (no) [that's the choose your own adventure part].
And then the aftermarket companies (in this case, Identifix) condense it down into a different form, and that's what we get in the aftermarket. What follows, strange numbering system and all, is the entire test, exactly how it appeared in Identifix.
Circuit/System Testing
*Ignition OFF, disconnect the harness connector at the generator*Ignition ON, test for less than 1 volt between the generator turn ON signal circuit terminal B and ground.
=X If greater than the specified range, test the circuit for a short to voltage. If the circuit tests normal, replace the ECM/PCM
3. Engine running, test for greater than 3.5 volts between the generator turn ON signal circuit terminal B and ground.
=X If less than the specified range, test the circuit for a short to ground, and open or a high resistance. If the circuit tests normal, replace the ECM/PCM.
4. If the circuit test normal, replace the generator.
These were the results of the test:
*The circuit was not shorted to voltage or ground.
*The Ignition ON test showed 0.16v
*The circuit was not open, nor did it have high resistance
*The engine running test showed 3.95v
Now, which component is bad?
Source
EDIT: If you want to check your answer (or don't want to think about it, but just want to know the answer), here's a link to a gDoc that has the answer.
So what do you do with the joint projects such as GM/Toyota and a few others
4590384
Usually each automaker has their own diagnostic procedures. So you'd follow GM's procedure if you were working on a Pontiac Vibe, and Toyota's if you were working on a Toyota Matrix.
Generally, the two are going to be mostly the same, but not always. Toyota might have programmed the PCM a little bit differently, or Pontiac might use different brakes . . .
From the information given, it would appear to be the alternator (I will not call it a goddamned generator).
Oh, wait. I see it. Yeah, that's... interesting.
4590390
And yet most people agree the Vibe is basically a re-badged Matrix which both use the Corolla engine and transmission. a number of the parts are stamped with the toyota logo
The Identix instructions are incomprehensible to me. Outcome: replace the PCM and the alternator. ??
I remember one of the Japanese manufacturers writing amazing service manuals. The English wasn't *perfect*, but they could describe a procedure with two diagrams and some numbered sentences what a GM manual would take fifteen pages to do. Can't say I'm sure which one it was
I love reading your mechanic stories almost as much as your fiction.
i.imgur.com/7IZKXFC.jpg
Instructions unclear, now there's a 6bt in my rx-8.
At least it runs...
I prefer my personal mechanic, my Uncle, becuase he doesn't charge an arm and a leg. Thanks to the family he will not work on Audi/VW becuase he and several family members feel that the designers and mechanics in Germany must have been blackout drunk on the job. Most modern cars I feel are designed and build to put the average and independent mechanics out of business. I feel for you man.
It's clearly the turboencabulator.
I've spent more than one entire christmas break working on a car in a stranger's driveway.
*Stares* No. This...makes little to no sense. Just no.
I found your problem... you have a pony in the engine compartment. How you are going to get horsepower out of that engine when its a pony!
As you know I have extremely little actual knowledge of mechanical stuff.
So...
...one tested normal (replace generator), and one tested abnormal (replace ECM/PCM)? Am I reading those test results correctly?
Sorry, about all I can do is look at something and tell you if it's a car or not.
That thing? Definitely a car.
Oh dear...
well, the car doesn't decide, it's still based on your input (unless you have one of those fancy auto dimming ones). It does let you do some pretty nifty stuff though. On my dad's car the lights turn with the wheels, so if you turn, the lights point where your going.
You'd be surprised. I don't have auto experience, but you can end up with some pretty bad stuff. The worst offenders in my experience are electronics companies providing mechanical info. There was one drawing where they used neon yellow for the part and neon blue for the text. Couldn't read a thing. Ended up going with a different company.
My guess is to replace the ECM/PCM. Both the ignition on and engine running test seem to support that. You're testing for a certain condition (below 1V for the ignition on, and above 3.5V for the engine running) the "normal" result would meet that condition. You got below 1V for the ignition on, and above 3.5V for the engine running test so the results are "normal".
But yeah, that is a crappy diagnostic chart, I would have gotten an F if I turned in something like that.
Fuckit. Replace both.
~Skeeter The Lurker
The quantum physicist in my acquaintance sez:
"(I think:) Replace the generator. The engine running test showed 3.95 V, so the "If less than the specified range [3.5 V]" condition doesn't apply, and you fall through to 4. The circuit is normal (not shorter, open, or having high resistance), therefore, replace the genny."
I see what you mean there! One step say to replace the ECM/PCM and the next contradicts it self by telling you to replace the generator. What the hay!! Somepony messed that up real bad. Thanks for the help, buddy!
I work on air conditioning and heating equipment and I've seen this type of thing, too. You defiantly need to know what you're looking have a very good idea of how things work. Otherwise, you're just a pull and replace technician; which is expensive to the customer and makes the tech look as if he/she is less intelligent than what they are working on.
I think it's the ECM.
And you're right, the older cars were simpler....I want my 1970 F-100 back.
I work for a certain semiconductor company, where I maintain and diagnose the equipment ("tools") that turn silicon wafers into CPUs. This sort of troubleshooting tech writing would not pass muster with the company. The way I read it, both the ECM/PCM and the alternator need replacing, even though logically, the readings show nothing is wrong. But, maybe that's just me. Also, I grew up helping my dad fix cars in the driveway and literally in the shade of a big tree, I do know my way around older cars, and I don't even try any more with things newer than 2000 (if then).
Unless I misread something, shouldn't it be the generator?
4590402
It is, but it's not safe to assume that the engine controls are totally the same. For example, the Toyota might use an air/fuel ratio sensor, while the Pontiac could be using a simple oxygen sensor instead. You actually see lots of cars that are re-branded, joint-developed, or even share things like axles or transmissions.
I guess what I'm saying is that I'd assume that they were different until I knew otherwise.
4590415
I've found it's kind of hit or miss. The Germans especially tend to write very detailed, exact repair information. Toyota, Honda, and Subaru tend to be pretty good, although every now and then translation errors creep in. Nissan . . . I don't know what they think. They're very hit or miss.
General Motors and Ford tend to be pretty good, as long as no one else has tried to re-write their instructions. Chrysler has been in bed with so many other manufacturers it's basically a disaster (Daimler, AMC, Renault, Mitsubishi, and Fiat over the last couple decades).
Ah, for the good old days when I was a TV repairman, and we had an entire room filled with file cabinets of Sams Photofact folders...
4590421
Thanks!
It's funny you should say that; I just showed a friend my page recently, and while he's not really into the pony stories, he loves my blog. I guess the moral is that I should write more blog posts.
4590425
That's exactly the problem.
And it reminds me . . . back when I was younger, I worked part-time one summer painting a farm. I was working on one of the barns, trying to get the lights down so we could spray paint them. Now, we didn't know where the breakers were, so I was doing it live, on a ladder, in a horse paddock.
While I was up on the ladder, live wires in hand, a horse came over and decided that my ladder was a new horse toy.
4590431
There are so many reasons why that's a very odd thing to build. And probably dozens if not hundreds of better motor choices for that project. Then again, he's probably got the only 6bt powered rx-8 in the world.
4590437
I suspect that their designs are probably very intelligent, but they're just different than how the Americans and Japanese do it (and their nomenclature is weird). I really think that if you're familiar with them, everything makes sense. It's just not a type of product I work on very much, so I have to recalibrate my thinking every time I do have to work on one.
Also, sometimes their instructions suffer from weird translation errors as well.
I actually disagree; I think that a lot of independant mechanics simply don't want to learn the new systems and invest in the proper tools for the job. I attend monthly training classes that keep us up to date on how modern systems work, new ways to diagnose them, etc., and that's really helpful and informative.
What's really telling is that while I don't know how many mechanics there are in the greater Lansing area, the last class I attended only had six people, which tells me that lots of people aren't keeping up to date. And I've run into that in lots of shops where I worked--techs who didn't understand computer engine controls and didn't want to learn . . . which is too bad, because ultimately they're not going to be able to get a job, or they're going to be stuck just replacing simple mechanical parts, while guys who do keep up are going to be getting the better jobs.
4590457
My brother, who's an engineer, once spent an entire weekend taking the dashboard of a S-10 apart and drawing a circuit diagram by hand in an attempt to figure out why there was a battery draw. He never figured it out.
I took out the little plastic gizmo that was supposed to make a beeping noise when the seatbelt was unbuckled or the key was in the ignition and tossed it away, because I could feel that it was hot and anyway the truck didn't need that piece. Took me five minutes, and the truck was fixed.
I don't know what the moral to that story is.
4590465
Isn't it great? And you've got to wonder what step 1 and step 2 are, which they basically skipped in the instructions.
4590632
I teach a class called "circuit theory" and it has all the math and whatnot, for engineers. I have an exercise like that where I have them deduce which of a certain set of circuits are inside a black box. Half of them never bother to pick up the boxes, and half the rest don't notice that one is screaming hot.
4590468
It just doesn't sound as impressive when you say your car has 200 ponypower, does it?
I do know that.
4590469
You're not reading the test steps correctly, mostly because the test is very badly written.
4590496
4590517
Given that we have to pay for these instructions, you'd think they'd be a little bit clearer.
4590608
I can understand that
4590522
It does, though. It can say 'screw you, you don't get headlights.'
While that's unlikely for a piece of safety equipment, there are many systems which the car can turn off if it choses to do so, like A/C. You can push the button, but if the car doesn't want you to have air conditioning, you don't get it.
I think if the automakers didn't have their own dealerships and required warranty work, they'd be a lot more sloppy when it came to diagnostic procedures. Since they do, and since they can also make a bit more profit selling those instructions to the aftermarket, they do have an incentive to make them at least somewhat useful.
What is frustrating is that they generally don't document their networks at all, at least in terms of what we get in the aftermarket. So network diagnostics are always fun, because you have to guess a lot about what the little black box is doing, or should be doing.
Sadly, it's not the worst one I've seen, either. There was one that led us to replace a 4wd module because it failed to indicate that the 9v signal we were looking for was pulsewidth modulated. So when we had an erratic 9v signal, we assumed that the module must have a failed driver that couldn't produce a continuous 9v.
It ultimately turned out to be a $10 relay that was at fault, and not the $500 computer.
4590523
That's hard to get customer approval on, since 'both' more than doubles the cost of the repair.
4590540
If you read through it a few times, you can make sense of what it's actually saying, but it's so badly worded that you really have to puzzle it out.
Yeah, that's the key, and that's the difference between an actual technician and some guy that puts parts on. Heck, just last week we had a Mustang in that the guy had had at three other shops for no idle, and they'd put a laundry list of parts on it and hadn't fixed it. We figured out that it had a vacuum leak in less than five minutes, and then five minutes later with a smoke tester, we knew where the vacuum leak was. Everybody else had just been guessing (hmm, it idles rough so clearly it needs the idle air control motor . . . oh, that didn't fix it, well maybe the spark plugs are bad, too.).
Although the newer cars make more power, get better fuel economy, and are much more reliable. My dad's 83 Caprice was worn out with 120,000 miles on the odometer; my new van's got 240,000 and it still runs great.
4590548
Yeah, somebody was having a bad day when they wrote that test. I assume that if I got GM's diagnostic flowchart, it would be much simpler and more straightforward (and it would also have the missing steps 1 and 2).
IMHO, the newer cars are simpler if you understand the computer systems and how they work. If not, you're gonna have a bad time.
4590612
I've got a bunch of old Motor repair manuals, which cover all domestic makes and models from the 60s to the early 80s. They don't really do me a lot of good nowadays, although I suppose if I wanted to work on classics, they'd be helpful.
4590637
There are actually a lot of techs who are starting to use IR cameras for diagnosis. Things like modules staying powered that shouldn't are really easy to find with an IR camera, even though the dashboard.
Back when my grandpa was teaching navigation classes for the Coast Guard Auxiliary, he had one exercise where you had to figure out another boat's course from a radar plot, and by the time you finished your calculations, it would have already collided with you.
Then he told them how to quickly work out whether you were going to hit it or not, which is the most important thing to know when you're using a radar for navigation.
4590400 4590415 4590469 4590522 4590530 4590540 4590590
Since I don't want to keep y'all in suspense (but I also want to give new people a chance to guess at it), I wrote a fancy new gDoc which has the answer, and explains a little bit more about the system and what the purpose of the steps in the diagnosis were.
Here's the link!
If I'd been smarter, I would have written that before I posted the blog post.
And if I'd been really prepared, I could have also gotten a circuit diagram of the charging system.
4590677
Now, now, didn't we learn from Suited For Success that doing exactly what the customer wants isn't always a brilliant idea?
Love reading your blogs.
Used to help my dad fix are trucks when I was a wee lad. My brother almost became a mechanic. Worked for a local company for awhile. I never took it further then basics. I did rebuild an engine for a ford bronco 2 I had. And yes I prefer the older cars to the newer ones. Much easier to fix. Unfortunately not as fuel efficient as newer cars.
My current car keeps telling me to check the oil when I start it. Ill have to check when I have time and a level space to do it. Doubt its low as the senor checks oil quality and not the level.
Current concern is the lack of working ac. Clutch doesnt engage on the compressor. I should look for the nearest junk yard to see if I can pull one at a lesser cost then it would to have it replaced by a shop.
You know I miss my old car, it was a mid 90's chevy beretta. I beat the crap out of that thing and it still ran.
I guess I survived a day as an auto tech. woohoo.
Also, like many others before, I really enjoy these blogs.
4590677
I don't think i've had that happen, is it just due to a low sample rate on the computer? If it doesn't do what I tell it to do, then I'd consider it broken, which happens with old stuff as well.
They still need it for internal design and testing.
are network problems really that common? I'm not really sure what kind of documentation you could provide for network stuff though, or what you can adjust at a mechanics.
Honestly a lot of the stuff i've seen wouldn't fly at my school. Some of the stuff i've seen would be in the C- to F range. I assume that you don't have oscilloscopes then, it'd be hard to have that in a mechanic setup. They're pieces of lab equipment commonly found in electronics setups and they show how voltage varies with time so a PWM signal is immediately apparent.
That's good at least. Why did it immediately jump to the ECM? The computer should be last on the list.
EDIT: Also, looking at the instructions again, both steps 3 and 4 say to do something if the circuit is normal. That's atrocious.
Laptops have repair woes similar to the auto industry, with their own wierdities.
Back in '99, Gateway computers we bought for work had two issues: hard drive failures and video card failures. In the end, they replaced *every* hard drive in that year's buy, and the video card fails were because of a goofy bios in the computer. Somehow, it would turn the card off and make it a coaster. Worse, Gateway had massive issues with these, so they'd take the returns shipped back to them, test them (Gee, this looks fine), and send them back out to us as 'new replacements' which did not work, of course. We twigged to it when one of the techs started writing on the cards we sent back, and when we started getting replacement cards with his handwriting....
Currently, HP laptops have a 90B fan failure that is saturating their replacement part channel. Fan does not spin, so laptop shuts down about 20 seconds after turning on. We can tell when they're out of parts and are stalling us because they email back, "Update the bios and tell us if that fixes the issue." Um. Yeah. No, that's a non-starter. 20 seconds isn't enough time to even get into bios, let alone think about updating it.
4590625
As far as I can see, he partially did it because his friends told him he couldn't...
Powerful motivation, that.
4590695
Oh, cheap parts...
I'm planning on installing a "BestTest" alternator in my '90 Bronco, cuz it's the cheapest O'Reilly has, and I'm pretty sure the rectifier in mine has failed. (no or very low charging, case of alternator is hot in that location after about 30 secs of engine run time from cold)
I'd replace the rectifier, but it's 50% of the cost, and the one in mine is nonstandard, which means I'd have to buy other parts to use the new one. Also, the connections to the windings are soldered, and I am so done messing with it.
Wish me luck!
4590663
I'd be pretty impressed you managed to fit all of them inside.
4590695
It does tend to help when your parts mach the specs.
As a former GM technician myself, I can sympathize. Alldata was my savior most of the time, because GlobalConnect (are they still using that?) is an absolute PILE OF STEAMING SHIT.
Also, it's the alternator, isn't it? I wanna say I've ran into that exact logic-puzzle-from-hell cup of bilge juice. It's almost as bad as the one to diagnose if your trailer brake relay or trailer brake module was bad. (Both Alldata and GC would always tell you to replace the one that wasn't broken, so you had to know to do the opposite.)
4590696
Well, yes, and I don't feel bad for turning down people like the guy who wanted us to hang up his exhaust with part he provided . . . which was literally a coat hanger. But I figure if I'm going to diagnose a car, I ought to get it right.
4590719
Heh, in some ways you might have dodged a bullet there. Working on cars is a fine hobby, but it kind of sucks to have to do it for a living. Besides all the pain and suffering, tool trucks are the new company stores.
Actually, very few cars (if any) can actually check the quality of oil. Change Oil Soon lights are based on either just mileage or some algorithm. Checking the oil level is something that a sensor can do pretty easily. In their simplest form, they're basically just a float like in a toilet tank.
Easiest way to see if it is the clutch is to disconnect the harness that controls the clutch, and then wire up a car headlight to it. Turn on the AC; if the bulb lights, than it's the clutch. (This proves out the entire control circuit, and also proves that the wires are capable of carrying enough amperage to actually engage the clutch.)
With the right tools, on many cars you can just replace the clutch assembly and not even disconnect the AC lines, which saves you from having to evacuate and recharge the AC system.
Out where I live, they're all dead now. Mostly due to body rust, I assume.
The weird thing is that when I started working as a mechanic, they were super common, and then it seems like they all disappeared overnight. I haven't seen one in years.
4590730
Yay!
Thanks! I should write more of them, shouldn't I?
4590797
I would assume that to be the case with pretty much any repair industry. Cars is what I know.
We get that sometimes, too--defective returns that then wind up being sold back as new parts. If the package has been taped back shut, that's a big warning sign.
At least with cars, I get to be on the pointy end of the stick, so I don't have to deal with someone (other than the boss, occasionally) telling me that I'm diagnosing the car wrong.
Lack of repair parts sucks. We run into that sometimes, and of course customers aren't always understanding about how it'll take a week to get the part they need to make their car go again. Naturally, they tend to yell at the bearer of bad news, rather than the supplier.
4590815
Yeah, it is. I've written a few stories just because someone either dared me to, or told me not to.
Good luck!
With the cheap cost of remanufactured (and sometimes new) parts these days, it's generally not even worth rebuilding parts on your own. Like, I can get a new wheel cylinder for $10, or I can get the parts to rebuild the old one (eventually) for $20. Kind of a no-brainer. Heck, we used to just replace some types of GM calipers rather than replace the bushings, because the bushing kit was $15 and a reman caliper was $17.
Older cars are much more tolerant of whatever part you stick on . . . on your Bronco, the computer doesn't control the alternator at all, so as long as it charges, that's really all you need.
4590819
Yeah, that would be a bit of a challenge, wouldn't it?
equineink.files.wordpress.com/2016/07/pony-swim3.jpg
Here's a somewhat related picture of the Chincoteague ponies swimming.
4590830
One of our trainers works for CarQuest, so he can't say that the new CarQuest part doesn't work, although of course to be a good trainer he has to call it like it is. So he'd just say that they 'exceeded factory specifications,' and leave it at that.
He also had a habit of covering the embroidered company logo on his shirt with his hand when he went a little bit off the reservation.
4590855
I'm not sure if they are still using GlobalConnect (been away from the dealer for 10 years), but I'd imagine that they are. Yeah, it was wonky sometimes, although I will say that generally their trouble trees worked eventually.
Yes (here's a longer explanation)
The one I had the biggest problem with at the dealership (and that was as much of a manager problem as anything) was a Buick with a bad battery. We were supposed to use the handheld Midtronics tester on batteries, and if it failed the test, it would give you a code which you'd use for the warranty claim.
Well, the Midtronics said it was good.
The dealership also had a VAT-40 load tester, so I hooked that up and watched the battery fall on its face. It worked with a milliamp current, but when you loaded that sucker with a carbon pile, it couldn't handle it.
The manager said that since the Midtronics said it was good, I couldn't put a battery in it.
The car got towed back in the next day with a dead battery again. Midtronics still said it was good.
4590743
Generally, it's because of some other blocking condition. For example, some GM cars won't turn on the AC when the engine is in danger of overheating, when it's too cold outside (if it's 50 degrees, you don't need AC [it will turn it on for defrost, though]), when the idle is too low and the load from the compressor might stall the engine, when it has a fault in the cooling system . . . or to give another example, on most late-model GM products, a fault in the brake switch will block the cruise, since the car can't be sure that it will know if you push the brake pedal.
Not really--they have to write the procedures using tools that a tech or shop can be expected to have. I'm sure that there's some kind of alternator load and circuit tester that GM has that just plugs into the alternator and turns it at a specific RPM range and gets exact results, but you can't expect a tech to own one of those. You can assume that he owns a voltmeter, however.
They're not super common, but like anything on a car, the network wires can get damaged, or a module can fail. The problem is that they don't always even tell you which module does something, or on which network. Not that long ago, I was diagnosing a module or network failure on a Ford, and the module in question communicated on four different networks. If it wasn't the module at fault, than it must have been one of those networks (all four failing is pretty unlikely), but which one?
[That one wound up going to the dealership, and it turned out to be a software issue in the module that was not documented in the aftermarket, so I didn't feel bad for not figuring it out.]
One example where you can run into trouble: it's possible on a newer car for the speed signal that your speedometer displays to take a route like this:
Sensor --> Transmission Control Module --> Engine Control Module --> Body Control Module --> Instrument Panel Cluster. On its way, it would probably go from a high speed CAN bus to a low speed LAN bus.
I think it's Toyota who very helpfully tells you who the sensor reports to, and then who also gets that information, and on what network.
We do have a four-channel labscope, and it's really good for checking all sorts of things. And if we'd known that it was a PWM signal, we could have easily scoped it . . . but the diagnostic instructions didn't say that it was, so we were checking with a voltmeter, and figured that an erratic signal could have been the reason that the 4wd didn't work.
That's my general thought, as well. And when I do suspect the computer, I like to do as much extra checking and verification as possible before condemning it--am I sure that it's got good power and ground? That the signal it's getting is uncorrupted? That kind of thing. Because there's nothing worse than replacing a very expensive part and having it not fix the car at all.
The rational is that if the circuit is normal, than the failure must be in the ECM/PCM.
And I should also point out here that this is one place where experience really comes in handy. While the diagnostic instructions may make it appear that a computer failure is as likely as any other cause, I have never seen a PCM fail to control the alternator on any vehicle. Every single one I've ever worked on has had either a bad alternator, or a broken control wire.
4590608
Don't forget BMW-Mini!
Alternator! The decision tree is stupidly laid out but it has to be the alternator.
4590522
Fun fact! One of the first cars that had this (if not the first) was a Citroën DS which had not a single computer in it. Amazing what you can make hydraulics work.
4590632
"Once engineering goes past a certain point, it's useless to try to understand the whole thing, and must instead employ the methods of the natural sciences and form theories on how various inputs and outputs correlate."
That's a pretty good moral, I think. :)
It's certainly one I find familiar given that I'm a computer guy and <conspiratorial-whisper>nobody actually knows how your computer works.</conspiratorial-whisper> Trying to do it from first principles is akin from staring at the Standard Model and trying to derive Bach from four fundamental forces and seventeen particles. I mean, it's there obviously, but...