• Member Since 2nd Nov, 2012
  • offline last seen 1 hour ago

Admiral Biscuit


he/him

More Blog Posts791

Oct
1st
2019

Mechanic: The Bug in the Buick · 1:58am Oct 1st, 2019

Have a short mechanic blog!

Couple of days agoAbout a month ago [because that’s how far behind I am at the moment], we got a 2012ish Buick Regal towed in. Looked like this, more or less:

Source

The complaint is that it barely runs, and stalls frequently.


Sometimes we get cars where it’s hard to verify the customer complaint; other times it’s easy. Case in point today whenever I actually wrote this blog, I got a Malibu where the customer said that the brakes were grinding only when he applied the brakes and now they’re grinding all the time. I like to test drive cars when I can, because you can find all sorts of issues that way, but of course if it’s potentially dangerous, I won’t ... this particular Malibu got a test drive right into the shop. Needed all four brakes; in the front three of the four pads had delaminated. One inboard had fallen all the way off, and the other inboard was most of the way off.

And such was the case with the Buick: the customer’s complaint immediately revealed itself. I turned the key, it started, surged, and stalled. Second time, same result.

Gave it a bit of throttle, and it got to about 1500 rpm before stalling hard. This is going to be difficult to describe in exact words, but normally when an engine stalls there’s a bit of wind-down; in the case of the Buick, it was like the engine hit a brick wall.

After multiple attempts at starting, feathering the throttle, and judiciously using the gear selector, I eventually got it into the shop under its own power, which was nice. Didn’t have to push it in.

The check engine light was on, as would be expected. The car also knows something’s gone horribly wrong.

I’m not going to give exact code numbers on this one, but the gist of it was as follows:
Mass Airflow value incorrect
Bank 1 lean
Bank 1 oxygen cylinder stuck lean
Random misfire
Fuel volatility low

To save armchair detectives a bit of time, I’ll tell you it’s an inline 4, so bank 1 is the only bank.


A lot of times, we can watch the misfire counter on the scan tool and see which cylinder is the odd man out. There might not be enough misfires on a particular cylinder for the car to say with confidence which one is bad, but nine times out of ten the one with the most overall misfires is the bad one. Thing is, the car’s gotta run to get good readings.

Likewise, we can check airflow by watching the MAF value (rule of thumb is at idle, greater than 1g/s per litre of engine displacement). If the idle isn’t stable and at about the target value (probably about 700-800 rpm for this particular car), that value reveals nothing.

What about the oxygen sensor? Can’t we check fuel trims?

Well, no, they’re not likely to be accurate. The car didn’t fuel trim itself into this situation, and besides, most cars tend to at least start and run for the first little bit using learned/programmed values. Oxygen sensors don’t work until they heat up, and while little built-in heaters and other technology has changed the on-time from minutes to tens of seconds, it’s still not instant.


A very simplistic way to look at an internal combustion engine is to think of it as an air pump . . . and at its most basic, that’s what it is. Air comes in, changes a bit, and then goes out. There are certain physical limitations of how much air you can pump through a given engine, and there are certain minimum requirements for it to be efficient. That’s called volumetric efficiency, and you want the number to be as close to 100% as possible. Or over, if you’ve got a supercharger or turbocharger, but this Buick’s got neither.

Incidentally, this car doesn’t run well enough to do a volumetric efficiency test on, either.


So in a nutshell, here’s what I’m facing. I got five codes which overall aren’t very helpful. They’re all the OBD-II equivalent of “something went wrong.”

The sensor which measures inbound air volume can’t be trusted.

The sensor which measures outbound air flow composition can’t be trusted.

The misfire counter can’t be relied upon.


Now it’s time to divide and conquer. Time to try basically poking things with a stick and seeing what happens.

There was a code for low fuel volatility, and therefore one bit of data that might be useful is the fuel composition sensor. They measure how much alcohol is in the pony fuel, and adjust the injector on-time (pulsewidth) accordingly.


Source

Sometimes it goes wrong, and overfuels because it thinks the customer has put E85 in the tank when they really haven’t. So I reset that and start the car and nothing changes.

This car is also gasoline direct injection--GDI. I forgot to mention that above.

Basically, that means that it’s got a low-pressure pump in the fuel tank that runs maybe 60psi, and it pushes it into a high pressure pump which might hit 2500psi. It’s essentially a common rail diesel, for readers who are familiar with those.

If the high pressure pump fails, the car will only have the 60ish psi that the low pressure pump puts out. It will usually run, but it will not run well.

That wasn’t the problem. According to scan tool data, fuel pressure was what the car wanted.


I reset the fuel trims--basically, a scan tool command telling the car, ‘hey, you remember all that stuff you learned while you were driving? Forget it all.’

No change.


Source

Alright, fine, on to the next idea. Since it’s right there, I unplugged the mass airflow sensor.

Car starts and runs perfectly.


Back in the OBD-I days (or what Ford called EEC-IV), Ford had a pid called FMEM. Ford loves their acronyms, which is why they have things like COPENPLAT errors but that was a subject for a previous blog post.

Anyway, FMEM is “Failure Management Mode,” and it’s what the computer uses if one sensor goes down.

While things work best if it’s got actual measurements, if for some reason it doesn’t, it resorts to its lookup table.

GM calls that backup fuel strategy, incidentally.

Also back in the OBD-I/EEC-IV days, Fords occasionally had mass airflow misbehaviors that you could easily diagnose by unplugging the sensor, forcing it into FMEM. That usually didn’t work on GMs, because their programming was different, and they were usually pretty good at ignoring the MAF if all the other sensors indicated that they probably should.

Apparently that was not the case with this Buick. Maybe GM had hired a Ford engineer to program code, who knows.

Point is that it needs a mass airflow sensor.

Incidentally, this explains all the other codes, too. It thought it was lean, because it wasn’t measuring incoming air correctly. It thought the fuel volatility was low because it wasn’t measuring the incoming air correctly. Whatever the other codes were, it was because it wasn’t measuring the incoming air correctly.


Well, at least I hope so. That’s my working theory. I tugged at the wires a bit; I’ve seen them broken near the connector on various GM models. No corrosion or obvious faults so it’s probably the sensor that’s bad.

Next morning a new one comes, and I commence to removing the old one.

I’m going to step back here a little bit and explain how these sensors work. The idea is that they measure airflow, but how? You might imagine a little propeller in the air duct and while that might actually work, it’s not what they do. In fact, there are two designs that I know of.

The first design is a conceptually simple arrangement. There’s a door that’s held shut with a spring. Any air flowing past it forces the door open; the more air, the more the door is forced open. Little contacts on the bottom of the door indicate how far it’s pushed open and potentially how fast, and air flow can thereby be calculated. That’s what a lot of Japanese cars had in the early days of computerized engine controls.

That’s a really bad system, because it has moving parts. The spring might get weak, the hinge points might get stiff, and besides, you don’t want to introduce any restrictions when the air comes in if you can avoid it.

Most vehicles use a hot-wire MAF sensor, and how it works is that electricity tries to make a wire hot. How much electricity you need to get the wire hot correlates pretty nicely with how much air is flowing across it.

Such wires do occasionally get dirty, and in fact you can buy MAF sensor cleaner to help clean it . . . but before you rush out and buy a can, the only vehicles it actually fixes are the ones where you’ve got some kind of oil leak that contaminates the MAF. Pretty much everything built in the last decade does not need the MAF cleaned ever.


The MAF only measures a certain amount of air passing it; there’s a little sampling tube that extends into the air inlet tube. Occasionally, bits of air filter get stuck in that and cause all sorts of problems. That’s really rare, unless you buy the absolute cheapest air filters on the market, or have a mouse chew holes in the air filter. Sometimes when that happens, the car will idle okay, but when you give it throttle, it moves that little bit of paper and changes the airflow over the sensor, and then things go bad.

This wasn’t what was wrong with the Buick.

The Buick had a bug.

Well, specifically, a spider.


I don’t know how conductive spider corpses are, but apparently they’re in the sweet spot where it’s not so much that the PCM knows the MAF value can’t be correct, but where it’s enough that it can’t correct the misfueling in the amount of time the engine stays running.

To preemptively answer questions in the comments, yes, we probably could have blown that spider out of there and the car would have been fine, but we don’t know that for sure--spider corpse melted on wire might permanently change its conductivity. The last thing we want is a comeback. Most customers aren’t happy with us telling them that we might have fixed their car or we might not have and they should just drive it and see. And in all honesty, as badly as this car was running, the customer was surely expecting a very hefty repair bill and instead it was only a couple hundred dollars, including diagnosis.


Now, there’s one more thing I should tell you about this car. It has nothing to do with the badly-running engine, nor was it a customer complaint.

You may recall in a blog post past, I mentioned how there are lots of sensors peppered throughout the car, and now that things are networked, problems are often reported to the nearest computer (physically), who then passes the message along the network.

This Buick had a headlight out.

You may also recall that way back when, we looked at a sad Cruze, and among its myriad complaints, it had a burned-out licence plate lamp. The Body Control Module had known that, and the BCM had set two codes--a B3883-01 (high resistance) and a B3883-04 (open)--as the bulb failed.

You’d be right to think that a newer Buick would know that a headlight had failed, and it did.

But you’ll never guess which module reported it.

The Tire Pressure Monitor.


Source

Comments ( 51 )

Stay tuned for next time, when I might talk about GM’s MOIST MOST network/bus. Y’all who are in IT are going to love it, especially when I give you data transfer rates.

With how much you complain about howAmerican car companies make bad decisions I wonder what you think of all the foreign car makes.

I took a computer programming class in college (back in the 1970s)
Prof mentioned that “bug” = “problem” dated back to the very early days of computers when
it was vacuum tubes & wires (looked sort of like a toaster rack) & was open because
you reprogrammed a computer by rewiring it, plugging & unplugging stuff.
Occasionally, a bug would literally get into the circuits & do just that

And spiders just keep spinning their webs......:trollestia:

As the spider is being sucked in to its death:

To the last, I grapple with thee; From Hell's heart, I stab at thee; For hate's sake, I spit my last breath at thee.

:rainbowlaugh:

5130772
the term "bug" goes clear back to middle english "bugge": monster or "budde": beetle, and has been documented as referring to engineering faults in the 1800s.

the later incident of a moth inside the Harvard Mk2 computer came with a note next to the taped moth "First actual case of bug being found." the people who found it were well aware of the term, and found a humorous irony in a literal bug being the cause of a bug.

5130791

To the last, I grapple with thee; From Hell's heart, I stab at thee; For hate's sake, I spit my last breath at thee.

“.....Oh, but who has the TIME for all that? You, sir, shall hear from my attorney!”

From my little bits of experience from working with cars, I feel sensors are the most fickle things.

But you’ll never guess which module reported it.

The Tire Pressure Monitor.

And this contributes to my amusement. :rainbowlaugh:

5130770
I lived in Japan for a good while, and grew up with a lot of Japanese cars. The models across most of the major car manufacturers are definitely reliable and sturdy--albeit some of them are pretty pathetic to look at and drive (looking at you, Cube drivers!)--and the third-party extensions integrate surprisingly well; the garages in Japan are no joke when it comes to aftermarket. I recommend checking out car shows there if you ever pass by!

a spider, i can say that is a new one for me.
now if you want to talk about squirrels or mice or the odd ground hog yep ground hogs seam to love wires and fuel lines.

love the art work grate drawings.

The complaint is that it barely runs, and stalls frequently.

Well... It is a Buick.

I remember tapping on Mass Airflow sensors to see if they were bad, if it altered how the engine was running it was the culprit. The Snap-On scanner with troubleshooter suggested it. GM was noted for it if I recall correctly. It’s funny that it was actually a bug 🕷!
The tire pressure monitor picked it up?! Huh. Ain’t that something?!

My first thought was a major vacuum leak.

Such wires do occasionally get dirty, and in fact you can buy MAF sensor cleaner to help clean it . . . but before you rush out and buy a can, the only vehicles it actually fixes are the ones where you’ve got some kind of oil leak that contaminates the MAF.

Eh, GMs and Fords get dirty ALL the damn time. Cleaning them is part of my tune-up routine. Usually they don't have an oil leak or anything, especially given that pcv tubes are usually much further down the line. Excepting Buick 3.8Ls where the MAF is literally right in front of the throttle plate :/ I've never had much luck with MAF cleaner spray. I just use brake cleaner. :ajsmug:

The more modern hot film type of sensors never seem to get dirty, though. Also, my favorite German word (not that I know many) is: heissfilmluftmassenmesser (literally: hot film air mass meter). It's almost as fun a word as antidisestablishmentarianism! :pinkiecrazy:

Spider.

A spider.

...

I am so staying away from that one. :applecry:

5130796
Well, it’s been 45+ years, so I stand corrected.

The prof said he got into computers because his first after school job was working on computers.
He had a pair of gloves & a bushel basket filled with vacc tubes. He went around pulling & replacing bad tubes.
According to him, he’d just about go through the whole basket every shift.

jxj

Huh, a literal bug is a new one for me. I mean, it makes sense but I wouldn't have thought of it.

You’d be right to think that a newer Buick would know that a headlight had failed, and it did.

But you’ll never guess which module reported it.

The Tire Pressure Monitor.

ugh, this still annoys me

Stay tuned for next time, when I might talk about GM’s MOIST MOST network/bus. Y’all who are in IT are going to love it, especially when I give you data transfer rates.

this one's going to annoy me as well isn't it.

The bug in the Buick

I knew American cars were big, but you found a Volkswagen there?

Spider

It’s worse! It sucked up an Alfa Convertible!

Speaking of common rail... remember the old days of mechanical diesel fuel pumps having to be replaced? Me neither.

Dan

5130772

Yes. And the term existed before Admiral Hopper famously found a moth in the Harvard Mark II.

FTL

The Buick had a bug.
Well, specifically, a spider.

Well, that is just more evidence that nature hates our cars.
I had a similar experience with a 70's telephone exchange. It was run by ITT3200 processors (these looked just like 60's sci-fi computers). It used a customised high voltage Fast Bus (well, fast for the 70's) and this bus ran from the processor room into the switch block in the main switching floors. The bus used wire wrap terminations in the top of the subracks and late on one Friday afternoon we lost the 'B' bus from CPUs 1 and 3. After several hours of investigations with logic state analysers and such I eventually was the one who found a medium sized spider who had a lot in common with a certain parrot. I got down from my ladder and told my trainee to go up and have a look and to see if he could see what the issue was... it was then that I found that he was an arachnophobe when he jumped off the top of the ladder with a cliché girly squeal and then ran back to the control room while swearing a purple streak. :rainbowlaugh:

You’d be right to think that a newer Buick would know that a headlight had failed, and it did.
But you’ll never guess which module reported it.
The Tire Pressure Monitor.

I know there is a perverse logic to how automotive engineers build these networks but at the same time it just makes me want to beat them with a stick when they do things like this. :facehoof: Seriously, if a module now monitors tyre pressures and front bulbs, probably along with yet more things, then it is no longer a TPMS but now a Body Management Module... call the spade a spade, not an egg-flipper with bigger duties. :twilightangry2:

5130826

My first thought was a major vacuum leak

'Twas my first thought as well... we both lose. :fluttercry:

5130861
I understood...I understtod that reference...

...fuel composition sensor. They measure how much alcohol is in the pony fuel, and adjust the injector on-time (pulsewidth) accordingly.

First, where is this sensor located on my pony? And an associated question: how do I access it? The alcohol injection system on my pony keeps alternating between "lean" errors and "rich" errors, with very little time between them. I'm assuming that's why my pony is so erratic.

And subsequently, what kind of readout should I expect to see, and what values are nominal? Because I'll naturally need to know in what direction(s) to adjust the sensor (or replace) to improve my pony's alcohol intake efficiency. (Inconveniently, it didn't come with an owner's manual, for some reason.)

EDIT: I reread your blog and found this:

Time to try basically poking things with a stick and seeing what happens.

So I tried that. My pony's rear-mounted security covering and Fly Dispersal Unit (FDU) lifted, but I don't understand the readout at all.

5130861

The bug in the Buick

My automobile had been running poorly. Barely running at all, to be truthful. So I took it in to a mechanic.

I told the mare it had been acting strangely lately. It would turn on, but there was always a delay and it was always jerky when it started out, though it would smooth out after a few minutes. The muffler was acting odd, too: there was hardly any smoke coming out, and what little did smelled oddly sweet.

She asked if I'd taken a look under the hood. I embarrassedly told the mare that I loved the vehicle like a family pet, but my beautiful wife had absolutely forbidden me from working on anything mechanical after the pedalthopter incident.

As I was wending my way through that anecdote, she mosied around to the front of the vehicle and popped the hood open, and a belch of green flames greeted her immediately.

"Exactly what I was afraid of," she said. "Changelings."

My brother has an '05 Honda Accord. We were fast asleep in bed one night, when suddenly his car alarm starts going off out of the blue. We look out the windows to see who might be messing with it, but see no one. We go out and investigate the car -- no signs of attempted entry or vandalism, and so we lock it up and head back in.

The alarm starts again.

We go through this same cycle once or twice, and finally get fed up and leave the doors unlocked for the night.

The next morning, we're out there investigating the car top to bottom for issues. We look under the hood, thinking maybe the battery was short-circuiting or something -- nothing we can see. We check the door seals and seams to see if something might be loose. Nope.

It was only upon closer inspection inside the doorjamb that we bothered to look and see something odd inside one of the latches. A little bit of digging later, we pulled out a curled up dead wolf spider that had built a part of a web inside it. Don't ask me how it got in there or why it started making problems all of a sudden, but this little tale reminded me of that incident.

Fuckin' spiders...

Weird. I thought they used a couple piezos to blast ultrasound accross the intake pipe as different density has different properties and flowrates change how far teh signal has to travel? Then again, depends on the sensitivity needed. As for spider resistance, try a resistance meter accross the terminals in place, then compare with propelling pencil leads or such?

Because Im old and purely mechanical or Cultrure class designs, my first though was failed oil pump, giving a thermal response where the engine heated up on running and stalled.But because it wasnt loaded, it was repeatable.

Airplanes use pitots which is two pipes and a piezo?

5130770

With how much you complain about how American car companies make bad decisions I wonder what you think of all the foreign car makes.

They make different bad decisions. You gotta be drunk (on German beer) to understand how VW/Audi group their data; Mercedes and BMW require special tools nobody’s ever heard of (although BMW does give you instructions on how to make the wet water you need to install the control arm bushings, so that’s something). The less said about Fiat, the better.

The Asians have their own special way of doing things. Honda makes big motorcycles with four wheels, Toyota innovates weirdly sometimes, Subaru’s inspiration comes from construction machinery, and Nissan channels Rube Goldberg anytime they make a moving part. Kia and Hyundai take most of the best ideas other automakers have and copy them, and they get better at that every year.

5130772

Occasionally, a bug would literally get into the circuits & do just that

Spider webs and insect nests have brought down aircraft before.

You can’t design a system that’s totally safe from insects, but what you can try to do is make the failure manageable.

5130788

And spiders just keep spinning their webs......:trollestia:

Yes, they do.

They sometimes clog evaporator case drains with their webs, actually.

5130791
And let’s be honest, in death the spider won that round.

5130799

From my little bits of experience from working with cars, I feel sensors are the most fickle things.

But you’ll never guess which module reported it.
The Tire Pressure Monitor.

And this contributes to my amusement. :rainbowlaugh:

It actually makes sense to have the network distributed and have the module closest to the sensor report, rather than run a dedicated wire all the way back to the computer you think should report . . . but it does lead to some really weird complaints from unexpected sources.

I lived in Japan for a good while, and grew up with a lot of Japanese cars. The models across most of the major car manufacturers are definitely reliable and sturdy--albeit some of them are pretty pathetic to look at and drive (looking at you, Cube drivers!)--and the third-party extensions integrate surprisingly well; the garages in Japan are no joke when it comes to aftermarket. I recommend checking out car shows there if you ever pass by!

Japanese cars tend to be really reliable if you do what they tell you to do when they tell you to do it, in terms of maintenance. They did have trouble in the US with understanding that we put salt on our roads, and a lot of early Japanese cars suffered from critical rust (faster than their American counterparts, which of course also rusted). It’s been my understanding that they’re generally not given to the American idea of replacing an old thing with a shiny new thing; they instead like to make incremental improvements to an existing product. That tends to leave them lagging a bit in terms of innovation, but they also don’t often experience the massive failures many American automakers do.

5130800

now if you want to talk about squirrels or mice or the odd ground hog yep ground hogs seam to love wires and fuel lines.

Oh, yeah, rodents are hell on wiring. Luckily, we don’t see a whole lot of it--it’s not overly common in daily drivers--but when we do, it’s usually a mess. A while back I posted some pictures of a Jeep Wrangler that needed a new main engine harness ‘cause the mice had decimated it.

I also had to pull an intake and injector rail on a GM 3.6 to replace a harness the mice got. How they even got down there boggles me.

love the art work grate drawings.

Thanks! Sometimes I spend nearly as much time finding the right pictures as I do writing the blog.

5130813

Well... It is a Buick.

Hey, now, them’s fighting words.

5130820

I remember tapping on Mass Airflow sensors to see if they were bad, if it altered how the engine was running it was the culprit.

Man, that’s way old-school. The new MAFs don’t fail that way any more.

The Snap-On scanner with troubleshooter suggested it. GM was noted for it if I recall correctly. It’s funny that it was actually a bug 🕷!

Most of the GMs I work on have two fuel strategies, and they tend to be good at ignoring the MAF if it fails . . . that might have been from lessons learned early on, now that I think about it. Unlike Ford, GM kept a MAP sensor even when they had a MAF, so they had a second option for monitoring airflow/engine load.

The tire pressure monitor picked it up?! Huh. Ain’t that something?!

Like, from the engineering standpoint it makes sense, but it was weird to actually see it in practice, especially for the code to be in that module where you wouldn’t expect to see it, rather than it being passed to the BCM and coding there.

5130826

My first thought was a major vacuum leak.

That isn’t a bad thought. Usually, you can hear those, and often if you give the car more throttle (to overcome it), it’ll run--the Buick didn’t. I didn’t put that in the blog, ‘cause it wasn’t totally germane.

Eh, GMs and Fords get dirty ALL the damn time. Cleaning them is part of my tune-up routine. Usually they don't have an oil leak or anything, especially given that pcv tubes are usually much further down the line. Excepting Buick 3.8Ls where the MAF is literally right in front of the throttle plate :/ I've never had much luck with MAF cleaner spray. I just use brake cleaner. :ajsmug:

I almost never clean ‘em, and haven’t had any troubles. Maybe it’s the difference between living in the city and living in the country. Maybe corn pollen helps keep MAF sensors clear, who knows?

The more modern hot film type of sensors never seem to get dirty, though.
Except for the GM 3.6L engines that fill the air snorkel with a nasty water/oil mix . . . we’ve got a customer who didn’t want to buy the updated valve cover to prevent that, so he rigged up his own special arrangement to catch it and dump it out. Amazingly, he did it well enough that the vehicle not only runs like it should, but it doesn’t set any codes.

Also, my favorite German word (not that I know many) is: heissfilmluftmassenmesser (literally: hot film air mass meter). It's almost as fun a word as antidisestablishmentarianism! :pinkiecrazy:

German words are fun, ‘cause you can scrunch lots of things together to make a word.

I really like the french “lames de turbocompresseur” . . . there’s a bilingual warning on the old 6.5 diesels to not stick things you want to keep down the turbo inlet, and it includes those words in the warning.

5130828

Spider.
A spider.
...
I am so staying away from that one.

I know, right? At least it was dead, so that’s something.

5130857

Huh, a literal bug is a new one for me. I mean, it makes sense but I wouldn't have thought of it.

It’s one of the few failures I’ve seen where an actual insect is the cause.

ugh, this still annoys me

It legit makes sense, even though it’s weird. The TPMS not only runs near all four corners, in GM vehicles, it can control some of the lights.

this one's going to annoy me as well isn't it.

Probably some yes, some no. There’s stuff in there that would make a modern network engineer cry, but at the same time, there’s stuff in there that’s brilliant because of the nature of the network.

5130861

I knew American cars were big, but you found a Volkswagen there?

i.imgur.com/j1iJMGu.jpg
Cars carrying cars is an American hobby.

Spider
It’s worse! It sucked up an Alfa Convertible!

:derpytongue2:

5130880

Speaking of common rail... remember the old days of mechanical diesel fuel pumps having to be replaced? Me neither.

Heh, I did one in a Cummins a few years ago. Customer provided the part, and luckily it worked.

They did to tend to be pretty robust, though, I’ll give you that.

To be honest, we’ve seen lots of GDI cars recently, and I’ve done a fair bit of work to them, and I think I’ve only had to replace one or two high pressure pumps so far. Still do lots of electric in-tank pumps, though.

5135186

Amazingly, he did it well enough that the vehicle not only runs like it should, but it doesn’t set any codes.

Oh that's cool.

5130951

Well, that is just more evidence that nature hates our cars.

Can you blame her?

I had a similar experience with a 70's telephone exchange. It was run by ITT3200 processors (these looked just like 60's sci-fi computers). It used a customised high voltage Fast Bus (well, fast for the 70's)

So, like two bytes/second? :derpytongue2:

After several hours of investigations with logic state analysers and such I eventually was the one who found a medium sized spider who had a lot in common with a certain parrot. I got down from my ladder and told my trainee to go up and have a look and to see if he could see what the issue was... it was then that I found that he was an arachnophobe when he jumped off the top of the ladder with a cliché girly squeal and then ran back to the control room while swearing a purple streak. :rainbowlaugh:

Most auto techs are okay with insects, but there are lots who are not okay with snakes. I’ve heard lots of fun stories mostly from guys in the south about finding snakes in cars. To my mind, that’s a solution to the mouse problem. :rainbowlaugh:

I know there is a perverse logic to how automotive engineers build these networks but at the same time it just makes me want to beat them with a stick when they do things like this. :facehoof: Seriously, if a module now monitors tyre pressures and front bulbs, probably along with yet more things, then it is no longer a TPMS but now a Body Management Module... call the spade a spade, not an egg-flipper with bigger duties. :twilightangry2:

The thing is, it gets confusing when you start to move off the reservation when it comes to naming modules . . . . there are actually standards for that, at least to an extent (moreso with the coding; OBD-II codes have to have a specific format, and they’re universal across platforms [a P0300 is a generic misfire no matter what kind of car sets that code]), plus you’ve got legacy systems and legacy scan tools.

Part of it might be to save on bandwidth, too, especially since in the last few years it’s been commonplace for scan tools to scan all modules with one command, so it’s no big thing to get a list of all the trouble codes, wherever they are. Ten years ago when you had to do each module individually, that was a different matter.

Back then, Chrysler especially had informational codes--the PCM would have a P0700, transmission fault, which meant you ought to check the transmission control module to see what faults it had.

My first thought was a major vacuum leak

'Twas my first thought as well... we both lose.

It was also mine, but I quickly dismissed it since I didn’t hear it, and more throttle didn’t improve things.

Fair warning, a lot of times there’s stuff I’ll leave out of mechanic blogs because I ruled it out quickly.

5135197

Yeah Cummins are great engines. The 5.9L 12Vs were practically indestructible. Same with the Mercedes OM617. I have 6.7 Cummins now. It's great, but I do worry about the common rail system in it. It is extremely expensive to fix, and unlike the OM617, I don't think the injectors just come out with a ratchet. lol

5130959

First, where is this sensor located on my pony? And an associated question: how do I access it? The alcohol injection system on my pony keeps alternating between "lean" errors and "rich" errors, with very little time between them. I'm assuming that's why my pony is so erratic.

Hmm, yes, that is a common problem. Proper pony fuel is a delicate balance, and of course varies from pony to pony. My first thought would be to apply as much friendship as possible, as that often serves a buffering function. Hugs and ear-scritches are most effective, at least for quick results, but you might also find long conversations over dinner or while watching movies to be effective. Walks on the beach are good with pegasi especially--they like the ocean--or through largely-open parks with lots of greenery, and ideally flowers. You could also consider several smaller meals spread over the day, rather than the traditional three humans prefer.

Kirin are the most sensitive of the pony tribes, and if you’ve got one of those you’ll have to be very careful with topics of conversation; they switch from friendly to on fire at the drop of a hat, often with very little warning.

And subsequently, what kind of readout should I expect to see, and what values are nominal? Because I'll naturally need to know in what direction(s) to adjust the sensor (or replace) to improve my pony's alcohol intake efficiency.

Typically, you’ll want to see fairly equal parts whimsy and solemnity, with the ratio varying off-center by tribe. Unicorns trend towards solemnity, while pegasi are more whimsical. Earth ponies tend to be centered, although of course there are variations from pony to pony. Observation over time is essential! Many novices make the mistake of believing all ponies are the same, and they’re not.

Ponies in general tend to be self-adjusting when it comes to alcohol intake, and extremes are typically a sign of a fault in some other system. Most often the cuddling matrix; that should be checked before making any other diagnosis.

(Inconveniently, it didn't come with an owner's manual, for some reason.)

That would be genre rule #5, prohibiting user guides.

So I tried that. My pony's rear-mounted security covering and Fly Dispersal Unit (FDU) lifted, but I don't understand the readout at all.

Was it a hard poke--a sharp jab--or more of a firm pressure or a gentle touch? There are right ways and wrong ways to poke a pony, and using the inappropriate method may lead to inconclusive results.

5131414
:rainbowlaugh:

Still, there’s something to be said about getting horsepower just by really loving your car. Could be the tech of the future, you know?

derpicdn.net/img/view/2017/7/24/1493788.png

5131590

It was only upon closer inspection inside the doorjamb that we bothered to look and see something odd inside one of the latches. A little bit of digging later, we pulled out a curled up dead wolf spider that had built a part of a web inside it. Don't ask me how it got in there or why it started making problems all of a sudden, but this little tale reminded me of that incident.

Man, I wouldn’t have expected that. I assume it messed up the door latch switch just enough that the car thought it was being broken into as the signal changed, when a piece of the web or the spider touched just the right thing in just the right way. . . .

Now that y’all have read this and are commenting on it, I’m working through my memories, trying to think of the weirdest thing I’ve ever seen malfunction on a car--either the oddest root cause, or the oddest failure. I can’t think of anything in particular, and I can’t help but wonder if that’s because I’m old and jaded, or if it’s because my memory’s shot from too many years of breathing brake cleaner fumes. :derpytongue2:

Fuckin' spiders...

I know, right?

5131632

Weird. I thought they used a couple piezos to blast ultrasound accross the intake pipe as different density has different properties and flowrates change how far teh signal has to travel? Then again, depends on the sensitivity needed.

You could do that, I think, but as far as I know, that’s not done on cars. Nothing I’m aware of, anyway. They do have barometric pressure sensors which they use to calculate the expected density of the air, and depending on the type, you could potentially confuse the car by driving from the bottom of Death Valley to the top of Pike’s Peak without ever shutting off the engine . . . but in normal usage, that flaw isn’t a problem.

As for spider resistance, try a resistance meter across the terminals in place, then compare with propelling pencil leads or such?

I suppose I could have used my meter to check the resistance of the spider. Didn’t think of trying that, but I can probably find another dead spider to try it with.

Typical MAF, there’s also the heat load to consider, and that would be more difficult to measure with the equipment I’ve got.

Because Im old and purely mechanical or Cultrure class designs, my first though was failed oil pump, giving a thermal response where the engine heated up on running and stalled.But because it wasnt loaded, it was repeatable.

Usually a failed oil pump would cause catastrophic mechanical failure before it affected any engine sensors, at least on an internal combustion engine or a compression engine.

Engine heat does play a role in lots of engine management, and as a result vehicles have open loop (cold) operation and closed-loop (warm/hot) operation, and there are problems I’ve seen that only appear when it switches modes.

Airplanes use pitots which is two pipes and a piezo?

I don’t know what’s in the tube that makes the magic happen, but they do use tubes, I know that for a fact. Airplanes also have problems sometimes with insects building nests or webs in the tube.

5135199
It’s a contraption, but it works, and sometimes that’s all you can ask for.

Speaking of which, just today I built a modified TPMS sensor (for reasons) which involved shaping two washers to fit, shortening the stem of a rubber TPMS valve, and also shortening the bolt, in order to make a sensor that was supposed to attach to a metal valve stem work on a rubber one.

I can’t speak (yet) to the long-term viability of my sensor, but in the short term, it works.

5135209
I’ve got a somewhat uncommon 3.9L 4BT, 1980 vintage. As yet, I have no idea what I’m gonna do with it; I have no real use for the vehicle it’s in--other than as a storage shed, and it doesn’t need the engine for that.

I’ve been considering getting motivated and installing it in a full-size pickup.


I’ve also got a Suburban with a 6.2L Detroit Diesel that I really ought to get back on the road one of these days.

5135247 That's cool. All in a day's work for a mechanic, eh?
I've got a guy coming in tomorrow with something I will similarly need to finagle. The reverse light wiring was missing, ordered a harness, dealer swore up and down that it came with bulb sockets (also missing). Guess what it didn't come with? :trixieshiftleft:

5135649

That's cool. All in a day's work for a mechanic, eh?

Yeah, although I’ll be honest, I’d rather just install the correct part and be done with it. I mean, it’s not you can’t get a proper TPMS sensor for a Town and Country, my manager just forgot to make sure we had enough of them. :derpytongue2:

Yesterday, I got to modify a socket for the ratcheting nut in the rear axle of a F350 . . . that was fun times.

I've got a guy coming in tomorrow with something I will similarly need to finagle. The reverse light wiring was missing, ordered a harness, dealer swore up and down that it came with bulb sockets (also missing). Guess what it didn't come with? :trixieshiftleft:

I freaking hate that. NAPA and CarQuest have gotten better at having pictures of what you actually are supposed to get in the box in their online catalog, which is nice. Dealer parts, you’re at the mercy of what the parts guy tells you it comes with or what you have to actually order to get the part you want.

Not that long ago, I had a guy at a Chrysler dealer try and tell me that there are seventeen(!) bolts that hold the exhaust manifolds on a 4.7L, said bolts are $13.00 each, and that there are no manifold-to-block gaskets on that engine. I thanked him, and then called another dealer to get the correct parts and prices.

5135851 LOL. Dealers, man. But you know, I feel like there was one car I worked on that did not have an exhaust manifold gasket. Can't remember. No way in hell it was the 4.7 though lol

Login or register to comment