• Member Since 2nd Nov, 2012
  • offline last seen 3 hours ago

Admiral Biscuit


he/him

More Blog Posts825

  • 1 week
    Signal Boost: Autumn Blaze, Formula One Driver

    Sock Puppet's at it again, giving us another entry into my Not-A-Contest . . . and giving ponies another shot at driving Formula 1 cars. Before too long the whole F1 field will be ponies (or pony adjunct), and I for one would embrace

    Read More

    5 comments · 158 views
  • 1 week
    August State of the Author, with bonuses!

    It's August! And I'm secretly working on a secret project which I can't tell you about because it's a secret.

    I also can't find a good, topical opening image.


    Source

    Read More

    37 comments · 338 views
  • 3 weeks
    Power Outage Followup

    I promised a second blog with more pictures, and this time you're not gonna have to wait a month or two to get it!

    There'll even be a video which isn't exciting but it does have pretty flashing lights.


    Source

    Read More

    35 comments · 517 views
  • 3 weeks
    My Bad Neighbor*

    I was gonna start this one out by telling you about my bad neighbor–you remember, the one who crashed a car across my driveway and left it there?

    Then I remembered that I never blogged about that, so you’d have no idea what I’m talking about.

    Read More

    99 comments · 607 views
  • 3 weeks
    After Trotcon!

    Hey, I'm way more on the ball with this than I was with my SeaquestriaFest blog.


    Source

    Which I still haven't written yet. But I'll get to it in due time!

    Read More

    27 comments · 353 views
Feb
7th
2018

Fun with pulse-width modulation, and bare wires · 2:16am Feb 7th, 2018

What follows is the diagnostic procedure on three different vehicles, with three different results.

We're going to start with a Cadillac.

Y'all may remember from my network blog that various computers on the car are networked together, and how if there are communication problems, you can have issues that you might not expect?

I can't say for certain exactly what symptoms this car presented with.

See, my manager is a Cadillac expert. He worked at a Cadillac dealership for somewhere between ten and a hundred years, and I don't question him on that. I personally hate Caddys, especially of the era that they were putting the Northstar engine into them, so if he wants to diagnose every one that comes through the door, that's fine with me. And to be honest, I'm not all that familiar with the GM computer we've got, and he is.

So when this car first came in, I didn't really have to think about it or worry about it at all. He worked his Goodwrench(TM) diagnostic magic on it, and concluded that it needed a driver's door module (DDM [not to be confused with DDR]).


From an episode via Derpibooru

We didn't have one, and couldn't get one right away, and it wasn't something that rendered the car inoperable, so the owner took it away while we were waiting for parts to arrive.

By last Monday, the parts had arrived, and the customer brought the car in.

Now, I was totally out of the loop on it. My manager had decided to take an extra day off work so that he could sit in the woods and shoot at Bambi, and he figured that the Monday after Thanksgiving would be a perfect day, because nobody would be bringing their car in for service.

Which turned out to not be the case.

The customer dropped off the car, and there was only one revision in our system, for the DDM. And the customer helpfully informed me that the manager had called him and said that the module had arrived, and indeed, when I rummaged around on the parts shelf, there was a DDM. Used, but that's okay; those parts rarely fail, and a lot of times with computers the difference in the price for a new one vs. a used one is in the hundreds or sometimes thousands of dollars.

So I pulled the car in and installed the module, then I tried to figure out how to program it.

It turned out that I couldn't get the computer to log on and access the scan tool—somebody had changed the password and hadn't bothered to write it down, and the laptop locked me out. But that was a moot point; I thought that maybe I could program it with our Snap-On tool, since according to Identifix, there wasn't much you had to do, and that's when I found out that half the Class 2 network—including the DDM—was down.

I printed out a network topography diagram, and figured out that there was no particular logic to what was offline, at least in terms of the network wires. They went through two different splice packs, and there were some modules on the splice packs that worked, so it wasn't that kind of failure.

Since we had a lot of other stuff to do, I just backed the car out of it and figured out that I'd let him do it.

🦄🦄🦄

When he got back, he asked me about the car, and so I told him that I'd gotten the new module in but I hadn't been able to program it. I could have told him that I couldn't communicate with that part of the network at all, but I figured I'd let him figure that out. Maybe that part of the network didn't function before the module was programmed—it could be that the other modules who were down needed a wake-up call from the DDM.

Well, as I'm sure you can guess, he couldn't program it either. And he said that he hadn't actually diagnosed it; he'd sold the job but planned on diagnosing it after the part arrived.

It didn't take him too long to figure out that some fuse was blown (I don't know which one; I was working on something else). When he plugged it back in, smoke started coming out from the door panel.

Now, I've never liked condemning computers unless I can find no other cause. For all the griping people do about them, in truth they're generally pretty robust, and I'd at least like to verify that I have power and ground and communications to the computer before I'm going to start assuming that it's bad. That's not always fun, especially when you have to do something like pull a door panel off to get to the wires.

On the other hand, it saves you from looking like an idiot when you discover the actual cause of the failure.

In this case, you wouldn't have to be very knowledgeable about vehicle electrical systems to identify the problem.

🦄🦄🦄

Some of y'all might be wondering why I didn't see that when I replaced the module. Because I'll be honest with you, those wires are right behind the module, on the other side of the weather shield on the door.

Well, Cadillacs have these special clips that hold the door panel on, and they're a bastard to remove. People complain when they go to shut the door and pull the handle off in their hand, so Cadillac made sure that that wouldn't happen, and the door panel stays on.

The way I've always gotten them off is after I pop the plastic Christmas trees around the perimeter is reach in with a long pair of needle nose pliers while somebody else tugs on the door. And I was going to do that, until I realized that I could see and reach the module from under the door, without taking the panel off. So that's what I did—I reached up, unclipped the old one, clipped the new one in, and then popped the Christmas trees back in place. I couldn't have seen it from where I was working.


Now onto another electrical problem, yay!

This particular truck is the sister of that white F250 that I blogged about a while back. It's a little bit newer, and doesn't have as many miles; it's also bigger. It's got dual rear wheels.


Source
It's also got about ten thousand dollars worth of add-on accessories that this particular truck doesn't, but you get the idea.

It came in for the four wheel drive not working. Sometimes when the customer selected four wheel drive, the light would flash but nothing else would happen. Other times it would work normally.

I'm going to step back a bit here and give a brief overview of four-wheel drive control systems. See, you typically can't have a four-wheel drive system active all the time. When a vehicle goes around a corner, all four wheels have to turn at a different rate. If they're all locked together, they can't do that unless one of them slips [the same applies to two wheels on the same axle, which is why differentials are designed to allow for this movement]. Not a problem off road or on slippery road surfaces, but a major issue on dry pavement.

Back in the day, you'd engage the system by pulling a lever, and usually getting outside to lock the front hubs, too. And that was fine when four-wheel drive was only used on farm trucks and Jeeps. But when there started to be a demand for it on everyday cars (or when the automakers thought that there was), they wanted to make it more automatic.

GM was one of the first out of the gate with a truck transfer case that you could shift on the fly. The system didn't last long—only one model year, I think—because it turned out that you could put the truck into four low at highway speed, and unfortunately the drivetrain could take it. The engine overreved, and all four wheels locked, and if you weren't belted in, you went flying, too.*

Not as fun as it looks. Wear your seatbelts, kids.

After that, automakers started including ways to not let you put it in the wrong gear, and normally required that you be stopped before you could put the truck into low range.

As things progressed in the automotive world, we went from mechanical linkages and vacuum pots to a computer. When you turn the dial on the dash, the transfer case control module (TCCM) decides if everything looks good, and if it does, it locks the front axle and engages the transfer case. If not, it flashes the light.

Luckily, most of those TCCM modules can be communicated with, and when it doesn't work, you can see what the fault is with a good scan tool.

And that's often a blessing on intermittent conditions, because even if it isn't acting up when we go to look at it, the modules usually remember what went wrong, and it gives us a direction to go.

Naturally, this module had no diagnostic codes.

However, we could still follow the troubleshooting guidelines, and verify that all the inputs were what they should be, and that the outputs looked good.

It didn't take too long for the tech to find a problem. The diagnostic instructions said that the yellow wire should have battery voltage on it, and it did not. It showed up on the meter as nine volts and change, and it kept fluctuating.

He reasoned that it was possible that sometimes the voltage was high enough for things to work right, and other times it wasn't.

Since this was an output from the TCCM, he accessed the module behind the dash, and found the same condition at the module. All other powers and grounds were good, so obviously the module was bad, right?

Well, you know it wasn't. If it had been that easy, there wouldn't be a blog post about this truck.

We put a new module in, programmed it, and at first the problem seemed to be fixed, but then it decided it didn't want to go into four-wheel drive. Of course, it didn't set any codes. When we shut it off and started it up again, it did go into four-wheel drive, and it did it for a hundred times in a row (I know, because I was the one test-driving it) before it malfunctioned again.

After further experimentation, it seemed to not like going into four-wheel drive when there was any kind of load on the drivetrain. Say, you were turning a corner and wanted four-wheel drive, you wouldn't get it. Driving straight, no problem.

Sometimes it wouldn't come out of four wheel drive, either. Especially if you had a load on the drivetrain. It would try; the light would flash for a bit, and then it would return to normal.

If it did go into four-wheel drive, you could switch between high and low with no trouble.

🦄🦄🦄

After an embarrassingly long time, we eventually figured out that the component that was actually bad was the relay.

The easiest way to think of what this relay does is think of it like a power window switch. When you push it one way, it puts the power to one wire and runs the motor in one direction; when you push it the other way, it reverses the polarity, and the motor goes in the other direction. Window goes up, window goes down. Four-wheel drive goes on, four-wheel drive goes off.


Source
Simple.

"So what about that yellow wire?" I can hear you asking. "The one that had nine volts on it when it should have battery voltage?"

Well, let me tell you about that yellow wire.

But first I'm going to tell you about your microwave, and believe me, the two things are related.

On a conventional oven, there is a heat setting, because different foods cook at different temperatures. That doesn't work on a microwave, though; you either have microwaves or you don't. You can't change the size of them depending on what you're trying to cook. All you can do is change how many of them you have.

And how you do that is turn the transmitter on and off. So that power level button? When you're at full power, it blasts out a steady stream of microwaves. Set it at 50%? It turns itself on and off rapidly, so that it's only on fifty percent of the time.

That's called pulse-width modulation, and there are actually a lot of things that use it. Some computer turns a component on and off rapidly in order to get the output it wants.

Now, if you were to measure this with some kind of meter, what you'd find is that you aren't getting what was advertised. To go back to our microwave, if we had a thousand watt unit, and we were running it at fifty percent, and if there were some sort of tool you stuck into microwaves to see what the output was, you'd find that you had some kind of reading that wasn't a thousand watts. It would be less. How much less would presumably depend on the sample rate of the testing tool; theoretically, it would be five hundred watts.*

That yellow wire was pulse-width modulated. It sent out battery voltage, yes, but not all the time. Sometimes it sent out nothing. And voltmeters average the voltage they get.

________________________________________
*I think; I'm not sure how long the microwaves bounce around in there before they leak out or whatever it is that they do.


If I'd been my manager, I'd have yelled and screamed at the guy who diagnosed the truck for not knowing that he was supposed to be getting a pulsewidth modulated signal. What I did, instead, was read through the diagnostic instructions that he'd been using. Because the fact is that we do make mistakes, and we overlook things, and even though I now knew what had actually been wrong, I thought that it was important to know where the process had failed. That's the only way that we're going to learn how to prevent that kind of mistake in the future.

And what I found was eye-opening. The instructions never said that the signal was pulse-width modulated.

In fact, I wouldn't have known that if I hadn't managed to get my hands on actual Ford instructions. Ford was concerned that a tech might make this mistake, and had in fact very specifically stated that the signal was pulse-width modulated. They'd even put it in a little gray box with a yellow warning triangle, to ensure that you noticed.


Source

Mitchell, however, hadn't felt that that was important enough to include in their instructions.

Sometimes all you can do is cry.


On the plus side, this did pay dividends later on. I was diagnosing a heater circuit failure in an oxygen sensor on a Chevy truck a few months later. It was supposed to be battery voltage on the wire, but it wasn't, and it kept fluctuating.

While I knew from experience that oxygen sensors are more likely to fail than the PCM (who controls the heater), a bad oxygen sensor can kill the drivers in the PCM—and I have seen that happen.

But then I got to thinking that maybe it was pulse-width modulated, and I did some more research. While the diagnostic instructions once again failed to mention PWM, in the theory and operation section, it did say that the computer pulse-width modulated the oxygen sensor heaters.

That was good enough for me, and so I put a new oxygen sensor on it, and that made the truck happy.


Source

Comments ( 51 )

*While I haven't had the joy of experiencing a vehicle going into Four Low on the highway, I did have an old S-10 that one day decided it wanted to downshift two gears while I was doing 70, and it also wanted to keep the torque converter locked while it did it. I heard and felt the transfer case slam into the cab floor as the drivetrain wound up, and the tach went 2,000 rpm over redline. I can't believe that nothing came apart when that happened, especially given how worn-out the whole truck was.

Also, does anyone know what that happy truck is? It seems to be some kind of Russian off-road vehicle (at least, every site I found with it was in Russian).

4791238

That is one happy roly-poly wagon. All it needs is some googly additions to its headlights.

At least you havent been faced with bi phase error corrected pulse width modulated signals yet. You know, when you want to put two signals down and save money by using a single wire? :pinkiecrazy:

jxj
jxj #4 · Feb 7th, 2018 · · ·

So I pulled the car in and installed the module, then I tried to figure out how to program it.

do they not come flashed from the manufacturer?

Now, if you were to measure this with some kind of meter, what you'd find is that you aren't getting what was advertised. To go back to our microwave, if we had a thousand watt unit, and we were running it at fifty percent, and if there were some sort of tool you stuck into microwaves to see what the output was, you'd find that you had some kind of reading that wasn't a thousand watts. It would be less. How much less would presumably depend on the sample rate of the testing tool; theoretically, it would be five hundred watts.*

And that's why I just use a scope 90% of the time.

Mitchell, however, hadn't felt that that was important enough to include intheirinstructions.

yeah, that sounds pretty important

jxj

4791286
I haven't run across that one before. That does not sound fun.

It sounds like the Mitchell books will soon be in the trash or recycling bin after that level of an error.

4791279

Thatisone happy roly-poly wagon. All it needs is some googly additions to its headlights.

Dude, that would be so awesome.

Actually, I kind of want one of them now.

4791286

At least you havent been faced with bi phase error corrected pulse width modulated signals yet. You know, when you want to put two signals down and save money by using a single wire?

Pfft, that's old hat for the automakers. They multiplex all sorts of things (although for critical systems, they use a twisted pair with a redundant signal rather than just a single wire). Old GM LAN is all one-wire stuff, and so is Chrysler's I-can't-remember-what-it's-called. :derpytongue2:

The really cool thing with the one-wire system is that if one module goes, the whole network dies.

4791303

do they not come flashed from the manufacturer?

Depends on what it is, but the general rule these days is that they don't. Which makes sense; the same DDM might work in all Cadillacs for a given model year, but it needs to know things like how many doors the car has, whether it's got heated seats, memory seats, and so forth.

Sometimes you can get modules pre-programmed from aftermarket sources, but that's pretty rare for new stuff.

In the early days of PCMs, there was an EEPROM chip that you transferred from your vehicle to the new PCM--that was how it was 'programmed.'

And that's why I just use a scope 90% of the time.

That's not a bad idea, although usually it's not worth the trouble to hook up and configure. Now, if I had a good graphing multimeter, that would be nice--I'd use that a lot. But they're expensive.

yeah, that sounds pretty important

I could do a whole blog post about Mitchell screwing things up when they re-write instructions. Sometimes you've got to get creative to figure out what they're telling you.

4791334

It sounds like the Mitchell books will soon be in the trash or recycling bin after that level of an error.

It's an endemic problem in the auto industry. Mitchell isn't the only one that gets things wrong; Identifix sometimes has bad instructions, and occasionally even the directions from the automaker aren't right. Heck, not that long ago, I was trying to program a PCM on a Lincoln Zephyr, and I had to lie to the Ford scan tool and tell it that the car was a Mercury Milan or else it wouldn't let me get to the next screen.

I had a CTS a few weeks back. The shift interlock would not release, and the key interlock would not let the key out of the ignition. The scanner could talk to every single computer on the car, but all of the serial data ones complained of comm errors. Long story short, those things are handled by two different body computers, both which get the 'you pressed the brake pedal' data from the ABS computer, which also happens to be the gateway between the serial data network and the canbus network. The ABS computer talked to the scanner via canbus, but simply wouldn't talk to the serial data computers at all. Imagine trying to explain to the customer why he needs an ABS computer so he can take the key out of the ignition.

GM was one of the first out of the gate with a truck transfer case that you could shift on the fly.

Also, AMC/Jeep was one of the first to have an electrically switched transfer case, if not the first.

Sometimes all you can do is cry.

If there's one thing I hate, it's when fucking Mitchell tells you it's supposed to be 12v, and it's really PWM! :twilightangry2:

I had a 1976 Subaru 4WD with manual pull-the-lever, but when you did, *every* wheel turned exactly the same. No sharp corners unless you were on gravel or ice, and then you could hear the tires squeak, BUT after you pulled that lever up, you could drive it up the side of a building like Spider-Man, I swear. Drove it to Kansas City once and quite literally passed one hundred cars on the side of the road in -20 degree weather. Literally. We counted through the little one square foot space on the windshield we could get de-frosted.

The happy truck just needs some horns and spiky bits welded to the sides and it'll be fit to trample hapless mooks underfoot.

4791238
I got curious and ran some sites through google translate. That one seems to be called "Хрен" (horseradish?) according to this blog.
Edit: it also called it "Хренов" once by photo 39, which is pronounced Khrenov. So one might be a nickname and the other an official name?

Made by RENT. I think it's for all terrain cargo transport.

FTL

4791397
I actually still have my grandfather's '83 Toyota Tercel and it is just like the old Subys in that all wheels are locked when in 4WD mode. As you said, once you pull that lever you can drive almost anywhere and in abysmal conditions as long as you don't want to change direction fast and you keep off dry surfaces. It does not have low range like the Suby but instead has an EL (Extra Low) gear that you can only engage once in 4WD mode. In that gear it is so slow you can get out and walk around the car as it crawls forward at idle and then get back in without issues. I still use the old Turtle as a small town runaround instead of having to drive the Landcruiser or Falcon all the time.

4791350
Fow PWM outputs, or trying to confirm if one is PWM, I usually revert to an old analogue meter. A steady 'twitching' needle is usually a firm sign of a PWM output, the longer the PWM cycle, the wider and slower the 'twitch'. Cheaper and more portable than a 'scope (although I still have my old 1970s 20MHz one for other electronic work).

4791368
You got to love having to explain those situations... I remember years ago having to explain to a whole court (cul de sac) that the reason none of their ADSL services would work and none them could listen to AM radio below 950kHZ was because of one neighbours' DVD player! Seriously, the DVD player's crap switchmode power supply had become a 2400W broad spectrum transmitter that was flooding the spectrum with noise and that was even being picked up by the phone lines. The powerboard it was plugged into had turned brown on that outlet from the power draw and the player was far too hot to touch at the back. Much longer and it would have thrown a breaker or set fire to itself.

4791343

I once deisighned a flash digitiser with decoder circuit at the logic level for a real time 4:2:2 standard TV definition decoder, just the composite video stream, because I hadnt time to handle the NICAM, DTS encoded audio streams as well. That was fun, involving swapping pairs of samples as it goes. I suspect car companies, in order to try and make their standards as totally unique and broken as possible, so they dont have to pay prior licenses, use all sorts of nasty variations.:pinkiesad2:

I printed out a network topography diagram

...you are sure that you are a mechanic, non a network administrator, are you? Because I've worked as a programmer and I wouldn't know how to do that or why I'd ever want to.

*I think; I'm not sure how long the microwaves bounce around in there before they leak out or whatever it is that they do.

Microwaves are easily absorbed by even the thinnest layers of metal, as a consequence of how the electrons in metal behave. Microwaves cause what's called diaelectric heating: they rapidly spin water around by its dipoles, which are caused by an uneven distribution of valence electrons in the outer atoms of the molecule. This agitates the other molecules around it and since heat is basically just the vibration of atoms (under what's called the kinetic model of heat), you get hot food where there previous was cold food. It also means that only substances containing molecular dipoles can be heated by microwaves.

Well, metal doesn't have those. It has electrons, but it kind of shares them across the entirety of a piece of metal in a diffuse cloud, which is what makes it conductive in a way that's different from the conductivity an ionic solution of salts (like salt water or the human body) happens to have. It is also rigid, making any dipolar molecules impossible to spin even if it had them, which it doesn't. As a result, instead it just slurps most electromagnetic radiation right up and grounds the induced current, so microwave ovens, with their metallic insides, are basically their own radiation shielding and also the reason why you should never run them while they're open, unless you want to get a face-full of microwaving.

4791503

...unless you want to get a face-full of microwaving.

This tan is nice, tho. :derpytongue2:

4791279
Mustache. It needs a mustache.

I can rebuild a engine and starter motors. Figgering out what is wrong with the computer not so much.

Could someone please point out for us inexperienced types what was wrong with the wiring bundle in picture 2? I mean, it doesn't look good, but could someone please point out what I should be seeing? Thank you.

4791368

Imagine trying to explain to the customer why he needs an ABS computer so he can take the key out of the ignition.

That's the worst! Luckily, the one major network failure I've had to explain to a customer (the transmission won't shift right because the BCM isn't being a proper gateway) went alright, because his brother is a network engineer, and explained to him that I wasn't full of crap; networks do work like that sometimes.

Also, AMC/Jeep was one of the first to have an electrically switched transfer case, if not the first.

I'd bet on them being the first in the US (probably Citroen did it in the 20s, but nobody cares).

If there's one thing I hate, it's when fucking Mitchell tells you it's supposed to be 12v, and it's really PWM!

Seriously, how many billable hours and incorrect repairs can be chalked up to the diagnostic instructions being wrong or misleading? I bet a lot.

4791397

BUT after you pulled that lever up, you could drive it up the side of a building like Spider-Man, I swear. Drove it to Kansas City once and quite literally passed one hundred cars on the side of the road in -20 degree weather. Literally. We counted through the little one square foot space on the windshield we could get de-frosted.

Subaru makes a good four wheel drive drivetrain. Maybe not as robust as some actual trucks, but for what it's for, it's more than competent.

And yeah, the advantages of locking differentials (which is probably what you had) is that it sticks to slippery pavement like a mountain goat.

Less good on dry pavement, though. Hard to turn without a tire slipping, or something in the drivetrain breaking.

4791415

The happy truck just needs some horns and spiky bits welded to the sides and it'll be fit to trample hapless mooks underfoot.

And it would look so excited while it's doing it!

4791425

I got curious and ran some sites through google translate. That one seems to be called "Хрен" (horseradish?) according to this blog.

I feel that 'horseradish' is probably a bad translation, although maybe not.
(I once google-translated an article about the Chemain de fer de la province (not sure I spelled all that right), and it kept on calling it 'the train pine nuts).

Edit: it also called it "Хренов" once by photo 39, which is pronounced Khrenov. So one might be a nickname and the other an official name?

That's possible, or it could be the manufacturer and model (a Khrenov Horseradish?)

I think it's for all terrain cargo transport.

That's what it looks like to me, too. Probably like a wheeled version of a Snow Cat.

Here's a video of one of them pulling its friend out of the mud.

4791454

It does not have low range like the Suby but instead has an EL (Extra Low) gear that you can only engage once in 4WD mode. In that gear it is so slow you can get out and walk around the car as it crawls forward at idle and then get back in without issues. I still use the old Turtle as a small town runaround instead of having to drive the Landcruiser or Falcon all the time.

Up into the 80s, some GM trucks with a manual had that--creeper gear + 4wd low = about a 1mph top speed, and it would go over anything.

I didn't know that there were 4wd Tercels. Around here, any Toyota made in the 80s has turned into a pile of rust.

Fow PWM outputs, or trying to confirm if one is PWM, I usually revert to an old analogue meter. A steady 'twitching' needle is usually a firm sign of a PWM output, the longer the PWM cycle, the wider and slower the 'twitch'. Cheaper and more portable than a 'scope (although I still have my old 1970s 20MHz one for other electronic work).

I'm not sure that that would work on an automotive system, although it might. They're getting fast enough on the PWM that the needle might not twitch at all.

You got to love having to explain those situations... I remember years ago having to explain to a whole court (cul de sac) that the reason none of their ADSL services would work and none them could listen to AM radio below 950kHZ was because of one neighbours' DVD player! Seriously, the DVD player's crap switchmode power supply had become a 2400W broad spectrum transmitter that was flooding the spectrum with noise and that was even being picked up by the phone lines. The powerboard it was plugged into had turned brown on that outlet from the power draw and the player was far too hot to touch at the back. Much longer and it would have thrown a breaker or set fire to itself.

Ooh, that sounds exciting!

I've always wanted to build a broad-range jammer just for the fun of it. I've heard that some old ignition coils were really good for that purpose.

Actually--and this is true--I've got a spark tester that works on the spark that leaks through the wires. Just touch the ignition wires, and if there's spark in the wire, the light flashes.

If they're really bad wires, it can pick them up a few feet away.

4791482

I suspect car companies, in order to try and make their standards as totally unique and broken as possible, so they dont have to pay prior licenses, use all sorts of nasty variations.:pinkiesad2:

It's hard to say how much of it is to avoid licensing fees, and how much of it is because it's done in-house. There are standard protocols for some things (CAN bus, for example), but for other things they can pretty much do what they want. They also some times have to deal with backward compatibility issues, especially when it comes to communicating with their own scan tools. Even then, though, you occasionally run into problems or one-offs, especially if it's a vehicle that multiple carmakers sell as their own (such as the Mercedes/Freightliner/Dodge Sprinter van).

4791503

...youaresure that you are a mechanic, non a network administrator, are you? Because I've worked as a programmer andIwouldn't know how to do that or why I'd ever want to.

You probably would rarely (or never) want to, because your networks mostly work differently than car networks. From what I understand of computer networking, you've got x number of computers that can talk to a server and maybe each other, possibly directly, or possibly not, and you've got routers to make all the data go where it's supposed to.

On a car, it's generally a bunch of computers tied together in a ring network that yell at each other, and whoever yells the loudest gets the message through.

You also typically have multiple, independent networks that communicate at different speeds, and a few modules that need to translate--do things like kick a message from the low speed network to the high speed network, or vice versa.

When there's a problem, knowing which modules communicate with which other modules is vital.

Microwaves are easily absorbed by even the thinnest layers of metal, as a consequence of how the electrons in metal behave. Microwaves cause what's called diaelectric heating: they rapidly spin water around by its dipoles, which are caused by an uneven distribution of valence electrons in the outer atoms of the molecule. This agitates the other molecules around it and since heat is basically just the vibration of atoms (under what's called the kinetic model of heat), you get hot food where there previous was cold food. It also means that only substances containing molecular dipoles can be heated by microwaves.

Well, metal doesn't have those. It has electrons, but it kind of shares them across the entirety of a piece of metal in a diffuse cloud, which is what makes it conductive in a way that's different from the conductivity an ionic solution of salts (like salt water or the human body) happens to have. It is alsorigid, making any dipolar molecules impossible to spin even if it had them, which it doesn't. As a result, instead it just slurps most electromagnetic radiation right up and grounds the induced current, so microwave ovens, with their metallic insides, are basically their own radiation shielding and also the reason why you should never run them while they're open, unless you want to get a face-full of microwaving.

Huh. I didn't know any of that!

4791713

I can rebuild a engine and starter motors. Figgering out what is wrong with the computer not so much.

I've never rebuilt an engine or a starter. :rainbowlaugh:

4791879

Could someone please point out for us inexperienced types what was wrong with the wiring bundle in picture 2? I mean, it doesn't lookgood, but could someone please point out what I should be seeing? Thank you.

A bunch of wires are missing bits of insulation, or are melted entirely. That green wire that curves around the bottom is broken, the copper strands next to it are another wire that's melted out of its insulation, and you can kind of see the thick blue wire is missing some of its insulation.

It's not the best picture, to be honest. Here's a slightly better one of another door wiring problem (this one on a Ford):
i.imgur.com/0XezI8q.jpg
(insulation broken on black wire with yellow tracer; green wire completely broken).

4791992
Long as you don't mind the oil and grease Oh it still sucks . Its nasty and hard

4791994
I noticed the blue, but I wasn't confident that it was the problem. Funnily, I also saw the bare wire, but assumed it was the ground! :twilightsmile:

4791997

Long as you don't mind the oil and grease

That doesn't bother me.

Oh it still sucks . Its nasty and hard

That's what she said.

4792013

I noticed the blue, but I wasn't confident that it wastheproblem. Funnily, I also saw the bare wire, but assumed it was the ground!

I can't say for sure which wire was responsible for the smoke, but it was one of them.

There are some uninsulated wires used for grounds on cars, but they're usually big straps. And that's generally only on older cars; turns out that ground straps corrode and break if they haven't got insulation on them.

jxj

4791350

Depends on what it is, but the general rule these days is that they don't. Which makes sense; the same DDM might work inallCadillacs for a given model year, but it needs to know things like how many doors the car has, whether it's got heated seats, memory seats, and so forth.

Sometimesyou can get modules pre-programmed from aftermarket sources, but that's pretty rare for new stuff.

In the early days of PCMs, there was an EEPROM chip that you transferred from your vehicle to the new PCM--that was how it was 'programmed.'

Ok, I think i figured it out. Are you just programming in info about the model year? Based on what you said, it sounds like it's flashed from the factory and your just selecting settings.

That's not a bad idea, although usually it's not worth the trouble to hook up and configure. Now, if I had a good graphing multimeter, that would be nice--I'd use that a lot. But they're expensive.

I'm pretty lucky in that regard, the stuff I work with is small desktop stuff (at least during testing) so I always have one nearby. Plus a lot of the time, i'm looking at signals so i need a scope. yeah, a graphing multimeter would be useful for you.

FTL

4791984

I didn't know that there were 4wd Tercels. Around here, any Toyota made in the 80s has turned into a pile of rust.

Yup, the wagons from 83-88 were part time 4WD units. Picture the little red wagon Jesse drove in Breaking Bad, that is pretty much exactly my car, same colour with almost the same brown 80's interior. ( I never watched the show but one of my trainees spent a week trying to work out why my weird little car was familiar to him when he first saw it and he told me about it when he worked it out). Oddly most Toyota's survive pretty well over here, not just the Landcruisers and Hilux but there'd still be tens of thousands of 60s/70s/80s cars (especially Corollas) on the road every day. My 1983 Corolla had 730,000kms on it when I handed it down to my brother and the Turtle has 485,000 on the clock now with just one small hatch rust spot needing attention.

I'm not sure that that would work on an automotive system, although it might. They're getting fast enough on the PWM that the needle might not twitch at all.

Sometimes the twitch is only 1/10th of a volt or sometimes the needle seems to be almost a thicker 'fuzzy' line on fast PWM signals (moving so fast you can just see the 'blurring' of the needle) and that is the giveaway as it is consistent blur instead of a stable needle. Have used this on Bosch/Ford/Toyota/Mitsubishi systems. Not guaranteed (like most things in life) but it has led me to suspect dodgy diagnostic charts before, then again, remember I am an amateur not a professional... you probably have seen more different systems in a month than I have seen in years.

I've always wanted to build a broad-range jammer just for the fun of it. I've heard that some old ignition coils were really good for that purpose.

You are right, they do make good noise generators. Unfortunately they can only cycle up to the 2-300kHz mark before even the best start to breakdown but with harmonics this still can make AM radio unlistenable. Basically you generally need to match your jammer to a spectrum or frequency range to properly 'jam' and not just interfere. You can make some pretty wide ones with short range and reasonable power consumption but the 'Jam anything' jammer would need to be powered by kilowatts of power and would not be very small or just one circuit. Just an 'annoy the crap out of the neighbours TV/Radio or WiFi' unit could be done though reasonably easy though. :twilightsheepish:

On networks, your auto systems actually are much like computer 'networks' from the 70s and 80s. Different systems that talk through intermediaries at different speeds, protocols and physical cabling. Any old-fart network tech/engineer who has worked on old combined Token Ring, Arcnet or HP/IBM proprietary systems will know exactly the pain you go through today. Almost drove you to drink somedays. :facehoof:

4791988

On a car, it's generally a bunch of computers tied together in a ring network that yell at each other, and whoever yells the loudest gets the message through.

You also typically have multiple, independent networks that communicate at different speeds, and a few modules that need to translate--do things like kick a message from the low speed network to the high speed network, or vice versa.

That's really incredibly interesting. I knew cares were getting more and more complicated, but I had no idea that they contain entire networks of components that you need to map out before you can even begin fixing them. It's really impressive to me because it really sounds like one big, chaotic and unadministrated mess, but somehow, it still keeps working, often for years at a time without any real maintenance required. That shows some real attention to detail.

Huh. I didn't know any of that!

I hope you find it as interesting as I do. You spend so much time educating us about mechanics and I always really enjoy these tech posts of yours, so I thought I'd share something from my own field right back. :pinkiesmile:

4791984

I've always wanted to build a broad-range jammer just for the fun of it. I've heard that some old ignition coils were really good for that purpose.

4792233

You are right, they do make good noise generators. Unfortunately they can only cycle up to the 2-300kHz mark before even the best start to breakdown but with harmonics this still can make AM radio unlistenable. Basically you generally need to match your jammer to a spectrum or frequency range to properly 'jam' and not just interfere. You can make some pretty wide ones with short range and reasonable power consumption but the 'Jam anything' jammer would need to be powered by kilowatts of power and would not be very small or just one circuit. Just an 'annoy the crap out of the neighbours TV/Radio or WiFi' unit could be done though reasonably easy though.

Be careful about building or using one of these. They are super illegal, and the FCC will fine you tens of thousands of dollars if they catch you. They're not kidding about this, either: prisons have asked for permission to install them to prevent prisoners from using smuggled cell phones to manage criminal enterprises and were given a flat "No."

Of course, if you're just screwing around a little bit at home, nobody will probably call them and you'll go unnoticed. On the other hand, if you're like this guy, who didn't want his boss tracking his work truck, you can buy a GPS jammer and install it into the vehicle...then go to work at Newark International Airport, where they have GPS being used to land aircraft. The FCC will put a great deal of effort into finding you, and they will fine you $32,000 dollars when they do.

Even for smaller uses, they might eventually track you down. There are plenty of stories on that FCC link where somebody was using a GPS jammer on their commute to work, and after people's cell phones stopped working at the same time of day every weekday, the FCC got called in and tracked it to their car.

FTL

4792685
I actually used to track spectrum interferers, both accidental and deliberate devices, here in Aus for a living so I am aware of the consequences and the levels we went to to chase them down (mainly focused on mobile networks and licensed spectra). As I was saying to Admiral B, the movie 'jam everything jammer' is a near myth unless you are building a big system with lots of power. I was also highlighting that if you want to do thought experiments or create targeted jammer devices they are actually quite easy and can be made from some unexpected objects. The challenge of taking something like an automotive ignition coil and making a device that can obliterate radio reception in a small area is a fun little design exercise. They are also, as you mentioned, available on the Interwebs. That said, I am pretty sure everyone knows that all of this is completely illegal despite some individuals I met who tried to claim that the Internet told them it wasn't. Pro Tip - That excuse will not get your charges dropped. :raritywink:

On the other hand, if you're like this guy, who didn't want his boss tracking his work truck, you can buy a GPS jammer and install it into the vehicle...then go to work at Newark International Airport, where they have GPS being used to land aircraft. The FCC will put a great deal of effort into finding you, and they will fine you $32,000 dollars when they do.

Yup, that is the classic dumbarse. :facehoof: Over here there are several acts you can be charged under and several offences and it is a max of a $100,000 fine and/or 7 years imprisonment for the favourite one that is used for aviation infringements. As you can imagine there are limits on what I can talk about but I can say that the smallest penalty I saw for this was $47,000 and the worst was $80,000 and 3 years jail with no parole.

4792173

Ok, I think i figured it out. Are you just programming in info about the model year? Based on what you said, it sounds like it's flashed from the factory and your just selecting settings.

Really depends on the module. I'm sure that there are some settings that are done at the factory--I don't think that they come completely blank, but they might. It generally takes a little while to program them--IIRC, labor time is generally an hour to program a module (although some of that is setup; you need to have the car attached to a stable power supply, and you might have to pull fuses or disable part of the network).

I do know that GM used the same PCM in all their trucks in the late 90s/early 2000s; we put a used Blazer computer in an Express van, and it was able to run and drive with that computer. It didn't like it very much, since the Blazer had six cylinders and the Express had eight. . . .

I'm pretty lucky in that regard, the stuff I work with is small desktop stuff (at least during testing) so I always have one nearby. Plus a lot of the time, i'm looking at signals so i need a scope. yeah, a graphing multimeter would be useful for you.

Snap On used to make a really nice one that didn't have a lot of features. They still make a nice one, but it's really expensive. I think it's over a grand.

The Modis does have a graphing meter function, as well as a four-channel lab scope, but it's kind of a pain to set everything up (and to be honest, it's not something I do a lot, so that slows me down, too).

4792255

That's really incredibly interesting. I knew cares were getting more and more complicated, but I had no idea that they contain entire networks of components that you need to map out before you can even begin fixing them.

Well, luckily you usually don't have to map them out before fixing things. In network failures, though, it's really useful, especially if you've got a partially functional network--when you know how it's wired, you can often eliminate a bunch of things as potential causes. For example, in the case of this Caddy, I bet if I'd dug into it, I would have found that every module that wasn't communicating was either on the same power fuse as the DDM, or was behind it in the network (I think that part was single-wire network).

It's really impressive to me because it really sounds like one big, chaotic and unadministrated mess, but somehow, it still keeps working, often for years at a time without any real maintenance required. That shows some real attention to detail.

It is one big, chaotic and unadministrated mess. But in a way, it's also simpler than a normal computer network: pretty much all that the computers do on a car is take inputs and use them to decide whether to turn something on or off. Plus, they're not flexible (like normal computers); all they can do is run the machine they're attached to. You probably couldn't write code to make a PCM into a word processor, for example.

I hope you find it as interesting as I do. You spend so much time educating us about mechanics and I always really enjoy these tech posts of yours, so I thought I'd share something from my own field right back.:pinkiesmile:

:heart:

4792685

Be careful about building or using one of these.They are super illegal, and the FCC will fine you tens of thousands of dollars if they catch you. They're not kidding about this, either: prisons have asked for permission to install them to prevent prisoners from using smuggled cell phones to manage criminal enterprises and were given a flat "No."

Interestingly, there seem to be short range cell phone jammers available (although I don't know what the law is about using them).

I did know that they were super illegal; that was one of many things that a friend and I discussed when we were thinking of building a pirate radio station. We never wound up doing it, although we did have some ideas on how to covertly attach an antenna that would be very difficult for the FCC to find . . . mind you, this was with 20 years ago tech.

Even for smaller uses, they might eventually track you down. There are plenty of stories on that FCC link where somebody was using a GPS jammer on their commute to work, and after people's cell phones stopped working at the same time of day every weekday, the FCC got called in and tracked it to their car.

Yeah, that seems like a bad idea where you'll get caught eventually (heck, for the most part, illegal stuff in your car is a bad idea, because from what I know of the law, it's much easier for the police to search your car than, say, your house). I'd never really considered what I'd do with one if I did build it. Probably something dumb and rednecky, like annoy the manager by making the shop radio go to static every now and then. :rainbowlaugh:

4792968

The challenge of taking something like an automotive ignition coil and making a device that can obliterate radio reception in a small area is a fun little design exercise. They are also, as you mentioned, available on the Interwebs.

Yeah, I was just thinking for funzies, rather than for any malicious purpose.

That said, I am pretty sure everyone knows that all of this is completely illegal despite some individuals I met who tried to claim that the Internet told them it wasn't. Pro Tip - That excuse will not get your charges dropped.:raritywink:

Honestly, taking legal advice from the internet isn't a good idea.

Yup, that is the classic dumbarse.:facehoof:Over here there are several acts you can be charged under and several offences and it is a max of a $100,000 fine and/or 7 years imprisonment for the favourite one that is used for aviation infringements. As you can imagine there are limits on what I can talk about but I can say that the smallest penalty I saw for this was $47,000 and the worst was $80,000 and 3 years jail with no parole.

I'd guess that interfering with aircraft navigation is about the worst thing you can do with a jammer. Maybe military is more serious . . . over here, we've had some problems with people thinking it was fun to shine lasers into the cockpits of commercial aircraft, and the FAA takes that very seriously.

There are also occasionally chemtrail conspiracy folks who talk about taking matters into their own hands and shooting down an airplane or something stupid like that, which is also something the FAA takes quite seriously.

jxj

4794704

Really depends on the module. I'm sure that there aresomesettings that are done at the factory--I don't think that they comecompletelyblank, but they might. It generally takes a little while to program them--IIRC, labor time is generally an hour to program a module (although some of that is setup; you need to have the car attached to a stable power supply, and you might have to pull fuses or disable part of the network).

I do know that GM used the same PCM in all their trucks in the late 90s/early 2000s; we put a used Blazer computer in an Express van, and it was able to run and drive with that computer. It didn'tlikeit very much, since the Blazer had six cylinders and the Express had eight. . . .

i'm not really sure. The actual programming is pretty fast (at least the short programs i've done) and i'm not sure how long it would take to upload the code. what made me think you might just be doing settings was that you said that it used to be done on EEPROM. EEPROM isn't used for program storage (at least as far as I know), it's used as nonvolatile RAM, memory that doesn't get lost when it loses power, like system settings. Plus you can get the MCU pre-flashed from the manufacturer and that's something I can see companies taking advantage of.

4794719

Interestingly, there seem to be short range cell phone jammers available (although I don't know what the law is about using them).

They're illegal. Any variety of jammer which operates by transmitting a signal is illegal. To my knowledge, the only people who have gotten an exception in the US are military contractors testing jammers used by troops overseas, and they can only use them in designated testing areas. I don't think that even troops conducting training are permitted to use actual jammers, only "notional" jammers that are just an empty box with wired-up switches and LEDs.

Non-active methods of blocking a signal are OK. You could, for example, put up metal mesh in the walls of your business that acts like a faraday cage and that would be legal.

This is why I jumped in. I agree that it would be cool to build something like this. Just like it's cool to build bottle bombs and set them off in the swamp. I've done that, and it's great fun! But do not be under any illusions about what could happen if you get caught. I don't know if the FCC would stomp on somebody who just built one for fun and only used it in the woods just to see if he could do it (if you happened to get caught, which is unlikely), but remember that if you do get caught and if they want to, they can hit you with huge fines.

4794935
Yeah, I'm not really sure on all the technical specs, either.

Probably the best way to think of it was that the early computers were like a console game system, where you put a cartridge (the EEPROM) chip in to get it to do what you wanted. GMs even had a slot in the top for this purpose, IIRC. Newer systems, it's more practical and sensible to program them in a more conventional manner, by downloading the program on a scan tool, and then putting it via data wire to the appropriate computer. Plus, that way you don't have to physically access the computer to change/update the programming on it.

Some cars, such as the Tesla, are programmed wirelessly.

4795669

They're illegal. Any variety of jammer which operates by transmitting a signal is illegal. To my knowledge, the only people who have gotten an exception in the US are military contractors testing jammers used by troops overseas, and they can only use them in designated testing areas. I don't think that even troops conducting training are permitted to use actual jammers, only "notional" jammers that are just an empty box with wired-up switches and LEDs.

That makes sense. So the ones I see for sale every now and then are either illegal, or legal because they don't actually work (given that a lot of them seem to come from oversees sellers, I'd bet that the latter case is the more common one).

Non-active methods of blocking a signal are OK. You could, for example, put up metal mesh in the walls of your business that acts like a faraday cage and that would be legal.

True fact: I used to live in an all-aluminum mobile home, and it worked very well as a faraday cage. I lost two or three bars on my cell phone when I went inside, and we also couldn't reliably pick up a local TV station (antenna was inside, on top of the TV) despite the fact that we could actually see their broadcast antenna from the house.

This is why I jumped in. I agree that it would be cool to build something like this. Just like it's cool to build bottle bombs and set them off in the swamp. I've done that, and it's great fun! But do not be under any illusions about what could happen if you get caught. I don't know if the FCC would stomp on somebody who just built one for fun and only used it in the woods just to see if he could do it (if you happened to get caught, which is unlikely), but remember that if you do get caught and if they want to, they can hit you with huge fines.

I'm sure that there's some discretion on what, if any, charges are filed (and surely some of that is based on what kind of a mood the agent in charge is in that particular day). I think you're right that if you just build some super low-range thing to play with they're probably not going to come down too hard on you, but that's probably not something to rely on. Might depend on the device in question, and it might also depend on what else they find when they come looking.

jxj

4796951

Yeah, I'm not really sure on all the technical specs, either.

I'm just speculating at this point.

Probably the best way to think of it was that the early computers were like a console game system, where you put a cartridge (the EEPROM) chip in to get it to do what you wanted. GMs even had a slot in the top for this purpose, IIRC. Newer systems, it's more practical and sensible to program them in a more conventional manner, by downloading the program on a scan tool, and then putting it via data wire to the appropriate computer. Plus, that way you don't have to physically access the computer to change/update the programming on it.

yeah, I get how it works. It's a pretty novel idea that probably worked fairly well when you couldn't really get a hold of programmers easily. What I was getting at is that EEPROM isn't used to store programs on modern microcontrollers, it's used as a more permanent version of RAM. It's possible that EEPROM is used for program storage on older microcontrollers or specialty ones used by the auto industry. So that led me to speculate that your just giving the chip car info, not uploading the program from scratch. Both could be called programming.

Login or register to comment