noise."Mechanic: the BCM that lost itself · 1:33pm Jun 10th, 2019
Y’all are going to think I’m running out of things to talk about here besides network diagnosis, because here comes another one. This one’s a 2005 GMC Envoy, which is essentially a pretentious Chevy Trailblazer.
So, y’all know the drill:
The story with this one is that the wipers don’t work. The customer thought it was the wiper motor, and had the skills to replace it . . . which didn’t fix it.
Our most recent new guy, Earwig,* looked at it first, and then he got pulled off it for something else and I got put on it.
The first thing I noticed was that the wiper motor was not new. I questioned the manager about that, and apparently when the wiper motor hadn’t fixed it, the customer took it off and returned it, and put the old one back on.
Assuming that the customer’s not an idiot, and that the new part was actually correct and functional out of the box--something I can’t necessarily assume--the next question is how to go about diagnosing it.
___________________________
*Who as y'all know has quit since I wrote the draft for this blog post.
Back in Ye Olde Days, the switch was just an on/off switch, and it turned the wipers on and off. Then things got complicated with two-speed wipers (high and low), and then again with delay--delay used to be built into the switch, but now it’s built into either the wiper control module, one of the other modules on the vehicle, or both.
Since the cowl was already off, testing whether the wiper motor was getting power when commanded would be a good start. However, there were a bunch of wires running to the motor (six, I believe), and I didn’t want to guess. Unfortunately, our computer was in use, so I couldn’t look up a diagram.
What I could do was get the scan tool and see if there are any codes, and if the body control module knows that the wipers are being commanded on, or if I can control them manually.
That’s actually one nice thing about computer controls and networks. A lot of times you can use a good scan tool to control something, sometimes even in a way that the manufacturer didn’t intend to happen in normal use. For example, on some old Dodges, they let you turn off fuel injectors individually for testing purposes, and you can often turn on cooling fans and things like that.
Just recently, I worked on a Chevy truck with a broken power window switch. What I like to do, since customers never tell us if the window works or not, is roll them down just a crack and then back up before lowering them all the way.
This truck had auto-down (as many vehicles do), and when I pulled up on the switch to stop it, nothing happened. The ‘up’ side of the switch was broken.
Back in older times, I would have had to take the door panel off, access the wires, and then raise the window that way. In this case, I just hooked up the scan tool, accessed the Driver Door Module, went to functional tests, window motor, and clicked on “up.”
Turns out I wasn’t so lucky on the Envoy. Not only did I not have the option of controlling the windshield wipers with just the scan tool, I also didn’t have any data indicating what position the switch was set in.
However, since I was there, I figured I might as well check the codes.
It would have been really nice to find a code like “Wiper Control Circuit Open” or “Wiper Ground Fault” or “Wiper System General Malfunction”* or really anything that might be a giant arrow towards the fault, but I didn’t.
So I scanned all the codes, and not surprisingly got some communication codes.
_________________________
*You scoff at ‘general malfunction,’ but just wait.
Now, GM has issued more than one service bulletin stating that communication codes should not be diagnosed unless there is a complaint**, and the Trailblazer/Envoy platform is a poster child for why this is so. The Liftgate Control Module is so finicky that you can watch in live data as it cycles between having a fault and not having a fault (typically, about once per second), and given the nature of how secondary networks are designed on vehicles, it’s pretty rare that every computer on the network always responds when queried.
On most makes of car, a U1000 is a general bus fault; two modules on the network had set that in the past. The HVAC had a high voltage network code, which was a bit odd.
I sort of struck gold with the Body Control Module--the BCM.
B1440 Power Mode Mismatch
B2530 Reverse Lamps Circuit
B2580 High Beam Circuit
B3811 Rear Washer Relay Circuit
U1305 Class 2 [network] low or high
U1160 No communication with DDM
U1161 No communication with PDM
U1162 No communication with liftgate
U0164 No communication with BCM
U1016 No communication with VCM/PCM
Also worth mention, the transfer case control module had set a C0374 “general malfunction”
None of those seemed to address my problem. Sure, the rear washer didn’t work, but the customer hadn’t complained about that, and for all I knew it hadn’t worked for years.
Still, I dutifully documented all the codes, in case the owner wanted us to take a look at them, too, depending on what we found with the wipers.
____________________________________________
**recently, I worked on a vehicle that the BCM had set a U1003 code: Serial Data Traffic garbled--no diagnosis required
My manager was off the computer, so I went up front and got a wiring diagram for the wipers. I was going to test at the wiper motor, as I’d initially intended, but then I realized that checking the fuse first would be even easier. On these things, there’s one fuse box under the hood, which is where you’d assume the fuse for the wiper is. There’s also a fuse box under the left rear seat, which is where the fuse for the wiper actually is.
Since the Envoy has folding rear seats, that box is easy to get to, so I stuck my test light on the fuse and it didn’t illuminate.
At this point, I have already determined what the problem is. The wiper motor is not getting power, and a new wiper motor isn’t going to fix that. Nor would I have been able to control them with the scan tool, if it had had an option for me to do that.
So, of course, the next question was what the fuse got its power from, and the answer to that was the ignition switch. The wipers are only supposed to run when the key is on, so when the key is on, a circuit is powered which makes that leg of the fuse box live.
I found that out from the power distribution wiring diagram.
Now, the ignition switch is in the steering column, which I didn’t want to have to take apart unless I had to. And I have seen fuse boxes which fail internally--even just one wire, sometimes--so before I got under the dash, I consulted the power distribution for what else might be powered on that circuit.
Nothing on that circuit worked. Every single fuse on that circuit was unpowered.
Trailblazers and Envoys were known for breaking wires from their ignition switches. The bundle of wires made a sharp bend to get under the bottom of the column, and while it was in a plastic wire guide, those things aren’t flawless. I’ve replaced them before, and I think there might have even been a recall on some of them.
Luckily, it didn’t get to that point. I got the knee bolster panel off, and found this:
I know GM’s quality control isn’t great, but I do know that they don’t tend to use duct tape for wire insulation.
In fact, I also know that this vehicle used to have a remote starter or aftermarket alarm or both, and that’s why all those important-looking wires have been taped over. The customer might not be aware of this; it might have been gone before he got the Envoy, but I know.
Just to further prove my diagnosis, I turned the wiper switch on, and when I wiggled around those wires, the wipers would work. So, we got him a new ignition switch harness, and that not only fixed his wipers, but also his power door locks and a few other things that didn’t work on the Envoy.
Now we get to the fun part of the blog post. Because this is one that I would have ordinarily skipped, to be honest. The customer tried to fix it and couldn’t; I diagnosed it properly and methodically and did fix the problem.
Sure, it wasn’t what I was expecting to find, so I suppose there might be a lesson there in failures having unexpected origins, but this vehicle did something special, and those of you who have really been paying attention might have caught it. For everyone else--including me; I didn’t notice this at first--look back up at that code list. One of the codes the BCM set was U0164 No communication with BCM.
Now, I’ve told y’all that when you have a module go down, it’s the one that doesn’t know. Every module on the network is asking where the radio went, and the radio is just doing it’s thing, completely unaware that anyone is looking for it. Every module who cares sets a code for lack of communication with the radio, and the radio’s oblivious.
Not this time.
The BCM couldn’t find itself.
Source Unfound
See, there are parts of the BCM that have to go live right away. As soon as you hit the remote or put the key in the door lock, it’s doing things. It’s getting ready, waking up modules, and so forth.
But to save power, it doesn’t turn itself all on at once. Other circuits are powered from other sources, and some of its functionality might not come online until the key is in the run position, let’s say. And, as it happens one of the BCMs power sources was on the same leg of the fuse box as the windshield wiper.
I can’t say exactly how GM programs these things, but I bet that when the vehicle was started, the BCM knew it was running. At that point, the BCM would expect the part of itself that only powers up when the key is in the run position to be powered up, and it wasn’t. Had I gone further into the diagnosis of the codes, I bet I would have found that that’s exactly what B1440 Power Mode Mismatch would indicate, too. As far as the BCM knows (since it’s not programmed to think about broken wires), the key is both in the run position and in the off position, and the BCM knows darn well that can’t be.
We may not have flying cars, but we do have ones that can experience existential crises.
Yay?
Always love these stories. I suspect that it is due to your skills as a storyteller.
You know, back in my days of shadetree mechanicking, I had a 'speaker wire' of some sort with alligator clips that I could run back to the battery and clamp onto "Does this work" stuff. Dome light doesn't come on and the fuse isn't blown? Attach clips and say "Oh, no light. Must have burned out. Time to make a store trip." That was actually how I expected you to test the wipers.
Now, I'd be too afraid of making all the smoke come out of some module. (once all the magic smoke escapes, the widget will quit working for sure)
You need a new title: Auto Therapist.
"My car doesn't work."
"Does it exist?"
"It doesn't think it does..."
"Ah. Too much Nietzsche. Insert more pone."
Adventures in Modern Mechanicing! by Admiral Biscuit
I both love and dread it. Having done it in the past, I understand it enough to follow what is going on, being the past I'm starting to understand how far behind I am if I ever have to start doing it again. I'm seriously considering that my next vehicle will be a '70s Chevy Pick up done up by a friend who does those things for a living because having module to control a windshield wiper motor is overkill. The more complicated something is the more that can go wrong.
I hope the customer was beyond pleased that you not only fixed the problem, but fixed other things he didn't know were problems.
This sounds about as janky as my old Aerostar which could change radio stations just by opening and closing the driver side door.
You lost me at the picture of sunset on the beach
I read the book The Reckoning (Japanese car industry, especially Nissan vs Ford). IIRC, before they let you design cars they make you work on the assembly line for 6 months. Then, every so often, you get a 6 month refresher course either building them, selling them, or FIXING them. They tend to have fewer screwy problems.
I also remember that, when I was young, a VW Beetle had the spare tire up front & the windshield wiper motor ran off compressed air from the spare tire. This had 2 results:
1) The wipers tended to wimp out after awhile
2) The spare tire was always flat
Oh yeah. I read somewhere that if airplanes had improved at the rate that computers have, we'd have the Starship Enterprise
If railroad locomotives had improved that fast since the first one in 1804, we'd have the Tardis
Customer: My car has power locks???
So… I'm guessing you're a mechanic?
5072405
Yeah. For most of the lifetime of silicon transistors, they were in the more or less unique position of having a path for improvement with no trade-offs. If you could find a way to improve the resolution of your photolithography and make the transistors in your IC smaller, it was all upsides and no downsides.
5072327
I once turned an Chevy Express (full-size van) into a Chevy Blazer (mid-size SUV) by changing the PCM.
Also, I can’t remember if this ever made it into a blog, but I worked on a Ford Econoline van that thought it was a F-150.
5072328
Thank you!
5072334
You can still do that for some stuff.
i.imgur.com/I1w4T7c.jpg
You just have to know what you’re attaching jumper wires to rather than just willy-nilly apply them. Because...
If you do it wrong, you can make all the smoke come out of some module. And you’re right, once the magic smoke gets out, you can’t put it back in. Gotta replace that component with a new one that’s still got the smoke in it.
Ooh, I like that.
5072342
Legit I worked on a Ford Econoline that didn’t know what it was. The PCM said one thing, the ABS module said another thing, and the IPC was unprogrammed, so it really had no idea. The factory scan tool said it was a F-150 pickup...
5072349
Not your grandfather’s auto repair, to be sure.
The good news for Joe Average is a lot of the fundamental stuff is still conceptually similar to what it used to be, it’s just controlled with electronics rather than mechanical and vacuum stuff. And most of it you can diagnose at home with patience, a good voltmeter, and an understanding of wiring diagrams.
Although in my experience, despite popular opinion, computer-controlled stuff is typically more reliable and more capable of dealing with problems than the old mechanical stuff. Harder for the backyard mechanic to work on, though, without a good understanding of electronics.
Also, FWIW, I have a 70s Chevy pickup. 250ci inline 6, 3-speed Saginaw transmission, manual brakes and steering. Can be run off a gravity bottle.
i.imgur.com/yNMLulwg.jpg
5072376
I hope so, too, but my manager never said if he was happy that his power locks worked after I fixed the wipers.
5072621 I can believe that. Some of the Ford vans were built with a bunch of pickup parts. My old Ford Aerostar was rear-wheel drive, and whenever you hit ice, it exhibited pickup behavior by the rear end trying to get out in front.
5072396
He said the A word! Shun him!
Those weren’t terrible vans, although they sucked to work on. Rock solid drivetrains.
5072397
While I aim for my blogs to be entertaining and educational, if nothing else you can just scroll through and look at the pretty pony pictures.
5072403
That’s not a bad idea, really. Although I would take anything Nissan says with a grain of salt; of the Japanese automakers, they’re the ones that think of Rube Goldberg solutions to problems rather than something simple and elegant (their variable valve timing system gives me headaches just trying to understand how it’s supposed to work). Also there’s at least one vehicle where the diagnostic instructions not only labeled the terminals in a connector incorrectly (mirrored), and if you followed the instructions exactly, you’d replace the PCM every time, regardless of which component had actually failed.
When I worked at a Firestone in Kalamazoo, we actually had an aspiring designer work in the shop for a while, and I did think that was smart of him. See how things work in the real world, rather than in the design headquarters.
I have a hard time believing that they actually ran the wipers off the spare tire. I’ve heard of pneumatic wiper systems, they weren’t all that uncommon back in the day, but why use the spare tire when the engine always makes vacuum?
HAVING SAID THAT, I did a bit of googling, and in fact they did run the windshield washer pump off the spare tire, just like you described. In theory, the valve in the washer pump assembly wouldn’t let the spare go below 26psi, although I imagine that the pressure regulator valve only worked properly when the car was new...
I’d imagine that the reason that neither of those things happened was that at first there was a hard limit to how we could practically build a chip, while more powerful ones were possible in theory (and all the advances since the first ICs were invented have been putting theory into practice). Automobiles and trains also have theoretical limits, which were largely achieved much earlier . . . but I think it’s worth considering how in the early days of cars they went rather quickly from something that only had a top speed of a jog, to hundreds of miles per hour (a couple decades, IIRC), and locomotives likewise rapidly went from a poor substitute for a horse to something that could haul hundreds of tons quite efficiently, to then sort of plateau out as all the low-hanging fruit was picked.
5072437
At least if he was that clueless, he got a feature he didn’t know his car had. Versus the customer who thinks their car has a feature it doesn’t, and I’m the bad guy when I explain that their Honda Accord never had air conditioning, and thus I can’t “make it work.”
5072499
Yes....
Or as 5072334 proposed, an Auto Therapist.
5072503
Other tech has been in that position, just not in most of our lifetimes. I mean, think about going from a kite with a lawnmower engine to flying to the moon in 50 years, for example. Or a self-propelled wagon that could hardly compete with a horse-drawn one to hundreds of miles an hour in a couple of decades.
Or for that matter, a cell-phone system that required carrying around a bag phone, was limited to about six calls per cell tower, and had maybe an hour battery life to what we have now.
LED light bulbs that used to cost $120 each, and now they’re $9.99 for a three-pack.
5072627
Yeah, the Aerostar was a Ranger driveline wedged into a doorstop.
You should have gotten the 4wd version. That was pretty good, or so I’ve been told.
5072639
Well, The Reckoning was published in 1986, so that WAS the way Nissan did it, but now?....it's at least 35 years out of date.
No Clue Whatsoever if Nissan still does that
Crap, AB, you gave me a flashback to troubleshooting the armament control system on an A-4...
5072641
The more you know, I guess…
5072644
No, actually. Flight, internal combustion engines, and batteries were all plagued by the normal engineering trade-offs. In fact, internal combusion is fairly infamous for how much brilliance had to go into it to balance the trade-offs.
Communications and LEDs did have meteoric improvements, but those benefitted heavily from what I was talking about with silicon microelectronics fabbing where every refinement to the transistor fabbing process that allowed smaller transistors didn't invoke new trade-offs in the design stage.
Also, bear in mind that the transistor scaling bonus ended over a decade ago. Since then, they've had a lot of trade-offs to deal with, such as more difficulty in preventing or compensating for increasing electron leakage as transistors continue to shrink.
On that note, the size-based names for the CPU fab process nodes (given in nanometers) don't correspond to actual sizes of components anymore. Rather, they're "equivalent to" numbers which represent attainable transistor density achieved through any combination of process and design improvements, comparable to how the model numbering scheme AMD used for the original Athlons wasn't actually a measure of clock speed, but rather an indicator of what clock speed on the old progression would be needed to match its performance. (eg. an Athlon XP 5200+ was claiming that improvements in architecture would allow it to match a 5.2GHz on the old progression at a lower clock speed.)
That code for 'no comms with BCM' might be a code in a sub module that has no access to the main network. And the BCM reads its codes and stores it. I've seen that before a couple times, though admittedly, I've never seen the BCM say it's lost comms with itself lol.
I did have a Saturn just recently that had a strange issue. Complaint was no shift lock release. Of course when I got it, basically none of the BCM functions were working. I unplugged the BCM to inspect its PCB and the plugs, and to check for power and ground. Everything checked ok, so I put it back. It magically started working. There were all kinds of comms codes, but I don't recall if the BCM had any codes of note. I couldn't get it to mess up again, so I sent the customer home, and explained how to manually bypass the shift interlock. It was a week later when she came back, same issue. This time though, I thought I would check the ground wire source, which was one of those 20-pin plug-looking ground wire junctions that bolted near the strut tower. It was covered in grease like someone had just cleaned it and didn't want it to corrode. I yanked on the wires and 3 or 4 popped right out.
So I chopped that shit off and put some eye crimp thingies on the wires, and bolted them back. Now the BCM was awake, but would cycle on and off once a second, easily visible by turning on the map lights and watching them blink. I wound up putting in a used BCM, programmed the keys, and it worked fine. Strange BCM failure. I guess from a weak ground for a long period of time? Nothing was visibly burned on the PCB. And when I had tested the ground at the plug it was perfectly strong. That's actually the reason I test with an incandescent testlight and not a power probe or volt meter. Because you can tell a weak connection by the light bulb's lack of brightness. I've had a voltmeter read 12v on a connection so weak it can't power a tiny bulb. All around strange situation.
(I added an extra ground wire to the wire at the BCM plug, just in case.)
I drive a school bus. Every day before we start transporting the little
hellspawn... uh... darlings, we have to do a pretrip inspection. One day recently, the pre-trip inspection didn't reveal anything, but one of the strobe lights on my stop sign quit blinking during my run. After returning to the barn and parking, I shut down the engine and turned the IGN back on and activated the reds. I walked around the side and got reeeal close. The one good light was accompanied by the usual strobe light whine, but the bad one was heralded by a pop noise. How did I write up the malfunction report?One stop sign strobe lite inop. Controller making smoke signal noises.
Our school bus co-op has a good crew of mechanics. They were able to interpret my story and have my bus back in road-ready condition by the afternoon run.
5072629
That was straight up diabeetus inducing cute.
5072628
Yeah, even though it had all the aesthetics of a Soviet civilian family transport, the 4.0L engine and electronically controlled 4wd could make that thing scoot.
Too bad the speedometer only went up to 90. Before it stopped working altogether. (This made the question "Do you know how fast you were going?" really awkward to answer)
5072646
I hope they still do it that way. Their designs are weird, but I guess they work for Nissan.
I wouldn’t buy one, though.
5072647
I would hope it’s not as weird as an automotive network, but I’ve got a terrible feeling that it is, and probably not as well documented.
5072649
5072652
I’d argue that in the early years, at least, it was all ups and few to no downs. I mean, you’re not wrong that there were normal engineering trade-offs, but in my limited knowledge of computer there were also engineering tradeoffs.
Mind, I could be totally wrong, and some of it’s gonna depend on where you set the starting line on things. I mean, the first automobile might have existed in the 1700s (if not earlier), and there wasn’t much advancement for quite some time, but then you could view computers broader than just ICs and consider vacuum-tube systems to fall into the mix, or if you want to be really frisky, mechanical difference engines (and that could lead down a rabbit hole to ankytheras and other mechanical computers). I think that automobiles [for example] suddenly got to a point where they were practical enough that a lot of people innovated and there was a burst of meteoric improvement, followed by a flattening of the curve as the technology matured.
In regards to the fabrication, though, can we consider that as a separate thing? I personally don’t think so, but if you include it, then you’ve got to include other, potentially unrelated tech. We couldn’t build nukes until we’d at a minimum discovered radioactivity and come up with some theories about how it worked.
Dumb electron leaks.
I’ve actually fixed a car where that was a problem. Mind, it was due to bad insulation on the wires (physical damage) rather than cramming things too close together . . . but unsurprisingly, a 7v communications system doesn’t do well when 12v is overlaid on top of it.
In other words, largely marketing hype.
I can get behind that; EMD used to number locomotives based on horsepower, then Alco beat EMD in the horsepower race, so they had the GP30 which had ‘30 improvements’ over the old model (horsepower wasn’t one). Likewise, there’s probably a marketing reason why Chevy went from the C 10, C 20, and C 30 to the 1500, 2500, and 3500.
And even the Feds do that for certain ratings; a PZEV rating is, if I remember correctly, a car which outputs about the same emissions as an average power plant making electricity for an electric car would.
5072700
That’s possible, I’ve seen that before. One car it was the TPMS, and the BCM dutifully stored all the codes the TPMS set as it went belly-up. The gist of it was something like “things went terribly wrong and here’s what the TPMS told me in its death throes.”
There are legit a few repairs that involve forcing the computer to reboot and stop being dumb. Some Chevy trucks in the mid-2000s for example would turn on the ABS light, but there were no codes in the ABS module. The solution was to disconnect the battery, touch the leads together for 30 seconds to clear all the memory, and then reconnect them.
Intermittents are the worst. Of course the car never malfunctions in the shop, and it’s funny because that’s the one time where the tech winds up being really excited that it did break. It’s rare that customers can provide enough info about when it happens--if there is a pattern--for it to be useful. I got really excited once when I was test-driving an Impala for the fourth or fifth time and it actually did what the customer said it did.
Also, my former manager once went bajaing in his backyard with the boss’ Yukon to get it to act up (which it did, yay!). That was a bad ground on the back of the cylinder head, but it only went bad when there was a lot of flexing going on.
Those are the worst. Visual inspection, it looks good, but there’s nothing there.
Freaking computer failures. I love it when stuff’s melted, ‘cause I can see that with my luddite eyes, but when it’s not? Always second-guessing, is it the computer, or is there a test I should have done that I forgot about? Not that long ago, did a theft module in an Excursion, and did all my tests twice, just to be sure that there wasn’t something I was forgetting . . . and still wasn’t confident, until the new one programmed successfully and the vehicle actually started.
Obviously, we both know you’ve got to be careful what you toss a big load on (like, you wouldn’t check the CAN bus by throwing a headlight on the circuit), but you’re right, the DMM can lie, ‘cause it tosses such a low current down the wires that a single strand will pass, and when you put an actual load on, it fails.
Had that happen once at the dealer. Car towed in with a dead battery, charged it and didn’t like how it looked, you know? Put the GM-approved Midtronics handheld battery tester on it, pass. Put the VAT-40 carbon pile tester on it, failed spectacularly.
Manager said that since it passed the Midtronics test, it was good.
Guess what got towed back in the next day and what I put a battery in?
5072732
I aim to please!
5072765
It’s like the Astro van--I can’t really sing the praises of the asthetic of the design, nor the comfort, nor the ease of maintenance, but in terms of doing what a van should do, they were good at that.
I’ve had vehicles like that.
“Did you know your reverse lights were still on?”
“Oh, sorry, forgot to flip the toggle switch.”
“Why do you have a toggle switch for your reverse lights?”
“Because the part my truck needs is only available from museums.”
5073255
*boop*
5073254
No, the documentation was pretty good - it was just the size of a couple of phone books for each subsystem...
5073555
derpicdn.net/img/view/2019/5/30/2053406.png
5073696
Well, that’s a bonus at least. I mean, wading through pages and pages of documentation sucks, but at least it’s better than just not having it.
One of my favorites is instructions that say to use “special tool,” but don’t say what that tool is or what it does.
Also, some of Subaru’s instructions are . . . interesting.
5073265
Nice!
5074063
It’s always important to know your tools and what they can test and what they cannot.