noise."It's not Lupus · 10:48pm Jul 22nd, 2017
Source
It's never Lupus.
Get your juiceboxes or your favorite adult beverage ready, 'cause this is kinda a long rambly one. And I should also say here that we're not actually talking about Lupus, that's just an appropriate metaphor because everybody knows House.
Now, I'm sure y'all remember a couple of weeks back I posted a blog about the fun of using diagnostic trees and how much more fun it was when the diagnostic tree was chopped down, put in a woodchipper, and rearranged into an exciting new form.
And there was something that I didn't tell you, but which all techs who have been in the field for a while would know--most of the time, engine computers on cars are pretty robust. They normally don't fail. So when I see a checklist like that one, the absolute last thing I'm going to consider is the PCM, and I'm sure as heck not going to condemn it until I do every test two times at least. On a car like that Buick, I might even run new wires to the alternator first, just to be sure that there isn't some problem with the wiring that I'm not finding when I test it. After all, Heisenberg's uncertainty principle is kind of in effect when you're diagnosing a car: did disconnecting that wire change the results? And it's not like I've never found problems with terminals; they corrode, melt, and get loose all the time, and sometimes that's really hard to identify, especially when you have to do disassembly to get where you need to be to access the wires.
So last week, the first car that I worked on was a Lincoln MK-S. The customer thought that there was a problem with the transmission, because the car had a bit of a shudder. It was obvious to me as soon as I drove it that it actually had a misfire, and when I pulled the codes, sure enough, #5 was dead as a stick.
Luckily for me, that one was on the front. So I pulled off the coilpack, and it had gone from regular to extra crispy. I can't describe the smell, although anyone who's worked with melted electronics enough will know it. Down at the bottom of the spark plug well was a little puddle of some of the innards; I assume that the rest went up in the form of smoke.
A new coil pack and a tune up ought to have fixed her right up, but . . .
Basically, all a PCM can do is turn things on and off. It 'sees' engine performance data, and then it goes to a lookup table and decides what it's going to turn on or off on the basis of that. Generally on domestic vehicles, things are ground-side controlled, so in its simplest form, inside the PCM there are little magical pixies holding a wire and when they're told to, they touch it to ground and hold it for a little while, then they pull the wire back up until they have to do their job again.
For those of you who don't know, electricity flows much like water in a pipe (that's how they were able to build the internet as a series of tubes), and the bigger the pipe, the more electricity you can get through it. There's a reason why your battery has big, fat wires going down to the starter, but most of the rest of the wires on the car are pretty tiny.
Coils are high current voltage devices. At the business end of them, they can easily kick out 50,000 volts or more. They have to, because it's really hard to ignite gasoline vapor under pressure, and if you can't do that, you might as well get out of the engine business and go back to building horse-drawn carriages.
On one end, you've got 50,000 volts; on the other end, you've got a practically microscopic wire that a pixie is holding. If something goes wrong with that system--if the smoke gets let out of the coil--there's a chance it's game over for the pixie. PCMs have current-limiting devices in them which is supposed to save the PCM when something goes horribly wrong out on the engine, but they don't always work. It depends on the failure mode.
In the case of this poor PCM, the pixie didn't stand a chance. I tossed a new coil and a new spark plug on the thing in the hopes that it had pulled through, but no joy. The PCM might as well be a brick for all the good it does this car now; it can no longer fire the #5 spark plug no matter how much it wants to. That pixie's dead.
Just to make sure, I tested the computer every way I knew how, made sure that none of the control wires were shorted or grounded, but this is a failure I've seen before across multiple platforms, and when I ohmed the control wires for #4 and #6 and got a much different resistance value than for #5, I knew what it was gonna be.
I really wanted to pull open the PCM and look inside. A lot of times with driver failure, you can see a melted chip, or even better, scorch marks on the case. Heck, I once put a PCM on a Nissan that had such a spectacular driver failure, you could see scorch marks on the outside of the case. It must have been fun times when that one went.
Sadly, they'd made the PCM childproof, and even when I took all the screws out, the covers stayed stuck on. I assume that they're glued on somehow, and since we haven't actually sold the job yet, I can't get too destructive on the old unit. I might have to put it back in the car, if the guy decides he'd rather have a 5-cylinder Lincoln instead of spending the money to fix it right.
After I'd diagnosed that one, my next mission was an older Ford Ranger that was a no start. It didn't take too long to figure out that it had no spark coming out of the coil pack, which was brand new.
It also looked like a really cheap part.
So I shoved that thing into a bay and went to work with my test tools. It didn't take too long to figure out that it had all the inputs that it needed, and it should have been happily throwing out sparks at the right time.
I focused right in on that coil pack. When I was a younger tech, Wells made GM coilpacks that didn't work, and I've also seen a lot of cheap imported parts that don't meet specs at all. And I did find out that this particular coil pack didn't meet specs . . . but it was pretty close. It should have worked. Maybe not long-term; maybe it's one of those coil packs that you're gonna be replacing every six months or so, but it should have still been working.
So then it was time for more circuit testing, and after about a half hour of very thorough diagnosis, I came to the conclusion that the Ranger also had a failed PCM. One of its coil drivers had also given up the ghost.
That one did not have a childproof case, so I pulled out the PCM and took a look inside. Unfortunately, I didn't see any crispy components or smell any burned electronics, which was a shame. Besides being fun to see, it's also an absolute confirmation of the diagnosis.
This isn't out of the Ranger (although it is a Ford PCM), to give you an idea of what I was hoping to see:
Source
I did find out that on the older Fords a lot of times you can identify when they're failing, because the aluminum case of the PCM gets hot by the affected driver. If you've got a thermal camera, apparently it's really easy to see. In this case, that wouldn't have been a good diagnostic path to take, since it didn't run at all.
The owner of the Ranger did decide to fix it.
I'm always a bit nervous when I diagnose a PCM. There's almost always an underlying cause of failure, and it's not always obvious what it was. Obviously, if you don't fix that, then the new PCM isn't going to last very long . . . but sometimes you can't test for it when the PCM doesn't work.
Wiring diagrams aren't all that helpful, either--they generally just show computers as mystery boxes with lots of wires going to them. The Ranger wound up being a fairly clean diagnosis; testing revealed that one of the circuits was always grounded at the PCM even though it shouldn't have been. Why it wouldn't run the other coils in the coilpack, I don't know; it seems like it should, but . . . well, mystery box.
There are no specifications for what the resistance on the terminals ought to be . . . I thought that on the Lincoln I could compare--I could run wires to all the PCM grounds and test for continuity through the coil trigger wires. While I didn't know what the spec was supposed to be, I did know that five of the coils worked like they should. Therefore, if they all had one similar value, and #5 was different, it would be proof. Unfortunately, I didn't get continuity through any of them, which only tells me that my test was no good. But I guess since I checked and double-checked everything else, it could only be the PCM.
So thus it was in the space of a morning, two vehicles in a row not only had legit PCM failure, but both had the same component fail inside the PCM. As Dr. House knows, sometimes it is Lupus.
You can bet that when the PCM arrived, I was a little bit worried that I'd plug it in, and the truck still wouldn't start. And then I'd be wondering if my diagnosis had been wrong after all; if I'd overlooked something. Maybe I'd been wrong about the coil pack. Maybe it was far enough outside of spec to not work, and maybe there was something inside it that was confusing the PCM.
Or maybe the rebuilt PCM was no good. I'm sure that sometimes they make bad ones.
Luckily, that wasn't the case. I bolted it in, connected the wires, crossed my fingers, and turned the key. The Ranger fired right up and ran perfectly.
I did suggest replacing the coil pack with a better-quality one, since it's possible that the one that's on there could take out the new PCM, but the customer didn't want to.
Stay tuned for more adventures in the auto repair world! At some point in the future, we're gonna talk about blue car theory, and since I've been inspired by a YouTube blog, I'm also going to admit some of my spectacular mechanic failures in an upcoming blog post. Sooner or later.
An update on the Lincoln--it's now been two weeks and the car is still sitting, disassembled, in the shop. I think I'm going to put it together on Monday and move it outside, because I'm getting tired of looking at it.
4609351
what do you do about vehicles the owner abandons? When my neighbor (he's a mechanic) has a car sitting in shop uncollected for two weeks he calls the owner that if they don't come and get the vehicle that he'll start charging storage fees. That usually makes the vehicle disappear in a hurry. He's only had to "repossess" vehicles for unpaid storage a couple of times.
If a horse can't ignite gasoline vapor under pressure, what are the chances it'll be any better at drawing carriages?
derpicdn.net/img/view/2012/9/25/106307.png
Reminds me of the coilpack that kept going on my old Jetta. It never mattered new or old after a set time the same cylinder would mess up another coilpack. Another reason I will never fully trust German engendering.
The patience one must have for your line of work.
Also
Luna!
4609361
In Michigan, at least, there's a couple of processes you can use to get rid of an unwanted car. The first, and easiest, is just have the local police department tag it as abandoned; after a waiting period, they'll get a tow truck to take it away, and then it's not your problem any more. That's really useful for when you've got a vehicle that's not worth fixing, and that the customer doesn't want back because he doesn't want it clogging up his driveway. But, if you've got any investment in it (let's say that you did fix it, and the customer decides that he doesn't have the money for the repair), then you can go to option two: a mechanic's lien.
I don't know exactly how that works, but there's a legal process where you can basically take possession of the car in lieu of time/materials invested in repairing it.
You can also charge storage, although we rarely do. We have a sign saying that we will, just in case it comes down to a situation where we have to. I honestly can't think of a single time that we did, though.
We have bought junk cars from customers, or just taken them . . . basically, exchanged the service fees we had on the car for the title, and gotten rid of the car that way. A few of them went off to the junkyard, some got fixed and sold, and I own or have owned several that came to us that way . . . in fact, thinking back on it now, my last four cars have all been customer rejects.
As an aside, it's kind of a delicate balancing act to buy cars from your customers, since if you tell them that the car's not worth fixing, and then a week later they see you driving around town in it, they begin to wonder if maybe you're ripping them off. There are certainly people I wouldn't even consider buying their car, even if I knew I could get it real cheap . . . and one that we bought from a customer years ago got sold real cheap to a friend of mine on the condition that it never, ever come back to my hometown for any reason.
4609365
That's a somewhat horrifying picture.
These are enjoyable
4609367
Maybe the PCM was bad. Or there was some problem with the wiring that caused repeat failures. Or the replacement part (if you weren't getting genuine Volkswagen coilpacks) wasn't made as well as the original.
After posting, I noticed that the new coil looks to have a thinner top than the old one, which makes me think that that particular aftermarket coil isn't as well built as the Ford one. That could just be an illusion from the camera, but maybe when I'm back at the shop I'll put them side-by-side and see.
4609370
You can swear a lot to fill the free time.
And actually, it's usually not so much patience that's the virtue you need, it's a tolerance for pain. Especially when you're looking at something that you know is going to hurt you but you've got to go in anyways.
4609379
Thank you! I'm surprised that people like my odd little rambles, but since they do, I might as well keep on going with them. Lord knows I've got enough amusing mechanic stories.
In the medical world it is called finding a unicorn.
4609382
4609379
Absolutely. I'm not a car person myself, but I still find these stories interesting.
I find it to be a unique smell that I just classify as that burnt circuitry smell.
Magic Smoke.
Uh, no. Coils are LOW current devices. I mean, each individual spark is incredibly high current, but for a tiny instant in time. As someone who's both taught a circuits class and spent way too much time working on cars... if there's a lot of current running through your coils, you've got a problem.
But I suspect you said the wrong thing, rather than you mean the wrong thing. It seems you know what you're doing, in general.
You wanna come work on my '63 Chrysler? I can't quite get the timing and carburetor to get along.
4609404
I could use more unicorns in my day job
i.imgur.com/KPZdTkI.jpg?1
4609407
4609424
Yeah, once you know it, it's unmistakable. I often smell Ford coils, because that's as good a test as any on them.
4609438
That's a perfectly legit description!
4609464
You're right; they're supposed to be high voltage, low current. Although this one was an exception--when the driver stuck on, it became a high current device, even though Ford didn't intend for it to be.
I hope so--I've got a shiny ASE patch that says I do, anyways.
Also, for the record, professional mechanics make up descriptions all the time. I call vacuum hoses 'vacuum wires,' and the little pipes on water pumps 'antennas.'
Oh hell no. I generally hate carbs and working on older vehicles (as in, not computer controlled). I have very little experience with the older, carbureted stuff.
This was an interesting read
I think you mean high voltage.
Coils actually use little current, hence the tiny wires. They just step up the voltage.
It's probably filled with potting. You know, the jello-like substance that car manufacturers like to drown their PCM PCB's in.
omg I fucking hate those so much. It's always the later model cars too... you know, the less understood ones? On older car wiring diagrams, the computer terminals are usually somewhat labelled, but on the older cars, it's usually also easier to guess what they would be.
Odd about that Lincoln. Any guesses as to what the hell cooked the coil? Not really surprising that it fried the driver. I mean, if it was drawing enough current out of the driver to melt the coil... but like you say, the root cause is baffling.
4609438
Clarke's Third Law...
Anyone else ship Dr. House and Dr. Bones? (Bones Brennan, not Bones McCoy)
Dr. Cox can stick to Rosie Palms.
I'm liking these mechanic blogs
It's pretty unique.
Pipe flow and electrical circuits are actually mathematically identical.
If you don't treat them as black boxes, they get really complicated really fast. The entire microcontroller may also be a single piece of silicon as well.
This talk of coilpacks reminds me of a repair adventure I helped my dad with on the family's '98 15 passenger Ford Club Wagon (V10).
Long story short, it began with pulling the intake manifold to replace the Stupid Heater Pipe; continued with us, being used to iron engines, not using anti-seize compund on the spark plugs that we changed while we were in there; and ended with mom calling us out because the van had lost power and sounded like a lawnmower, one coilpack and plug having been ejected from the head
Ended up needing helicoils on two cylinders for the stripped threads.
As far as PCMs go, the only car in the extended family to need a new one, so far as I know, was my grandfather's '86 bronco, equipped with a carbureted 5.8. Being a transitional year, and a California spec vehicle, replacements were in short supply when it failed, and so currently,(assuming my aunt hasn't scrapped it) it has a near-final prototype unit from out of a desk drawer at Motorola.
I feel the need to rewatch The Magic School Bus S2E08 and S4E10.
Lovely that netflix is rebooting the series. Terrible that Lily Tomlin isn't reprising the role. There's no one else who can or should do it.
Of course, I was into the original books well before the cartoon came around.
Ah, tales from the shop floor... Let me share my story.
The shop I used to go to wound up scoring a major contract, meaning that they didn't need customers anymore. They would still work on some people's cars (people they liked), but turned most people away. The shop was always full with new Toyota Hi-Lux's they were fitting out for their contract, so I didn't bother them with minor stuff.
Cue the miss fire in my Mazda MX-5 (Miata).
I was unconcerned, as I figured it was just the plug-leads. As they age, the insulation breaks down and they arc out the side to the cylinder head. They burn out slightly faster than timing-belts, so I figured they were about due. So I run it down to the guys. "Filters, fluids, inspect the brake pads, and it's got a miss fire that is probably the plug leads. I'll call you in the afternoon."
The afternoon rolls around and the car is ready. I settle up the bill, and drive off. Where it starts missing again. 'Round the block and back to the shop. "Hey guys, guess what..."
It's intermittent. Idles fine, heavy load is usually okay, full-throttle is good, the miss just comes and goes. "Leave it with us, we'll double check in the morning."
I call in the morning. Coil pack is on it's way out. It was built before coil-on-plug was common, so you have to buy the whole pack that bolts to the firewall. Settle up again in the afternoon (they cut me a break on the labour), and drive off.
Still with that random miss fire.
More diagnosis. Injectors are fine. The engine earth is checked, and rechecked. Swapped out the air-flow sensor (because the miss went away briefly when it was unplugged)... I'll lay odds that they tested every electrical connection in the engine bay.
Eventually they open the ECU. No fried components, but a tiny crack and a little soot on the printed-circuit-board. Enough connection left for it to be mostly fine, or arc through, but also enough for it to lose connection randomly...
But that's not quite the end of the story. You see, this particular MX-5 was the very first of the 1.8 litre motor. It was built for about 3 months before they changed almost everything in the engine bay. ('93-'94, I think about the same time the USA started on the OBD-1.5 standard.) It's an orphan, parts are a real pain in the arse to get. There are no replacement ECUs to be found...
So my broken ECU is sent off for repair, with the understanding that when I get it back it could fail at any time. The bill is paid again (just parts, no labour or diagnostics, they REALLY cut me a huge deal). And I drive off with a butter-smooth running engine.
It's never lupus, except when it is.
4609505
"Vacuum wires" actually makes sense, though. The function of the vacuum system in a car is entirely analogous to an electrical circuit.
4609554
4609564
I did. Oops.
Of course, in the case of this one, it turned into a high current device, even if Ford hadn't meant for it to be.
I haven't seen a PCM that's full of potting yet--just the little glazed coating to protect the circuit board and components. Although I suppose it's possible. I think the case is just sealed shut, and if the owner does decide to buy a new one, I'll do my best to pry it open and see.
In the service information to one of my older vehicles (I think the s-10) it says that the computer takes sixteen inputs to control the engine.
Those were simpler times.
I'm pretty sure on my Suburban, the only thing the engine computer can control is the EGR valve. It's a 6.2L diesel.
The combo of cooked PCM and dead coil seems to be a common one on those, at least according to Identifix. I'd venture a guess that the driver went first, stuck on, and fried the coil, but there's really no way of knowing. It's kind of a chicken or egg problem.
4609585
I'm perfectly satisfied with a computer being filled with magical pixies, so long as I know what those pixies are supposed to be doing.
4609654
No, I think that's just you.
Dr. Cox is awesome.
4609711
Huh, I didn't know that. So the internet is a series of tubes.
That's true; however, some diagrams do show partial circuits in the PCM, which can be handy for diagnosis. Just knowing what pin the driver ultimately grounds through would have been useful, since I could have tested that. Without that information, I was just guessing on one of my tests.
4609728
Which Stupid Heater Pipe, the one under the intake, or the one in the back of the intake? (I've done both, and the one under the intake is actually the easier one to replace.)
You were screwed from the get-go on that; the OE Ford plugs only had three threads and were prone to being ejected. In many cases, it happened with the OE plugs that Ford installed at the factory.
The 'good' news is that the new 3-valve design of the Triton motor now has spark plugs that don't come out, whether you want them to or not.
Of all the crappy cars I've had (and currently still do), I haven't had a PCM fail yet. I'm sure it's going to happen sooner or later, but it hasn't happened yet.
4609748
I'm not sure that I ever saw a single episode of The Magic School Bus.
4609771
Man, multiple failures are the worst. Probably you could have gotten away with just the PCM (if the shop had known, although I can't blame them for not replacing it first--I'd also go for the obvious other stuff). Once ran into a Charger that had both PCM and throttle body failure, and we had to replace the PCM to find that the throttle body was bad . . .
I'm curious how they fixed the PCM. Did they re-solder the questionable contacts and hope for the best?
There are some shops that will fix electronics in-house. A few cars that we've worked on, I've thought about it, although I don't have a steady enough hand with a soldering iron.
4609861
Yeah, it's basically the same thing when you get down to it. On HVAC systems, the vacuum tubes are often color-coded, as well, just like regular wires.
The only big difference (and it sometimes makes diagnostics easier) is that unlike electricity, you can't reverse the 'flow' of vacuum, so on bidirectional controls, you've got to have a vacuum line that 'pulls' each way.
4609983
The one under the intake. Sure, it wasn't too difficult, except for the lack of room under the dash. Had to sorta glide the intake straight back through the hatch. Also we didn't have a tool for the fuel line connectors, so we made one out of pvc pipe (worked perfectly).
I only call it stupid because the idea of having several feet of unsupported tube mounted closely to a vibrating, ridged chunk of metal strikes me as stupid. Especially since the tube isn't straight, giving it more flexibility. We had to replace it because it had a hole worn into it halfway along its length. If it could be replaced without r&r of other major systems, I'd give it a pass, but it can't.
4609987
Since there was a crack, they probably used (a) piece(s) of wire to connect the components to each other directly, bypassing the damaged traces completely.
If they were good, they would have drilled through the board at the point of the crack, to keep it from growing, and used a saw or small file to turn the crack into a slot, to prevent the sides rubbing.
4609969
Really? A lot of them are. Caravan pcms are. It's kinda cool when you're the first person to open one of them, and the potting is perfectly clear reddish, legit looks just like a fresh jello mold.
Hmm, I can't find a picture of one. Oh well.
4610616
I've got two Caravans, and the PCM's pretty easy to get at on them. Maybe I should pull one open just for giggles.
4609983
well, obviously the physics are different, but in terms of modeling and energy storage they are the same. You can make a fluid capacitor and it stores energy the exact same way. An electric capacitor accumulates electrons and a fluid capacitor accumulates fluid particles.
yeah that stuff definitely is useful and I feel your pain. I've been working on a student project over the summer and they didn't give me anything about the wiring. I have no idea what's hooked up where, or even what type of sensors they're using.
4611944
My buddy and I ran into that when we fixed the broadcast console at a radio station. Apparently, over the years, whenever they added a new piece of equipment, they didn't remove the wires for the old piece. We took out a bushel basket (literally) full of unneeded wires, and even discovered a live 110v circuit that dangled down near the DJ's seat. Good times.
Reminds me of something weird with my sunfire, but I think its more of a slow gas leak
4612692
wow that's bad.
4613196
Gas leaks aren't good--you should get that looked at. Besides the obvious extra cost of irrigating the road with your gas, fiery crashes are only cool when you see them from a distance, not when you're involved in them.
4614032
To be fair, it was a college station that was mostly maintained by volunteers, but still.