MECHANIC: the B-Van and the Sable · 12:34am Oct 5th, 2023
Alrighty, kids, today I've got two different vehicles for you. Neither of them were particularly complex diagnoses, but both have cool pictures with lots of diagnostic wires added.
First, though, y'all know what to do.
We'll start with the B-van. Many automakers have some kind of platform code for particular vehicles; for example if you're familiar with the older GM pickups, you might recognize them being called C/K trucks. Ford likes animal names for at least some of their cars, so you've got Fox-body Mustangs, and I legit have no idea what Dodge does. Back before they had ProMasters, their full-sized vans were 'B' (Ram pickups were 'R'), and then the weight rating: the one in question is a B2500, or a 3/4 ton van.
I don't remember the model year; it was either very late 90s or very early 00s.
The customer's complaint was crank no start, and he even said that he thought the ignition coil was bad.
You might remember from a previous blog that ignition timing was first controlled by a distributor which might have various mechanical advance techniques, then just the computer deciding when to fire the coil (but still with a distributor) and then coil packs for multiple cylinders, and finally a coil for each. This van falls into the 'distributor and computer-controlled timing' category.
Regardless of which specific model year it was, it was over 20 years old and the coil might have been original; if not, it hadn't been replaced in a long, long time. Initial tests showed that it had no spark when cranking, and the ignition coils were a common failure on those, so the easiest first thing to do was hang a new coil on it.
I did that, and nothing changed except the coil was new and shiny instead of old and rusty.
I then commenced to doing additional tests on it. The coil had good power (it would light a headlight bulb) but it never had a trigger signal—at the appropriate time, the PCM is supposed to ground the signal wire, which will collapse the electrical field the coil makes and fire spark down the coil wire to the distributor, where it'll be distributed to the appropriate cylinder.
While the scan tool can't tell me if the computer is commanding the coil to fire, it can tell me if the computer is getting signals from the cam and crank sensors, which are the ones it uses to determine when to fire the coil. In theory, it only needs one of those to work, since the cam and crank should be connected by a chain and if that chain is broken, you've got bigger problems than a lack of spark.
Identifix says that the most common cause of lack of spark is a failed PCM. Now, I know (based on previous repairs) a bad coil is actually more likely. We're dealing with biased information, though. Any tech experienced with these things is going to put a coil on first, and only look up tests and whatnot if the coil didn't fix it—I've put dozens of coils on engines in this family, and never a PCM.
Way back when, I blogged about a Mercury that fried the coil driver in the PCM when the coil shorted, and I expect that there's a good chance the same thing has happened here. I had some more options on the Mercury; it was coil on plug so it would run on five cylinders, proving that the cam and crank sensor were good and giving nice, sensible signals to the PCM.
Luckily, even if I can't make it run, I can test those signals. The scan tool says that the PCM says that they're good but I've seen weird failure modes on a Ford when the signal is mostly okay except where it isn't. That one didn't set a code for bad signals.
I can also test the coil trigger wire at the PCM, and that'll tell me if I've got an output and a broken wire somewhere, or if the PCM has an internal failure. So I go and hook up some wires:
Both the cam and crank sensor should be making square wave signals. Both are 0-5V signals, and the crank sensor should have a particular pattern to it, while the cam sensor should just be a steady square wave. Both signals should increase in frequency with engine RPM, although since this one won't start the only RPM I'll get is what the starter can produce.
The signal should look like this [In the interests of full disclosure, this might not be a post-repair signal from the B-van, it might be from something else, but you get the idea.]
If I were to have matched up the crank and coil driver on the scope (I didn't), I should have seen four pulses of the coil for each revolution,* which I would have known because the crank signal has a specific pattern. If I thought it was firing at the wrong time, I would have looked that up, but it wasn't firing at all . . . and the test proved it. I've got good signals going into the PCM and no coil trigger coming out; the coil driver is cooked.
*Because it's a four-stroke engine, the crankshaft rotates twice before it gets back to the start, so only half the cylinders should fire with one crankshaft rotation, hence only four firings rather than eight.
I would have liked to open up the PCM to know for sure if there was an obvious visually melted component, but the case was glued shut, and the coil driver didn't fail hard enough to leave scorch marks on the outside of the PCM . . . I've seen that before, and it's a real easy diagnosis when you do. Delicate electronics shouldn't have burn marks on them.
It took us a week or more to get a PCM, but when we did the vehicle fired right up. We left the new coil on it, since it's cheap (compared to a PCM) and the van probably needed it.
Which brings us to the Sable. The customer's complaint was 'lack of power,' and the check engine light is on. I pull codes and it's got two, both of them for the Output Speed Sensor (OSS). [I think they were P0720 and P0722.]
Before I'm even out of the parking lot, I know it's got a problem; it starts in the wrong gear and the speedometer doesn't always work. Starting in the wrong gear always causes a 'lack of power.'
Most automatic transmissions have certain failure management modes; with the codes that are set on this one, it'll only work in park, reverse, neutral, and fifth gear. It's a way to get you home or to the shop if something goes wrong, but it's not ideal.
Incidentally, that's one reason that people thought computer controlled cars were unreliable; a lot of times they'd failure-manage themselves as long as they could, and then they'd get beyond what they could help and just quit completely. Since I've posted the blog by now, remember the Mercedes that had managed to adapt itself to a hole in the intake manifold and ran pretty okay except for a high idle on startup before it settled down.
Also, it's possible I've already said this before.
Transmissions rely on a bunch of sensors to make sure everything's working like it should. I know I've talked about the OSS before, but for a quick refresher, at the very least transmissions usually have an input speed sensor (ISS) and and OSS which will tell the PCM (or TCM if that's a standalone module) what gear it's actually in, how much torque convertor slip there is, and also how fast the car is going. It can usually cope better with the loss of the ISS, since it knows from the cam and crank sensors what the engine RPM is, and the input shaft speed can be inferred from that.
Diagnosis is reasonably straightforward (it's a nine-step process). I also find a technical service bulletin (TSB), which applies to this car. Some of 'em were built with sensors which could intermittently short out. The TSB doesn't offer a huge amount of advice, except to say in step 9 of the workshop test, replace both the ISS and OSS instead of the PCM.
Step nine is basically 'if you didn't find any problems, replace the PCM.'
And a second test drive bears out this advice; this time the car works normally.
As with the B-van, I can hook up the oscilloscope and see what kind of pattern the two sensors are making. Just like on the B-van, it should be a nice square wave, and it should increase in frequency as the speed of the shafts increases.
Before I get to the pictures, though, I'll tell you that sometimes what things are doing over time is just as important as what they're doing in the instant. Years and years ago, we had a Dodge Caravan with a misfire on one cylinder. Only at idle, and it wasn't real bad, but you could feel it.
We checked the usual suspects, and none of them were the problem. Often an issue only at idle will be an airflow problem; at higher RPMs enough air is moving that it makes the problem go away.
We did a compression test with a manual gauge and it was good. Then we did one with an pressure transducer and it was good, too . . . except when we zoomed out on the whole captured waveform, the compression went up and down in a wave-like pattern; the lifter had a bad spot on its roller, and didn't always open the valve as far as it should.
So now on to the pictures. The sensors are making a nice square wave:
. . . except when they aren't.
The green trace is the OSS, and it's not failing enough for the PCM to flag a code for it, but it certainly has a problem.
And if you thought that was weird, look at what the ISS did:
That is very much not a good square wave.
(If you noticed that it says '2004 Chrysler Sebring' at the bottom of the screen, that's 'cause you can use the oscilloscope without closing out of whichever vehicle you were using the scanner on.)
I put in the sensors, which on this transmission means pulling the front cover, the solenoid body, and the valve body. Not exactly my forte (I think I can still count on one hand the number of times I've done it). I also probably should have hooked the scope back up to it when I took it on its next test drive, but I forgot. Still, it seemed to be okay . . . I guess I'll find out if it comes back for the same problem.
[Narrator: it hasn't.]
So today's lesson is that as fast as the scan tool can be at updating information, if you really want to know what kind of signal it's seeing, get out the oscilloscope and observe.
Also if you have an oscilloscope, it'd be wise to practice with it on known goods, 'cause that way you get experience setting up the scale and the sweep and recognizing what you should be seeing—while it was obvious on both these that something was wrong, I've been in the situation before where I have gotten the data but I'm still not sure if it's showing me anything useful or not.
It's been awhile since I've owned a car. But, I've dealt with mechanics of the "replace a part & see if that fixes it" school of thought. That can get expensive REAL fast.
Voice of Experience
1) If you've got a complaint about the repair:
A) Older folks get taken more seriously than teen ages
B) Men get taken a LOT more seriously than women.
In other words, get your father to do the complaining.
Oh, & you won't get them to give you the parts but you MIGHT get the labor costs reduced.
Sounds like a fun day at work lol
The School of The Problem and The Symptom may take some time (and luck) to connect. Been having heck for the last week with building workstations off the network. Darned thing gets to a random spot and dies with what looks like an 802.1x issue (network version of "You ain't on da list so you can't come in.") Then one (not all) of the office printers starts throwing random 802.1x issues. Was partway to writing Those Upon High to ask what was going on with the overall 802.1x when somebody took a perfectly good working computer and plugged it into the cable I was using to build workstations only to fail, fail, hm, I wonder. Backtracked both cables to the same switch in the cabinet. Moved them to different switches and they worked perfect. J'cuse! So nice to be able to point the finger of blame at a specific chunk of malfunctioning hardware.
Now we know where Twilight got the idea for her sweep music - her oscilloscope!
Had an easy one on the farm a couple days ago. For given values of easy.
Claas 960 forage chopper had a belt squeal starting when I tried testing the chopper system. Turned out to be the serpentine belt and a locked-up alternator. My guess is that the power demands of turning on the forage chopper which probably included a number of solenoids and whatnot was the last straw for the alternator. The local alternator/generator/motor shop noted that plastic innards had melted, which is how the thing ended up being frozen.
Best part? The machine was parked in the shop when it happened. I could just do a quick removal so that the new part could be gotten in the morning. It took less than an hour to reassemble, and everything worked nicely.
5749196
Man, you have got to coast on the feelings from repairs like that.
5749145
It can, but at the same time, sometimes you gotta do it that way. Sometimes one failure masks another one.
Depends on where you are and the shop. We are on the edges of Big Three territory, so we have a few customers who work for the automakers one way or another; I remember one time where a woman (who was an engineer for one of the Big 3) was explaining to her father what the FPDM was on his truck and why he needed it.
Also depends on where you are. Here in Michigan, we are required by law to give you back your old parts, if requested. The only exceptions are parts with a core charge--although if you pay the core charge, you can get it back--and parts that are too heavy or too big to transport . . . like, if you were to get a new rear axle put in your Neon, we wouldn't give it back to you in the Neon (won't fit), but if you brought a pickup truck or trailer to put it in, you could have it back.
5749147
Several fun days at work! Wait until the next blog post when I tell you more than you ever wanted to know about the LIN bus on a 2010 Chrysler Town and Country! What do the windshield wipers and compass have in common? You'll find out!
5749158
It's nice when a baffling complaint can finally be narrowed to a common cause (and I think you'll really appreciate the next blog, since it's up your alley--it's a network problem).
Our shop computer was running really slow recently, and we eventually narrowed part of its problems down to the little WiFi dongle it's got malfunctioning. Internet speeds were in the kb/s range . . . I installed a new component in a car while waiting for it to load a single webpage.
We moved the dongle to a new USB port and it's happier, but I don't understand why the computer didn't get hardwired into our network. There's no reason not to, it's only got to run about ten feet to get to the router. Taking the longer, easier way (so it doesn't have to go into and out of the ceiling), it could go over a door, then along two walls, could even be zip-tied to the conduit that's already there.
I have a feeling the boss didn't want to spring an extra 40 or 50 dollars for the cable, not realizing that he would up spending more than that in unproductive tech time while waiting for pages to load, or swapping a part that didn't need to be swapped.
5749162
That's not an unreasonable assumption.
While my own headcanon says ponies don't have oscilloscopes, there is also something delightful about a pony world like this:
derpicdn.net/img/2017/1/16/1340742/large.jpg
Sauce
5749196
One of my favorite 'we eventually found it' malfunctions was also a bad alternator. Chrysler minivan owned by an old dude, and every now and then the battery light would come on; when it did, we'd find that the alternator fuse was popped, put a new one in, and the problem was gone. Happened very intermittently, and the guy never really wanted to leave it for full diagnosis . . . one day, we finally did get to spend some time with it. Put in the new fuse, did the usual battery of tests, all good. Old manager took it for a test drive, and managed to get the fuse to pop when he was accelerating out of town, which lead us to think it was RPM related.
Sure enough, once you got over 3500 RPM, the alternator would all of a sudden start fountaining out sparks, and the fuse would pop. Old guy hardly ever wound it up that hard, which is why it was so intermittent (and as I recall, the test RPM for our old alternator tester was 2000, so we wouldn't have wound it that hard, either).
It's really nice when it breaks in the shop, makes it much more convenient. We've had a few of those; my all-time favorite was years ago when an Aerostar came in for a front end inspection. The tech pulled it into the service bay, and when he hit the brakes, one of the tie rods fell off.
5749288
My favorite from back in the day was EGR valve stuck open due to carbon on GM 4.3L engines. They idled horribly, smelled like sulphur, and made unhappy engine noises. The valve was right on the front, easy to access. Pull it off, if you were feeling frisky, rev the engine a couple times to dislodge any carbon rocks in the passage, chip the carbon out of the plunger, slap it back on with a new gasket that has a screen to keep the rocks out, and it's fixed. What most customers thought was likely to be an expensive engine repair took about ten minutes and cost under a hundred dollars.
Usually didn't even charge a diagnostic fee on those; it was faster to pull the EGR valve and see than it was to hook up the scan tool.
5749405
I grew up in a REAL small town H L Menkon said "Everyone knew whose check was good & whose wife wasn't.
They only bought the newspaper to see who got caught." near Interstate 80 PA exit 1.
I meant "give you the NEW parts free". Yeah, I knew to tell them to give us the old parts back. My grandfather was a carnie grifter turned auto mechanic. He was the spiritual descendant of P T Barnum.
& this was about 50 years ago. Until 1974 there were only 2 states in the USA California and Tennessee that would give a woman a bank loan without a man cosigning it.
5749407
For that distance, you’d pay less for a run of cat6 then for a replacement dongle.
Unless the router has no available ports.
5749528
I'm sure it's got available ports. I don't know how many they usually come with, but AFAIK the only things plugged into it are the other computer and maybe the printer (based on how slow the printer prints from the shop computer, I suspect it's plugged into the main computer and gets a wi-fi from the shop computer).
Management doesn't always do the smart thing in the shop. I'm grateful that after only a year or more of asking, we finally got new lights for the shop.