Mechanic: Comebacks · 1:54am May 10th, 2023
This one's gonna be a quickie . . . probably. 
You know what to do.
One of the 'fun' things about being a mechanic—or in any of the trades, I'm sure—is comebacks. They come in various forms: the 'you said you fixed it but it's still broken the same way it was before' (which may or may not be true; I'll tell you about blocking codes sooner or later); the 'you put tires on my car and the wheels fell off;' the 'it was fixed but the new part broke after some period of time;' or the one we're gonna talk about today, the 'you fixed the problem I had but now it has a new problem which might be related.'
I'm not going to be super specific about the vehicle, 'cause I don't remember exactly what it was equipped with, and it doesn't matter for our story.
The initial complaint was that the cooling fan didn't work right. As we mentioned in a previous blog, electric cooling fans run at one or more speeds to pull air across the radiator and any other coolers the vehicle might have (in this case, also a condenser and an intercooler for the turbo; the trans cooler is built into the radiator). The check engine light was also on.
To save money on the engine computer, GM figured out a clever way to run the three-speed cooling fan with only two drivers in the PCM. It took four relays to accomplish, but I guess relays and associated wires are cheaper than a board-mounted driver? Or else they were putting late-stage improved functionality on an existing PCM, who knows.
I won't bore you with diagnosing it. Figuring out the wiring diagram and what worked when was its own challenge.
Not to mention that both our service manuals indicated that the resistor for the cooling fan is under the driver's side of the dashboard. Which would be weird if true; it's actually mounted by the radiator.
The resistor was bad, and the only way you can get it is to buy the entire cooling fan assembly. Nobody local's got one, and the fan still works on one speed, so the customer takes the car with the intention of bringing it back when we get a fan.
Last week, we got a fan. It wasn't all that fun to change; everything's stuffed into the engine compartment tightly. I had to pull a couple of shields from the bottom, drain the cooling system, remove the degas bottle, both radiator hoses, the transmission cooler lines, and a few other sundry parts.
I don't know if I beat the clock on the job (part of the issue was I thought I could cheat it out without removing a couple parts, found out I couldn't, and then had to lift it up again to take them off), but it got done. I refilled the system with an Vac-N-Fill, which pulls a vacuum on the entire cooling system so you don't have air pockets when you're done.
Usually.
It's also a good check to see if you have leaks. If it won't hold a vacuum, you have a leak.
It held vacuum, it took less coolant than I expected (but it's a very smol SUV that looks like a roller skate), all the fan speeds worked when controlled by the scan tool, and I was gonna take it for a test drive but my manager said he'd do it, and that was that.
Until Monday when it was back! The customer had made it less than ten miles before it overheated. We'd worked on the cooling system, so this might be our fault.
It wasn't the Dorman fan assembly we got (although I didn't test it); it overheated due to a lack of coolant.
For some reason, my manager didn't believe me, even though the degas bottle was empty.
I hit it with the pressure tested, and nothing happened. It held pressure like it should.
I looked around for leaks, and didn't see any from the top. Aside from the oil leaks (it's a GM product), the engine and engine bay were dry.
But, I saw what looked like it might be coolant on the underbody shield. After lifting it up, I saw drips of coolant on the shield, and on the front of the engine. I also saw where the serpentine belt had been slinging it . . . and yet, it held pressure.
He thought maybe I got some coolant on the shield, and I reminded him that when I had opened the cooling system, the shield was already off the vehicle, leaning up against my toolbox.
Due to the engine design, easy disassembly didn't make it obvious where the coolant was coming from, however, the only things that could be leaking coolant on that end of the engine are the water pump, head gasket, or a hole in the cylinder head or block.
Might I mention it was still holding pressure?
As it turned out, the water pump only leaked when the water pump was turning. I was finally able to see the leak with the engine running and the pressure tester cranked up to about 15psi [103kpa] (system's rated to 20psi [138kpa], but I have to be careful because while the radiator cap will vent excess pressure, the pressure tester will not).
It got the water pump, and this time it got three test drives over the course of two days to make sure.
We could have charged the customer for the water pump and labor; after all, it was pure coincidence it broke just after we fixed the cooling fan. It might have started to leak before that, but if it had, the coolant hadn't made it all the way to where it was noticeable yet, nor lowered the level in the degas bottle enough that I noticed it. But it's a small town and we have a good customer base and a high rating on Rotten Tomatoes among our customers, so we just charged her for the part and installed it for free.
Since it held vacuum and would have held pressure when I tested it, there was no way I could have known when I first looked at it, if I'd even done a pressure test.
There are some jobs that go sideways where you can look back and learn from it. To use an obvious example, always torque lug nuts to specification, and if you're not sure if the lug nuts are good, replace them. Yes, this is the voice of experience speaking.
Always look up the oil filter, even if you're taking off a company oil filter and could just match the part number. The last guy might have gotten lucky . . . this is also the voice of experience.
But this is one of those cases . . . what could I do? Assuming that the water pump was leaking very slightly when I put it together, should I have done additional disassembly just to check? Maybe done a block check, too, in case the head gasket was leaking coolant into the combustion chamber?
There's some stuff where you learn by experience, things where you know on a particular car or engine that if X is bad, Y might be, too.
But there's also some stuff where you just don't know. A car has a lot of parts, any of which could go bad at any time. Sometimes it's just after you replaced something . . . and even if the tech knows better, it's gonna be hard to explain that to the customer. Might be best to just bite the bullet—you know you're right, and you know that's not the hill to die on.
Bonus! I found too many images to use, so. . .
derpicdn.net/img/view/2023/3/2/3055852.png
Source
The car companies don't even call you a "Mechanic" anymore. You are a "Service Technician".
That says so much about American cars...
Another "fun" thing is when the water pump is driven by the timing belt for who-knows-what reason, on an interference engine, and the water pump decides that its time is up...
That looked like a Grumba brand auto-vac with Trixie on it.
It also looks a lot like one of those skylanders bases.
Tidying up in the charities machine shop last week, was trying to sort nuts and bolts in the stackadrawers. take a nut, put a bolt in, spin it through, take it out, put in a drawer etc etc.
Then I found something intresting. a bolt that went in the nut no problem the first couple turns, which would do a fast pass, but after only anothre half turn, it stopped dead. Very carefully picked out another two apparently identical bolts of the current one and the ones previous. All the current ones jammed in the same place reliably, the prior lot all went clean through no problem. At a glance the bols were identical, M3s 10-15 mm longs?
They had very sligtly different thread pitch, havent yet gone to town with calipers or count by pencil but wouldnt suprise me if one was so many TPI and the other was so many TPCM? or 9/10 etc as I dont have a lot of detail experience with the use etc.
Maybe the water pump seal decided it was time to fail because of the pressure cycling on it etc? Lots of variations in modern cars when being serviced, especially given the stories of the torque nuts. Hopefully things have been got before engines fully reached the point where once an engine is built in the factory, thats it, no repair possible, because all the parts get stressed up on finilisation, like Prince Ruperts Tear, taking it apart breaks it so badly that the only thing can be done is the shred it and put it back in the ore processing plant?
Reminds me of when I replaced a thermostat on an 01 Sunfire, and vapor locked the system.
340 Magnum tractor saga update!
The new machine runs nicely, and has a lack of minor issues the old one had. I made sure to learn how to keep from trashing clutch packs (pull throttle to idle, downshift at 15mph or lower).
The loosening bolts, not much to do different. I'll just be as careful as I can be when using the dozer blade.
And now, back to work replacing the bearings on a disc cultivator wheel bearing that went out last night. I discovered the problem when the wheel tried to pass me after entering the farm driveway.
5727481
British vehicles are notorious for oil leaks, too. A (somewhat dated) nerd-humor joke from the 90s was:
Q: Why don't the British build computers?
A: They haven't figured out how to make them leak oil yet.
5727481
I loved my old Saturn, but if you BREATHED on it wrong the oil would start to leak.
5727475
Depends on who you ask. I think my title was Service Technician when I worked for GM, but now that I'm at an independent shop, it's "mechanic." I also say that I'm a mechanic on my tax forms.
5727481
The good news is that some of them self-rustproof as they get older.
It's been an advantage on some of my cars. The starter likes to stick on some of Chrysler's minivans, including the one I have. Luckily, since I wasn't too worried about the leaking valve cover gaskets, when I had to change the starter it fell out as soon as I got the bolts loose--no struggle required!
5727491
The reasons are that it's a more central location in the engine, and it simplifies getting coolant to the engine (for example, a lot of V engines had coolant passages through the front cover, which gave you another couple spots that might leak coolant. If the water pump's right on the block, there's only one gasket to worry about.
Another advantage is if you break your serpentine belt (or v-belts on an older car), the engine won't overheat; it'll run normally until the battery gives out.
Usually, if you replace the water pump at the same time as the timing belt (as well as the drive pulleys and tensioners) with good quality parts, you'll never have a problem with that type of cooling system.
5727499
I think it's the first. I know that there were videos of cats riding Roombas back in the day.
5727505
There are a few that are deceptively close, and they can get you in trouble. Unfortunately, the ones I know of are in the 'both are used on cars' category . . . the nastiest one IMHO is 3/8-24 or 10mm brake line fittings. They're almost identical when they're new--you can thread a 3/8 male into a 10mm female, but it's a little loose. You get some rust on the threads, the 3/8 feels good, tightens down . . . for a while. Then you've got a brake line leak.
You can also thread a 1/2" lugnut on a 12mm stud, and it might get tight, but it won't feel right. Sadly, that 'feel' comes with experience.
This was one of those engines where the water pump is known to leak from the shaft seal, and it's just our bad luck it happened right after we did a different repair on the cooling system. If this had come in for a coolant leak and we'd seen coolant on the front of the engine, would have just assumed it was the water pump without tearing anything apart.
5727523
Lotta those have little bleeder screws here and there to let the air out. Some of them have more complicated instructions. . . .
Huzzah!
I forget, can you see them without getting too far into the machine? Paint marks could be your friend if nothing else works.
That's something you never want to see. I've gotten lucky and never seen one of my wheels pass me, but I know people who have.
Does remind me of a close call I had years ago. Hauling a trailer behind my pickup, everything went fine. Started to unload it and the tongue suddenly came up as I shifted weight to the rear of the trailer . . . the nut fell off the hitch ball. No idea when it had happened; I could have been hauling that thing down the highway with just gravity holding it in place.
5727531
Ford built a PCM that leaked 'oil.'
Not really oil, it was whatever magic juice went into capacitors at the time. They were easy to diagnose, though: if you thought that was the problem, you'd pull the PCM cover and look at all the caps; if one of them had a puddle around its legs, it was bad.
EDIT: I've also poured engine coolant out of a PCM, but that of course wasn't the PCM's fault.
5727553
Most of those were more known for burning it instead of leaking it.
Those were good cars, though. Simple and reliable and mostly easy to work on.
Coincidences happen. Nothing can be done about that. As for when when to stop looking for problems, that is an undecidable problem.
5727627
I agree. Despite problems with the oil, that SL-1 lasted from 2000 to Feb 2021 and very nearly made it to the moon (It was about 20,000 miles short.) before the transmission randomly exploded on I-215 when I was driving home. I miss that car dearly.
5727611
Travis McGee used to call himself a "Salvage Consultant". Simon Templar had his occupation listed as "Gentleman" on his passport. He was British & pre WW2 they let you do that.
...because even bikes apparently complexify themselves by now
https://www.lowtechmagazine.com/2023/02/can-we-make-bicycles-sustainable-again.html
I think I'll remain strictly pedestrian (in Russia/Spb).
I love these, even though my car knowledge starts and stops at replacing my specific car's headlights, putting on new wipers, changing a tire, and adding gas... you explain this very well even for people like me who spend ages reading the manual to figure out which bit is the engine!
5727636
There's a lot of back-and-forth (some of it contentious) with 'did you check' and then the explanation of why or why not.
5727638
Transmissions randomly exploring is a problem, especially on older cars where it might not be worth the money to fix. One of the good ones I had (bought cheap, was better than expected) was either gonna go with the transmission or the brake lines . . . it was the brake lines that went first, but probably not by much. In the time I owned it, the transmission strategy was in the 'this is what I have left, I hope it works' mode.
5727641
I wonder what I could get away with listing on IRS documents? They don't exactly have a drop-down menu for 'occupation'; aside from fear of an audit, there's nothing to prevent me from putting "car whisperer" in the 'occupation' field.
5727784
I have a somewhat complexified bicycle, with an aluminum frame, suspension on both ends, and disc brakes. It probably isn't as reliable as an old faithful . . . then again, as a kid, I often rode my dad's old 3-speed (fifties?) which was great until it wasn't and I stopped my fall with my face.
That was also when I learned that you can hit the ground hard enough to knock all the change out of your pockets.
Old tech is good and tends to last forever, but it also has its drawbacks. I think it's up to a person to decide what they want and need; I don't know if my new bike is as well-built as my dad's old bike nor do I know (yet) if it'll last as long, but I do know that so far all its failures have been more benign, at least in terms of a face-hitting-pavement standpoint.
5727816
Thanks!
If I do 'em well, the beauty of these is that you don't have to know much going in, and you can at least follow along. Maybe you'll learn something new, and as a longer shot maybe it will serve you in the future. If not, you'll at least have some cute pony pictures to enjoy, so that's a win, too.
And I'll be honest, there's also a bit of skillbuilding and knowledgebuilding that happens on my end; if I'm doing it well, I need to explain to an audience who often has no idea what I'm going on about in a way they'd understand--not only is that a good tool for talking to our actual customers who might not know why their engine has a water pump or what it does or why that's important, but it also sometimes leads me out of the technician path and into the basic concepts path which can be very useful as a diagnostic tool. "What should this do, and why isn't it?" is something that can get me wrapped around the axle when I'm hours deep in a diagnosis. Literally just this morning I was doing one of those where the end repair was something that could have been diagnosed quicker if the customer had mentioned one other symptom the vehicle had, or if the previous tech working on it had checked the codes.
I also at first thought the wrong part had gone bad, but then I did some additional testing and found that the problem was different than I'd originally thought and was able to find the root cause.
5727835
One of my brother's summer jobs was working for a firm that rebuilt generators (among other things).
He said he got so he could guess what decade it was built in just by looking at "How much copper is in the old coils?".
If the failure mode is silent and invisible even on a test drive then it's one of those failures you're not going to catch, it's that simple.
5727845
it's a lot easier in my field; the external appearance of the car gives it away.
If I was just looking at an engine not in a car, I could probably guess the decade, though, at least for most engines, based on what materials were used to build it (cast iron, aluminum, plastic) and what kind of gaskets it has.
5727877
Yeah, that's true. We've got one of those in the shop right now; odds are the failure is customer error, since none of us can think of a failure that could happen like the customer described and not leave any evidence behind . . . but the fact is we don't know at this point. Could be something sneaky and weird. Just yesterday I diagnosed a failure that fooled the first tech that looked at it; now that I know what actually went wrong, I understand why he was misled (and why I was at first, too).
That one might be the subject of a future blog post
5727835
I think the "old tech lasts forever" trope is often survivors bias. People built a lot of junk back in the day as well, but thats all broken down and been scrapped by now. The stuff that's still around was probably of higher quality.
5730801
Yeah, I think that's fair.
And there was a lot of older stuff that was overbuilt by today's standards. If you don't have good engineering models, you can make up for it by adding more material, which is usually going to make it last longer.