Sometimes, it is just easier to install new parts.
I mentioned before how the seatbelt switch on the driver's seat on the Mercedes Sprinter van went bad - or more precisely, the wire connecting the switch to the SRS (Supplemental Restraint System) computer was severed when Mr. See rotated the seat around.
It was kind of appalling, actually, how it was put together. Two tiny thin wires were snaked through a hole in the seat base and just hung there, unprotected, inches from the seat slide mechanism and a host of sharp metal edges. To be fair, this was exacerbated by the seat swivel that Winnebago installed - providing a guillotine for this tiny helpless wire.
People think Mercedes are fancy cars, but once you look underneath, well, you realize they are just cars. And a commercial van? Acres of cheap plastic. For example, the top of the seat base is covered with a thin layer of plastic foam of some sort, that just slides in place. Granted, there are electrical components in each seat base (Mercedes relays and fuses under the driver's seat, Winnebago power relays and circuit breakers under the passenger seat). But to me, that would be all the more reason to cover this with something more substantial than corrugated cardboard.
Anyway, I contemplated soldering the wires back together and was not looking forward to splicing tiny wires under the seat. On a whim, I checked online for a new seatbelt buckle and found one, on eBay, for only $38 with free shipping. For that much, why solder?
The part came in a few days and I removed the seat (which was held in place with those star-shaped bolt heads - fortunately, I had a set of matching sockets) and cleaned out 10 years of dust and dog hair from the seat base. The new seatbelt buckle fit right in, and I tested it first by plugging it into the wiring harness and whoopee, the SRS light went out (but the error code did not reset!). I cleaned everything really well and re-installed the foam plastic cover, this time aided by a plethora of self-drilling, self-taping screws. Put the seat back (and the rotating base - heavy!) and all is good.
Or is it?
I splurged $110 on a new code reader - one that reads and resets codes for all the computers in the vehicle - and there are several, I learned. Sure enough, there was an SRS error code for "driver's side seat switch" which I reset and it went away for good. Error codes in auto computers are interesting. Some OBDII codes (On Board [Engine] Diagnostics II) won't set off the "check engine" light until they persist for some time, or if a certain number of codes are present. It all depends on the type of code and the manufacturer. I guess the idea is to eliminate false alarms.
And false alarms are a problem. I remember jump-starting one of my E36 BMWs (318iC) and it set off the check engine light. I pulled out my trusty code reader and it showed a litany of errors (including, not surprisingly, low voltage) which, when reset, went away. "Hard" errors usually come back right away, which is why resetting codes is a good first step (and maybe the last step) in diagnosing the problem.
For example, another E36 was throwing a "camshaft position sensor" code, which, when reset, would come back right away. The camshaft position sensor was bad and needed to be replaced (not an expensive part, nor a difficult job, either). On the other hand, our E53 (X5 3.0) would throw an "oxygen sensor adaptation limit reached" code, which, when reset, would go away (and the CE light go out) for months at a time - a classic "soft" error. The problem in that case was a tiny pinhole leak in the intake elbow, allowing un-metered air into the system, confusing the hell out of the mass air flow sensor and the ECU. The rubber elbow was a $14 part, held in place with two band clamps - a five minute job, once the problem is diagnosed. You'd be surprised how many dealer mechanics would replace the oxygen sensors (all four of 'em!) and charge the client $1500 or more, rather than replace this $14 elbow.
SRS codes are more strict. Like I said before, on our E36's, people would kick under the seat, loosening these seat switch wires which would set off the SRS light. An OBD II error (CE light) might reset itself if the physical error is cleared, after so many starts. Not so the SRS! So you need an SRS reset tool to clear the code in that case.
Speaking of false alarms, the "loose gas cap" code frustrated more than one car owner. OBD II cars (1995 and up) have a system to detect air leaks in the fuel system. If a gas cap is loose, the computer treats this as a leak in the system somewhere. By the way, over-tightening the gas cap is not the answer! As I learned the hard way, that only serves to compress the gasket too much, causing a leak. If the instructions say "tighten to one click" (as it does on our KIA) then one click it is - not three! Sometimes you just have to break down and buy a new gas cap, though.
Cars in the "good old days" had vented gas caps, and over time, gallons of gas would slowly evaporate from the tank. But hey, gas is like 35 cents a gallon - who cares? But "unburned hydrocarbons" are a big source of smog, and in 1960s Los Angeles, you could not see more than a few blocks on a bad smog day. So as part of EPA emissions requirements, gas tanks were sealed.
Emissions aren't just from the tailpipe, but include things like tire dust and brake dust - and evaporating gas (which is why lawnmower gas cans have these annoying nozzles these days). With regard to the former, look around a parking garage sometime and check out all that black dirt-like powder everywhere. Yea,, tire dust. And brakes? asbestos dust - back in the "good old days." Tires last a lot longer today, thanks in part to better tire compounds. And we no longer use asbestos in brake pads. You can have the "good old days" - I prefer to live in the modern world.
But getting back to loose gas cap codes, manufacturers realized they were losing a lot of customer goodwill when a loose gas cap caused an expensive trip to the dealer. So, many added stickers on the fuel door advising owners of the problem, or even a "loose gas cap reset" button on the dash. But the problem still persists.
But getting back to the topic at hand, why spend $38 on a new seatbelt buckle rather than splice the wires? Well, as I have learned over the years, both as a Technician and Engineer, every connector or splice in a wire is a potential failure point. Wires generally do not fail if left alone. They only fail if they are subject to repeated flexing or are not properly grommeted when passing through a sheetmetal hole, or some other form of mechanical malfeasance. But just sitting there they last forever.
Connectors, cuts, and splices, on the other hand, are failure-prone. And I've see far-too-often, people try to save a few dollars splicing wires on a generic oxygen sensor and wondering why it doesn't work - when a plug-and-play factory OEM part is only a few dollars more (and a lot less hassle to install!).
Of course, to help prevent a replay of the original incident, I sheathed the seatbelt switch wires in corrugated plastic wire loom, which I carefully wire-tied on each end to hold it in place to the seat and seat base. Then, I made sure that the shielded wire would not bind or get caught in the mechanism, regardless how the seat was moved or swiveled. Problem solved - hopefully.
This was a pretty simple setup, too. Some cars have several wires. On the E36 passenger seat were wires for the seatbelt buckle (to check that you were buckled in) another set for the "seat sensor" (to detect the presence of a passenger) and a third set for the explosive device (!!) that cinched the seat belt tight in the event of a crash. And that was back in 1997.
One final word on error codes. I was surprised how many systems there were on the Mercedes Sprinter van - or how many there were potentially. When I hit "scan all devices" the code reader tool went through a litany of devices (nearly a dozen as I recall), including some not on our vehicle, such as backup radar, radar cruise control, and the like. Each system has its own microprocessor and each has its own set of error codes.
If all this sounds daunting, it isn't. Like wires, most electronic components work just fine if left alone. In addition to loose gas caps or disturbed wires, the biggest offender is often loose or corroded connectors. A "wheel sensor" for the ABS (anti-lock brake system) is just a coil sending out a tiny signal generated by a notched part of the wheel hub to indicate rotation. Coils like this rarely wear out, but the connector is located in the wheel well, where it is splashed with water, snow, salt, sand, and road debris. Cleaning the contacts and applying dielectric grease often fixes the problem. Code-monkey mechanics, on the other hand, blindly replace the part, believing (falsely) that just because a part is mentioned in an error code, then that part must be replaced.
Such is not always the case.