Don’t let your diagnostic blind spots bite you in the rear

When it comes to cars (and life too, I suppose), things aren’t always as they seem. Two weeks ago, in a story about “Points vs. Pertronix,” I mentioned installing a brand-new condenser that turned out to be bad. The slightly longer story is this: I’d just had the distributor in my 1972 BMW 2002tii rebuilt. I outfitted it with a new cap, rotor, points, and condenser purchased at the dealership—e.g., presumably high-quality parts in BMW boxes with BMW part numbers. I test-drove the car, and after about five miles, it began to run horribly. I barely made it home.

I posted this on Facebook, and a good friend and professional mechanic offered his help (actually, what he said was, “Can I come over and play?”). When he arrived at my house, he was holding a distributor. “This,” he said, “is known-good. I just pulled it out of my working 2002. Let’s install it in your car.”

I was, candidly speaking, a little miffed. “How can you say yours is known-good and mine isn’t?” I shot back. “Mine was just rebuilt and all the parts are brand-new OEM. There can’t be anything wrong with my distributor or the components in it.”

My friend, the professional, looked at me, and asked rhetorically (and sharply), “Are you having a problem or aren’t you?”

“Well, yes,” I admitted, sheepishly, but grumbled under my breath that this was a stupid idea.

When we dropped his dizzy in my car, it fired up and ran smooth as warm butter. I was stunned.

“Now,” my friend said, “we’re going to transfer one component at a time from your distributor to mine until we find out what’s bad.” Sure enough, when we swapped my brand-new condenser onto his distributor, the problem came back. My jaw fell open. This meant that my new condenser, purchased at a BMW dealer, removed from a BMW-logo box, was bad.

It was then and there that I learned two extremely important lessons:

#1: There is no “known-good.” There is only diagnosis.

#2: Whenever you say, “It can’t be that,” a bell should go off in your head, warning you that you might have a blind spot in your ability to diagnose, and if you don’t remove the blind spot, it’s likely to bite you eventually.

Of course, learning the lessons and living the lessons are two different things, but I recently had a great opportunity to see if I actually could take off the blinders.

My other BMW 2002tii just spent 11 months in the BMW Car Club of America Foundation’s museum in Greer, South Carolina, as part of its “2002 ICON” exhibit. This was the decade-dead car I’d bought sight-unseen in Louisville in early 2017 and fixed it where it sat, then road-tripped it 1000 miles home. It was the subject of my book Ran When Parked. As part of that feverish initial sort-out, I replaced the clutch slave cylinder. I initially left the clutch master alone, as it appeared to work fine. Then I found that the clutch disc was seized to the flywheel, and I had to break it free by starting the car in gear with the clutch and brake pedals depressed. I then headed toward the entrance ramp to begin the long drive home, and the clutch master cylinder chose that exact moment to fail. A friend came to my rescue by producing, on a Sunday, a new old stock (NOS) clutch master he’d had on the shelf. I installed it, and the next day I was on my way home for real.

Me and “Louie” at the “2002 ICON” exhibit at the BMW CCA Foundation museum.
Me and “Louie” at the “2002 ICON” exhibit at the BMW CCA Foundation museum. Rob Siegel

The fact that the car’s original 45-year-old clutch hydraulics died, particularly after sitting for 10 years, isn’t surprising in the least. But after replacing both the clutch master and slave cylinders two years ago, I expected them to be good for the next 15 years. Well, funny story…

A few months after I drove the car 900 flawless miles to deliver it to the museum last winter, the curator informed me that, when they tried to take it out to exercise it, it had become difficult to shift into gear with the engine running. At some point last fall, “difficult” transitioned to “impossible.”

As I prepared to go down and retrieve the car when the exhibit closed in January, I needed to deal with the apparently non-functioning clutch. The curator said that the fluid reservoir was above the fill line and there was no fluid under the car, so the problem didn’t appear to be caused by fluid loss. That left three possibilities: Air in the line, the clutch disc stuck to the flywheel, or hydraulic failure of the master or slave cylinder. The idea that two-year-old clutch hydraulics would fail seemed remote, and since the Foundation museum is a climate-controlled facility, the possibility that the clutch had re-seized to the flywheel seemed equally unlikely.

The photo I was sent of the fluid reservoir showed the fluid clearly above the level of the clutch hose on the left, indicating that the clutch hydraulics weren’t out of fluid.
The photo I was sent of the fluid reservoir showed the fluid clearly above the level of the clutch hose on the left, indicating that the clutch hydraulics weren’t out of fluid. Rob Siegel

I did my best to remove the “I just replaced the clutch hydraulics two years ago so it can’t be that” blind spot. If they were indeed bad, there would be no way for me to procure the parts on a weekend, leaving me with a disabled car and no way home. Looking online, I found that I could purchase both a clutch slave and a master for $110 and have them shipped to the museum, so that’s what I did. I figured that the odds I’d need them were slim, and that I’d probably wind up putting them in inventory in my garage, but better safe than sorry, right?

Replacing the clutch hydraulics in the back of the museum.
Replacing the clutch hydraulics in the back of the museum. Rob Siegel

When I got to the Foundation museum, I grabbed a friend and had him pump the clutch pedal while I looked under the car. What I saw was quite surprising. When the clutch pedal was depressed, the clutch slave extended and pushed the clutch release lever as it should, but then it immediately relaxed back instead of staying extended as long as the pedal was depressed. I bled the hydraulics to remove any air in the line, but it made no difference. Clearly either the master or the slave cylinder had gone bad.

Although I suspected the master, the slave was easier to change, so I replaced it first. That made no difference, so I moved onto the master. With it replaced, the clutch action was restored, and the car got into gear and could drive perfectly well.

Out with the bad (left), in with the good (right).
Out with the bad (left), in with the good (right). Rob Siegel

I would not have predicted a two-year-old clutch master cylinder would be the failure point. It’s possible that the seal in the NOS part I’d installed two years ago had dried out while it had been sitting on the shelf, but that’s just a guess.

I’m just glad I was able to take the blinders off, admit that needing to replace two-year-old clutch hydraulics was a possibility, and order the parts and have them with me, because if I hadn’t, dealing with the car would’ve been much more expensive. It wasn’t trivial changing both the master and the slave, but it wasn’t that bad, either. It took perhaps four hours from start to finish, and I was blessed to be in a heated, well-lit environment. Doing it in the winter in some cold parking lot would’ve been pretty miserable.

So, whenever your car has a problem and you catch yourself thinking, “Well, it can’t be that,” give that a good hard look. Maybe you’re right. But maybe your own blind spot is about to stick it to you. Don’t let it.


Rob Siegel has been writing the column The Hack Mechanic™ for BMW CCA Roundel magazine for 30 years. His most recent book, Just Needs a Recharge: The Hack MechanicGuide to Vintage Air Conditioning, is available on Amazon (as are his previous books). You can also order personally inscribed copies here.

Click below for more about
Read next Up next: These 6 oddball cars are the heart and soul of Paris Rétromobile

Leave a Reply

Your email address will not be published. Required fields are marked *