Shortly before Thanksgiving in 2015, I worked for Bentley Publishers, writing a vintage ignition systems book. Most of the photography used my own cars. There was one more chapter I needed to shoot. I decided that the best car to use was the Bavaria since its distributor is at the front of the engine, whereas the distributor on a 2002 is at the back of the head and partially overhung by the cowl, making it more difficult to photograph. Unfortunately, I’d already moved the Bav out to Fitchburg for the winter. So, after verifying that sunny weather was forecast, I threw a spare battery in the Z3, drove out to Fitchburg, put the battery in the Bavaria, and prepared to swap the cars.

I do miss those Fitchburg garages.

But the Bavaria had trouble starting.

Well, hey. The car had been sitting for most of the last six months since I drove it to The Vintage, and it was cold out. Oh, and one of the automatic chokes didn’t rotate to the closed position unless you did it by hand. Still, the degree of trouble was unusual. I’d gone out to Fitchburg two weeks before to exercise the cars before putting them away, and the Bavaria didn’t have this problem then.

The no-start was bad enough that I ran the battery down. I maneuvered the Z3 nose-to-nose with the Bav—I store the cars with their battery end at the garage door for this very reason—and jumped it. I pulled the Bavaria out of the garage space, pulled the Z3 in, yanked out its battery, threw it in the trunk of the Bavaria, and headed for home.

The next day (two days before Thanksgiving), I drove the car from home to Bentley Publishers and photographed what I needed to on the ignition. Nothing too involved; I had the cap and rotor off the distributor and took a few shots of a multimeter measuring the primary and secondary resistance of the coil, and the capacitance of the condenser.

Over lunch, I checked the weather and traffic online and learned that the heaviest traffic volume of the Thanksgiving weekend was expected Tuesday evening. Additionally, rainy weather was expected to move in on Saturday. I thought that if I buttoned things up and left before 3:00 p.m., I could hit a window in which I could get the Bavaria back to Fitchburg and be done with it before the holiday. So, at 2:45 p.m., I went for it. I re-started the car—which again started a little hard but then seemed to be okay. I pulled out of the driveway at Bentley and made two right turns…

And the car died.

Cambridge. Where you REALLY don’t want to be wrong about where you leave a disabled car.

The symptom was that the car seemed to want to start but would only idle for short periods of time, and when it did run, the idle speed was extremely low, and it wouldn’t rev up or stay running.

Fortunately, I was on a relatively quiet side street and was able to coast to the curb, but  I was in a tow-away zone. Unfortunately, I had very few tools with me.

This posed an interesting dilemma. Should I run—literally—back to Bentley and get some tools? If I did, I ran the risk of the car being towed in the short time I was gone. The City of Cambridge is absolutely vicious about towing; their motto might as well be “Cambridge: One Wrong Move, And We’ll Ruin Your Whole Damn Day.” In fact, the car died only a few blocks from the Great Towing Incident of 1981, when Maire Anne, her friend, and I all met for lunch, all parked on the wrong side of the street during a day that happened to be a street-cleaning day, and all got towed. (I had borrowed my boss’s 733i, and when they towed it, they picked it up from the front, which mashed in the tailpipe in the back, and I had some ’splainin’ to do.)

I poked around under the hood of the Bavaria. Nothing looked amiss. I reseated the plug wires, including the fat center wire to the coil I’d had off. I pulled the cap and made sure nothing looked wrong inside. But I didn’t have my remote-start switch, which is one of my number-one tools in this sort of circumstance.

I  made a quick check of the fuel situation. The tank was about half full; I wasn’t out of gas. I pulled the top off the air cleaner and thought that I could smell gas in the throats, but when I exercised the linkage, which should move the accelerator pumps and squirt fuel down the throats like little squirt guns, I didn’t see anything.

My quick conclusion: The float bowls had run dry. Either the fuel pump had died, or there was an obstructed fuel filter, or air was getting sucked into a line between the gas tank and the fuel pump.

It was now 3:00 p.m. Both the sun and the temperature were about to drop rapidly. I’ve spent hours wrestling with cars and my Hack Mechanic pride, have found and fixed many minor things on the spot, and gotten cars back on the road again. But bereft of a voltmeter, a remote-start switch, and can of starting fluid, I opted for the easy way out.

Hello, AAA?

I logged the call at 3:15 p.m. Remembering the Great Towing Incident of 35 years ago, near this very spot, with Nick’s 733i, I requested a flatbed tow. AAA asked me if I was requesting one because the car was a four-wheel drive. No, I politely explained; the reason was that the car was a vintage car. I’d previously experienced damage, I wanted it up on a flatbed, and I’d be happy to pay the extra $25, as is often the case with AAA. That’s fine, they said, but it would take longer, this being the busiest travel evening of the year and all that. I said I’d wait.

About an hour later, a tow truck arrived. It was not a flatbed. I politely reiterated that I’d requested a flatbed. The tow truck driver called his dispatcher, who said that one should be there in about twenty minutes.

About two hours later, a flatbed arrived. A young man began maneuvering the truck into position in the front of the Bavaria. I interrupted him, asking if he could ramp it up from the back since this would allow the car to be dropped nose-first in my driveway, letting me roll it into my garage in the easiest orientation to do work under the hood. He complied, but when he began winching the Bavaria in reverse up onto the ramp, the tailpipe came dangerously close to the bed of the truck. He noticed it immediately and whipped out a couple of wooden boards to change the angle of the rear wheels to the bed. The Bavaria was then easily and safely loaded.

Oh, the ignominy!

The driver piloted the often rutted streets from Cambridge to West Newton slowly and deliberately, showing a great deal of care over the less-than-glass-smooth surfaces. When he dropped the Bav in my driveway, I gave him a big tip, told him how much I appreciated the care with which he had treated my car and wished him a happy Thanksgiving.

The following morning, I had a quick look at the Bav, figuring that if it needed a fuel pump or other component, I wanted to order it as quickly as possible. But when I pulled the fuel line off, stuck it into a bottle, and cranked the starter, to my surprise, oodles of fuel came out. Clearly, the fuel pump was not the problem.

I moved the accelerator linkage to actuate the accelerator pumps on the Webers, as I had when the car had died. I again looked down the carb throats and again saw no gas being squirted therein. But then I looked closer. Weber 32/36s are progressive two-barrel carbs; the accelerator-pump nozzle is a two-prong spigot, with one prong aimed down each barrel. I realized that, since I was doing this from the left side of the engine, I was looking at the left barrel, which happens to be the secondary barrel. When I looked closer at the primary barrel, I could see that fuel was, in fact, being squirted down the venturi. So my hasty diagnosis by the side of the road in Cambridge was simply incorrect.

But were the accelerator pumps supposed to be squirting down both barrels? I didn’t know. So I pulled the tops off both of the Webers and was surprised at what I saw: Both float bows had a significant amount of rusty sediment at the bottom.

If you saw that, you’d think it was “the problem,” right?

Aha! Well, that’s certainly enough to muck things up, right? I wiped the sediment out of the float bowls, removed the main jets, and blew them out. I also unscrewed the accelerator pump nozzles and learned that even though there’s a nozzle hanging over the secondary barrel, it’s blind; it doesn’t even have a hole in it. The accelerator pump squirts fuel only down the primary barrel. I assume that, on a Weber 38/38 (non-progressive), the same nozzle part is used, but both sides are drilled with a hole.

I dumped fuel into the float bowls, verified that the squirt looked good, put both carbs back together, turned the key—

And it made no difference. The car still seemed to try to start but couldn’t sustain an idle.

Hmm. I wondered if this really was a fuel-delivery problem at all. I grabbed the can of starting fluid and gave a good blast down the primaries of both carbs. This is a great, quick, easy diagnostic test; if the car starts right up, but then dies about five seconds later, it indicates that the ignition is fine, but there is a fuel-delivery problem.

And again, it made no difference.

Well, then. Both my thinking that no fuel was being squirted and the rust in the float bowls appeared to be red herrings. This probably wasn’t a fuel problem. I turned my gaze to the ignition. This made perfect sense, of course, since I had had the ignition partially disassembled at work.

The coil resistances all looked good. The condenser capacitance was fine. I hooked up my remote starter, spun the engine, and measured the dwell with a dwell meter. It was about 50 degrees, which was in the ballpark. Anyway, it wasn’t zero—points always open—or 90—points always closed.

I checked for spark the old-fashioned way, pulling the center wire from the distributor and holding it with rubber gloves near an engine ground while cranking the engine with the remote-start switch. Then, I did the same with the plug wires. The spark looked fine, both coming out of the coil and into the plugs. This observation was in direct odds with the fact that the car wouldn’t fire up with starting fluid.

Now, I’ve had weird problems with vintage cars with condensers, sometimes with brand-new BMW parts in BMW logo boxes. I’ve sometimes had these problems go away when I replaced the condenser with a “known good” one pulled from another running car. So, even though the condenser tested out okay in terms of direct capacitance measurement, I was suspicious, but I didn’t have another known good one at the house to swap in.

So I turned to the next possibility: the plugs. I pulled them out and found that they were a set of Bosch W8DCs, old but serviceable-looking, neither black nor white nor burned, and crusty.

Then I thought back. When I was about to drive the Bav to the Vintage that past spring, I measured the car’s dwell and found it to be greater than 80 degrees out of 90, meaning the points were barely opening. When I discovered this, I pulled the points out and found them to be a corroded, pitted mess. At that point, I ordered a Pertronix electronic ignition for the car. After all, I’ve used them quite successfully in my 2002s, and I feel that they have substantial advantages over points and condensers. You can forget about points closing up and pitting, you don’t have to deal with the garbage quality of new points and condensers, and you never need to set the dwell again. But now I recall that when I installed the Pertronix in the Bav that spring, the car wouldn’t rev above 5,000 rpm. Needing to leave the next day for the Vintage, I never figured out if the Pertronix was bad or if my installation was somehow mechanically or electrically impeding revving; I simply removed the Pertronix unit, reinstalled the original condenser, and put in a new set of points I was fortunate enough to find, unopened, in the trunk of the car.

Now, there are certain things you can find at your neighborhood Autozone on the day before Thanksgiving and certain things you can’t. NGK BP6ES plugs for a Bavaria are among the former; points and condensers are among the latter. So, after verifying that the NGK plugs were in stock, I ran down to Autozone, bought them, and installed them.

The car fired right up—I mean fired right up, immediately, no hard starting, no hesitation. Dang, I guess those plugs really were old.

So, I thought, I guess that’s that. It wasn’t the diagnosis I’d expected, but with 22 people coming the next day for Thanksgiving, I had plenty of other things to do. Not every repair has to be an iron-clad exercise in determinism. Fixed is fixed.

On the day after Thanksgiving, I was determined to get the Bavaria back out to Fitchburg before rainy weather moved in. I threw the very tools I’d wished I’d had earlier in the week—voltmeter, remote-start switch, starting fluid—into the trunk and headed toward I-95, expecting to go from there to Route 2. But as I approached the entrance to the highway, the car began hiccupping, first a minor mid-rpm stumble, then a bit worse. I wasn’t certain anything was drastically wrong, but prudence seemed to dictate heading in the direction of home or at least putting a few more miles on local roads before pulling onto the Interstate.

Then, as I was heading home, the running condition suddenly deteriorated badly. Though the car never died, it wouldn’t sustain any engine load. I got trapped at a stop light about a mile from my house; as I waited for the light to change, I kept my foot in it, trying to sustain engine rpm, but neither holding a throttle position nor pumping the pedal seemed to keep the revs up.

However, I noticed that when the engine stumbled, even though the car never died, the rpm on the tach dropped completely to zero. That’s a major clue. On a vintage car, the tachometer is fed directly off the secondary winding (terminal 1 on the coil). So if the tach drops abruptly to zero, but the engine is still actually spinning, it means that it’s very likely the secondary winding on the coil is cutting out.

I got through the traffic light, hiccupped, and stumbled my way onto a side street, and pulled the car over. It was still running, but horribly, maybe at an erratic 300 or 400 rpm. I opened the hood, put on rubber gloves, and began wiggling wires on the coil and the distributor.

Then I found something. When I moved the wire connected to the condenser—mounted on the side of the distributor—in one direction, the car suddenly began running fine. When I moved it the other way, the car ran horribly. It was very repeatable. Then, I noticed that when I moved the wire, the condenser itself was actually moving.

Gotcha!

The little screw holding the condenser to the side of the distributor was loose. This is more than just a mechanical mounting; the condenser is grounded through the body of the distributor. That small screw is what keeps the condenser’s metal case touching the distributor’s metal body; if there’s not a good connection between the two, the condenser has no electrical ground connection, which means that the primary field can’t collapse and cause the secondary field to create the spark that jumps the gap at the spark plug. And one can imagine how certain engine vibrations at certain speeds if the screw is loose, can cause the ground connection to be made, or to be broken, or to be intermittent.

Sorry to pull a photo of a 2002 distributor off eBay, but I’m not able to get into the garage right now (photo eBay).

I tightened the screw, wiggled things around, and could not make the problem recur.

As I drove the Bavaria the mile back to my house, I remembered that it had started hard in Fitchburg. The loose condenser screw was likely the cause of that as well.

But I wondered: Had I loosened the condenser at Bentley? Was this part of what I had disassembled to photograph? I didn’t think so. In fact, I was certain. Oh, well, I thought. It must’ve just loosened up over time. You can’t find a direct cause and effect for everything.

Now, those of you who follow me on Facebook know that I post a lot of this stuff in near-real time. Someone commented, mostly tongue in cheek, that that’s what I get for still running points and a condenser. I explained how I was, in general, a big Pertronix fan, but that when I’d installed one in the Bav, the car wouldn’t rev over 5,000 rpm. I never figured out why, so I threw the points and condenser back in. My friend Tom Schuch offered something that I’d never have thought of: “Hmm.” he said. “Could that be how the condenser screw got loosened?”

Bingo.

Regarding the removed Pertronix that wouldn’t rev over 5000 rpm, or the rust in the float bowls, or the new spark plugs—you know, maybe next Thanksgiving, we’ll just have red herring for dinner.

—Rob Siegel

____________________________________

All eight of Rob’s books are available on Amazon. Signed, personally inscribed copies can be ordered directly from Rob here.