I am veering back to my vintage Honda car for this story, but most of the steps and procedures will apply to motorcycles and cars.
For those of you who recall a previous story concerning my various woes encountered with the 1987 Honda Civic Wagon, which I purchased right before Christmas, most of the problems stemmed from failed vacuum diaphragms in the distributor and the EGR valve. Honda packed the little cars with numerous vacuum controlled diaphragms which operated carburetor functions, air control systems, thermostatically triggered functions and other EPA related tasks.
When I first test drove the car, the engine was pretty smooth and idled properly around 800 rpms. After two trips to different smog stations, after “repairs” were done at a local garage, the engine wouldn’t idle down below 1200 rpms and the car was sluggish and fuel mileage was down in the mid-twenties. Once the distributor advance unit was replaced with a new one and a good used EGR valve installed, I reset the ignition timing to specs and the engine settled down to a usually normal idle after warm-up. Still, after a run on the freeway, coming to a stop light afterwards, the idle would be high again, slowly settling down after the engine cooled a bit.
I don’t know how many times I had the air filter off to get limited access to the idle speed functions of the 3bbl carburetor, but it was a LOT. It never seemed to hold a steady idle for more than 5 minutes and gradually I became aware of a “miss” in the engine. You can feel when an engine is going weak on one cylinder by a regular “shudder” and vibration that occurs.
One way to isolate a dead cylinder is to pull one spark plug wire off at a time (at the distributor!) and see if the miss/vibration seems to change. A strong running cylinder will drop power and idle speed when that plug wire is removed or grounded momentarily. A weak, non-running cylinder will yield almost no noticeable effect on idle speed or power delivery.
On the little Civic, the distributor cap has all of the plug wire outlets lined up in a nice neat row, so it is easy to grab each wire with a pair of insulated pliers and remove it to see if there is a change or not. In this case the number two cylinder appeared to be the weakest of the four, with little change whatsoever when the wire was removed from the cap. Fearing the dreaded “burnt valve” syndrome of a 135k mile original engine, I removed the spark plug and tested the cylinder with a compression gauge. I disconnected the ignition coil wiring so the ignition wouldn’t function during cranking and then held my breath about what I would find. RRRRRrrrrrrrrRRRRRRRrrrrrrrrrrrRRRRRRRrrrrrrrrrrRRRRRRRrrrrrrr went the starter motor as it spun over the 1.5 Liter four’s crankshaft, rapidly cycling the pistons up and down on compression, power, exhaust and intake strokes. I couldn’t see the compression gauge from inside the car, while I turned the key to start/crank position. When I stopped and inspected the gauge, I was relieved to find 180psi showing on the dial! That is a pretty normal finding for a good running engine.
The other “clue” that the engine had good even compression is to listen to it spin over with the ignition disabled. A “dead” cylinder will have less compression resistance when the piston starts up that stroke, so the load on the crankshaft and starter motor is temporarily relieved. You will hear the engine speed up for a moment and as you listen to the cadence, you can be assured that one cylinder is low, when the speed interruption occurs. As I cranked the engine over with the key switch, I listened intently to the “sound” of the starter motor spinning and heard nothing but a steady strain on the motor as it turned. The compression gauge confirmed what I was hearing; four good cylinders with even compression.
Okay, so if you don’t have a burned valve or holed piston, why is the engine “missing” on the number two cylinder? The spark plug wires were all fairly new, as was the aftermarket brand distributor cap. On vehicles with a distributor cap (Honda actually used them on some of the 1958-59 motorcycles!) you have to look for the possibility of “carbon tracking” inside the cap. This usually only happens with very old and high-miles ignition components (or ones with a factory defect or poor build quality). I did remove the cap, just to check and found some carbon “dust” fluffed around inside the areas between the high tension terminals, but nothing that stood out as a “carbon track.” Carbon tracks are the result of arcing inside the cap when oil or dirt builds up inside, leaving a thin layer of material which can conduct electricity in certain cases. These are often apparent when there is high humidity in the air or the engine has been running during a rainstorm and the cooling fans have thrown moisture over the top of the engine’s ignition components. Arcing and carbon tracks can also occur down the sides of ignition coils and from unsealed high-tension terminals of the cap or even spark plug caps.
I sprayed some cleaner inside the cap and wiped it down with dry cloths to be able to more closely check the cap for evidence of arcing, perhaps from one terminal to the adjacent one in the firing order. Nothing obvious was seen, so I closed it back up for the night and let my brain go on troubleshooting function for awhile. When you have all the elements required for an internal combustion; air, fuel, compression and spark (all at the right time and amounts, of course), then there is no “reason” for a misfiring cylinder in your logical mind. There is just one carburetor feeding all four intake runners on this model, so if the issue was fuel metering, it would affect all four cylinders.
After a night’s sleep and some good daylight available for visual inspection, I looked at the intake manifold runners and saw that a large hose fitting was inserted into the #2 cylinder runner, which lead to the brake booster! A-HA! Pinching off the large hose with some duckbill pliers made an immediate change in the idle speed and quality of the engine sounds. In all my years of buying, fixing and selling cars (and a lot of motorcycles), I had never replaced a vacuum brake booster, but it appeared that I was about to do just that (on the day that an interested buyer was coming later in the evening!).
I had replaced the brake master cylinder the day before, due to internal seal leaks and noticed that the master cylinder had a thick rubber seal which fit into the brake booster opening, to seal the two parts together. It felt a little “loose” when I was tightening down the retaining nuts, but I didn’t give it that much thought, although the engine roughness seemed to increase just a bit afterwards.
I removed the master cylinder again and fished out the brake booster from a very limited space between the vacuum line bank and intake manifold casting and hustled it down to the only shop in San Diego who rebuild brake boosters and promised to do it in about 2 hours! I spent time with some local friends in the area and then picked up the newly painted and rebuilt unit for $125, then raced back home to install the booster, master cylinder and then bleed the brake system again. I had it all buttoned up, started the engine and was dismayed to hear the same misfiring sound from the engine after all that work and expense. I’m sure that the booster probably was failing because it was the original part and was 25 years old, subjected to 135,000 miles of driving and brake applications. Knowing that the only remaining place for a vacuum leak to be was the fit between the master cylinder and booster, I removed the master cylinder once again from the booster, but left the brake lines connected. I was able to pull the master cylinder out sufficiently to remove the supplied cup-style seal and replace it with a nice fat metric o-ring from my Harbor Freight metric o-ring kit. When the master cylinder went back into the booster opening, there was resistance to it entering the hole for the last 1/8” of movement. It felt like it was a tight fit this time and when the job was completed the engine went back to a nice even purr again. I had to readjust the carb idle diaphragm screws and finally the engine settled down to an even cadence, idling down to the suggested 750-800 rpms when warmed up. It was a long 24 hours of head scratching, diagnosing the cause, chasing down possible trouble areas, rounding up the parts and getting it all installed before my buyer showed up.
“All’s well that ends well” wrote William Shakespeare and so it was. The troubleshooting and repairs were completed just in time for the car to be happily spirited away by a new owner. Again, great lessons learned in the great mysteries of the internal combustion engine and all of its associated parts.
Bill “MrHonda” Silver
Comments