Okay, here’s the symptoms.
97’ manual Honda Accord.
On occassion, maybe 3 times a week, it has a tendency to “kill” around 3000 rpms when the clutch is pushed in. In any gear.
I will be in gear doing say 40 mph at 3000 rpms, push in the clutch, and the car just dies and the tach needle falls to zero.
Strangely it sometimes pops back on a second later.
One time I was rolling at about 30 mph with the clutch in and the car would turn off-on-off-on which made the tach needle jump wildly from zero to 3000.
I had the car in recently for this problem and because it wouldn’t start. They replaced the spark plugs and distributor cap which fixed the starting problem but a week later the “kill” happened again. They of course could not find a problem.
It sounds like an electrical problem to me. Short maybe?
If I go in to another dealer what should they look for since “duplicating” the problem is difficult? A complete electrical system check??
I’m thinking a ground problem…a short in part of the switch that prevents the car from being started unless the clutch is depressed.
Possibly, it is shorting out when you activate the clutch. It think there is a simple way to defeat it that was discussed on the boards in the past 45 days.
If it only happens with the clutch pedal depressed, the logical suspects are a clutch pedal switch (and its circuit–see below), as Philster mentioned, or an electrical wire being hit by the clutch pedal.
The clutch interlock switch connects electrically with both the fuel injection main relay and the powertrain control module (PCM, “computer”). Those items and the wiring to them are on the suspect list.
I’ll add to this, only because this is failry easy to do, and you might as well eliminate variables that are easy pickins:
Check all wiring harnesses under the hood. At 3000 rpms, when dropping the clutch, you might be in that zone where the engine lurches the most on it’s mounts. If a wiring harness or wire has loosened, you just might be finding that RPM range where the engine moves enough to loosen the wire, and the restart might be related to the fact that since the engine has shifted back, the connection is re-established…only until the engine reaches 3000 rpms again, lurchesand the connection is lost…etc.
I mean, what the heck, it’s worth a shot, and it was mentioned on a board I visited recently.
Oh wow, I really like Philster’s take on it. I can even picture it. I could maybe even see a broken motor mount causing electrical contact to be lost. The engine torques enough at 3k rpm to lift off the motor mount, the engine quits and the torque drops off, and the engine drops back down, re-engaging the contact again until it lifts off the motor mount…lather, rinse, repeat.
I love a good mystery. Let us know if you see anything.
I won’t say it’s impossible, but the evidence we have doesn’t support the torque theory. It happens when the clutch is depressed, which lets the engine relax into position just like taking your foot off the gas would–but it doesn’t happen when taking the foot off the gas. And this: One time I was rolling at about 30 mph with the clutch in and the car would turn off-on-off-on which made the tach needle jump wildly from zero to 3000. tells us it occurs when there is no change in engine load.
Hampshire, the best thing when taking it to a shop is to describe your symptoms as clearly, completely, and succinctly as you can. Let them decide how to test for it. This one is potentially very tricky, especially since it’s intermittent. I’d say it’s going to take a better-than-average shop to have a good chance of resolving it, and even they may get stymied if they can’t duplicate the symptom.
Darnit Gary, you’re raining on my parade. And you’re right. One entertaining theory shot all to hell. 
:eek:
You will need a young priest, and an old priest.
Are you completely certain that the engine is ACTUALLY STOPPING? Tach problems (where the needle jumps erratically) are so common, it’s not funny. But in all cases that I’ve seen, the engine speed doesn’t differ, it’s just that the tach itself is malfunctioning.
Not to imply that you’re so dumb you can’t tell when your car is running, but I’m sure your Honda Accord runs very quietly.
Make sure that the engine has actually died, and it’s not just the tach doing crazy things.