I asked my wife if she had to turn the switch back to the "accessory" position before cranking and starting the engine again, when it happened to our car on the 7th. She could not remember, and since it's never happened to me personally, I cannot say either.

From what pointer said, I understand thathis stepson'scar is rolling along running fine and the ignition turns off (engine stalls) exactly the same way that it would if you grabbed the ignition key and turned it ccw from the "on" position back to the "accessory" position. So in effect the ignition turns off, butwithout the key being turned, and remains off even though the ignition switch remainsin the "on" position. Definitelyfaulty behavior.

His car rolls to a stop and I assumethatthereare none of the usual dash lights that indicate that the ignition switch is in the "on" position (ie that the ignition system is on), nor will thestarter crankif the switch is atfirst turned cw to the "start" position (that is, directly from the "on" position it was in when the engine suddenly stalled).

It is only when the ignition switch is firstturned back to (at least)the "accessory" position and then turned cwback to the "on" position that the ignition system energizes. From that position the ignition switch can then continue to be turned to the "start" position and the starter will then engage and the engine will run even after the switch is returned to "on". I guess things run OK for a bit and then the ignition mysteriously turns itself off again.

Because the switch must be turned back to the"accessory" posittion after ignition failure, it sure sounds to me that it'sthe ignition switch, too, but I would think it's the mechanical parts of the switch (the moving contact surfaces) that are faulty.

If it is indeedthe ignition switch thatiscausing the problem, I doubt if there would beanyOBD code, but maybe I'm wrong on that.If there were one, I doubt if it would be a15. In the case that faulty switch contacts are causing a code, I would think that every time you turn the ignition off in a perfectly-functioning car,you'd generate the samecode. (For all I know, maybe it does!)

For obvious reasons, I'm very interested to hear what pointer's car's problem was.

 

Well I didn't get the OBD code today, my mech. wasn't at his shop when I went by there. I'll check tomorrow. So keep checking back. I'm very desperate to fix this problem. The first time I took the car up to the mech. he drove it a solid week and it never messed up one time. I brought it back home and within two days it started, crank and start to back up and it would die, one day my stepson made it a couple of miles and it died while driving, he ground on it for appr. 10 min's and it wouldn't crank. I went back two hours later and it cranked right up and I made it back home. My mechanic has one of the little portable devices to check codes with I'm pretty sure. He's not a shade tree mechanic by no means, he was the senior mechanic at a very large Chev. dealership until he decided to go out on his on. All I can really tell ya is that he said that about turning the switch back off and that he was losing fire to the injectors when the car would die. It don't always just crank back up even by turning the switch off either, he told me one day last week that it would crank for two days and then it ust cranked up. Yesterday he did replace some small wiring harness in the switch but he didn't replace the ignition switch itself. He thought it might be security problem but after some reading he said the 95 accord didn't come out with a security system until 98. And if it was a security problem it wouldn't crank at all. I'll post more when I learn more. I'm sure he has probably checked the main relay but who knows whats happening.

 

I am currently experiencing the sam problem pointers son is. The code I have is 15. My ICM went 2 years ago but there were no symtoms like the ones I'm having now. I went to a freinds shop and when I was ready to leave the car would crank but not start. I replaced the ICM and that solved the problem. This is a new dizzy I bought last year when I changed my engine.

 

I fogot to add that the car idles pretty low and I'm unable to turn it up anymore using the adjusting screw. The adjustment screw is set to max and the car is barely at 5000rpm. The car will start and drive fine, then just cut out. The only thing I can do to restart it is to turn the key totally off and then start it again. In most cases, I have leave it off for a couple of mins before starting it again otherwise it starts and just dies immediately. I've tried playing with the key, I've bypassed the key and hot wired the car so the key is not an issue and still have the same problem.I am going to stop by the shop I bought the distributor from and see if I can get another ICM to try tomorrow. I will let you guys know the results later.

 

Ok guys I've got a little more info., the mech. said that there wasn't a OBD code, so nothing there. He has replaced the ignition switch, no fix. Replaced main relay, no fix. He told me today that there is a small box in against the firewall ( the box is appr. 3" square with fins for cooling I guess) on the drivers side that 12v goes to from the switch and from there to the computer or whatever the part is called under the carpet inside the car. When the car dies he says he still has 12v between the box and the computer with the switch on but is losing the 12v from the computer to the injectors. One might think that the computer is the problem but he don't thinks thats it. He said there are four circuits running from the computer to the injectors and when the car dies all four crcuits show no voltage. He said most of the time when the computer goes bad it will be just one circuit that loses voltage and causes problems and when that happens it don't just fix itself. The car cranked up today and was running fine but we know the problem is still there. Buying a computer not really knowing that is the probllem is a expensive test. Does anybody have any suggestions on something that might be causing the problem that is between the computer and the injectors?

 

Thanks for allthe additional information about parts replaced and the fact that there is no code. I'm actuallyhappy thatthe cardoes not have a code 15, while our car's problem generates it. That means that the problem your stepson'scaris havingis likely different than the one in our car and, after reading what you've gone through so far, I don't want to trade places with you.

Hopefullythe HAF heavyweights will have some ideas, or someone else who has experienced the identical problem will chime in with a possible solution.

If it turns out that your tech is completely stumped by the problem, perhaps you should consider taking it (and certainly a list of the parts that have been replaced)to a reputablelocal Honda dealer. Your stepson's car is not unique, soHonda techs will most likely have seen the problem before and have a fix for it, or, if they don't, they can certainly get it sorted out with Honda engineers.

Please continue tokeep us informed of what happens with the car. I'd really like to know what's wrong with it, and it will no doubt be valuable information for anyone having the same problem in the future.

Good luck to you, pointer.

 

As you described, the situation is normal. When an injector is actuated (grounded by ECU), you would not see voltage on the wire from injector to ECU, because it's at ground potential. When injector is not being actuated, you would see 12V on the wire from injector to ECU, since circuit is opened by ECU! Since injectors are firing several times per second, the voltage is constantly changing in the wire from injector to ECU, but mostly it's 12V (open) compared to 0V (grounded). An oscilloscope is really needed to see the on/off pattern.

The box w/ cooling fins is a simple set of four resistors that reduces drive voltage to each fuel injectors. They should be hot all the time engine is running or keyswitch is ON. The fuel injectors are fired by the ECU providing ground to the injector for a few millisecs at the proper time for fuel to enter the engine. If power to the injector resistor box is lost, then it's a wiring fault or Main Fuel Relay (MFR) fault. I would suggest rigging a tell-tale light to the 12V lead to the resistor box and watching that when no start occurs.

If you have an old or spare MFR, you can hard wire the relay easily by pulling the cover and jumpering the relays. I did this when I thought I had a bad MFR. Turned out to be the fuel pump. W/ jumpers the fuel pump and injectors are hot as long as the keyswitch is On. The MFR is probably a safety device to reduce probability of a fire when car is stopped violently w/ keyswitch left On as in a major accident. So don't leave it hard wired if you do this. But for diagnostics, it's not a problem.

Frankly, I don't think the resistor box is the problem. Your mech is missing something fundamental. I think you may need some better shop help. This problem should not stump a good shop.

First step in resolving no-start is to determine whether loss of fuel supply or ignition is the cause of no-start. I don't see that this fundamental step has been resolved.

good luck

 

Well my car seems to be fixed for the moment. All I can do is just wait and see if the problem returns.

I went back to the guy I bought the distributer from and he told me that I should change the rotor brush and cap and this usually resolves the intermittent shut down issues for his other customrs. Well I went out and did that, but the problem got even worse. Now the car wouldnt start at all. I had to remove all the new parts recheck voltage going to the ICM which was good, then reinstall the old parts and the car started . I went back to the shop and the guy gave me a new distributer. I installed it right there and so far things seem to be working great.

------------------------------------------
1995 Accord EX

 

Great! Please let us know if any symptoms reappear in the future.

I'm curious, though. Did youhappen to reset your OBD codes, and, after that, did you seecode 15 again?

 

I hope he's aware that the ECU does NOT supply voltage to the injectors. The injectors will have power all the time at one of their wires. The ECU fires them by switching the ground side of their circuits (the other, non-powered wire).