I have a 1997 Honda Accord EX that is running me nuts. I was driving home one day and it simply died and would not restart. I rebuilt the disto with new ICM, coil, rotor button, cap, internal seal and external "o" ring. The car ran like new for a few months. The car started dying for no apparent reason. No code either. I replaced the ignition switch as it seemed most logical. I started receiving fuel system too lean. I read the Oxygen sensor might be the fault. The downstream sensor failed about three years ago on the heater side. I replaced the upstream Oxygen sensor. The car ran like brand new. Amazing acceleration and no more cel. I thought all was well. I drove the car and it did well until I noticed it was hard to start two or three times. I drove the car around the block and it died multiple times. I was able to put the car in neutral and it would restart. One time I noticed the tach was a zero rpm and I attempted to restart the engine. To my surprise the engine was still spinning down. It restarted immediately after all the idiot lights finally came on. The starter is less than one year old, napa rebuild, the starter solenoid finally failed and it was easier to replace the entire starter. I pulled the main relay and I did not see any cracked solder or burnt or hot spots on the circuit board. I also noticed the temp gauge showing different levels within a short amount of time. Any chance it might be the neutral switch on the tranny? I have a jasper rebuilt tranny that is about one year old. I suppose it might be either the positive or negative battery cables breaking down. The ECU components might be an issue. I think this was built before the bad electrolytic caps that plagued PC motherboards, however, I am not certain.

Any suggestions will be helpful. I have searched many threads on various sites and found similar complaints. Most have not followed up with the final outcome and/or repair.

Thank you again for your suggestions.

 

fuel, air, spark. and last... the ecu.

What kind of dying? Violent shaking? Quick stall? Do the lights stay on when it dies or does everything shut off?

 

Normal shutdown as if one had turned the key off. No violent shaking. The dash lights remained "normal" ie all idiot lights off. The only odd thing I noticed one time was the tachometer going to zero. The idiot lights finally came on. I pulled the gear selector to neutral, cranked the engine and it restarted. It was hard to remember all that was happening. It was dark, the headlights were on the bright setting and were putting out a normal amount of light for the bright setting. I was fighting the steering wheel around a curve without the benefit of power steering or power brakes.

I forgot to mention that I replaced the spark plugs, with genuine Honda plugs from the dealer (NGK), at the same time I rebuilt the distributor. The plug wires are original 177K. I have looked for spark coming from the wires at night and have not noticed any sparks. The air filter is not new, however, it has less than five thousand miles. I will try replacing the air filter as it is easily changed and not very expensive. I doubt it is the cause, however, it is quick rule out. I cleaned the intake/butterfly with cleaner. This removed a good deal of gunk and grime. I also removed, cleaned, reassembled and re-installed the fast idle valve assembly. I replaced the three "o" rings on assembly. I have the figure 8 shaped seal for electronic air control valve. I planned to remove, clean and reattach the engine air control valve, however, there is very little room in which to work. I have not removed, cleaned the unit or replaced the "o" rings. I do not want to remove the entire intake manifold unless absolutely necessary.

I thought about running a jumper wire from the positive lead on the fuel pump plug to a light or LED and grounding the negative lead (lamp or LED). This would let me know if the fuel pump is receiving voltage while running and if 12V+ is dropping out to the fuel pump when the car dies for no apparent reason.

I checked the crank position sensor with a LED, when I rebuilt the distributor, and the feedback from the ECU caused the LED to blink. The crank position sensor sends a signal to the ECU and the ECU returns a signal to the ignitor. I think .... hope .......the distributor is good. I added CPU/cooler thermal assembly paste to the ICM/heat sink assembly when I replaced the ICM/ignitor in the distributor. The ICM is under warranty and I suppose it might be breaking down when the engine reaches operating temperature. I doubt this is the case.

Thank you again for your thoughts, assistance and time.

Sincerely,

Brían S. Du Bois

 

Ok. Sounds like either transmission or perhaps ignition.

Does the car shut off when you are moving or when you come to a stop?

 

if when the car dies the dash lights turn off with it its likely the ignition switch ... other likely culprit is the main relay ... ive had a situation before when a really bad oxygen sensor would shut my car off but its not common

 

I suppose I could have received a defective unit from the dealer. I replaced the unit and I have not started the car more than two dozen times since the replacement. I found a video on youtube that included a part number for the recall kit. The kit included the electrical half of the ignition switch assembly and two special twist head bolts.

The headlights and the instrument lights remained illuminated. The engine eventually spun down even though the tach was at zero rpm. I pulled the main relay and looked for cold solder joints, cracked solder, oxidized areas, or hot spots on the circuit board and found nothing. I plan on changing the main relay as it is relatively simple and not too expensive. I will pull the relay apart and leave the old plastic case attached to the car. It is too hard to get to the bolt/screw to remove it as a unit.... unless you have a trick to easily remove the unit.

Would a VTEC solenoid that is temporary shorting to ground cause the engine to stall and the tach to go to zero? I have engine oil leaking out the VTEC solenoid and coming through the insulation on the wire and into the electrical connector (one pin). I replaced the oddly shaped seal between the solenoid and the mounting base. This stopped a small oil leak between the adapter and the solenoid, however, it did nothing for the leak at the connector.

I started the car this afternoon and it started fine. I allowed it to warm up and I moved it up in the driveway. No stalling problem, however, I did not go around the block. I was concerned it might stall and this time I would not be able to get it restarted.

This is why I am wits end. I replaced most electrical items that could be causing a problem. This is the most frustrating repair I have ever attempted. My next step is to find a local auto electric and see if they can run down the cause.

Thank you again for your help.

Sincerely,

 

I would think that a short to ground would likely cause a fuse to open, but anything is possible.

When you say you had problems starting the car, what was happening when you turned the key to the start position? Did the starter crank the engine? Or did the engine not fire up (that is when it is running and you release the key)? People use different terms, so I want to be on the same page.

I like your idea about hooking up a light to the fuel pump. I did that on my accord and it was pretty simple to do. I just back-probed the power wire for the fuel pump near the fuel sending unit and put a 12V bulb to ground. The wire color was yellow (I think).

You could do the same thing to the blk/yel wire going to the distributor. That would indicate a loss of power from the ignition switch.

I would recheck your codes and post any codes you find on here.

 

First and before I forget it...... currently no check engine light and absolutely no codes. I checked for pending failures in the event an out of range event occurred only once and did not set the CEL. No pending failures or incidents are stored. This is what makes it so difficult to run down.

I had been receiving two error codes. One was a Honda code and one was not. I don't remember the code numbers. They indicated mixture too lean. I cleared the codes and all was well for a few days. I noticed my gas mileage was about half what I had been getting for months. I researched several Youtube videos and this source and found a reference to defective oxygen sensor. It mentioned the bad sensor might not be throwing a code. I decided to replace the sensor. I replaced the sensor and took it for a drive. It was like the old car was new again. The acceleration was amazing and the engine ran even smoother than it had been running. I drove my family to a nearby town about 25 miles away. I dropped them off and proceeded to the local superstore to kill time. The car restarted fine. I parked the car and shut it off while I waited for their meeting to end. The car did not start immediately. It cranked fine. Starter speed sounded good. I turned the key off and turned the key back to the start position. The car started and ran fine. We returned home without incident. Later that evening my Son and I decided to go around the block. We live in a rural area and that is about two miles. The car died without warning three or four times. It cranked well each time. Twice the car restarted immediately. Twice the car did not run on the first attempt and one time it required three tries to get it running. I suppose it could be flooding and my Son noticed the smell of gasoline. I don't understand why it did not set a CEL if the mixture was too lean (which increase the amount of fuel provided by the injectors). I remember reading the ECU is receiving a false signal of too lean a mixture and it increased the amount of fuel to compensate. This floods the engine. That is all well and good, however, it does not explain the tach going to zero. One would think the RPMs displayed would drop as the engine speed decreases and finally stops (at which point the idiot lights illuminate).

The starter cranked the engine each time without fail. No issue here. When the vehicle is cold it cranks and starts/fires/runs without issue. It is the most bizarre feeling. The engine is running, trans shifting and one is driving down the road without a care in the world. Suddenly the car begins to slow down slightly without warning. Then one notices the steering is extremely difficult to turn and the brakes require more effort as the vacuum assist stops working due to lack of vacuum. This happened at night and the instrument lights remained illuminated and they remained at the same brightness level. I only noticed this once because I was going slowly enough to look down at the instruments. The tach was at zero. I pulled the gear selector into neutral and moved the ignition switch back to position II. I did not want to go to position I and lock the steering column. I moved to past position III to start. The engine cranked and engine started briefly and did not continue to run. I repeated the process two more times and it started and continued to run. I tromped the accelerator while in neutral and the engine rpm climbed effortlessly to the red line. I dropped the trans into D4 and we headed home. We continued home, about two tenths of a mile, without incident. The car died two or three times prior to the event I just described. We were glad to be home and decided to remain home and sleep on the symptoms and possible cause(s). I decided to reach out of help and see if anyone else had experienced similar problems and found a solution.

I noticed the lower end of the ground cable, that attaches to the transmission, has definite corrosion/oxidation. This is a stranded wire and the exposed end is starting to unwrap/untwist. I am seriously considering the replacement of both battery (positive and negative ) cables. The negative looks worse. I plan to check all connection points on both cables, clean any corrosion, apply dielectric grease, re-install the cables and torque the bolts. I don't remember if I mentioned this earlier, however, I replaced the battery last summer with a genuine Honda Battery from the dealer. I try to use genuine Honda Parts or OEM parts (ND Oxygen sensor at Napa was $56.00 and three hundred ten dollars at the dealer). The parts were not 100% physically the same, however, the treads and connector were exactly the same. The wire was about two inches longer. I have used ND OEM replacement Oxygen sensors in the past without an issue.

I an going back through the troubleshooting guide in the Helm shop manual and see if I overlooked something. I will post any and all additional attempts to resolve the issue and I will also post the final solution.

I am not certain I really want an new vehicle. I know there are many safety improvements, however, I don't like the throttle by wire/stepper motor arrangement. I like the older cable system in my current Accord. Sorry I digress too much. Sorry for getting off the topic. I suppose it is a sign of my age!

I apologize for the length of this message. I hope the additional information might provide a clue to what is causing the strange stalling of the vehicle. The tach acts as if the ignition switch has been turned off, however, the engine is still rotating until it finally spins down. The idiot lights do not come on with the tach going to zero. They only come on after the engine finally spins down to a stop. I remember depressing the accelerator when I first noticed a small decrease in speed. The engine did not respond to the additional throttle.

I think I remember the main relay contains two relays in one enclosure. One side activates the starter motor solenoid via the ignition key. The "other half" of the main relay supplies power to the ECU. If the ECU's power is cut off... then the ICM/ignitor would not receive power and the tach would go to zero. I need to review the wiring diagram in the Helm manual. This is only an uneducated guess on my part.

I have no problem with replacing parts as the car has some age. I don't want to simply throw parts at the car. The parts I have replaced to this point were defective. I believe the Oxygen sensor was failing, or would have failed, based on the mileage and previously stored P codes.

I sincerely appreciate everyone's suggestions, help, aid and assistance.

Sincerely,

 

Is this a 4 or 6 cylinder accord?

The main relay sends power to the fuel pump, ECU, and the injectors. The main fuel relay has nothing to do with the starter, that is a different circuit.

You have a lot of information on here, so I am going to ask some basic questions. How many miles are on the maintenance items, like distributor cap, distributor rotor, spark plug wires, and spark plugs? If you have the original Honda wires, they seem to last forever, but you should still inspect them.

When I first read your description, I was leaning towards ignition switch, and you replaced that. So now it sounds like either the ICM or coil in the distributor, or the ECU (PCM), because you had a code come up that does not exist. Replacing the PCM is usually the last step in the diagnosis. You may want to pull the PCM and open the cover to see if there are any obvious problems.

I would find a used PCM from a junkyard. I like u-pull-it yards, because these accords are pretty common. I would pick up an ICM and coil from an EX accord as well.

The only way to diagnose the problem is to test the car when it will not start up. Since this is an intermittent problem, this would be tough to figure out IMO.

 

I apologize I forgot to include it is a four cylinder VTEC engine. I replaced ICM, Coil, Cap, rotor button, internal seal and "o" ring about four or five months ago and they have less than four thousand miles on them. I found numerous Honda Accords at our local pull a part, however, finding one for a VTEC is more difficult. I need to make another trip to see if they might have one in by now. I am seriously leaning toward the main relay as it is not too expensive and relatively easy to replace.

Thank you for your time.