Georgie

Hi,

My 94 Honda Accord is having a problem with starting and immediately dieing. The problem is intermittent and has been getting more frequent over the past few months. Seems to get worse in hot weather.

Here's what happens:
I turn the key to postion III, it starts, runs for maybe 0.5 seconds and immediately dies. CEL light stays on when this happens. It does not stall in traffic or when the engine has been successfully started. Only occurs when starting and makes no difference whether the engine it is hot or cold.

Car details:
94 Honda accord ex, sedan, 4 cyclinder, vtec, 140 some miles, ignition coil within distributor.


History:
A year back when this first happened, I picked the low hanging fruit and installed a new main relay. That seemed to fix it, but within days the problem returned.

A few months after that I replaced the ignition switch. That did not fix the problem either.

In the past weeks, having learnt a little more about error codes, I pulled the error code 15 by shorting the 2 pin connector on the passenger side and that gave a problem description of "Ignition output signal missing or defective output signal".

With the code, I checked the voltage across the Ignition Control Module (ICM / igniter) using the shop manual/Haines and it did have battery voltage.. Also checked the resistance across the coil terminals and that was ok too. Then checked the spark at the plugs using a tester and that appeared to be ok. I can't quite remember the quality of the spark but I think it was yellowish. Will have to repeat that.

Having checked the ignition as best I could and thinking it was not an igniton problem, I checked out the fuel pump and convinced myself it was that. So I replaced the fuel pump 06/16 and guess what, it isn't the fuel pump. The car is still not starting. Nasty job that fuel pump replacement.


What to do now?
I am thinking it is back to the ignition system and having another look at possible causes for that error code 15.
My question is, is it likely to be the igniter (ICM unit) or ignition coil that is causing this or are there other things to check out next. I have not yet checked the voltages at the fuel injectors and i do need to re-check the spark quality at the plugs.

Appreciate any help and thanks for going through that long history

G

holmesnmanny

"My question is, is it likely to be the igniter (ICM unit)"

Yes, usually it's the igniter. You can try pulling one from the junkyard instead of a new one if you don't want to spend the money.

Georgie

Thank you Holmesnmanny. I'll go with the igniter. That's a good idea about the junkyard since that particular ICM is kind of pricey.

Georgie

Here is an update on progress or lack of ....arrrgh.

I replaced the ICM (igniter) and still the car won't start. Still doing the same thing - starts for a short time, 0.5 seconds and dies. The error CEL code is still showing error code 15 - ignition output signal.

I also pulled the fuel injectors out, cleaned them and checked the resistance across the terminals and that was within spec at 2.1 ohms.

Re-checked the spark at the plugs using a test spark plug and I am getting a blue spark on all four cylinders. Since I have spark, I am not sure why I am getting that error code 15. Could the computer ECM usnit I think it's called, be the problem here?

As for fuel, I have a new fuel pump in there and the last check I did on fuel delivery seemed to be ok. I tested it by undoing the banjo bolt on the end of the fuel rail and turning the ignition switch to position II. That caused a bit of a spray of fuel from the loosened banjo bolt.

Onwards I guess and appreciate any and all help.

street_accord94

The computer could be causing the issue. Assuming your test across the 2 pin connector at the ICM was correct, the next step is to test at the ECM to determine whether it's an issue with the ECM or wiring.

Connect your voltmeter between pins A21 and A26 with the ECM and harness connected and the ignition switch on (you can find pin out diagrams on the web to help). If there's battery voltage across the pins, you should swap a known good ECM - if the problem goes away then the ECM was the issue. If there's not battery voltage, and you've already replaced the ICM with a known good part, then you need to search out an open in the A21 wire (yel/grn) between the ECM and ICM.

Hope this helps.

Georgie

Thank you for the reply street_accord94. Yes it does help.

The ICM I replaced earlier this week and in fact the previous one passed all it's voltage tests. So should be good on ICM.

I looked at the ECM unit and I'm not certain I can locate those specific wires into the unit, despite having the diagram for it. It's a bit of a tangle of wires. As I was already considering getting a replacement, I've ordered a used ECM off ebay.

Will update when the new ECM is fitted.

thanks

stolebrothers

Sounds like a bad ignition switch to me. The ignition switch is the electrical part on the backside of the key switch and it has wires that run down to the fuse box.

Georgie

Thank you solbrothers. It's a good point you made about the ignition switch, but I replaced that some weeks back. That was a complete unit, the barrel and all the electrical connections. The ignition switch was kind of warn anyways, so needed fixing.

My wife thinks I'm going for a new car...one new part at a time I can see the humor but it's getting frustrating now. Just hoping it's not a short or bad wire connection between the ECM and distributor, cos I see that as a car killer.

thanks

Georgie

It looks like the Haines manual lists some stuff I have not checked so far. Would be grateful for any thoughts on these items.

"engine starts but stops immediately"
1. loose or faulty electrical connections at distributor, coil or alternator
2. insufficient fuel reaching the fuel injectors
3. vacuum leak at the gasket between the intake manifold and throttle body

ITEM1
I think I've pretty much covered item 1 apart from the alternator. Have been all over the distributor, tested, replaced the ICM, tested the ICM and coil and generally messed with it several times. But still, the error code 15 is there.
Have not looked at the alternator at all. So must do that**

ITEM2
I've checked the ohms at the injector electrical connections. Replaced the fuel pump last week and the fuel filter is within 2 years old. Don't have a pressure test gauge but when I undo the banjo bolt at fuel rail and turn the ignition to step II, fuel does spray out.

ITEM3
Have not looked at this at all, so another action item**

holmesnmanny

You should check your grounds first. I believe one is at the back of the throttle body and the other should be at the transmission. Don't just look at it. Take the ground off and clean it really good then put them back on.

However, this doesn't normally turn out to be the fix. This would be a good time to have a friend who has the same car so you can swap his ecu into yours. If not, and you've tested the distributor connector properly... http://techauto.awardspace.com/ ...you should pull another ecu as has been suggested already

Georgie

Thanks holmesnmanny. I checked and cleaned the main engine compartment grounds, particularly the throttle body one, G101. Using the Shop Manual checked these ones: G2, G1, G101, G351. The shop manual didn't list a transmission ground though.

Friday I ordered a used ECM computer/ECU so will see next week whether that works.

Also, as per the items to check from the Haines manual (listed in an earlier post ITEM1), I checked the connections to the alternator and they look good.

Georgie

Hi street_accord94. I did manage to figure out how to test the pins on the back of the ECM. I was able to flip up the sides of the plastic connectors taht plug into the back of the unit and see the pins. With ignition on II, I got a reading of 8.74V across the yellow green A21 and brown A26 pins. That's with all the connectors connected to the back of the ECM unit.

The replacement (used off ebay, reputable seller) is arriving Monday 06/24, so will swap out the ECM then.

thanks

Georgie

Some positive news

I did a continuity test using the fluke multimeter, from the yellow/green stripe pin wire at the back of the ECM computer unit (pin A21) and the yellow/green stripe wire that connects to the igniter (ICM unit) within the distributor. It passed, I got an audible signal.

I'm not sure if that test rules out a short in the wire? But at least it is passing current.

holmesnmanny

I'm not sure why you'd get only 8.74v on any test. Reference volts would be around 5 and battery voltage should be 11+ at the least. Do a reference volts test on the map sensor connector and tps sensor connectors with the ignition on and the car off. You should get around 5v on one of the three pins at each connector.

lfdelar2

have you tried cleaning your intake or iac valve or finding out if theres something thats not letting the engine get enough air? might be a stupid suggestion but you never know

Georgie

Hi all,

I took delivery of the ECM computer unit this afternoon (used item from Ebay, $70), fitted it and my beloved Honda fired straight up No hesitation at all. It's starting just fine, for the first time in two weeks!

Thank you so much to all those who wrote back and helped. I am very grateful for your great support and prompting me as to what to check and rule out, particularly Holmesnmanny. But also thanks to street_accord94, solbrothers and lfdelar2.

Thank you all.
G

street_accord94

Glad to hear the ECM swap resolved the issue.

If you plan to keep the car for a while, you may find a Helms/Honda service manual to be a great resource (much better than a Haynes manual).

Georgie

Thanks for tip the street_accord94. I'll look into getting a pdf copy of the Helms manual. And thank you again for responding and the great help.