[dude911]

Hello,

I am working on a 91 Civic DX hatch, 1.5 in our shop, a GM dealer. The tech working on it is somewhat familiar with the car but we cannot figure out what is going on.

First of all it has what seems to be the common code 16. Initially it would run bad on hard accel, but progressed to the point of stalling and then a very long drawn out starting procedure had to be done. The customer mentioned that sometimes cycling the ignition would get it to restart. We replaced the cap, rotor, wires and plugs, as well as the fuel filter before beginning the task of the 16 diag chart, all to no avail. Actually it seemed to run great, but if you stabbed the throttle to quickly it would bog down, backfire and die. It has good fuel pressure, but now it idles rough as well.

Anyway we went through the diag chart checking various things along the way until we were led to the replacement of the ECU, which I don't have a problem with if it fixes the car, obviously. My concern is that we aren't sure it is going to fix it and I have found prices range from low 200's to 1000 dollars, and all have to be ordered, a little too much to guess with. Can anyone give me any advice, are there things we are missing, or is it to the point of ECU replacement and crossing our fingers? Any help appreciated.

[dude911]

Have you tried this, yet?

For DTC 16, start engine, allow it to idle, and listen for the clicking sound from each fuel injector, using a mechanic's stethoscope. Turn the ignition off, disconnect the 2P connector from the fuel injector that does not click, and measure resistance between the 2 terminals of the fuel injector. Is there 10-13 ohms? If NO, then replace the fuel injector(s) that are not 10-13 ohms. Turn the ignition switch ON and measure the voltage between YEL/BLK (+) terminal in the 2P connector and body ground. Is there voltage? If NO, then repair open in YEL/BLK wire between fuel injector and the main relay. If YES, turn the ignition switch OFF, reconnect the 2P connector to the fuel injector and measure for voltage between A23(-) and the pin for each injector (No. 1=A1+, No. 2= A3+, No. 3=A5+, and No. 4=A2+). Is there battery voltage? YES, substitute a known good ECM. NO, repair open in wire between ECM (A1, A3, A5, or A2) and the fuel injector.

I know he told me he checked the fuel injectors and they are working properly, he also said he removed the iac valve and it was full of buildup. Cleaned it but no change. We are at the point of EcM Replacement but I was just wanting to be sure.

-----------------------------------------------------------------------------------

Try this next:

Disconnect 3P MAP sensor connector, turn ignition ON, and measure the voltage between the connector's reference wire (probably YEL/GRN)+ and ground. What voltage did you record?

I don't know for sure if he checked this or not, but I don't see that step on the diag for code 16, is that something you have seen that can cause the issue?

[dude911]

This simple "K Test" tells you whether the ECM is delivering 5 volts to the MAP sensor, the "key" sensor on Hondas. If it is not, chances are your ECM is bad. If the MAP sensor doesn't get the necessary voltage, it can cause catastrophic failure on Hondas, affecting ignition and fuel. The MAP sensor controls timing and air/fuel delivery on Hondas. Make sure the MAP sensor is functioning properly, as well, on both the electrical side and vacuum side.

Honda ECMs are starting to go on vehicles over 8 years old. It's a bigger problem then many are willing to acknowledge. They are not "bullet proof," as some would try and lead you to believe. Odds are that if you don't replace it now, it will have to be in the not too distant future. I've already replaced the ECM on my 1993 Civic DX and it lives a "charmed life."

Good deal, that's the kind of info I was after. I am a service advisor at a GMC dealer so my knowledge is mainly directed at GM vehicles, although the basics remain the same across the board. I will have that sensor/circut checked. One last question, you say if the map sensor isn't getting the correct voltage etc. etc. The ECM supplys voltage to the sensor, so if the sensor is getting the right voltage, is the ECM good, and the sensor bad? Or could it be both, shouldn't a code set for the map sensor if in fact its bad?

[dude911]

To directly answer your question, if the sensor is getting the right voltage the ECM has only proven itself for this test. However, this is the most crucial test it must pass, since failure can be catastrophic for the ignition and fuel systems even working.

Both the ECM and the MAP sensor could be bad. Like ECMs, MAP sensors are, generally, very reliable components. Both are very expensive items to replace, so thorough testing is imperative. New MAP sensors from Honda cost approx. $325. You are right that a code 3 (electrical) and/or a code 5 (vacuum) would normally be set for a failing MAP sensor.

I appreciate the help, I will make sure that the circuit gets checked. The tech working on it is a master GM tech in all areas except body, as well as a Mazda Master Tech, so I kind of feel bad for going behind his back and asking, but it doesn't hurt. What is your opinion of buying a ECM for this car through, say, auto zone, versus the dealer. The is a HUGE price difference of course.
My feeling, since they require a core is that they were at one point Honda ECM's they they refurbish, so I guess it's a matter of the quality/skill of the tech doing the repair. If it were only a few dollars I always recommend staying with OEM, but we are talking approx 700 dollar price difference between Honda and Autozone.

[dude911]

Miscellaneous thoughts:

A. Did you check the timing? It could be retarded too much.
B. Did you install NGK plugs? I haven't seen cross-reference plugs work well in Hondas.
C. Did you clean the injectors? Try Regaine in the tank.
D. Did you adjust the idle speed to spec.?
1. Start the engine and warm it up to normal operating temperature.
2. Connect a tachometer, normally to the BLU wire coming from the distributor. It is often buried in the cluster of wires near the master cylinder reservoir. Can be difficult to single out, because Honda does a great job of hiding it. I eventually found mine.
3. Disconnect the 2P connector from the Idle Air Control (IAC) valve.
4. Start the engine with the accelerator pedal slightly depressed. Stabilize the rpm at 1000, then slowly release the pedal until the engine idles.
5. Check idling in no-load conditions: headlights, air conditioning, etc. not operating.
6. Adjust idle speed to spec. (normally 450 rpm). The idle adjustment screw is "leaded in," to prevent owners from adjusting them. You will find it on the top end of the throttle body. Take a small screwdriver and remove the lead seal and make your adjustment.
7. Reattach the 2P connector to the IAC valve and everything should be fine.
8. Disconnect negative battery terminal or pull fuse to reset ECM.

A: Yes, that was actually the first thing we did because about 400 miles ago we replaced the water pump and timing belt.
B: We put Denso plugs, the placard under the hood has three types listed with #'s, they match.
C: No, didn't clean them initially but they were taken out and inspected and all looks OK.
D: No, but its worth a shot. Another thing, the ecm to my knowledge has never been "reset" will that change anything other than erasing the code?

[PAULIE]

Could the ecm be the reason my cooling fan isn't coming on? I also have a 91 honda civic 1.5L, but mine is a dx 4 sedan giving almost the same kind of trouble as well

[dude911]

Something new that we noticed yesterday, the IAC motor runs all the time if the key is on. Does that mean anything. By the way we checked tha map sensor and its getting 5v and operating to spec.

[dude911]

I would think that is normal. Were you successful in following the procedure and reseting the idle speed? Are you going to replace the ECM?

No the car won't stay running now, it starts and idles OK, after a minute or so it sputters out and dies. The ecm will be replaced soon, I'll let you know how it turns out.

[PAULIE]

Ok,can somebody please tell me where the ecm is on a 91 honda civic dx?

[dude911]

Ok,can somebody please tell me where the ecm is on a 91 honda civic dx?

Passenger side under carpet between botom of dash and floor.


Anyway I ended up taking the Civic to the Honda dealer down the road from us, I figured it was worth the $80 to be sure it needed a ECM. Turns out it had a bad ignition igniter and coil, which is internal to the distributor, and an intermittant main relay. I let them do the work, that way if it didn't fix it I had them by the balls. Picked it up today and it runs like a top.
Thanks for all your help greasemonkey, it got us in the ballpark anyway.

[dore]

Turns out it had a bad ignition igniter and coil, which is internal to the distributor, and an intermittant main relay.

I see this is an older thread, but I'll ask anyway. I'm also dealing with a code 16 on a 91 Civic, and was puzzled by the eventual solution to Dude911's problem. Code 16 indicates "fuel injector", so I'm wondering why the problem(s) ended up being elsewhere. Are these codes not as accurate as I'd wanted to believe?

My symptom is that after driving the car for 10 or 20 minutes, shutting it off, and then trying to restart it a few minutes later, it sometimes doesn't start. If I remove and replace the fuse to the ECU, the car starts right up.

The code 16 appears to be on basically all the time, though the car seems to run well (aside from the problem described). The only other problem I've noticed is some pinging at full throttle.

So I'm wondering if I should suspect the distributor and main relay, in addition to the normal suspects for an injector problem? Would these problems normally be fixed by removing the ECU fuse? Or does that cure point to a problem elsewhere?

Thanks in advance for any thoughts! :)

dore

[TxGreaseMonkey]

dore:

. Clean the under-hood BACKUP fuse, going to the ECM, with #240 abrasive cloth. Also, remove and clean the 3-wire brass connector, going to the main ECM ground, on the thermostat housing.

. Perform the K-Test on the ECM:

The K-Test: Remove the MAP Sensor connector and turn the ignition switch to ON (not start). Using a multimeter, check for 5 volts going between the MAP Sensor connector's reference wire (+) and ground. As you look at the connector, this is the socket on the right. Really press the black test lead into the cleaned main ECM ground on the thermostat housing. If the voltage is low, it's probably indicating ECM failure. Most failed ECMs will record a fraction of a volt. To me, the K-Test is simple, elegant, and accurate.

. Check all under-dash and under-hood fuses with a test light or multimeter:

https://www.askmehelpdesk.com/cars-t...tml#post252145

. If the problem persists, start the engine, and, using a mechanic's stethoscope or large-handle screwdriver, listen for a clicking sound from each fuel injector. Turn the ignition switch off and disconnect the 2P connector from any fuel injector that does not click. Measure resistance between the 2 terminals of the fuel injector. Replace any fuel injector that doesn't measure 10-13 ohms. Should the problem still persist, replace the main relay.

[dore]

Should the problem still persist, replace the main relay.

Thanks so much for the advice, txgreasemonkey -- and Merry Christmas to you, too! :cool:

After some searching, I found the factory manual for the dual point fuel injection system. Between that and your advice, I'm feeling pretty optimistic about solving this mystery.

Here's a link to those factory manual pages, if anyone else is having trouble finding them:
Honda-Tech.com: Honda CRX / EF Civic: [FAQ] ECU Codes and trouble shooting

Thanks again!
dore

[dore]

Here's a link to those factory manual pages, if anyone else is having trouble finding them:
Honda-Tech.com: Honda CRX / EF Civic: [FAQ] ECU Codes and trouble shooting

As I'm going through the diagnostic steps, I notice that the resistance values for the injectors in the manual are a little different from the ones in your post. For the 88-91 dual point system, the manual calls for .6 to 1.6 ohms for the primary injector, and 6 to 10 ohms for the auxiliary injector. I guess it varies from year to year and model to model. I see the 88 to 91 multipoint system calls for a resistance of 1.5 to 2.5 ohms.

I follow the link and found all the helpful stuff you've posted about Honda maintenance -- what a wonderful resource! I'm forwarding the link to my mom's Honda mechanic, so he can maybe fix the idle fluctuations that have defied diagnosis for 15 years... :rolleyes: