I had some of your issues the the front O2 sensor was cause all my problems. JFI

 

I bit the bullet and put in a new ignitor/icm. I've got to drive it some more before passing judgement, but I don't think the new ignitor is going to help. On my 20 min test drive, I noted a couple misfires (but not so severe as to give me the tach needle dropping to zero). My wife drove it a bit this afternoon and also reported one severe chug ("felt like someone hit me").

I tried shaking/hitting the key and the surrounding area to see if I could get it to happen at will - no dice. I also tried stopping the engine with the key while driving and this does not produce the same symptoms. When I turn the key off - both the tach and speedo drop to zero. And when I turn the key back on, the engine starts running again (I kept the clutch in - no starter) and I get the typical 2 second cycle of the CEL before it turns off. With the misfire/loss-of-ignition episodes, only the tach drops to zero; and I don't get the 2 second cycle on the CEL.

I'm at a loss.... no idea where to go from here...

At this point I have:
replaced: plugs, distributor cap, rotor, coil, ignitor/icm
checked: ignition switch, main relay, timing

 

With the tach dropping out I'd suggest checking the ICM connections.

Is there any sign of oil inside the dist?

The main relay is still in play, I know you checked it but if it get hot (as the car is running) then the joints could loose contact.....has the ign switch been replaced under the recall? Still doesn't rule it out either

 

Since you've replaced most everything in the distributor and have no CEL (right?), the loss of tach suggests somehow the power or ground to the distributor is being lost.

Suggest checking the ground lead from the distributor to ground point on the thermostat housing. Make sure it's clean and tight.

If that's OK, then an intermittent power supply is possible. Checking this is more difficult. Suggest rigging a light from blk/yel wire (crimp splice) and local ground to the cabin where you can see the light and verify it's staying on or going off when problem arises. You may want to rig the crimp splice between the distributor and distributor connector to allow checking of the connector w/ this test. A good place for bad connection is in the distributor connector. This connector get's made/broken many times and damaged wire would not be surprising.

good luck

 

Still working this problem...

TexasHonda: I do have a CEL, but the code is P0420 (cat or second o2 sensor dead). When a bad hiccup occurs, I lose tach, ignition, and the CEL goes off and stays off, even when the engine comes back to live after the hiccup (code is cleared).

I think the ECU is losing all power - that's the only way I can explain the loss of both ignition and the CEL code getting cleared out of the ECU. I gather the ECU has both switched power (from the main relay) and non-switched power (from fuse 13 - feeding enough juice to the ECU to retain CEL codes). Losing juice to both the ICM (which no doubt relies on the switched circuit) and the ECU code memory (fuse 13) suggests it might be a bad ECU ground - which might be common. Either that, or a harness (or something) is intermittently shorting, and bleeding all power out of the ECU - although I'm not blowing fuses...

Does anyone have a good wiring diagram for the ECU and associated connections and harnesses? And where is it grounded? does it rely on the ECU case being grounded, or are there ground wires in the connecting harnesses?

Thanks!

 

Does your car have an security system/remote starter on it?

I would suggest buying a pdf factory shop manual from a site like Ebay. I got one for my TSX and it was worth the $20 I spent on it.

 

The only way I know to reset ECU is to 1) disconnect battery or 2) disconnect BackUp Fuse (7.5A) in underhood fuse box. If ECU is resetting, I would first check battery cables for security and inspect underhood fuse box, BackUp fuse (make sure it's securely plugged in) and no damage to fuse bus.

If all is OK, this suggests the ECU has a cracked circuit board that is interupting power to the ECU.

I had a cracked bridge connector from harness connector to circuit board on my 94EX. It was causing a persistent code 9. After eliminating defective CYP sensor, connector, and wiring I replaced ECU and fixed problem.

If you're sure the ECU is resetting, then I would start looking for a good deal on a used ECU. Check ebay and car-part.com.

good luck

 

PAHonda: no security system or remote starter.

TexasHonda: yeah... I'm in denial... I've been hiding from the prospect of having to buy a new ECU.

But, another day, and another discovery and a new problem/symptom. My wife had been driving the car since I last worked on it and had reported that it had again become undriveable. I test drove it last night and sure enough it was persistently rough. I pulled the plugs and found one had black carbon - and I found that the crappy spark plug cable had a bad connection at the plug end...

Throughout this troubleshooting effort, spark plug cables are pretty much the only ignition component I have not replaced. I replaced them a few years ago and had assumed they were good to go. I've ordered a set of NGK blues.... in the meantime, I've recrimped the bad wire. Now it runs smooth again - great. But now I have a new code - P1382 Cylinder Position (CYP) Sensor No Signal.

Now this has me rethinking the whole sequence of symptoms.... I think the original chugging problem was fixed by the replacement coil & icm. But in the process, I must have damaged the cheapo spark plug cable. I'm thinking that it was occasionally able to get spark to the plug, but that when it couldn't the coil's charge was perhaps arching somewhere inside the distributor and occasionally that spike was getting back to the ECU and causing it to reset. The CYP sensor is in the distributor on this engine...

I've tested the CYP sensor/coil resistance - 950 ohms, which is in the expected range. I've also checked for shorts to ground - none. I've also checked the harness, measuring the same 950 ohms back at the ECU plug. So the sensor tests fine and the harness/connections seem fine, yet I've still got this P1382 popping up. If I clear the codes (pulling fuse 13) and restart the engine, the code pops back up within a couple seconds of engine run time. So it's a hard code.

Yet the engine runs smooth now. I drove it all day around town today without any misfire or chug - with the CEL on with the P1382. What's the risk of leaving it like this? The car runs smooth and I noticed no performance issues... too early to tell if my mileage is suffering. Do I have to worry about the ECU being in "limp home mode"?

And what options do I have from here? I think P1382 with a sensor and harness that test ok means the ECU is damaged (back to my jolt theory). Looking at car-part.com, it looks like used ECUs can be had for $50-200... but then I'd need to hand this off to someone who can program the immobilizer, right? so maybe I'm at an end for what I can do myself.

Thanks!
Jeremy

 

I had a persistent CYP circuit fault on my 94 Accord EX. I finally determined that ECU had to be the problem. ECU replacement fixed the problem.

ECU's are not that expensive as used part, probably < $100 for your model. Check ebay or car-part.com.

good luck

 

What exactly does the CYP sensor do? It looks like its a coil that senses rotation of the cam shaft... does this thing drive the tach? The odd thing is that my tach is working just fine - but P1382 suggests the ECU is getting no signal at all from it.