OK, so now that the "new guy" post is out of the way, I could use a bit of guidance. I'm trying to fix a misfire on my 96 Accord EX. Here's the low-down.

Basic Info
1996 Accord EX 2.2L VTEC I4, eng. code F22B1
~220,000 miles
No customizing has been done. Vehicle is ugly (Minnesota rust), but it runs & drives well and is generally healthy.

Issue
Few weeks ago I got in the car to go home from work. Started up fine, but ran really rough. It was rhythmic and I suspected it was missing consistently on one cylinder. It had driven fine on all four cylinders just that morning. Since then I've been trying with my limited repair experience to track down what the issue is.
I have a bottom-of-the-line scanner, and I had four codes: P0303 (#3 misfire), P0401 (EGR flow), P0420 (Cat efficiency), and P1491 (EGR lift). With the P0303 in there, I operated under the assumption that the cylinder misfire on #3 was potentially causing the other 3 codes for EGR and cat efficiency. I cleared the codes and have continued to check them periodically. The only code that keeps showing up is P0303, so that's what I'm working on.

Troubleshooting History
I've generally followed the misfire steps at troubleshootmyvehicle.com. I've found this guy's articles to be very helpful and used the steps in his ignition issues article to troubleshoot a bad ICM this summer.
SO, I checked spark at all four wires & they all passed. That eliminated the wires, dist cap, and rotor.
Next checked compression on all four cylinders. They also all four passed. Compression was right around 160 PSI for all four and pretty even across the board.
So, that left fuel injectors.
I checked resistance and it was within a tenth of an ohm (provided I did that one correctly). Regardless, I got an approximately equal reading for each injector.
Next I swapped the #2 and #3 injectors, then did a power balance test. That identified the 2nd cylinder from the left as the dead cylinder. Here's where my inexperience shows: I assumed that was cylinder #2, and therefore the injector was clearly the problem. I then replaced all four injectors.
Of course, I performed the swap under an incorrect assumption: I thought that when standing in front of the vehicle, the cylinders on an F22B1 were 1-2-3-4. I know now that is not correct, and that they are 4-3-2-1. I know this because replacing all four injectors did not solve the problem. So, my swap test didn't identify the injector as bad, it actually eliminated the injector as the problem. Ugh.

So, figuring it was still a fuel issue, and had to be cylinder specific (since I'm consistently getting misfire on one specific cylinder, not bouncing around or random), I decided to test the injector signal. I didn't have a noid light, so I MacGyver'd one with a spare 194 bulb, spare dash bulb socket, a couple pieces of doorbell wire and some extremely amateur soldering. (Of course real solder actually disqualifies it as MacGyvering, since he would have actually used chewing gum.) Surprisingly, the device worked. I tested on known good cylinder #4, and got a pretty little flashing light. Next I tested on bad cylinder #3, and it was dead. "Aha!", think I, "I have found you!" Well sort of...

Stuck
SO this is where I'm at. I've identified the bad cylinder for certain with a 2nd power balance test. I have eliminated spark and compression, eliminated the fuel injector itself. I am guessing that since I have 3 operational cylinders, it's not a global issue like a bad main relay or bad fuel pump. I have identified the affected injector is not getting it's switching signal/injector pulse.
What I actually need help with is where to go from here? Have I done enough investigation to eliminated everything but the injector connector? Or is there some other component between the PCM and the injector that could be bad? What are the possibilities for one injector not getting it's pulse signal but the others are fine?

Thanks for any guidance you might have. And if this is TL;DR, I apologize!

 

The injector wiring should be hot (12v) all the time. ECM sends a ground signal to the injector causing it to fire.

From your description, I would suspect that 12v is missing and ground signal is available on injector #3. You can confirm w/ test light to ground side of injector and test light hooked to positive battery. You should get a pulse when engine is turned over. A noid light should be used to avoid excessive current draw w/ your home-made rig.

If ground fire signal found, the problem is a break in the wire or injector connector fault supplying 12v through the injector resistor bank (front firewall on driver's side). Disconnect resistor bank and check for 12V on the #3 injector wire. If 12V available, wiring fault between resistor bank and injector. If 12V missing, likely a faulty resistor bank assy. replace resistor bank.

good luck

 

Thanks TexasHonda. Please bear with me - I'm a computer programmer, not an automotive technician, so there are things I need to clarify.

Right, so if I'm reading this correctly:
1) If I turn the key to II and probe either or both wires above the #3 connector, I should get 12v on both - either with a standard test light, or a multimeter - correct?
2) If the 12v is missing, I can double-check to make sure the ground signal is coming by probing the ground side wire as you describe connected to the pos term on the battery while the engine is cranked; this serves to eliminate me having two problems here instead of just one. Are those statements correct?
3) What if the 12v is there, but the ground is missing? Would that suggest a problem with the ECM (ouch)?

OK, I think I get this. Another question: Can I perform these checks on the resistor box with just the key turned to II or do I need someone to crank the engine while I test? I assume I can just turn the key to II since I'm looking for the power supply to the injector, essentially.

Thanks again!

 

OK, doing this in bits & pieces since I don't have a warm garage and it's 25 degrees out.
I verified I am NOT getting 12v at the connector, but I AM getting the ground pulse when I test it like you suggest. So, on to the second part then - testing at the resistor bank. After my fingers warm up. Might have to brush some snow out of the way too...

 

I think the wire color for the 12V side of the injector circuit is red/blk (or blk/red). The 12V power wire to all four injectors is the same color. The injector resistor box has 5 wires to the connector (all the same color). The numbering is 123 on the top row, then x56 on the bottom row. The x is a missing wire. You want to check the resistance with a volt meter on the resistor box side. You keep one volt meter lead at wire 1, then test the resistance at 2, 3, 5, then 6 with the other volt meter wire. The resistance should be between 5-7 ohms between 1-2, 1-3, etc.

The EGR codes you are getting will take some troubleshooting. The first thing I would do is clean the EGR ports on the intake manifold, because that is a common problem. There is a good ericthecarguy video on how to do this on your model accord. If codes return, you can do more troubleshooting. Do not replace the EGR valve just by the codes you are getting.

 

Thanks for the reply, PAhonda.
Right, so this looked pretty much exactly as you described except that the connector rows were vertical. So, I had:
RED - X
BLK - BLK
BLK - BLK

My ohm readings were as follows:
RED - X
7 ohms - 16 Mohms
7 ohms - 7 ohms

I'm going to guess there's a resistance issue within the resistance box, yes?


Just cleaned the EGR passages this summer. None of the codes have repeated except the Cylinder 3 misfire, so I will fix that and see if I still get any error codes afterward.

 

The 16 milliohm reading is out of spec and likely causing your problems on cylinder 3. You may want to retest just to be sure you get the same reading.

This may be the first time I've seen the injector resistor fail on an accord on this site. If you have a u-pull-it junkyard in the area, you can probably pick one up for pretty cheap and have little chance of a bad unit. You could always test the unit before you buy it. I think all 94-97 EX accords use the same part.

 

Looks like you answered your own questions. Good work1

Looks like one leg of resistor bank is open.

good luck

 

Thanks. Found a used one for cheap - I'll post back when I get a chance to install it and see if corrects the issue.

 

Wahoo! The resistor bank assembly replacement fixed the problem!

Just to verify, once I removed the old bank I tested the resistance from the connector side. Results were:
RED - X
6.8 - "0L"
6.8 - 6.8

I am interpreting that "0L" as no current getting through to even test, since that's what it shows on the display before you make contact with the leads.

I pre-tested the replacement bank and all showed 6.8 ohms. Plugged it in and it started up nicely. Engine hasn't been that happy in a long time.

Sweet! On to the next set of problems!

Thanks TexasHonda and PAHonda for your very accurate and helpful assistance!