Rep Power: 0

Hey guys,

I’ve recently finished putting the engine back together in my 2004 Civic EX (1.7 L) after replacing the head gasket, piston rings, timing belt, water pump, and a bunch of other odds and ends. With great effort, I’ve managed to get it started once or twice since the rebuild, but it takes a lot of cranking, and it has to be tethered to my Accord which I keep running. Sometimes, I can't get it started at all.

Using an inline spark tester, I was able to determine that the spark plugs are not firing consistently. When they are firing, they do so with a consistent rhythm, but most of the time, I’ll connect the tester and see nothing as the engine turns over. I have had times when only one cylinder will fire and show up on the tester. In these cases the engine will “catch” just a bit as that plug fires but will of course not start because of the lack of support from the other three. I’ve tried moving coils around to no avail; when only one is firing, it fires no matter which plug and coil are attached.

I've also had some symptoms that were a little more odd. There was a time after the engine reassembly when I couldn't get the code reader to "link" at all, so I couldn't see engine codes or live data, but I'm able to see both now. There was a time or two when I got the engine started with a code reader plugged into the OBD-II port. Once I had the car started, I pulled the reader out, and the car shut off immediately.

Another thing I find strange: when I finally decided to take the car apart to get to the bottom of the head gasket issue, the gauges on my instrument cluster weren't working at all. Thinking it was just a fuse, I proceeded with the repair anyways, planning to use to my reader's live data to watch RPM, ECT, speed, etc. to complete the break in. About a month later, as I was finishing up the rings, head gasket, etc., I plugged in the battery, and the gauges worked. However, when I would use the battery for something, like lowering a window or turning on a light, the RPM and tach. would jump a bit, then drop sometimes and stay flickering near that level other times.

Now, when the car is running and I'm not pushing the gas, all gauges act normal and move smoothly, regardless of speed. When I start to push down on the gas, the tach. and speedometer start to jerk around. The needles flicker back and forth at a higher value than shown on the code reader. For example, doing ~50-55 MPH at ~2K RPM will show normally on the gauges, until I push the gas to accelerate... then the gauges might jump to ~70-80 MPH and start to flicker around. The tach. does a similar thing, sometimes raising the "RPM" up past 8K. Moving windows, turning on headlights, etc. have no effect on the gauges.

Honestly, I don't really care about the instruments, but they might be an indicator that something else is wrong. With these symptoms, I’m thinking the issue is electrical because of the clear evidence of electrical issues with the plugs in the engine and the gauges in the car. I also believe it originates before the coils, because of the lack of spark on the tester, regardless of coil condition, age, or position. I don’t believe this car has a distributor, so would that mean the ECU is having issues? Would this be a bad ground? Wire harness issue? Water in a connector somewhere? I’m not too familiar with troubleshooting electronics, so I'm struggling a great deal to come up with theories.

I see no evidence the engine isn’t getting enough air. I can smell gas in the cylinders, I know the timing is correct, and compression is good. I can’t see any sensors that are not plugged in. I have no engine codes. I use jumper cables as I crank the engine over and test things to keep the battery in good shape.

Thanks in advance for any ideas or help, and let me know if I can clarify any points.

 
Rep Power: 367

check and clean camshaft and crankshaft position sensors, they usually cause issues after headgasket job

 
Rep Power: 0

Thanks for the idea. I pulled both sensors and inspected the sensor and connector for each. Both appeared to be pretty pristine and in good condition. I cleaned them anyways while I had them off just in case. I reattached everything, but no dice. The engine's still cranking for sure, but no spark on any coil.

 
Rep Power: 367

disconnect camshaft sensor and try again (when it is disconnected and runs, the sensor is a goner)
then try again with crankshaft position sensor disconnected (can't remember which makes the engine run disconnected...)

https://www.civicforums.com/forums/3...rt-solved.html

 
Rep Power: 0

I gave it a try... I started by unplugging the crankshaft sensor. I plugged in the code reader, then put the battery back in the car (13.57 volts). The car started up immediately. As expected, it threw a crankshaft sensor code. I let it run for a few minutes, varying the RPM at low levels while the engine was unloaded. It would not idle below 1,200 RPM. I also threw on the jumper cables since I had removed the alternator belt to get the crankshaft pulley off. I then got out and plugged the crank sensor back in, and the car kept running. I then unplugged the code reader, and the car shut off that second. (By the way, I did replace the crankshaft sensor when I did the head gasket and everything else.)

The car wouldn’t start without the reader in. It turned over, but no audible combustion or “firing”. Then, it wouldn’t start even when the reader got plugged back in. What it took was me disconnecting the battery and jumper cables, letting the car sit powerless for a few seconds, then plugging everything back in with the reader already plugged in under the dash. Again, the car started right up, this time with the crankshaft position sensor in.

Reading around, it seems someone else had this issue, and it was caused by a bad ground for the ECU/PCM. This is because the OBD-II reader grounds the PCM when it is plugged in, but stops when it’s unplugged.

https://www.reddit.com/r/MechanicAdv..._plugged_into/

Do you guys think there is any weight to that theory? If so, how can I check my grounds? I have a multimeter and free time.

If it’s not the grounds, perhaps I should check my fuses? Any ideas are appreciated.

 
Rep Power: 367

check your alternator bolts... If they are loose, it can damage the ECU

 
Rep Power: 222

​​​​​​This, and I hope you used OEM sensors

 
Rep Power: 0

posted in error

 
Rep Power: 0

posted in error

 
Rep Power: 0

I did end up finding a loose ground wire, so I cleaned it with electrical parts cleaner and a steel brush. I tightened it down, and the measured resistance is 2.4 ohms. When the car's running, the resistance is basically zero. Now the car can run with or without the code reader.

They are all tight, except the bolt that runs from left to right on which the alternator pivots. Do I need to tighten this one as well?

You were right. The crankshaft sensor is bad, so the car only starts with it unplugged. I haven't noted any issues with the camshaft sensor.

I didn't, but I realize now that was a mistake. I've heard a lot about how these old Hondas don't like aftermarket sensors. I kept the old sensor, so I'll be putting that one back on. It had no issues, and I only replaced it so I could get it done "while I was in there".

As of now, it looks like my electrical issues are solved. Thanks for the help!