onp2005

2003 civic coupe MT w/177k miles. DTC values move around between P0301 - P0304.

Engine started to misfire right after I changed the timing belt. Checked, double checked, checked and checked and the crank and cam are correctly aligned.

Replaced spark plugs, coil sticks, injectors, MAP sensor, PCV, checked/set valve clearance, disconnected battery for throttle reset. Back probed TPS and voltage is linear. Did not detect a vac leak using propane.

Cleared the CEL light while on the freeway and it didn't come back on until +100 miles later when I idled the car. Seems to be a problem only at idle.

When P0304 was stored (most recent) I removed the #4 coil stick and there was no change in the idle speed. Removing any of the other coils caused the engine to nearly stall.

My OBD app shows the ignition advance for the #1 cylinder drifts between 3% and 12% at idle. My other civic is steady at 15% at idle. The exhaust is leaving a mess on my garage floor.

What would cause the ignition advance to drift at idle and only at idle?

Mayday

holmesnmanny

may have jumped a tooth or crank sensor may be bad

onp2005

A friend recommended a new belt tensioner and spent the last couple of hours replacing it.

Before removing the existing tensioner I examined the camshaft and crank and they were ***** on aligned. Also examined the crank position sensor and didn't notice anything unusual about it. Wiped the sensor head with a shop towel. Image below.




The good news is the CEL did not come back on after a maintenance check ride. The bad news is the engine idles only slightly better than before. The OBD app still shows the #1 cylinder ignition advance pct drifting between 2% and 15%.

I figure to order a crankshaft position sensor in the next 48 hours. It doesn't make sense to me why this would be bad and yet have a smooth running engine/no CEL if the RPMs are anything but idle.

+$1000 in parts so far. This is killin' me.

BTW, the best tools for removing the crankshaft bolt are:

thisguyaves

Check TDC by actual piston position, not the crank mark and cam mark. Use a long screw driver or bit of wire, rotate crank until the piston (shown by screwdriver sitting on piston in the sparkplug hole) reaches it's apex, then compare cam alignment. I agree with holmes, it's probably bad timing. 1 tooth is all it takes

Ignition advance is determined by a lot of things, one of which being O2 sensors. A clogged cat could be throwing off O2 values, which would throw off AFR and cause a bad idle and tail pipe sludge

onp2005

Just finished replacing the crank sensor. Still idles like crap. P0300-4 codes being generated now.

Took pix of the crank and cam alignment when doing this service. Alignment looks spot on to me.

cam:


crank:


Here's a screen grab of the OBD app showing ignition adv % vs RPM. Notice the ignition adv chart line is all over the place. The image below this is another screen grab showing the same thing but at light throttle.

@idle:


@light throttle:



For comparison here's the same info from my good civic@idle:




Any ideas why the misfires continue? There's not much left to replace.

thisguyaves

Injectors? *shrug*

onp2005

Already replaced the fuel injectors. All new, all OEM. I've also replaced the coil sticks, spark plugs, PCV, crank sensor and MAP sensor.

The only 'reasonable' component left to consider is the cam sensor. Will probably order that late tomorrow. If a new cam sensor doesn't fix this then I'll most likely sell it cheap on craigslist. There's only so much time and coin I can put into this car.

I do appreciate the feedback/suggestions.

- john

---
Snoqualmie, WA

thisguyaves

That is fairly ridiculous. Hope you get it figured out. Might also check to make sure your cat isn't clogged up, bad O2 readings could skew timing.

onp2005

Below are the results of five leak-down tests I performed yesterday. Air was escaping the throttle body and other spark plug holes. No air was observed escaping the crankcase, radiator or exhaust.

Bad intake valves or the head gasket blew out between cylinders. Since this problem happened all of a sudden my bet is the gasket blew out between cylinders.

Hope to have the head off by this afternoon.

onp2005

Six bent intake valves is the word from the machine shop. Parts dropped off yesterday. Hope they can get this finished today before they close up for the holiday weekend.

For reference it took about 3 hours to remove the cylinder head which includes time for a short break.

The EGR passages were fairly clogged up with exhaust soot. See below.




I would expect damage to the pistons if the valves are bent. I certainly can't see any damage. And I would also expect all 16 valves to be damaged.




And in the ten years I've owned it the engine has never overheated. Still trying to determine the problem source. Really don't want this to happen again.

If I get the head back today I hope to be out test driving it tomorrow.

thisguyaves

Wow. Unfortunately this is a common issue with D17s.. Glad you got it sorted out!

onp2005

Cylinder head back today. Put the intake manifold back on the head and then put the head w/manifold onto the block. Tightened bolts. New OEM gaskets used.

Run leak down test

Call machine shop

Below is the latest leak down profile. It includes two previous tests. I've also transposed the layout.

The head is going back to the machine shop tomorrow. The owner/manager is as curious as I am to understand why it's leaking so much from the exhaust valves now.

I have two other cars and a motorcycle which all need reg maintenance. The motorcycle maintenance is prep for a tour later this month.

This project is clogging up my schedule/garage.

onp2005

The owner/manager wasn't in but the two other machinist watched as isopropyl alcohol came streaming out the exhaust valves in embarrassing amounts. They believe the guide used to grind the seats is no longer true. I suggested if one guide wasn't straight then there is a likelihood others are bent as well. Think they use a Sunnen VGS 20.

ALWAYS PERFORM A LEAK DOWN AFTER YOU GET A HEAD BACK FROM A MACHINE SHOP. Generate a baseline and protect yourself from sloppy shop work. If this shop released this I can imagine there are other shops out there doing the same sloppy work.

Let's see there's the rework, two 90 minute trips to the machine shop and another $40 for a new head gasket.

onp2005

Head back and just tested, see results below. Looks dang good for a car with 177k miles. The machine shop provided a Fel-Pro head gasket but after comparing it with the OEM I'm going to wait until I get another OEM gasket before finishing this job. Might have it on the road by next weekend.

onp2005

Engine back together and it idles as good as the day it came from the factory. Spent $290 on valves, gaskets and seals and $588 on machine shop work. About 6.5 hours to assemble, set valve clearances & add fluids.

I cleaned one piston head with acetone and a cloth rag but decided against doing this on the other pistons. The pistons will just get gunked up again and the process of cleaning them was causing flecks of gunk to fall into the coolant passages.

Also used acetone and a cloth rag for the gasket surfaces. For a couple of spots on the cylinder tops I used a brass brush. Compressed air was used to blast out any loose material in each cylinder just prior to assemble. The cylinder head was removed and installed with the intake manifold attached. Easier access to the fasteners in the back.

Permatex Ultraslick Engine Assembly Lube was applied on the cam lobes, cam journals and valve stem tops. I had been a fan of moly disulfide but discovered moly accumulating in the cracks and crevices of an engine I had used it on.

This problem really stressed me out. I thought for sure I had the timing belt off but repeated examination disproved this. I then thought it was a component I had touched or indirectly touched. $1400 later I had run out of ideas and components to replace. That and my bank account wasn’t going to take much more abuse.

The misfire and CEL only at idle baffled me. I was beginning to believe there was a chaffed wire or bad ECU. The latter seemed completely unreasonable but I didn’t have any other ideas. I’ve since seen a youtube vid (South Main Auto) where a weak cylinder will misfire and generate P030x codes only at idle.

Once I had run out of ‘reasonable’ components to replace did it occur to me to perform a leak down. I was dumbstruck when the leak down test indicated three cylinders with leaky intake valves. Still haven’t determined how this occurred. I rarely drive this car so perhaps there was a problem before I did the belt service. But the CEL wasn’t on before replacing the belt nor were there any complaints of a bad idle. So it happened when I replaced the belt? How? I rotated the crank pulley 720 degrees after assembly and there was no resistance. And the pistons do not show any damage. Image of piston #4 below. And why only 6 intake valves? If the timing belt was off I would expect all intake valves or all exhaust valves to be damaged, not 6 out of 8 damaged.

In the future I’ll do a check drive and OBD scan prior to any service. And I’ll be doing leak down tests on all vehicles as part of their regular service. Get a baseline and notice problems earlier.

And I’m still cursing the machine shop for their poor workmanship. The rework added another 13 days to this project while I waited for the second OEM gasket to arrive. Just another example of why I have a problem letting other people touch my cars.

thisguyaves

Very in depth, bravo! Here's to another 200k miles of reliable service.