I did not completely remove the rockers to install the new lifters, just loosened them enough to spin them aside and pull the pushrods. Put the pushrods through a V-shaped piece of cardboard to maintain the order and orientation After installing all six lifters on the bank, I poured motor oil on top of all of them and reinstalled the pushrods where they had been. To tighten the rocker arms, I started in the center and worked outward tightening each to 11ft-lb +90 degrees. I then poured oil in the cup end of each rocker to add some oil inside each rod.

For the new lifters, I submerged each in motor oil but only momentarily. I did not leave them overnight or pump them up by hand. They were not full and stiff when I installed them. When I installed the old lifters this spring, I pumped them up by spinning the motor on the stand with a mongo 1/2" drill until I had oil pumping out the top of each pushrod into the rocker arm cup, but I got the tick anyway when I fired it. Figured I just had a worn plunger that was just leaking down too fast. It never occured to me that it might be a bad rocker arm.

That'* easy enough to do. I should be able to disconnect the rear/and/or front and see what happens. On the advice of a friend, I had disconnected both sensors before I had a scan tool, and although that gave me a MIL, the car did run a bit better. Supposedly, because the default KR with no sensors is only 10 degrees. There is also the overtorque possibility I need to check out. Not gonna be easy on the rear, though.

Those rocker arm bolts are TTY so they might be getting loose.

Hmmm....

If you have a tick then it'* either exhaust or valvetrain. If we figure the exhaust is licked then lets focus on the valvetrain.

I'm sure if you knew but the rockers bolts are one time use, torque to yield (TTY). New they come with a thread locker coating on the threads. In your case, since you didn't remove the bolts then we know the holes are not full of oil thereby giving a false torque reading.

We can suspect the lifters might not have full enough of oil when you torqued the rocker bolts. You could even go slightly tighter on them, say 22 Ft. Lbs and then 100*. But still, in your case it'* possible the lifter wasn't (isn't) sufficiently preloaded or the bolts backed a bit.

You might need to check the rockers agin to see if they are loose. Best way is to start the engine and let it run for 30 seconds then start taking the valve covers off. The run time is to pump up the lifters. You may need to step through the firing order by checking the cylinder half way through the order to get the lifters on the base circle. Might be a good idea to get new bolts and install them. Make sure to clean out the bolt holes with q-tips and solvent.

As for the rear knock sensor, it'* not that bad actually, you can reach between the transaxle and the block to get a socket in there. I use a long thin screwdriver to disconnect the wire.

Cheers,

Thanks for the advice, Foghorn. I'll check the rockers tomorrow for looseness on the base circle. The rocker bolts have been tightened at least 3 times, twice by me, but never more than 11 plus 90* I hope that is the problem as it is so cheap and easy to address. I will let you know.

As the flywheel turns.........this morning I removed the valve covers and covered the rocker arms, valves, and pushrods with a piece of plastic wrap, tucking the edges well down into the cylinder head. Fired it up; watched and listened. It took about 30-40 seconds for oil to flow up through the pushrods and fill the rocker cups. I let it run for about five minutes as I touched each rocker arm and listened. Under these conditions, the noise (which seems more like a light knock this morning than a sharp tick) seems to be coming from within the motor, or at least somewhere lower than the cylinder heads. The valve train seems normal - just emitting the quiet swippety-swippety sounds of the rockers working. I could feel no impulse from any of the rockers relating to the sound I was hearing.

I hand cranked through the firing order several times, to check for looseness, and could get any of the rockers to slide left and right when unloaded, but could get none to move up and down. The valve lash seemed to be zero.

Put it back together and took it shopping with the wife. By the time we got home it was nice and hot and there was no knock, just the same sharp valve tick sound I have had. I probably have some clearance issues that give me some light knocks until things warm up to normal operating temperatures. I think I will have to check the rockers when they are hot to get a really good test. Tomorrow probably.

Keep at it! We'll be your cheering section.


You'll find the problem yet!

This morning, I got underneath and disconnected the rear knock sensor. Disconnecting either sensor sets the MIL and the PCM defaults to 10* knock retard. Even with the rear disconnected, the KS counts were still rolling up pretty quickly. As Foghorn said, to check the torque on the rear sensor, I was able to wiggle a 3/8"-drive 7/8" socket with universal joint onto it through the space between the trans and the oil pan/block. I put the torque wrench on it and it only took about 8 ft-lb to start undoing it. I re-torqued it to 10 ft-lb and left it alone. It did not leak any coolant. This rear sensor is new. Unfortunately, the front sensor is blocked by the starter and I did not feel like removing it to check the torque. I reasoned that as I had installed both sensors at the same time this spring, I doubt I would have torqued one much differently than the other. So I am thinking overtorque hypersensitivity is not likely the problem.

I still need to try the rear sensor with the front detached and watch the KS counts to make sure the front sensor has not become oversensitive due to some other problem.

So, I got to thinking about how the ticking noise doesn't show up until the engine is hot. The small change due to thermal expansion is enough to just make it tick intermittently. It makes the rocker arm bolts look more like a likely culprit. My favorite aftermarket house (O'Reilly) just said huh? when I asked for the bolts, so I ordered them from the dealer. The dealer will have to order them for me, they did not have them on hand. (Kind of makes you think that they are not commonly replaced even if they are TTY. ) Anyway, when they come, I'll clean up the holes, inspect the rockers carefully and torque them up.

In the meantime, since I had the connectors off the sensors anyway, I got out the ohmmeter and measured a resistance of 3880 ohms from each knock sensor output pin to ground. Went up to Radio Shack and bought a pack of 3.9KOhm resistors for 99c . They measured 3820 ohms. I used two of these shorted to ground as dummy loads for the wire leads to the sensors from the PCM. Fired it up and guess what? No KS counts. No knock retard. No MIL. On the down side, no safeguard against engine damage due to detonation, either.

Now, I do not recommend that anyone else try this, because with these dummy loads, the PCM is now unable to adjust timing for any detonation that might occur. This may seriously damage - even ruin my engine. But, until I can find and repair the source of my false KR, I may just run "easy" with this "test" setup. I did run a scan tool test up the same on-ramp as the earlier one in this post for comparison. I heard no detonation, even up a pretty good hill and under pretty heavy throttle. Of course, that doesn't mean there wasn't some detonation occuring that I could not hear, but it does indicate that any detonation that is occuring is not severe. I filled the tank, drove the car 81 miles and returned to the same pump and filled the tank the same way - 2.85 gal - not enough miles for a real accurate test, but 28 mpg ain't too bad, especially running 10% ethanol with 15W-40 oil.

Here'* the scan tool data "after"



My goal is still to get rid of the audible engine noise so that I can run the car as it was intended, with the knock sensors connected, and the engine protected..

Still no rocker bolts today, so I checked the KS counts with the rear sensor connected and the front disconnected. Same response from the front sensor alone as the rear alone. You can actually watch the counts go up along with the tick tick noise. I think the knock sensors are fine. It'* the engine that needs work.

Must.... find.... and kill ..... noise.