Starter/Flexplate noise

OK, I'm going to put this one on the board...

In the process of de-griping our 155Kmi '94 (4.0, 4wd, AW4), one of our gripes was a scraping noise in the engine/transmission interface. Basically, there's this metallic "chatter" noise at the frequency of crankshaft rotation that manifests itself mostly when idling in park. It's usually minor, and worse if it's facing downhill. If the A/C compressor is running, the noise is really objectionable, and sounds like it's ready to do some damage.

I originally suspected the much-discussed loose torque converter bolt problem in the flexplate, but I pulled the inspection cover and checked, and all 4 bolts are tight. There is no evidence of cracking near the crankshaft flange either. I can't check the crank/flexplate bolts without removing the engine or trans of course, but I can't detect any play in the plate that might indicate a problem there.

Curiously, the dust cover shows old scratches in an arc at the radius of the TC bolts, indicating that at some point in the past, a bolt came out and rattled around in there. The scratches were dirty, that's why I know they're old. My guess is that the flexplate was replaced at that time (it's quite clean, no marks).

I've scoped the noise right at the starter/belhousing point. Put a rebuilt starter in because the bendix looked a little worn, and the bushing at the tip was getting bad (maybe this is why). This helped, but only slightly. With the A/C on, it's still a problem. And the weather is warming up...

I'm stuck now. I don't believe that the flexplate is a problem, but I don't know for sure. Stuff I suspect:

* Crankshaft flange slighty bent from prior loose bolt damage such that it's runout is slightly out of spec, so that the flexplate doesn't rotate in a true plane. I would think this would destroy the tourque converter though.

* From what I've seen, one of the main bearings has a "thrust surface" to keep the crankshaft in position. It could be that this bearing is worn to the point that the crankshaft "floats" forward when the TC isn't loaded, or when the A/C is on and the belt is heavily loaded.

* Starter did not have, nor new unit come with any kind of gasket. Not required anyway, there's no need for a seal. However, I'm thinking about fabbing one to go between the starter and the bellhousing to pick up .010" or so of clearance. This would really just be a workaround for the real problem.

I'm thinking about going ahead and trying a starter shim gasket for now. I'm also thinking about pulling the engine and doing several things (bearing check, new seals, can't get to the oil filter flange bolt for new rings, needs coolant drain anyway, I could see the bolts on the flexplate, etc). I'd prefer to wait until fall for this though, and I'm fantasizing with the idea of bulding up a stroker to go it in anyway, which will take a few months.

Any thoughts? Thanks...

Shouldn't be the starter since it isn't supposed to be engaged other than when cranking. If it is engaged at engine speed, the starter normally explodes. I guess the starter could be not retracting the gear all the way and it is hitting the flex plate ring gear. How does the edge of the ring gear look. Is it polished and shiny in places?

Most likely is a cracked flex plate. The crack is most likely close to the hub where you can't see it, or one of the welds that holds the ring gear to the plate could be cracked.

A front bearing in the tranny could be the problem due to a previously broken flex plate, but not real likely. A shot bearing normally causes the seal to fail and you would be looking at a lot of tranny fluid leaking.

I have actually seen a broken bolt boss on the torque converter, but not on a jeep.

I have also seen bolts that were tight, but a touch too long so they didn't hold the flex plate well. Visually they looked OK, and were torqued.

Have you checked that all of the bolts are tight on the bellhousing?

Try using a short piece of garden hose as a stethescope to track down the noise. Try starting it without the inspection plate to see if the noise is much worse. If not, then it isn't comming from the flex plate, TC, or the tranny.

I did the garden hose trick and localized the sound to the starter/bellhousing interface, therefore I originally suspected that the starter gear wasn't retracting all the way, thus the reman starter (which did help, but only mildly).

To other points you so kindly raised:

*Bellhousing bolts: checked only next to the starter ('cause I was there...). Good suggestion.

* No trans leaks, inside the bellhousing is clean.

* I didn't look at all the ring welds, good idea.

* The visible face of the flexplate teeth are worn bright on the outer 1/4 or so. Looks to be all the way around.

Running it without the inspection cover is a good idea, and I ought to look at all of those welds. Maybe poking an inspection mirror in there wouldn't hurt too. Will be a couple of days before I can get to this.

If the inside of the flexplate is starting to crack where I can't see it, how much trouble am I in? (e.g. get it out now, or will it run a month or two?)

Thanks for the input...

-sc

Jeeps are a lot like women. They require a lot of maintenance, and can cost a lot to maintain. They are a lot of fun when everthing is working right, but just as soon as you think you have them figured out and everything is going fine, you're in trouble.

Now if you ask my wife of 30 years, she will have a whole different perspective about this.:hang:

My advise is to not wait until it breaks, but to spend a little now to fix it instead of a lot of $$$ and hassle later.

Oh, it's not the $$$ that gets me so much as the time involved. The $%^&! Dakota that sits next to it in the garage gives me more trouble and costs way more...

I'll take a look at the other easy stuff later in the week. Thanks for the help.

-sc

Check the bolts that hold the flexplate to the torque converter. I had the same sound your describing on mine when the bolts mysteriously became loose.

OK, this one is SOLVED. Tom, I owe you big time...

Problem was one of Tom's suggestions, that the bolts were too long and bottomed out in the converter itself. They were torqued properly, but the plate had about .015 or so of play in it towards the rear. You had to really look close to see what was happening. A key clue was (if I was smart enough to see the first time) that each of the 4 bolt heads had this "halo" of rust dust trailing behind them, it was being generated from underneath the bolts. If the bolt heads were truly snug against the plate, this couldn't have happened.

Bolts that were in it were M10 x 1.50 with 17mm heads, grade 10.9. Length measured at about .620" or so, and the tips were bright where they'd bottomed out. My crude measurement of the depth of the converter bosses + flexplate was about .590".

Best replacements I could find late on a Saturday afternoon were some GM comverter bolts with 15mm heads and wide skirts, grade 9.8, threadlocker already on them, length .560". I don't know what grade the factory used, but 9.8 at least seems reasonable for this application. If anyone knows what grade the factory bolts are, please post for reference.

Seems that checking torque on these every 10K mi or so might be really wise.

Put 'em in, torque 'em down, start it up, and it's just as quiet as can be, even with the A/C on. In fact, it's amazing how much quieter it runs now. On the road, I can hear gear noises that were masked before.

Thus, this is the last "gotta get fixed NOW" problem on our new-to-us '94. Last things to fix are a couple of oil leaks, saggy springs, and sticky door locks, and then we're good to go for awhile. Thanks everyone for the suggestions, I'm thrilled that this turned out to be something so simple. Tom, if we cross paths sometime, beer's on me...

-sc

Good call Tom. I was helping a friend put some Jeep coil spacers in the rr. of his Mustang the other day and he had me look at the valve covers that were pouring oil down the heads. I started looking at, asked him if he retorqued the bolts after heat cycling it, and then checked them myself. All tight, but then grabbed the cover and could rock it back and forth. He had removed a spacer under the vavle cover for some different rocker, and the long bolts used with the spacer would bottom in the hole just before the cover was clamped tight. I had a friend also who damaged the thrust main while idling for an extended period pointing downhill @60* while stuck. Only found out a few weeks later when the torque converter bolts were hitting the inspection cover and you could move the crank front to back 1/4".