I agree with Constantine and all his theories. I have no idea why Porsche went to the non-circlip design. It is harder to setup, but when it's adjusted it's maintenance free and there is no thrust bearing failure. There is a lot of discussion about TBF and I heard all kinds of theories. Mostly about the increased torque and power of S4's and younger models. Flexplate movement has nothing to do with the power output of your engine. It is only caused by the missing circlip design and a clamp design that was never improved.

Why did Porsche go this way? My theory is that they wanted to safe costs. Not production costs in material but man hour costs. It takes a couple of attempts to get the circlip and the washer adjusted to spec even when everthing is out of the car. Try this with the engine still in the car if you have to change the transaxle.... :)

I went the Loctite 290 way, because people tried it and it worked well and is cost effective. The only time that I heard about movement after application of Loctite was when it was not left long enough to cure. Releasing the pressure annually is OK but if you drive your car every day there is a problem. When is the pressure high enough to cause increased friction on your TB? Nobody knows! If you release the flexplate annually you are NOT eliminating the pressure completely, you're releasing the pressure temporarily.

Can anyone provide me with the pinch bolt torque? I don't have a manual although I'm starting to think it might be a good idea...

You need a manual if you are going to work on these cars, period!

Jim Morehouse has them on CD, money well spent!

I have Mitchell on demand at the shop which is a pretty good source of information on car repair, but it is no where close to what the manuals are.

I think that Constantine has the right info on this subject, for the price of a TBF on one of these cars, it makes his clamp look cheap as compared to what it will cost if you have a TBF.

Constantine's clamp really is a very nice looking piece, BTW.
I feel this is another area where my money was put to good use, along with his recommendations on having three bearings put into my rebuilt TT.

Hi Steen,

There are two, one that is in the manuals, 59 ft lbs., and the other is the one now used by many and has become the new standard, 65 ft lbs.

Interesting how this new rating came to be really, it morphed from a long ago post from a gentleman in the UAE (?) who was one of the first to publicly talk about the driveshaft pullout at the front flexplate on the Rennlist e-mail list. This was before we had this wonderful 928 Forum in addition to the list.

If I remember correctly the fellow had his mechanics call to Porsche AG to find out more about this pullout problem that kept manifesting itself in his 928S4. His mechanics were told by Porsche AG mechanics they recommended to change the flexplate and coupler assembly, change the pinch bolt and torque it 10% higher than in the manual.

Over time these instructions have morphed into only torquing the pinch bolt 10% higher and maybe changing out the pinch bolt. I guess people just dis-regarded the new flexplate/coupler assembly due to its cost and the cost of installing it. Again, our testing did not find using a new pinch bolt and torquing it to the higher setting does any good to stop the driveshaft pullout. That Porsche suggested (if it's true) to change out the flexplate assembly also tells us the clamp is not meant to be re-clamped much.

Hope that helps,
Constantine
P.S. I second the recommendation that you should have the 928 manuals in some form if you are going to work on these cars or want to learn more about it.

I still think there is more to it that twist-shortening. My 89's TT shaft has less spline reveal than it did 6 years ago. I release it every 3-6 months. Something has moved. TC ballooning is a concern with other cars. Why not ours?

Is the problem more evident with 87-up, or the other way 'round?

IE, would the larger torque converter contribute more or less to ballooning? (IIRC, larger diameter = lower stall)

Disclaimer: I am not a mechanic.

Stupid Question: It seems the "twist-shortening" theory is the most widely accepted. Wouldn't it be easy to test this? Seems like if you made a tiny permanent mark - say with a small file or even a hacksaw blade - in one or two of the shaft splines just as they meet the clamp, then re-check after a year or so, wouldn't the mark then be a few mm farther back from the clamp?

Edit: In fact, wouldn't it be exactly the same distance as the flex plate change?

Ballooning is not a factor because the TC sits in it's own housing. Ballooning forces push against the TC bearings, not against the transaxle shaft.

Ballooning is definitely a factor if the tranny is connected directly to the engine and the TC sits on the flywheel. I replaced my TC bearings and there is no way for the TC to push against the transaxle shaft.

Maybe I'm not understanding things... (not exactly a news flash! :roflmao: )

But if twist-shortening pulls the shaft ever so slightly rearward out of the front clamp, wouldn't that be measurable over time?

Some info in this thread. Especially see post #43.

If I ever buy '85-87 automatic one of the first tasks is to check front clamp position and add shims into it. '88-95 version would require glue or preferably custom clamp from Constantine.

Constantine...thanks for the spec

I have bid on a set of manuals a number of times on ebay but have never been able to close the deal. I have had 2 928's over the last 15 years and don't plan on ever not having one. The first one was an 83 and aside from brakes and an alternator , it was trouble free for the 5 yrs I had it. I replaced it with my current S4 and I have let it go somewhat. Hence the BIG PROJECT. So far I've replaced the TC seal , TC brgs , trans filter , tt brgs , rear wheel brgs , rear shocks , O2 sensor, gas tank strap and fuel filter. I'm going to remove the air pump and maybe the CATS and am thinking about an OTT X-pipe. I also have to tend to a nasty pwr steering leak somewhere in the resevoir area.

All this work has been a challenge for a first timer , especially without manuals , but rennlisters have eased the way considerably. My thanks to those who have responded to my queries.

I'm still on the lookout for a set of manuals...

The theory of the torque converter (TC) bearings wearing out and causing the forward push is not plausible. There would have to be constant forward pressure being exerted to the driveshaft from the rear. The design of the TC housing would not allow this just as Shocki found when he tested this theory using a press on the TC housing. If this theory were true then the Loctite, circlip/washer/bearing, and our clamp would not work. But they do.

As far as using loctite there are drawbacks. Once Loctite is used it is near impossible to release the driveshaft from the front coupler and people have had to unbolt the front clamp from the flexplates to remove the torque tube from the car. There was also an instance when the driveshaft did move after loctite was introduced and the owner could not release the forward pressure he found at the front flexplates.



If this theory were true then the Loctite, circlip/washer/bearing, and our clamp would not work. But they do.


There was also an instance when the driveshaft did move after loctite was introduced and the owner could not release the forward pressure he found at the front flexplates.


There is a contradiction here.





Constantine...




How did you verifier/ confirm this twisting of the drive shaft under load which method did you use to come to that conclusion.

There is constant forward pressure being exerted to the driveshaft from the rear, the rotating mass (torque converter) and fluid pressure within causes the converter to expand/ balloons some much it distorts the converter. Also the bell housing is made of aluminum; toque converter bearings are steel, heat expansion is at different rates creating an unknown amount of end play where the bearing seats in the bell housing, there must to be some movement.





You can’t compare a torsion bar to a rotating shaft, In a car one end of the torsion bar is attached firmly to the vehicle chassis; the opposite end is attached to the suspension arm or A- frame. Vertical motions of the wheel cause the bar to rotate along its axis and are resisted by the bars torsion resistant.


We have different opinions as to the cause.

In drag racing it has been common knowledge for years that torque converters drive the front pump which creates line pressure, which tries to drive the converter & everything in front of it forward, hence creating force against the bearing. It's been proven through tests. It's no secret to drag teams.

Hammer

Hi Steve,

There is no contradiction about the instance when the driveshaft did move after loctite was introduced. It moved for some reason then was frozen in place in a tensioned state. Others who have used this method have not had anymore movement, although I personally don't advise using loctite as I mentioned before.

I understand there is pressure in the trans under operation, but if you are saying the torque converters (TC) are ballooning under this pressure and pushing the driveshaft forward then the following would be seen by the owners:

1. A lot of aluminum shavings coming out of the little grate of the TC housing.

2. No effect of the methods currently used to stop this movement.

People have reported 3-4mm or more of backward movement of the front flexplates on the driveshaft when the front pinch bolt was released. That is too much for bearings to be displaced by heat. Also, the driveshaft has not been known to move forward when it is marked in relation to the front bellhousing.

I can compare a torsion bar to our driveshaft since it does wind up under load. A long steel rod about 1" in diameter that is our driveshaft will twist and was designed that way. That is why Porsche placed a groove in the rear of the driveshaft that a bolt goes through, the circlip bearing and washers up front and placed flexplates on either end to allow for this contractive movement. Not because there would be a forward push from the rear. It is our opinion Porsche should have re-designed the front clamp when they took away the circlip, bearing and washers from the front.

Chevrolet uses a hollow large diameter tube for their driveshaft found in the C-5 and up Corvettes which also have a torque tube set up. Chevrolet uses a rubber piece in between the engine input shaft, the driveshaft and the trans input shaft to let there be a bit of a twist, or torsion effect, in their driveline with no appreciable contraction. However, I was told they also suffer TBF issues due to heating and expansion due to this different set up and they need to follow certain procedures when replacing a TT, trans or engine.

As far as us having different opinions as to the cause, I would agree. ;)

Great to see you back on the list by the way, a lot of people are having issues with their transmissions!

Kind regards,
Constantine

Hi Constatine,

I think we have 2 different topics here, I am not questioning that your stronger clamp fixes the problem. I am questioning your theory about the cause.

BTW, I am not sure why Porsche put a clamp on the front, I understand the TT is located at the back so why does the TT not float in the flex plate at the front? In the manual the clutch floats on a similar shaft?

I understand people usually release the pressure on a flex plate by releasing the clamp bolt then the flex plate returns to normal and so the clamp slides down the shaft by 2mm or 4 or whatever; the clamp does not need to be levered or pulled - for me that is then equilibrium.

So take the example of an auto in normal use with a flex plate that is buckled by say 2mm and wants to slide down the shaft to the correct position but is prevented by the clamp. Then a condition arises from your theory where the TT shortens by say 0.1mm (I do not imagine that this temporary shortening would be 2 or 4 or more mm?) - why then does the shortened TT pull out of the clamp by this 0.1mm to make the buckle 2.1mm? I would think the flex plate would happily reduce the buckle from 2mm to 1.9mm to accomodate this temporary TT shortening?


Cheers

Marton