'86 915 input shaft seal / guide tube advice

[DW SD]

Guys,
I had a lot of grease and dirt inside the 915. Not sure if it was from the rear main of the old motor or from the 915 input shaft seal? Planned to change the seal. However, the diameter of the splines on the input shaft (closest to the engine) is larger than the diameter at the seal surface itself. I started to remove the guide tube, but it doesn't budge (screws are out of it). Before I went too far, I figured I'd solicit some advice!

I have a few questions:
How do I change the seal? OR do I need to remove the old one?
How do I remove the guide tube? Should I change the guide tube, too? Is this necessary?

Should I do any other upgrades? I have an upgraded centerforce clutch and new TOB as part of the 3.6 install. Chassis is nearly prepped for the install.



As always, thank you!

Doug

[Gunter]

According to Bentley SC Repair Manual 340-9, the guide tube for a 915 needs to be removed with the help of a home-made tool before the new seal can be installed.
It describes the procedure with a picture.
Would a Carrera manual have the same info? '86 still has the 915.
Don't forget to r/r the cross shaft.

[Jim Sims]

If you mean dirt inside of the bell housing (where the clutch and flywheel operate) this is normal and is the remains of clutch facing material, flywheel surface and dirt from the outside making it's way in through the windows in the top all mixed with oil from the engine and transmission.

Make a "bail" out of hardened steel wire; the wire used in heavy duty clothes hangers works well. The bail is a loop of wire with hooks formed on the ends. The hooked ends of the bail engage the holes in the sides of the guide tube. The loop of the bail should project some above the plane of the edge of the 915 bell housing. Lay a strong wooden board across the bell housing and use this as the fulcrum to pry on the loop of the bail using a strong lever (a 1/2" drive breaker bar usually works well as a lever).

The guide tube contains the input shaft seal; note the depth and orientation (which way the sealing lip points) of the old seal, make or improvise a tool to push it out, Clean guide tube then push in the new new seal (Curil-T or transmission oil on the outside of seal and transmission oil on the inner lip) to the same depth. Note the guide tube base is sealed with an o-ring. Replace it. With a soft faced hammer (or a wooden board to protect the guide tube and regular hammer) tap the guide tube back into place. If you note the wear surfaces on the outside of the guide tube (where the release bearing has been sliding) you can flip it over and utilize a second set of surfaces; if the tube is badly worn replace it.

Cheers, Jim

[DW SD]

Gunter and Jim,
Thanks for the advice! I know dirt is normal, but I figure while I've got it apart, I might as well change the seal.

The input shaft is supported by the pilot bearing in the crank center, but how much play should there be in the input shaft (unsupported by the pilot bearing)? I don't have a lot, but there is some wiggle - more than I'd expect. Should there be any? How is the input shaft supoprted - inside the transmission?

Just to clarify: The OD of the splines of the input shaft are larger than the OD at the sealing surface. I'd guess they would tear up the ID of the new seal, the difference is not insignificant. Does the seal just give as the guide tube is driven back into place? To clarify: the input shaft seal resembles any normal CAM or CRANK seal that I've seen.

Thank you,

Doug

[Jim Sims]

Degrease the splines and cover them with vinyl or teflon tape as the seal is being passed over then. Remove the tape and re-coat the splines with grease.

The factory installed tolerance for run-out on the input shaft pilot bearing journal section is .1mm (.004") max. There is no specification for "play" versus side loading. The nearest supporting bearing for the input shaft is on the other side of the final drive (sometimes called the transmission housing) which contains the differential. This bearing is held by a retainer plate and the outer race of the bearing is thermal shrink fitted into the wall of the final drive. If this retainer plate is worn or the outer race is loose and has worn its bore then this can show up as play in the input shaft. To check this bearing race and retainer plate the transmission must come apart.

[Grady Clay]

Doug,

You are asking all the right questions and getting all the best advice.

The only 2¢ I can offer is to install the “dam seal” in the guide tube. Its function is to re-direct any leaking transmission oil past the input shaft seal from contaminating the clutch to harmlessly leaking out of the bell housing.

Check these:
“’72 915 Dam Seal”
Input shaft seal Problem with 915 Mag case
and
Shuddering Clutch 915
and
Replacing 901 Tranny Shift Shaft Seal
and
Early 915 tranny seal Questions

Best,
Grady

[DW SD]

Jim,
I thank you for claryfing! I'd guess the shaft can move 0.050" at the nose given a small side load maybe lifting 10lbs. This is pretty much a wild ass guess. I called this wiggle. I did not use a dial indicator. Having been around various machiens for a long time, I was surprised at the amount of movement, though.

The transmission shifted perfectly, made no noises about six weeks ago, before I pulled the motor and transmission.

This is a 915 with 65k miles, thus far only behind a 3.2 (3.6 is in the works).

Grady,
I appreciate your post and validation of my questions! I began reading several of those threads, which were indeed helpful. I intended to install a dam seal, after replacing the input shaft seal.

Grady and Jim,
Any way you could offer some measurable metric for how much in put shaft play is acceptable given a small side load at the pilot nose?

You think it should come apart to check it? I was planning to have a LSD installed sometime soon and could it be checked then? One installer said he'd isntall the LSD with the transmission in the car, though.

Doug

[Jim Sims]

If you're sure it's moving .050" under only 10 lbs of lateral force that seems excessive to me. Grady has more experience so he may have a different opinion. If reinstalled the transmission must come out of the car again and apart to check/repair the bearing races and retainers. This cannot be done from inside the differential housing. If the bearing race bores in the differential housing are worn excessively this can become a more expensive repair. I'm unsure how a LSD can be checked/setup properly (unless one is simply lucky and everything is the same in terms of size/tolerances) without taking the transmission apart. Maybe the mechanic has somehow worked up a procedure for shimming between the ring gear and differential. Cheers, Jim

[DW SD]

Jim,
Thanks a bunch! Can't thank you enough for your input!

Grady?

[Grady Clay]

Doug,

No No No, you shouldn’t install a LSD with a transmission
in the 911! !! !!!

Replacing the open differential with a Limited Slip Differential
(LSD) requires some careful steps. All are necessarily done
“on the bench” and not in the car. If someone tells you
otherwise – run like hell or be prepared to buy a new R&P
afterwards. I suppose it is possible to do this but it is false
economy.

I can think of very few transmission repairs that I would do
in the car. The minimal cost/effort to remove the transmission,
get it nice and clean on the outside before disassembly and
be able to properly work on it on a workbench makes for
successful results.

If you replace the open differential with a LSD, you should
also replace the tapered roller bearings and set the R&P
as original. You should talk to WEVO.

I have posted several instructions for this. Please review
them. I/we should go back and find the links.


There isn’t a specification of force vs. deflection for the
installed input shaft. I suspect we should make some
actual measurements and possibly develop one. I
propose loading the shaft one direction with 5 N (1.125#),
set an indicator to zero, apply the same force the other
direction and measure the deflection. Let’s specify the
input shaft seal is removed, the transmission is in neutral
and at room temperature.

The difficulty is that the input shaft is fairly flexible in bending.
We need to be able to separate the bending from movement
of the bearing if possible, certainly without damaging anything.

The spec for the radial runout is 0.10 mm as Jim said.



The issue with aluminum case 915s is the bearing outer races
in the main (differential) casting aren’t sufficiently supported.
Over time and particularly under severe load, the press-fit
races become loose in the casting.

Starting with December 16, 1983 production the input shaft
roller bearing was reduced in size (to the 930 Turbo bearing)
in order to allow for more interference fit and stronger
material (red) between the bearings. The earlier bearing
(999.110.032.00) was Ø62 mm OD and the later bearing
(930.302.391.00) was Ø55 mm OD. The real issue was to
better support the pinion bearings. This didn’t solve the
issue but helps.

"
"
© Dr. Ing. h.c. F. Porsche A.G.

This change occurred from transmission numbers:
915/67 -02340
915/68 -02192
915/69 -10081
915/70 -10217
(for reference 915/67/68 have 100 mm CV joint flanges and
915/69/70 have 108 mm.)


If the outer race of the roller bearing is loose in the casting
and the 4-point ball bearing is no longer firmly clamped then
the input shaft can move laterally. The pilot bearing keeps
the nose of the input shaft in place but the support in the
middle is loose. This doesn’t cause an immediate problem
but it is progressive,

The real problem is usually the pinion bearings are also loose
in the casting. This allows the pinion-to-ring contact to not
be as originally set. At best it just gets noisy. At worst it
dramatically shortens the life of the ring & pinion gears.

PCNA seems very proud that they can now offer new
replacement ring & pinion sets – for some $7000!

If this is your problem, Powerhaus II (303-442-4327) has
an inexpensive solution.

Here is the tool “of local manufacture” to remove the
TO guide tube:
“
"
© Dr. Ing. h.c. F. Porsche A.G.


Best,
Grady

[DW SD]

Grady,
I've been in touch with Hayden from WEVO, though not specifically about the bearing issue.

This 915 has 108 mm drive flanges. I believe it is an '86 AL 915, which would have the upgraded 930 input shaft bearings.

I thank you for your detailed response! It is invaluable! I will remove the guide tube by making a tool similar to what Jim and you have suggested. The picture is helpful, though my mental picture was similar, based on what Jim had offered.

Next, I'll see if I can setup something to put a dial indicator on the nose of the input shaft. Are you suggesting only a 1.125 lb force for measuring the deflection?

In parallel, I'll dig through Pelican archives looking for some threads about installing an LSD. I have never taken apart a transmission. I'm not afraid, but also don't have specialized tools for that purpose. This is another case where it would be great to have a spare to "play" with. I'm eager to have the new motor installed and running.

Thank you again to you and Jim.....I'm pleased to have some marching orders!

Doug

[Grady Clay]

Doug,

Get a spare 915 of similar vintage. They are available and that give you a lot of flexibility. There is a huge amount of info on 915 rebuilds in the Pelican and other archives so there isn’t any mystery. Of course my admonition is to accumulate all the technical information. You can make or borrow the few specialized Porsche tools.

Regarding the input shaft movement, the critical issue is to apply enough force to move the bearing yet not permanently deform the input shaft. I don’t know where that point is.

As I have said before, these are engineered and built by mortals and are easily repairable. The best part is we have 20+ years of hindsight and experience.

Best,
Grady

[DW SD]

Grady & Jim,
Finally got around to getting some material (3/16" round bar stock) to build "the tool of local origin" for pulling the guide tube.

I'm awaiting arrival of the rubber o-ring which is beneath the guide tube. I'd just as soon as have everything ready to re-install before I take it apart. Also, I will make some sort of fixture to attempt to measure input shaft side play.

Are the casting numbers in pictures 2 & 3 meaningful? #2 is taken from the outside of the bellhousing looking in towards the input shaft (if you could see through the casting). #3 is 915.116.613.1R

Thank you,

Doug

[Gunter]

Doug:
Here is some input from an old Mech. Engineer with loads of field experience:
It is very difficult to measure wear/play in a roller bearing at the other end of a shaft by deflecting the shaft; even a new bearing will show play when measured this way.
This bearing does not experience deflection when running because the other end is supported by the pilot bearing.
Unless you remove it for closer inspection, it's mostly guessing. Unlike taper-roller bearings, roller bearings are not really designed for side-loading in general.
Without abuse, and good lubrication, a roller bearing will outlast many other components. A new pilot bearing is a must here because of the increased torque coming from a 3.6. Inspect the fork closely for cracks (Magnaflux) and refurbish the cross shaft bushings.
It would be interesting to know the difference between a 915 and the G50 regarding wear/stress. What did Porsche intend/know when they changed to the G50 in '87?

[Jim Sims]

The issue is not wear within the bearing itself (which is likely small) but between the outer race and it's mounting bore (a thermal shrink fit) within the wall of the differential housing and to the bearing retaining plates. If the outer race is loose and turning then there will excessive lateral deflection (play) and run out and more as it further wears the soft aluminum.

The numbers appear to be just casting numbers; the important numbers are stamped on the bottom of the transmission where the may be worn off by inadvertent contact with the ground or objects near the ground.

Jim

[DW SD]

Jim & Gunter,
Thanks for the additional information. Gunter, your point does make sense. I'm really seeking something that suggests the free play is out of whack. If the bearing race is loose in the case, the free play at the end of the input shaft should be acute and abnormal.

Doug

[Gunter]

Industrial application for reducers/transmission:
All outer races sit losely with a sliding fit in the bore and are held with spacers/shims.
The inner race has a shrink-fit on the shaft obtained with the help of a little heat from an electric bearing warmer (Preferred method) or a propane torch.
Unless the bearing freezes up due to contaminants, there is little chance that the outer race will turn and damage the bore.
Besides, everything is submerged/lubed in oil.
Example: The cams (Steel) run in the cam towers (Alu-Alloy) without any bushings, or bearings.
Clean oil is the key.
In over 30 years of field-experience, I have seen a lot of worn bearings due to some breakdown/dirt/debris and very, very seldom an issue with the outer race damaging the bore.
Unless you found steel chips in your 915, why worry?
What makes you think that the outer race is turning in the bore?
Is there a history of this happening with 915's?
To check for any serious radial deflection/play, you would have to duplicate the scenario by shrinking a used bearing of the same size onto a shaft of the same length and find a way to hold the outer race solid. Then you would be able to determine the deflection/ play in the rollers.
Is it worth the trouble?
A phone call to JW on this issue would be better time spent. He has seen/repaired many, many 915's and would know.

[DW SD]

Gunter,
Thanks for sharing your real world experience. There is a history of 915s input shaft races coming loose in their bores. The Porsche factory went to a 930 input shaft which was smaller, but allowed more "meat" / material in the bore. According to Grady and Hayden, several products address the issue.

My own input shaft seemed very "loose" as I described early in this thread, which catalyzed suggestions for evaluation. I've assembled many machines over the years and just didn't "feel" right. Not an expert on Porsche transmissions, I asked for help and comments.
Getting another opinion is always a great idea!

Thank you,

Doug

[Jim Sims]

"All outer races sit losely with a sliding fit in the bore and are held with spacers/shims."

The 915 bearings under discussion utilize a thermal shrink fit for the outer race into the bore of the final drive housing wall. This is per the factory shop manual.

"What makes you think that the outer race is turning in the bore?
Is there a history of this happening with 915's?"

Yes, I had five 915's apart and 8 out 10 of the outer bearing races under discussion were spinning in the bore.

[Gunter]

Wow, experience counts; never had a 915 apart.
Now I have something to worry about.