 
     |
| Step by step F23 bellhousing swap info? (Page 2/2) |
|
wftb
|
MAY 07, 09:41 PM
|
|
The one I worked on had someone else swapped a LSD in and there was close to 1/16" slop in the differential bearings... because they didn't pay attention.
Thats a lot different than just swapping out bell housings.
|
|
|
|
fieroguru
|
MAY 07, 09:59 PM
|
|
| quote | Originally posted by wftb:
The one I worked on had someone else swapped a LSD in and there was close to 1/16" slop in the differential bearings... because they didn't pay attention.
Thats a lot different than just swapping out bell housings. |
|
Not really. Production tolerances happen. Shim thicknesses are based on the tolerance stackup from the differential carrier, both case halves, and both bearings. If you change a single part, it should be checked, even if that part is a quaife or a bellhousing case half made a decade earlier than the rest of the transmission.
|
|
|
|
pmbrunelle
|
MAY 07, 10:32 PM
|
|
I just had a look in the P-book, and I found the page that lists the available diff bearing shims for the MT2 Isuzu 5-speed.
Thinnest shim offered: 1.08 mm
Thickest shim offered: 1.96 mm
Therefore, we can assume that total variation is less than 0.88 mm. Assuming that the F23 is of a similar design, and that the tolerance stack is similar...
A good amount of that variation can be attributed to the tapered roller bearings... as anything tapered is axially imprecise, by nature.
In the bellhousing, the bearing pocket and mating flange are both machined from the same side, so there should not be a setup change. The distance between both surfaces is supposed to be accurate in this situation. (then again, I'm thinking of things that were supposed to be easy for a supplier to do, and it not happening)
At 1/16" slop, I agree with wftb that there must have been some other screw-up.
Nevertheless, for the home builder, any time you mess with any item in the tolerance stack, you should re-check it.[This message has been edited by pmbrunelle (edited 05-07-2019).]
|
|
|
|
fieroguru
|
MAY 07, 11:42 PM
|
|
The F40 shims have a range of 0.8mm to 1.95mm = 1.15mm = 0.045", which is pretty close to my 1/16" (0.0625") approximation.
I have taken apart several F40s and seen the range of shims used... I wouldn't blindly swap bellhousings and hope for the best...
|
|
|
|
Will
|
MAY 08, 09:00 AM
|
|
| quote | Originally posted by pmbrunelle:
A good amount of that variation can be attributed to the tapered roller bearings... as anything tapered is axially imprecise, by nature.
|
|
Bearings are very precise. You can replace them without worrying about shims.
| quote | Originally posted by pmbrunelle:
In the bellhousing, the bearing pocket and mating flange are both machined from the same side, so there should not be a setup change. The distance between both surfaces is supposed to be accurate in this situation. (then again, I'm thinking of things that were supposed to be easy for a supplier to do, and it not happening)
|
|
Well... that depends. If you look at the tool marks on the 282 case halves, the mating surface where the case halves come together was faced on a large LATHE. If that was a VMC, it may have been able to mill out the bearing pocket without a change of setup, but maybe the Muncie plant did have a setup change in between. We don't know at this point.
Also, such interfaces tend to be designed for a nominal shim rather than zero shim. That way the assembler can use a thinner than nominal shim when manufacturing tolerances dictate. The nominal shim ends up being a cheap shim, which means it's thicker than the minimum required by manufacturing tolerances. ~1/16" would be about right.[This message has been edited by Will (edited 05-08-2019).]
|
|
|
|
Will
|
MAY 09, 08:36 AM
|
|
| quote | Originally posted by Raydar:
Thanks. I had forgotten about that. That might be a consideration, too.
I might just keep the 3.94 / 2.2 box intact, and see what it feels like. I can always do something else, later.
Edit - After seeing that thread, I might just have to run with what I've got.
Still open to suggestions.
|
|
Your two options for a taller final are the 3.74 and then 3.63 used in the Cobalt XFE model over different years, as well as the 3.84 used with the Ecotec once the 2.2 pushrod was phased out. The 3.63 ends up gear-for-gear almost exactly the same as the original 282. Of course with an engine that revs like the 3.4, you could go the other way to a Saturn Vue transmission with 0.81 fifth and 4.41 final.  HHR transmissions have a 4.17 or so final gear as well.[This message has been edited by Will (edited 05-09-2019).]
|
|
|
|