the shift forks and rails are one of the variables when it comes to these cases. whenever i build one, i try the fork on the rail without any o-rings on the rail but the case set-screw tightened down to see where the detents and fork end up as far as the high,neutral and low is concerned. i've seen a few where the casting of the fork actually hit the case on the shift rail end. actually i've seen a few of them, some hit in low, some in high gear. i've also seen a few forks where the ends were warped during the heat treating process. if the high-low collar cannot wiggel a little from side to side, then there's excessive fork warpage.
to avoid these issues, i have a few forks and rails on hand. i've even welded up the hole in the shiftrail ( where the set screw locates it in the case ) and then ground a new hole in the rail to get the fork to end up in the proper location.
..
Looking back thru this thread, I see that you mentioned the fork possibility when the problem arose last year. I did not want to comment on the fork before speaking with AA about it, and confirming my observations.
1. The fork supplied with my case was indeed NOT heat treated. The new fork clearly is. So that should help....ME! What about the others who received units from the same time frame?
2. Because the shift fork was not heat treated, it acted as the fusible link, and apparantly ALL the wear was on the fork; the sleeve still looks like new, and holds to factory tolerances with the new fork. Whew!
3. Upon the initial test fit of the new fork, I observed as you note, that the forward edge of the fork was butted against the front end of the opening in the housing, raising the question of a bias on the detent. Not having EXTRA forks to play with, and wanting to address this in a way that would be helpful to all, I thought it would be appropriate to shave approximately 1/16th inch off of the side of the fork, which I did.
4. Having now created clearance between the fork and the side of the housing, I was able to determine that the fork's position in hi gear [the one that failed me] was now centered on the detent...or was it?
5. Here's where it gets interesting. After determining that the fork was no longer being 'influenced' by the housing, I turned my attention to the shift collar. I was surprised to find that even with the inboarded detents on the AA shift rail, the hi range detent still had the collar pushed so hard onto the hi speed gear that the collar was pushed up hard against the shift fork!
6. After having had a suggestion from AA about backspacing the shift rail on the case [yes, my era of case still uses the locktab, NOT the set screw] I installed a 1/16th flatwasher behind the locktab to index the rail back in the case.
7. This newly indexed position completely alleviated the side pressure on the shift collar in hi gear. Unfortunately, it also pushed the collar too close to the lo range gear when in lo. Not as bad as the hi range bias, but definitely not in a 'sweet' spot either.
8. As fate would have it, because I've disassembled and rebuilt over a hundred t-cases over the years, in my boxes of stuff I have some of the old Toyota hardware that includes the thin wave washers Toyota used to use before going to the split washers. And wouldn't you know it, when I replaced the flat washer with the wave washer, I achieved a 'neutral' collar position in both hi and lo!
Of course, it would require the extra work Georg has suggested to deal with this on the set screw version. So in a way I'm thankful for having the locktab version. And I've got another hunch about that particular aspect of my case that I will keep to myself for now.
I will start the re-assembly tomorrow.
Thanks to Matthew at AA for letting me bend his ear with my ideas.