I hope you're meaning 0.8mm play?
The top gear can move laterally up to 0.4mm without issues. The K4 and later engines had a spacer between teh bearing on that end, and the gear, selected to ensure that the engagement of the gear dogs are at least 3.0mm (or more) when shifted in.
With the appearance of the case-mating goo, it looks like the cases may not have been fully bolted together? If so, the final-drive shaft would wobble a bit, which never makes a constant-mesh tranny very happy...when I have to repair one of these where the chain was crashed thru the cases, it takes quite a bit of finessing of the weld-machine-weld-machine cycle to make sure the final-drive shaft stays parallele with the others. If it isn't, then there is a tight-loose-tight feeling to the final drive shaft when turning it. If the engine is run, it makes a slightl growling noise right there, too. That's because the teeth on the final-drive shaft are not mating parallel-ly (is that a word?) with the output of the mainshaft.
Thanks for the great info Hondaman.
It is more like 1mm lateral play of the top gear though.... It does look like the k3 mainshaft (according to cmsnl, no 7? but it's a typo I think, meant to be 17 - thrust washer) also has a washer inbetween the bearing and gear, no? Is it ok to shim this further to reduce the play?
Try the parts fiche at South Sound Honda instead of CMNSL...I have found many errors in the CMNSL images.
The story on the gear is: beginning in the K4 engines (injection-molded cases began at K4) this looseness was finally documented and characterized into a prodcedue (on the production lines). The whole 'picture' is this: when the gearshift fork presses the adjacent (4th) gear into the top gear, it must have 3mm or more of gear dog engagement - this will show as witness marks on the dogs of the moving gear. When the insides of these cases would get too wide, the last gear on the end of the mainshaft could move away from the engaging dogs, and when the engine was hot, it had 2.0mm or less engagement. This makes it wear really fast because the gears then tilt. So, the shim washers came into being (in documentation and parts fiche for the post-K5 models, that is: they were often found anyway in the early engines) in 3 sizes of 0.1, 0.5 and 1.0mm thickness ('No shim' is also a valid thickness). This shim washer went onto the end of the shaft to push the last gear inward a bit to ensure a good 3+mm dog engagement when shifted up.
This tolerance wandering is due to the manual molds they used (even the injection molds were hand-operated), and depended to a large extent on how cooled the cases were (or weren't) when hand-removed from their molds. I have actually seen finger-shaped depressions (wide, like thick glove fingers) in the K0-K3 cases, making me wonder what Superman was standing that close to this steaming chunk of aluminum? I have also seen hand-corrected flat 'dents' where one was picked up, probably too hot, and dropped: the little fins were then quickly scraped back up into a fin shape, but not regular in form like the others. These are seen on the front lower parts of the engines from time to time, on those rudimentary-looking fins. The early cases that didn't "make it" were thrown back into the pot, so to speak, and in many of the early engines you can see the used and new metal mixed together, under the paint!
