[Paul W]

The ball can't fall out by itself because it sits inside the mainshaft between the two pushrods. But if a previous person has worked on the bike and pulled out the pushrods the ball is easily lost, especially by someone unfamiliar with the bike.

The system is quite simple in design but from past experience, some mechanics (such as me) are even simpler. 

[charte]

Yea, maybe I should clarify my point...…….I blew up my gear box, basically had shards in there.

It's completely rebuilt, and somewhere along the way, I lost the ball.  without it, the clutch rod doesn't even extend for the clutch lever to push it.  I added 1 ball, and I didn't get enough throw.  So I added the second, and that did the trick.  There is a thread on this in the forum here, ACE even weighed in.

So I had a fundamental rebuild which is an obvious change.

It is weird that everything worked, and now it doesn't, and there are no changes.  Like Paul says, it's just 2 rods and a ball, nothing to go wrong.  I supposed these things can wear over time of course, but it doesn't seem like there are a lot of miles on the bike.

Nevertheless, if throw is the issue, I'd be tempted to add one to see if that does the trick.  If it does, I guess a) you could leave it in there like I did, but I don't know the long term consequences of that, so best to get advice, or b) you know something is wrong there and maybe take it apart and see what's up.

[Paul W]

I wouldn't want to fit "ball against ball" as it were  , they will tend to slide past each other and possibly cause a lateral wear problem. It would be better to compare the pushrods against new ones, comparing their length and replace them if necessary.

Some folk recommend cutting an existing pushrod in half, carefully cleaning off and squaring up the ends (important!) then re-hardening them by simple heat treatment. They are then re-fitted with another ball between the two halves, which makes them longer in practical terms. It takes up some wear and possibly prevents lost motion which can occur if there's internal wear in the mainshaft drilling, allowing slight bending of the original longer pushrod.

Problems can also be caused by the plates forming grooves in the slots in the clutch hub. The plates can catch against them, giving a hard or notchy action. Coupled with a stretchy cable this could be why the clutch plates don't release. 

There's a Technical Note about clutch problems in the Hitchcock's website.

 

[Bilgemaster]

Clutch plates were removed? Are you absolutely sure the "dished" plates are now oriented properly? I made that mistake at first when putting my clutch, etc. back together after removing that starter sprag business, with similar results: the clutch wouldn't engage. At first I'd thought maybe that ball between the rods had gone missing somehow, but it was in there. Neither the various service manuals nor the parts catalog I have contain particularly helpful or clear diagrams of the clutch assembly.

[ddavidv]

I put them back in the same orientation they were but I'm going to check them again to be certain.

[Bilgemaster]

For what it's worth, I once described the clutch plates and their various orientations, including excepts from Snidal's Bullet Service Manual, in this earlier posting: https://forum.classicmotorworks.com/index.php?topic=28239.msg322639#msg322639

After all, just because YOU pulled 'em out and kept 'em in the same order doesn't necessarily mean that the last guy in there put 'em in the right way, right?

[ddavidv]

Orientation of all parts is correct.
I measured the pushrod movement and it is 1/16". Sure doesn't seem like much. Can anyone confirm or deny that is correct? There is no measurement in the Snidal manual.

[Paul W]

That sounds like far too little. Ideally it should move almost the same amount as the inner cable at the lever end, given a bit of lost motion due to cable stretch. Have you checked the whole length of the cable? Are the cable ends sitting correctly in the adjusters? It might just need a new one.

The clutch release mechanism is fairly simple but there’s lots of places where lost motion can occur, it’s surely just a matter of carefully checking the whole thing downstream from the handlebar lever. I’d take the petrol tank off and examine everything in detail.

[ddavidv]

Well, this is interesting.
There are no balls in the rod linkage. Now before you say "Ah-Ha!" this is a 1973 which makes it pre-1999. The parts diagram for the earlier bikes do NOT show any balls. In fact, the rod from the clutch cable pivot is one 10" long piece and not two 5" pieces as the later bikes.
So, no balls (hahaha) is correct for this bike.

When I remove the trans cover (clutch assembled on the other side) this is how much of the rod is poking through:


I looked at some videos on YT of guys taking the trans cover off to replace kick start springs and it sure looks like that thing should poke out more.
At the clutch end there isn't really anything variable(?). The short rod and pressure plate look totally fine.

[ace.cafe]

Probably worn down.

Maybe a good idea to just buy the 2 rods and one ball set-up from a parts supplier. Very low cost items.

[Paul W]

I agree, after all it’s a very simple system. If the bike has very low mileage it’s likely that the original part was shorter than standard.

[ddavidv]

Or could I just stick a ball in between the long rod and the release rod?

[ace.cafe]

Or could I just stick a ball in between the long rod and the release rod?

If the ball will be contained inside the tunnel, that should work.
It is a 1/4" ball bearing for the 4-speed Albion box. 5mm ball bearing for the 5-speed box.
Grease it up.

[charte]

Yeah I can confirm that looks too short as well compared to mine. On my 1969 it was also that short so I put in a second ball. Though that may not be completely advised it works like a charm. The fact that you don’t have a ball ha ha and it is clearly about a quarter inch short from the oic I think you can put in a ball and see what happens. It cost me $.89. And now she runs like a champ. I also suspect that Ball provides A serious function by separating those two rods and allowing some spin.  A theory why it worked and now it doesn’t work therefore could be that the two rods were grinding on each other at the end and slowly lost mass. Then they got shorter. I think the ball is in there to prevent those two rod ends from grinding on each other.

[ace.cafe]

Sometimes, the rods will friction-weld to the ball bearing, if the clutch is held in too long at stoplights. That is why they recommend shifting into neutral at lights, instead of holding in the clutch. The ball was put there to reduce the problem from the non-ball system which were worse. The point contact of the ball in the center of the ends of the rods gave minimal friction in the system. But it still got heated with holding the clutch for too long.

We eliminated that concern by installing a ceramic ball bearing instead of a steel one. Ceramic has less friction for less heat, and it cannot weld because it is not metal. Works great!
One of our Fireball tricks.