First gear jumps at high RPM

maylar said:

Yes, but if you buy the bushing separately and press it in it'll be undersized and need reaming.

Click to expand...

I pressed the old bush out using a vice and spacers, pressed the new bush in and it didn't require reaming.

L.A.B. said:

I pressed the old bush out using a vice and spacers, pressed the new bush in and it didn't require reaming.

Click to expand...

I've done it twice in 40 years and both times the bushing needed to be reamed. A new gear with bush installed fit perfectly.

maylar said:

I've done it twice in 40 years and both times the bushing needed to be reamed. A new gear with bush installed fit perfectly.

Click to expand...

With mine it was a brand new layshaft , the pinion did fit but it was too tight to just leave so I reamed it very slightly

Also check that the shift arm (that protrudes through the intermediate plate) doesn't bottom or top out in the rectangular window of the intermediate plate. I've seen this problem and have had to file the top or bottom of the window to clear the arm.

I made a small stepped piece in the lathe for pushing out the old bushing and installing the new one. I use a vice to do the pushing. Works great. Have not needed to ream the bushing

To jump out of 1st gear, the engaging layshaft sprocket has to move to the left, ànd the camplate has to turn clockwise.
If you partially assemble the box ( pic.) you can notice 2 things:
1. The dowel that locates the selector fork sits not fully at the end of the groove (slot) in the camplate.
2. It takes little effort to make the camplate turn ( out of gear) by simply pushing on the selector fork.
It is only held in position by the tension of the plunjer spring.

I believe this is a design flaw.



It can be improved by grinding a small pocket ( 0.5 mm is enough ) at the end of the slot. ( 1) and by removing some metal ( 2 mm) at 1st gear indent on the camplate (2)
This will make the camplate turn a little further anticlockwise and make for a more positive location of the selector fork.

Blue line is the position of the selector fork spindle. Arrow is the direction of the selector fork when jumping out of 1st.

Doing this will also make the ratchet plate turn further clockwise, so about 1 mm must be removed from the upper step of the ratchet plate( 3), or you could get stuck in 1st:



To set the ratchet spring:
Hold, or clamp the outer cover in a vise in the same position as it is on the bike.
The gearshift pawl will fall under its own weight on the lower leg of the spring.
Bend the lower leg so that the pawl does not touch the ratchet plate.
When you move the rachet plate, it shouldn't touch the pawl.
Then bend the upper leg of the spring so that it just clears the pawl.
The less clearance the better.
The slightest movement of the gear lever must make the pawl tilt.
Operation of the pawl can be checked through the cover opening.

While in there, see that you have the later (left) version of the stop plate.
( wider opening)

The design of the shift fork is to move the gears into engagement.

At that point the gear should not be trying to push back out of engagement.

If the gears do push apart under heavy load -because of wear on the dogs or lack of engagement depth- and the shift fork is under pressure holding the cogs engaged, then the fork will be burned or bent. The fork to gear interface is not designed for constant pressure.

Most gearboxes are designed so that pressure on the fork will push the trans "out of gear" rather than damaging the fork.

ludwig said:

To jump out of 1st gear, the engaging layshaft sprocket has to move to the left, ànd the camplate has to turn clockwise.
If you partially assemble the box ( pic.) you can notice 2 things:
1. The dowel that locates the selector fork sits not fully at the end of the groove (slot) in the camplate.
2. It takes little effort to make the camplate turn ( out of gear) by simply pushing on the selector fork.
It is only held in position by the tension of the plunjer spring.

I believe this is a design flaw.

It can be improved by grinding a small pocket ( 0.5 mm is enough ) at the end of the slot. ( 1) and by removing some metal ( 2 mm) at 1st gear indent on the camplate (2)
This will make the camplate turn a little further anticlockwise and make for a more positive location of the selector fork.

Blue line is the position of the selector fork spindle. Arrow is the direction of the selector fork when jumping out of 1st.

Doing this will also make the ratchet plate turn further clockwise, so about 1 mm must be removed from the upper step of the ratchet plate( 3), or you could get stuck in 1st:

To set the ratchet spring:
Hold, or clamp the outer cover in a vise in the same position as it is on the bike.
The gearshift pawl will fall under its own weight on the lower leg of the spring.
Bend the lower leg so that the pawl does not touch the ratchet plate.
When you move the rachet plate, it shouldn't touch the pawl.
Then bend the upper leg of the spring so that it just clears the pawl.
The less clearance the better.
The slightest movement of the gear lever must make the pawl tilt.
Operation of the pawl can be checked through the cover opening.

While in there, see that you have the later (left) version of the stop plate.
( wider opening)

Click to expand...

comnoz said:

The design of the shift fork is to move the gears into engagement.

At that point the gear should not be trying to push back out of engagement.

If the gears do push apart under heavy load -because of wear on the dogs or lack of engagement depth- and the shift fork is under pressure holding the cogs engaged, then the fork will be burned or bent. The fork to gear interface is not designed for constant pressure.

Most gearboxes are designed so that pressure on the fork will push the trans "out of gear" rather than damaging the fork.

Click to expand...

Ok, maybe I shouldn’t have called it a design flaw. ( well, at least I didn’t use the word “error”..)
But for me it is a matter of priorities.
I’ll try to explain.

STD camplate, with guiding pin in 1st gear position.
( selector fork on the wrong side):

Contact point is at 8 o’clock, so the thrust of the pin is at an angle of roughly 30 deg.,
forcing the camplate clockwise and slipping out of gear.

Modified camplate:


Contact point is at 9 o’clock, so the thrust is in line with the selector shaft ( black line) .
Jumping out of 1st is now almost impossible, but like you said, with a risk of overloading the selector fork.

I am not proposing this as a fix for jumping out of first, but as an added security.
Of course minimizing end float, bushes.. should come first,
but even the best box can occasionally jump out of gear.
Because of the type of roads I prefer to ride on, this is a chance I am not willing to take.
Dropping the bike in a tight corner or over-revving could do more damage than a bent selector fork.

Excellent information, all around.
While the selector modification may seem overkill, and probably on the high end of skillset for garage mechanics, it is well thought out and similar to well-proven mods in H-D Sportster 4-speed selectors which had shift/popping out problems when racing. Personally for a street bike if my gears were 90% and not well-worn, I'd likely stop short of this much work. Great to see it, though.

I agree with this ^^^

I have had the issue of popping out of first under hard acceleration twice in the 44 years of ownership. Both times it was cured with a new first gear layshaft bush. The last time (this year), I also shimmed the kick start shaft... because I learned about that here.

This forum is a great resource. Thanks for your insights, Ludwig.

ludwig said:

a design flaw.

Click to expand...

One of the first issues that came up for me on a low miles 74