Yamaha XS650 Experimental Clutch Booster (mousetrap clutch)

[TwoManyXS1Bs]

I thought I'd share another project I've been working on. An over-center spring assist for the clutch actuator (worm). My ol' panhead had a moustrap clutch, and I thought that was a quite clever thing. So, wanted see if a similar principle could be used on my XS1B.

The stock clutch actuator (worm) sits all by its lonesome in a cavernous side cover.

A view up into the left engine (drive sprocket) area shows a lot of unused space.

Maybe we can stick a bunch of stuff in there?

 

[TwoManyXS1Bs]

One of the goals in this project is "No irreversible, permanent alterations to precious vintage castings". Within reason, of course.

What if we remove the return spring and its mounting post, then tap threads (#10-32 UNF, or M5x0.8) into the already properly sized post hole, and mount a wheel there.

Note, I'm doing this to a crusty ol' cover that I use for mockups.

 

[TwoManyXS1Bs]

Then, with the magic of hammers and sparks, make a bunch of parts.

Here's the sidecover with the whole collection of parts.

 

[TwoManyXS1Bs]

Assemble the sub-assemblies.

Mount those sub-assemblies into the cover.

Then affix the worm and linkages.

The pic shows the clutch actuator in the relaxed, off, clutch-engaged position.


This pic shows the clutch actuator in the tensioned, clutch-lever-pulled-in, clutch disengaged position.

Here's an askance view of that position.

 

[TwoManyXS1Bs]

The mainspring is barely tensioned (too long) at this stage, just to see if the mechanism works. It's already providing 4 lbs of assist force to the actuator's (worm) arm. The spring needs to be shortened about an inch to get to the operational force.

This video shows the mechanism in action.

 

[TwoManyXS1Bs]

If I intend to run with this thing for awhile, I'll need to make a cover for it, to keep all that sprocket grunge offa it.

Unfortunately, near the end of making this thing, I had an epiphany, and thought up a couple of other designs that should be simpler and more efficient.

It still needs more work, solving some design flaws and tweaking things, before it can be used.
So, I think I'll see this version through, install it, and play for awhile.

 

[gggGary]

wow 2many this is cool! What returns the worm once the assist spring is tensioned up???

 

[YamadudeXS650C]

I am impressed.
And recently, I have not been easily impressed.
However, like Gary said, it seems that some sort of return spring or mechanism is needed to get the worm back in the engaged position.

 

[Brassneck]

So very cool! I imagine you could still feather the clutch when needed, correct?

 

[Mailman]

Very cool 2M! Always fun to see your design ideas come to life. When I had my old Beemer they used to sell an easy pull aftermarket assembly that just added a pulley into the clutch cable. BMW's use a big dry diaphragm automotive type clutch. The stock cable set up pulled directly on a long clutch actuating arm, like so.

Here is the easy pull set up.



I never had one but a friend did, and it pulled as easy as a warm knife slicing through butter!
Moose racing also makes an easy pull clutch assembly that uses a fulcrum to gain a mechanical advantage.
Not nearly as elegant and discreet as your design.


I am not suggesting you employ any of this, I just enjoy innovative ideas.
Carry on! As you were!

 

[TwoManyXS1Bs]

Thanx, guys.

... What returns the worm once the assist spring is tensioned up???

...some sort of return spring or mechanism is needed to get the worm back in the engaged position.

Why, the clutch itself provides the return action.

As measured on mine, it takes about 370 lbs of force on the pushrod to move the pressure plate, fighting its 6 clutch springs. The translational efficiency of the worm mechanism (in clean, good, greased shape) is quite high, about 85%. The worm's ramp angle, plus the length of the actuator arm (1.5" to clevis hole), gives about an 8:1 leverage ratio (0.4" cable travel yields 0.050" pushrod travel), making the force realized at the clevis hole around 50-55 lbs.

My goal is to augment the clutch cable tension by about 1/3 of its 50 lb value, or about 20 lbs. Which would reduce the clutch cable tension to around 30 lbs.

My other experiments with clutch cables seem to show a non-linear friction relationship to cable tension. As the tension goes up, the frictional component skyrockets. I'm hoping for a reverse effect, reduced tension resulting in a dramatic decrease in cable friction, and hopefully more worm travel.

This is a complicated project, balancing sinusoidal leverages with diminishing spring rates, energy release curves, and fitting the whole shebang in there without damaging stuff.

I think I need another nap...

 

[TwoManyXS1Bs]

... I imagine you could still feather the clutch when needed, correct?

Yes, that's one of the design criteria. Providing assistive force in balance with the existing demand, so that it just feels like a weak clutch, feathering and engagement zones unchanged...

 

[Brassneck]

I like it! Sign me up!

 

[TwoManyXS1Bs]

...Moose racing also makes an easy pull clutch assembly that uses a fulcrum to gain a mechanical advantage...

Interesting note about that contraption. All it does is change the leverage up top at the lever, the cable tension remains the same.

By moving the worm arm's clevis hole outward a little, the same leverage change is realized. BUT, you also get a reduction in cable tension, adding to the overall clutch lever force reduction.

Gotta be careful there though. Too much reduction and the clutch pressure plate may not move enuff to disengage the clutch...

 

[YamadudeXS650C]

Thanx, guys.





Why, the clutch itself provides the return action.

As measured on mine, it takes about 370 lbs of force on the pushrod to move the pressure plate, fighting its 6 clutch springs. The translational efficiency of the worm mechanism (in clean, good, greased shape) is quite high, about 85%. The worm's ramp angle, plus the length of the actuator arm (1.5" to clevis hole), gives about an 8:1 leverage ratio (0.4" cable travel yields 0.050" pushrod travel), making the force realized at the clevis hole around 50-55 lbs.

My goal is to augment the clutch cable tension by about 1/3 of its 50 lb value, or about 20 lbs. Which would reduce the clutch cable tension to around 30 lbs.

My other experiments with clutch cables seem to show a non-linear friction relationship to cable tension. As the tension goes up, the frictional component skyrockets. I'm hoping for a reverse effect, reduced tension resulting in a dramatic decrease in cable friction, and hopefully more worm travel.

This is a complicated project, balancing sinusoidal leverages with diminishing spring rates, energy release curves, and fitting the whole shebang in there without damaging stuff.

I think I need another nap...

Did I have confidence that you had considered that variable.? Yes .

 

[xjwmx]

IToo much reduction and the clutch pressure plate may not move enuff to disengage the clutch...

What would happen if you put a little block and tackle in there for mechanical advantage, and lengthened the lever on the worm to give enough (enuff) pressure plate movement? Back where you started? Worse?

 

[xjwmx]

I've noticed on my bike that if you try to change gears with the revs really high, like a racer, it is hard to do due to what feels like dragging. It must be like a torque converter effect where the plates are still coupled by the oil between them.

 

[angus67]

May I suggest bronze bushings for the pivot points? It all adds up

 

[MaxPete]

This is a very nice idea 2M - watching with interest!

 

[TwoManyXS1Bs]

May I suggest bronze bushings for the pivot points? It all adds up

Good point, Angus. The mechanism would be quite sensitive to unwanted friction. Because of the small diameters of the pivots (0.142"), and the high loads, I went with polished stainless, then will probably use some moly grease.

The next iteration I'm cogitating won't have bushing/pivot points, ball bearings instead...