Sheared Flywheel Keys, ADC, & Electric Starting- a few long thoughts

[2019 edit: the ring gear can drive the worm gear forward or backward it turns out... if & when the surfaces are pristine. I believe the dish in the ring gear allows this]

Sheared Flywheel Keys, ADC, & Electric Starting- a few long thoughts

Hi-

I've been interested in the sheared key problems that x-lites exhibit since R_Little reported the fix about 5 years ago. I have a few observations and thoughts that maybe are not mainstream- jump on in if you disagree.

The sheared key problem is caused by: the worm drive starter.

Hear me out.

All dirt bikes can expect to run backwards a couple of times at least in their lives. By backwards, I mean the piston does not complete the compression stroke, and before TDC- goes back down prematurely. Lets call this a misfire, backfire (that ain't what it is really) or kickback.

Hell, in the '60s & early '70s, you could kickstart a (Pursang? CZ?) at TDC and sometimes it would run backwards, although it did sound funny. And another way for this to happen is for a bike to be rolling backwards (downhill?) and the clutch gets dumped to stop it- which happens more than you'd think.

Now, in the x-lites, this stall/kickback/key shear is caused by a few common factors:

• Low momentum (ie low flywheel weight and either slow ground speed, and/or slow idle speed);
• the ADC- valve lash is spec'd too big or is not working;
• the biggie-> the starter system uses a worm drive.... and worms cannot be driven- they just drive (IOW- it ain't going backwards. I've used this example before- think of a vise: you can't spin the handle by pressing the jaws closed).

The key shears (maybe) when the spark plug fires, but there is not enough momentum for the piston to make it "over the hump" and so gets driven back down (reverse) with much force. The sprag then engages and the flywheel turns the ring gear- which tries to turn the worm (which it can't do, and is like hitting a wall. instantaneous flywheel stop, but the crankshaft wants to keep going). So something gives: maybe the flywheel woodruff key shears, the sprag breaks, the ring gear breaks or the worm shaft gets pushed backwards which breaks the case, or the worm housing mount. The teeth of the ring gear and worm can get galled, too, if this happens. Actually, most of the time nothing noticeable happens ('cept the teeth thing). Rarely, 2 or more of these events happen.

A key shearing is just one of many possibilities happening during a kickback, including nothing. And none of these would happen if the shaft wasn't driving the ring via a worm.

Weirdly, depending on which way the key shears, you might do more damage to other parts of the starter drive system when trying to re-start the bike again with a sheared key (I'm thinking if the ignition is more advanced now). And even when kick starting. Something to think about.

And why did husky design it like this? to reduce weight and size... which may or may not have been a mandate from BMW. And I think they would've solved this issue in a few years had Pierer not bought them.

Whats the solution? well, a one-way clutch on the ring gear would do it. Or the use pinion gears or bevel gears instead of a worm shaft, 'cause driving a starter backwards should not be too big of an issue. This also makes it easy to put a clutch on one of the gear shafts (BTW, a clutch on the worm shaft does nothing). But we're not manufactures, so this is beyond the scope of what we're talking about.

so instead, whats our solution? well, there are a couple of work-arounds: let the key shear is one. cheap, easy but might not save the ring/worm teeth. Most people would not be happy with this (btw, to do this: low torque on the flywheel nut, no Loctite)

probably the best approach is a multi-pronged policy to reduce kickback:

• a high idle speed: 2100+ rpm's
• Tight exhaust valve lash: 0.004" even?
• strong ADC spring which keeps the ADC working at higher rpms (husky has done this)
• heat-treated gear teeth (husky has done this on the updated ring/worm set)
• a good spark plug
• gear your bike down so you are at higher rpms in slower situations
• keep one finger on the clutch when you're in first or second, especially when you're lugging it; pull it in before a stall.
• don't lug it
• learn to modulate the clutch, in low-speed situations especially, to keep the bike from stalling.
• A Rekluse clutch (I'm not a fan, but if you don't have the skill-set to modulate your clutch- WTH)
• don't let your bike roll backwards downhill and then dump the clutch.

I am also of the opinion that you want the ADC working good when you're kick-starting. That gear train seems fragile to me also.

The prototype 340x had a revised starter w/o worm. I think it was raced in 14 as well.

Here is SWM's ... http://enduro21.com/index.php/40-general/318-first-look-swm-rs-340s-enduro

What's cool is they lost the external oil line, so they improved two big issues. Surprised they shelved this.

I don't remember anything about this bike. WEC racing?

Check out the shots from a MX GP in Italy on another thread. Engine still has external oil line and left side exhaust where the SWM has a new right side exhaust... http://www.cafehusky.com/threads/mo...red-with-non-xlite-husky-engines.46437/page-2

Trench

Yes, that appears to be what happens.

You are correct about trying to to start a bike with a sheared key. The timing is way advanced and will kick back for sure! Violently so and I bet that is what is blowing up cases. Maybe the thin woodruff key is a sort of safety valve being the 1st (and cheapest) thing to break?

But let me ask a question:
Can not the starter simply spin backward when the worm gear runs backward? Why does a worm drive prevent this?

The best answer I can give you is the example of a crescent wrench or vise- both easily adjusted by turning a worm. But try to move the jaws so that they force the worm to turn.... you can't.

That's what I meant when I said a worm can drive but cannot be driven.

I agree. If you pay attention, you can hear the starter drive catch the engine when it "rolls back" when you shut it off. While my bike has been pretty much bulletproof, a one way bearing going to a worm gear drive is kind of asking for failure. I'm not sure they really is a solution.

Good analysis. The reference to Bultaco reminded me of something bizarre I saw in the 70's.

We practiced Wednesday nights at a motocross track on a club member's farm. Dave had a Sherpa T 350 trials bike and was a talented trials rider as well as racer. He stumbled on it by accident, then worked out how to make it happen at will - lug up a gentle grass slope to near stop, the bike would "hiccup" and start running backwards with assist from the slope. He would then putt around in reverse, doing figure 8's on the driveway.

yeah.

but really, it's not the sprag ("one way bearing") that's the problem, it's that a worm can not be driven- in either direction. If you want the worm to turn, you have to spin it's shaft.

AFAIK. any mechanical engineers out there wanna edjumacate me?

Here is Howstuffworks text:
Many worm gears have an interesting property that no other gear set has: the worm can easily turn the gear, but the gear cannot turn the worm. This is because the angle on the worm is so shallow that when the gear tries to spin it, the friction between the gear and the worm holds the worm in place.

https://science.howstuffworks.com/transport/engines-equipment/gear5.htm

Yup... Engineer with a physics degree.

great- I've "known" this intrinsically... but never heard or read it from any source. so thanks! nice to have it in writing.

I was scratching my head a year or two ago when some shop claimed the engine was over-speeding the starter motor (when a sprag failed) on the x-lites. huh?? but no one ever said anything.

So that is why the cases blow up rather than just spin the starter motor backward?

Well, I seem to recall when my sprag seized to the flywheel the starter motor turned when I kicked the bike over. I could here it going "shuuuzzzz".

Was that not the starter motor?

That would be my guess. If the one way bearing fails, bad bad things can happen, too.

When I bought my bike the starter motor had been removed completely and the block off plate from the Motocross version was in there. Probably the safest route but I just like having the electric start. With the updated Ddive gears it has been great and I'm hoping it will just continue to work that way without breaking anything.

When did Husky start installing the updated/hardened ring gears?

How would we verify this took place in our TE310s?

I don't know for sure; my best "guesstimate" would be sometime in the early/middle of the 2013 model year run (2014s were made directly after). The ring gear issues seemed to taper off in that time frame.

Again: guessing.

Rich, I'm a little slow here... but I believe the sound you're hearing is the sprag pawls releasing (or trying to release in your case)

Interesting. This is most likely why my bike seized. It worked fine for over a year until I started learning to scrape without pulling in the clutch. 2 days later, the stress of the lugging must've been too much for the connecting rod bearing.

Juice- I doubt that lugging your 510 caused your catastrophic failure. It might've been an oil pressure failure or a defective bearing, but I doubt you did that damage from lugging the engine. Those big blocks are solid motors... plus they have a pinion gear starting system, I believe.

The electric starter system on the xlites uses a worm drive. The worm can not be driven by the ring gear, it can only drive the ring gear. When lugging a bike, the chances of the engine stalling are greatly increased. A stalled engine goes backwards for a very short period- sometimes causing damage to xlite motors when the ring gear tries to drive the worm.