final fix - alton starter/yves mod-upgrade

The Tri-Spark (original, installed in '08) is another of the items, like the Alton, that seems to have problems yet mine has been totally trouble free for 13 years. I guess I must be one of the luckiest Commando owners around! Not sure why that should be but, as the old saying goes, "Luck is always better than skill!"

My Tri-Spark was flawless for 11 years until a failed voltage regulator burned it out. The new model is supposedly good up to 24 volts, so that's not likely to happen again.

actually bought two, new, complete assemblies - one for the norton and one for my BSA spitfire project. also went with the tri-spark regulator/rectifier assemblies. my original tri-spark came with the bike when i bought - i figure it was an early design - 2006 vintage.

hope i'm not jinxing myself here, but i think i've FINALLY resolved all the alton starter issues with my Mk2. seems there's a small percentage of us that have alton starter interface problems. personally, i think it's some sort of tolerance stack up between the norton components and the alton parts and assembly. mine seems to be just outside the window of being compatible. after 18 months, two woodruff key failures, a dozen sets of polyurethane shear pins, all seems good. the "Yves fix" is working perfectly but i had to go with SIX shear pins on the starter/sprag clutch interface assembly. a couple weeks now, several dozen starts, and not so much as a hic-cup. for the shear pins, i used an, off the shelf stock from McMaster-Carr, p/n 2656T58, 10mm polyurethane rubber rod, cut into 12mm long pins. as a redundant locking feature, also used a serrated bellville spring lock washer under the crank lock nut, p/n 90895A216, although with the Yves fix, it's really not necessary. if i have any further problems with the starter, i'm out of ideas. HOPEFULLY i've put this issue to bed and i can get back to some serious riding once again.

​

maylar said:

My Tri-Spark was flawless for 11 years until a failed voltage regulator burned it out. The new model is supposedly good up to 24 volts, so that's not likely to happen again.

Click to expand...

I have an 18V zener across mine just in case.

joe czech said:

...I had to go with SIX shear pins on the starter/sprag clutch interface assembly.

Click to expand...

Doesn't this over-ride the shearing "safety" function and potentially put other parts at risk?

grandpaul said:

Doesn't this over-ride the shearing "safety" function and potentially put other parts at risk?

Click to expand...

good question. i'm thinking, IN SOME CASES, alton has recommended bumping up the number of pins used to six. the poly rod i used has a shore hardness rating of 80A, whereas the original alton supplied pins are 95A. i might still have a safety factor. if i use 3 or 4 pins, i could only get about 6-12 starts, before i would shear the pins and have to tear down start assembly for repair. to be honest, after a dozen teardowns, you get pretty discouraged with something that costs that much and doesn't live up to expectations. as of right now everything seems good and i'm going with it. as a side note, i'm wondering what role those pins actually play in the system - a system safety valve or a cushion between the starter and sprag clutch, OR both - ?????

Aren't they an anti backfire device?

Mart UK said:

Aren't they an anti backfire device?

Click to expand...

well, i guess if you consider them as a backfire safety valve, i would say yes. can't say if any of my failures were caused by backfire, but i can definitely say i experienced failure just engaging the starter.

Mart UK said:

Aren't they an anti backfire device?

Click to expand...

They won't/can't stop the backfire, but shear to stop potential damage from a backfire, should one occur.
They are, essentially, a torque limiting device that, unfortunately, limits torque in both directions of rotation.
Unlike the original Mk3 3 device, "slipping" means having to replace "pins".

I added an extra sheer pin at the beginning of last year, prior to that I was getting roughly one failure of the three pins a year.
With the four pins I managed a year without any failures.
Winter strip down of primary to change sprocket and chain show no damage to any component.
I will however rebuild with new pins.

I have an Alton from 2012 - AFAIK, the first year they were in production. The original style rotor failed (exploded) twice - replaced very rapidly by Alton at no charge. They then changed the design of the rotor and sent me that version in early '13, after the 2nd rotor failure.

AFAIK, my system has no plastic shear pins. There was no mention of them in the instructions and I don't recall seeing such things when I installed the Alton/replaced the rotors.

But since I received/installed the "new" rotor in '13, the system has been completely trouble-free and works extremely well. By now I have at least many-hundreds of "starts," if not more. Heck, one day at a meet, I must have started it 30+ times at folks request. It has been as reliable as a Honda! As I said earlier, it's a mystery why there are a variety of experiences - some like mine (no problems), others like Joe... It may be, as mentioned, a tolerance stack-up issue though I have to admit that I can't see how tolerance variation (Norton or Alton) has any affect on the starter system.

MexicoMike said:

...as mentioned, a tolerance stack-up issue though I have to admit that I can't see how tolerance variation (Norton or Alton) has any affect on the starter system.

Click to expand...

MM - this is just speculation on my part, but here's the way i see it -- it's not Norton OR Alton, it's BOTH. the norton components had a +/- tolerance factor. lets say, the notron assembly went together and stack up ended up the shy side. alton comes along, designs and builds with similar +/- tolerances, and alton's final assembly tolerance stack up also ends up on the shy side. even though both assemblies are within design tolerance, the tolerance deviation of the two assemblies may be compounded to the point where the final norton/alton assembly just ain't quite right. this wouldn't happen across the board, but only in a few cases where both assemblies are at one extreme side of a tolerance limit. my hunch is the core issue lies within the alton spacer or a component of the sprag clutch assembly. a fix may be as simple as adding a washer to the crank nut. it's not an alton fault - in fact, it's nobody's fault. the problem is, two separate companies designing two complex assemblies that should bolt together, that were designed 50 years apart. did alton have access to original norton engineering drawings? my guess is no. alton may have designed it on the fly around an existing commando without the use of original norton design data. today, there a science and a spec to all this, it's called Geometric Dimensioning and Tolerancing, but for it to work, it would have to be incorporated by both.

Makes sense, Joe. I was thinking in terms of how tolerance variations could affect the "strength" of the starter sprag but, as you said, it could be that a relatively minor variation in alignment or something similar could have a major impact...

Re "manufacturing tolerances." I recall reading a few years ago on a Norton-related site - can't remember if it was this one - that sometimes the factory had trouble fitting some portion of the frame to other parts due to misalignment. One method used was for a worker to "jump up and down" on the frame section to get it to align with the part(s) being bolted in. This method was related by a person who claimed to have been a Norton worker back in the day. If true, one would conclude that there was a very wide "factory tolerance," at least for frames!

MexicoMike said:

Makes sense, Joe. I was thinking in terms of how tolerance variations could affect the "strength" of the starter sprag but, as you said, it could be that a relatively minor variation in alignment or something similar could have a major impact...

Click to expand...

You mention sprag strength, but my understanding is that the nub of the issue being discussed here is that Alton uses, instead of the self-resetting Belville/ball clutch on the Mk3, a set of plastic shear pins to protect the bike against damage from kick-back. With the Mk3 you hear "click-click-click" then have another go. With the Alton it's time to pull the primary cover off.
I suppose other significant weakness is their choice to pass starting torque through the alternator rotor key which, I believe Yves has eloquently solved. This was the original thrust of this thread wasn't it? (by @joe czech )

But, as has often been pointed out - these "problems" don't occur regularly...or, based on Alton sales...often enough to cause people to not buy them. Surely if more than the "normal" failure rate of mechanical components occurred, the system would no longer be purchased/in production.

MexicoMike said:

But, as has often been pointed out - these "problems" don't occur regularly...or, based on Alton sales...often enough to cause people to not buy them. Surely if more than the "normal" failure rate of mechanical components occurred, the system would no longer be purchased/in production.

Click to expand...

correct! problems seem far and few between. from a piece/part and quality standpoint, the alton kit is excellent, there's something just not quite right between my 74 Mk2 and my alton kit. anyone buying an alton starter, i would suggest, at a minimum, adding a serrated bellville lock washer, thread locker, and bumping the torque to 70 ft.lbs. on the crank nut.

Had my Alton starter on a 71 fastback nearly 4 years 100s of starts, no problems even removed kick start, it got in way of my leg.
Hope I stay lucky.

update and almost two years for the Yves fix. i can honestly say, seems everything is spot on with no other issues. for the poly shear pins, i'm using a slightly softer material due to availability, but increased the number of pins used. still convinced, the alton starter to commando interface has a minor tolerance issue that seems to affect a very small number of bikes. as they say, all's good that ends well.

My woodruff key has just failed on my Alton Starter on my 850 Commando. We spent 3 hours with a disc cutter to remove the sprag clutch assembly...horrible Does anyone have any suggestions to help me when I put the new parts back on? I cannot get the Yves fix. Have purchased new sprag clutch assembly, primary chain, starter chain and new stator woodruff key. Pullers would not remove the assembly as it was unbelievably tight on the crankshaft. The last time before this when I removed it, it came off easily.