Skip to main content
000000 000001 000002 000003 000004 000005 000006 000007 000008 000009 000010 000011 000012 000013 000014 000015 000016 000017 000018 000019 000020 000021 000022 000023 000024 000025 000026 000027 000028 000029 000030 000031 000032 000033 000034 000035 000036 000037 000038 000039 000040 000041 000042 000043 000044 000045 000046 000047 000048 000049 000050 000051 000052 000053 000054 000055 000056 000057 000058 000059 000060 000061 000062 000063 000064 000065 000066 000067 000068 000069 000070 000071 000072 000075 000078 000081 000084 000087 000090 000093 000096 000099 000102 000105 000108 000111 000114 000117 000120 000123 000126 000129 000132 000135 000138 000141 000144 000147 000150 000153 000156 000159 000162 000165 000168 000171 000174 000177 000180 000183 000186 000189 000192 000195 000198 000201 000204 000207 000210 000213 000216 000219 000222 000225 000228 000231 000234 000237 000240 000243 000246 000249 000252 000255 000258 000261 000264 000267 000270 000273 000276 000279 000282 000285 000288 000291 000294 000297 000300 000303 000306 000309 000312 000318 000321 000324 000327 000330 000333 000336 000339 000342 000345 000348 000351 000354 000357 000360 000363 000366 000369 000372 000375 000378 000381 000384 000387 000390 000393 000396 000399 000402 000405 000408 000411 000414 000417 000420 000423 000426 000429 000432 000435 000438 000441 000444 000447 000450 000453 000456 000459 000462 000465 000468 000471 000474 000477 000480 000483 000486 000489 000492 000495 000498 000501 000504 000507 000510 000513 000516 000519 000522 000525 000528 000531 000534 000537 000540 000543 000546 000549 000552 000555 000558 000561 000564 000567 000570 000573 000576 000579 000582 000585 000588 000591 000594 000597 000600 000603 000606 000609 000612 000615 000618 000621 000624 000627 000630 000633 000636 000639 000642 000645 000648 000651 000654 000657 000660 000663 000666 000669 000672 000675 000678 000681 000684 000687 000690 000693 000696 000699 000702 000705 000708 000711 000714 000717 000720 000723 000726 000729 000732 000735 000738 000741 000744 000747 000750 000753 000756 000759 000762 000765 000768 000771 000774 000777 000780 000783 000786 000789 000792 000795 000798 000801 000804 000807 000810 000813 000816 000819 000822 000825 000828 000831 000834 000837 000840 000843 000846 000849 000852 000855 000858 000861 000864 000867 000870 000873 000876 000879 000882 000883 1.slide1 2.slide2 3.slide3 4.slide4 5.slide5
English French German Italian Spanish

ES2 Clutch hub

Believe it or not, I'm still trying to get my rebuilt ES2 to behave properly.

I noticed that the primary chain tension changes as the gearbox main shaft rotates.

At first I suspected the clutch sprocket/chainwheel but on inspection it looks ok.

I found that the clutch center/spider is a loose fit on the splined gearbox main shaft, can be rocked side to side (when apart). So when pulled up tight with the nut, it locks into a out-of-alignment position. The shaft measures ok with a micrometer- across the spline outer surfaces. So the wear (?) must be in the center/spider inner splines, difficult to measure.

But how can this be? There is no movement between these two when assembled, how can one wear so much that it rocks on the shaft?

Unless of course that one of the components is from another model which wouldn't surprise me with this project.

Any ideas?


Check previous posts , your spider probably has failed!!,  Get a new one before it rides up the shaft and damages it and ruins your gear selection as it interacts with the sleeve gear face.. Look very carefully at the shoulders in the spider and compare with a new one from RGM  which is a better design . Use a very close fitting  strong washer under the clutch nut. Do not  overtighten 40 ft pds? + a little locktite. For a perfect clutch I measure clearance and fit 1 thou oversize rollers in the race  to reduce wobble.  Once the spider goes up the shaft  you may struggle to get the clutch off without a proper puller. One thou over rollers need 3+ thou clearance to fit.


... for this Robert, but what do you mean the spider has failed? In what way?

I can understand that it could be damaged by overtightening and may ride up the shaft but where does the clearance which allows rocking come from?

I will check previous posts.


It may be you have a missmatch of parts. I would have expected the clutch center to  run concentrically with the shaft. But it is possible for a  partially collapsed  spider register to climb over the spline. (as happened to me.)  .  


After some thought, It could be that the rock you are finding could be  clutch rubbers have partially disintigrated  ( a regular thing) and the center is no longer being braced to the spider and can rock. A good clutch action is  essential if you are to enjoy your bike ,well worth a full stripdown.


Hi Richard,

You haven't mentioned it, but I take it that you have checked that the gearbox mainshaft is straight? Doesn't have to be bent by much to have a big effect on chain tension.

Hope it isn't for your wallets sake!

Regards, George



... yesterday I did a lot more measuring, the following is of interest:

....clutch rubbers are new

... main shaft is out of true at end by 6 thou giving an oscillation of 12 thou.

... gap between rollers (bearings) and the inner face of their rolling track is 5 thou total, ie 2.5 thou each side

... gap between shaft splines and hub spline recesses is 6 - 8 thou.

... displacement of clutch back plate at the outer tip axially when rotating is about 5 mm.

... displacement of clutch back plate radially when rotating is about 3-4 mm.

My rough calculations show that the axial displacement of the outer tip of the back plate caused by the slop in the splines only will be about 50 thou (1-1/4 mm.), and about 100 thou (2-1/2 mm) when adding in the shaft bend.

I'm going to mark the end of the shaft so that I will see if fitting the hub to the shaft in different locations reduces or increases the run-out (the two different effects should add or subtract together if fitted differently).

I cannot see anything wrong with the clutch center/hub, nothing seems to be wrong to my eye - photos attached.

I think that maybe the problem lies with these two causes, but I still don't see how the center can be so sloppy on the shaft splines.

Thank you both for your help, next installment soon.!



As you say, the spider register looks good.  The roller race could easily cope with some +1 thou rollers which can be found from bearing suppliers (may have to buy 50).  Chains can wear unevenly to give a tight spot.I also see some bad wear on the back of the spider where its getting too friendly with the end of the sleeve gear. I ground some metal off my spider before i realised the problem was in the register. Thought!!, could the spider be tightening to the sleeve gear before the register contacts the end of shaft? A close fit washer inside the spider at the register may help ,but I'm also thinking something may be amiss with the shaft position , incorrect assembly/wear inside the box or missing/wrong spacer. If so get the shaft pressed true while in there. The spider /clutch is not supposed to axially load up the sleeve gear.


Hi Richard,

I had a similar problem when I rebuilt my dominator.

When I tightened up the nut on the spider the spider ran out of true.

This resulted in the clutch chain varying in tension as the drum rotated.

I ended up buying three new spiders from different suppliers before finding one which ran true.

From what I recall the good spider came from Andover Norton.

I put the problem down to poor manufacturing tolerances from some suppliers.




I had the same problem as Ross. One spider was worse than a clowns bicycle when spun on the shaft in the lathe. A second was only marginally better and also rejected.

Last year RGM were selling these on Ebay as "seconds" with a slight wobble at very reduced prices.

The spline was not central, axially and radially.


......for your useful comments.

You are correct Robert, I think that is my main problem.

The shaft out-of-true may be compounding the problem.

Another thing, turning the gearbox shaft by hand - it tightens up over about 1/4 of a full turn, is fine and loose for about 3/4 of the turn. Can this mean the outer bearing is out of line?  It won't explain the slight bend in the shaft.

Today I'm going to take the inner case off, try to get a better idea of how the center is fitting against the sleeve gear, to check that the two are not binding. It will also allow me to use a dial gauge to get better measurements for everything.

Which supplier would be best for a new shaft and center hub? Nv, AN, RGM, W?

Being in South Africa, I must make sure that if I buy new items, they fit together properly.

Regards Richard







If you have the spider out of the clutch you can bolt it onto the shaft and torque it up  and check with a feeler for clearance .  There will be someone who says its not important, but I suffered with poor action untill I solved this. I use RGM as they  post fast and have been fair to me.


Had a similar issue on a Manx, although the splines were quite a good fit, i seem to remember tightening the spider onto the mainshaft, machining the register at both sides of the spider true and then sleeving back to size. The splines are not there to aid true running, this is achieved by the OD of the shaft fitting well to major diameter of the splined hole in the spider. You need to straighten that mainshaft though. Check also that the mainshaft is abbuting the face at the bottom of the splined hole in spider and not jamming on the spline cutter runout


Yesterday I spent a lot of time measuring and checking things.

First, the gap between the face of the spider/clutch backing plate to the sleeve gear face without any washer/shim is about 35 thou. However a shim had been fitted behind the spider of 1/16”  (62 thou) thickness, so the gap was – 27 thou. The spider was tightening up against the sleeve gear – why they didn’t lock solid I don’t know. The tightening up as the shaft/hub were rotated as mentioned above has gone.

Second, I measured the shaft for out of true, it has a bend in it of 0.32 mm (12 thou) giving an oscillation of 24 thou.

Thirdly, I made a mark on the end of the main shaft, made a mark on the clutch center/hub, fitted the two in line, tightened the nut reasonably and measured the distance from the outer edge of the clutch backing plate to the engine mounting plate. Rotating the clutch hub/shaft assembly, I measured this at 12 positions of the hub.

I then removed the hub from the shaft, refitted it at 90 deg to the shaft  mark, tightened the nut and measured again.

I did this with the hub/center at 180 d and 270 d to the shaft.

The results show that with the shaft/hub aligned at 0 d the runout is 2.6 mm.

Aligned at 90 d it is1.6 mm, at 180 d 0.9mm and at 270 d it is 1.5 mm.

This confirms my suspicions that the bend in the shaft either adds to the offset in the hub or opposes it depending how the two are aligned. The peak runout in each case occurred at the same position of rotation of the hub irrespective of where it was mounted to the shaft.

The offset in the spider/hub is greater than the offset due to the shaft bend

This would explain why when I first installed the clutch, this large offset was not present.

I will now remove the shaft from the box, check the straightness, see if it can be straightened.

I think the spider cannot be corrected as, in addition to it fitting loosely on the shaft, it seems to be misaligned, judging by the posts above.

I suspect I will need to buy a new shaft plus spider (to ensure they fit correctly and are true)

Thank you all for your invaluable tips.

Wait for the next exciting episode!


Don't forget to look for some O/S rollers.  At a push you may be able to fit some +2 thou ones. But it will be a gamble.  I looked long and hard before I found some at a reasonable price.  Buy 50 and sell half to someone else!.  I think you are going to finish up with a lovely gearchange and a snick into neutral if you carry on the  way you are working.


.... interested, I have put the results of my measurements on the four graphs attached.

It can clearly be seen that the pattern of the run-out of the clutch is constant with the hub rotational position - always at 1,2 o'clock vs 7,8 o'clock irrespective of the shaft position. The rotational position of the shaft either adds or subtracts from the run-out.

When the hub is positioned on the shaft with the marks opposite the run-out is least, but still there (180 deg, 3rd graph).

This confirms that the inaccuracy in the hub/spider is creating a worse situation than the bend in the shaft.  It also shows that it is not a dent in the clutch plate face which is causing  the run-out, the hub/spider is misaligned.

I sort of knew this before, but it's nice to see it laid out, I enjoy doing it.

So I'm in for a new hub/spider, maybe I can straighten the shaft.

Thanks again for your help.



I have now measured the shaft and confirmed that there is a bend in it – about 8-10 thou giving an oscillation of 16-20 tho. The opinion is that it can be straightened.

However, before I do, I need to know if the splines on the clutch end of the shaft are ok.

It would help me enormously if someone can tell me if the measurements below match a good shaft and hub/spider which someone may have. There is no point bending the shaft straight if the splines are worn.

(There is some slight wear on one or two of them, but I don’t know if it is acceptable.)

The outside diameter of the splines on the clutch end of the shaft is 20.56 mm.

The id of the matching groove in the spider/hub is 20.65mm.  A difference of 0.09 mm (4 thou), enough to allow a noticeable  rocking of the hub.

The hub can also be moved (rotated) on the shaft slightly.

My dilemma is – which one is badly out, or is it both?

I look forward to your reply.

Thank you.


Hi Richard,

Splines can be ‘top fitting’ as you have measured or ‘bottom fitting’ where the bore of the spider fits the bottom of the spline on the shaft. I don’t know which is correct so try measuring the opposite way.  There shouldn’t be any appreciable rotation though. Regards



...with the information about size of shaft at the spider splines, and the corresponding id of the spider?

Thank you.


Accurate measurements in a lathe showed an offset of 11 thou (22 thou oscillation). in the shaft. It has now been straightened (how does a gearbox shaft get bent??) in a 5 ton press, giving an offset of 2 thou, (4 thou oscillation)  (We decided not to go any further!)

Now, with the clutch bolted up to the shaft and in position in the box, the radial run out of the backplate is 15 thou (30 thou oscillation).  Axial run out is also 15thou.

I think I could leave it as it is now, it would probably be useable but as it is apart, I'm going to strip the hub to make sure all surfaces are clean and the rubbers are fitted properly.

Thanks for all your help.


Norton Owners Club Website by White-Hot Design

Privacy Policy