Questions? Looking for parts? Parts for sale? or just for a chat,

The WD Motorcycle forum

WD Motorcycle forum
Start a New Topic 
Clutch conundrum

The BSA single spring and triumph 3 spring have friction plates inserted first and these then bear on the chainwheel and essentially the clutch centre and chain wheel are one lump under spring pressure.
The BSA 6 spring is similar in that the plain steel plate, inserted first, bears on the cork inserts in the chainwheel.

The triumph 4 spring version has a plain plate inserted first that rests on a shoulder on the clutch hub and “floats there with no direct contact to the chainwheel (no corks)? All the drive is then through the tangs on the plates.

Any ideas why this should be?

I have a 4 spring in my M21 which is a delight whereas the one I got for my B31 is a complete pig, driving my nuts!


email (option):

Re: Clutch conundrum

hi Pete,
this might help, or not...

a few years ago a friend and I were talking about clutches and he showed me how much of a difference it made when the clutch cable was routed correctly versus when routed even slightly wrong. The difference was as big as the difference between a single spring and a BSA 6 spring clutch.

With a Triumph 4 spring clutch and correctly routed cable, I could pull the clutch in with one finger. So it might be worth looking at the clutch cable to see if it needs oiling, re-routing or replacement.

email (option):

Re: Clutch conundrum

Have you got the right clutch springs?...For Triumph clutches there are two types..A short heavy one and a longer lighter one...You need the longer lighter one for the singles to get the nice light operation...Ian

email (option):

Re: Clutch conundrum

Yes, the various designs of the clutches are very interesting. All will have its pro's and cons I think

One thing that I realized a while ago is that the tongues of the plates wear the most from rocking in the basket not when the clutch is operated but when the direction of power is changed from acceleration to braking, visa versa.

When the chainwheel is outside the clutch packet, like the WDM20, the drive is always fully on the outside tongues of the plates, regardless of the clutch is operated or not. The tongues having a hard life on both sides as they freely rock in the basket when power changes direction.

When the chainwheel is included in the clutch packet, like Norton, BSA 6 spring, etc. not all of the power will have to be taken up by the tongues as part will be taken up by the friction on both side of the chainwheel.

So less force on the outside tongues but they still will rock in the basket when power direction changes as the clutch will slip a little until all the tongues are contacting the other side.

Cheers, Michiel

Re: Clutch conundrum

thanks for the replies.

Yes, I am using the softer springs and the cable is clean and routed OK. It's not that the clutch action is heavy on the problematic B31, just it has a tendency to crunch the gears. I think I may need to clean the plates again (already checked for flatness) and try a lighter oil in the primary.

Anyway, my main question in my post was more about the design of the 4 spring clutch and why it is different to the others. Why the "floating" first plate? Presumably there was a reason why, although, as Michiel says, the tangs would seem to have a harder time than in the other clutch designs.

I remember reading, some time ago, the suggestion to grind off the shoulder on the clutch centre and then use an extra friction plate inserted first. Any opinions?


email (option):

Re: Clutch conundrum

When you adjust the clutch, do not do it as per the BSA MANUAL UNLESS YOU ARE 20 STONE and you load the bike with another 20 stone .
The factory settings are way too tight.
The sprigs need to be just tight enough so the clutch does not slip.
Then there is the leverage ratio.
If the measure the levers there are several different lengths and that makes a bg difference as well.
My B40 & A65 are both 2 finger clutches

email (option):

Re: Clutch conundrum

I'm pretty sure I will get to the bottom of the B31 clutch problem. It's the same cluch as in my M21, and the Triton that I've had for more than 40 years, so I have pretty much got the hang of setting them up by now and they are all light to operate and free of slip or drag, just the irritating noises on the B31 - which was OK with the old 6 spring...maybe I should fish it back out of the duckpond and refit that.

My question is, as before, why the design of the 4 spring with that shoulder?

email (option):

Re: Clutch conundrum

Hi Peter, I never had a 4 spring clutch in hand but had a look at the exploded view.

If the shoulder was not there the plates would push on the chainwheel. So the chainwheel would then have to take up the spring pressure. Think this would mean:

1. a different kind of bearing which could handle this axial pressure, like the BSA Sloper.


2. or a strong back plate to take up the pressure, like the Norton and the 6 spring BSA.

Why Triumph chose this design? Don't know.....Think the clutches with the back plates are the most simple and logic ones.....

Cheers, Michiel

Re: Clutch conundrum

The fabricatded clutch drums were not particular stiff so they needed the backing plate.
THe backing plate is what made the drum run true and they bend or wear where they sit against the flange allowing the clutch drum & thus sprocket to wobble which can throw the primary on the down change ( OK the chain has to be too slack for this to happen ) .
Because of the wobble . particularly when the clutch is the dissengaged position also puts a lot of stress on the side plates of the chain.

When they changed to a single piece cast drum it is a lot stiffer and does not require any stiffening.
IT usually will run true engaged or dissengaged.
And yes I do understand tour point about the transfer of drive from the face of the stack to the face of the drum
Where as the latter design transfers all of the power via the tangs.
You might like to notice that the tangs on the 4 spring clutch are 4 times as thick as the tangs on the singe spring clutch.

You also need to remember that both BSA & Triumph made their owns clutches because they were manufacturers where as the rest of the British industry were primarily assemblers using bought in parts.
When Burman announced a new clutch, they had to match that to maintian sales particularly if the press of the day continually printed pages of praise for the new clutch.
Generally the "NEW" clutch was a copy of some one elses changed just enough as not to infringe on patient.

email (option):

Re: Clutch conundrum

Hi Trevor, thanks for your interesting reply. nice to share views.

But think I can't agree with much on your first alinea :relaxed:

Don't think the backplate was needed because of weak drums, it is just a whole other principle of design with the chainwheel as part of the clutch packet.

In case of the 6 spring BSA clutch the backingplate is very weakly supported so that will bend, wear and wobble.

But the Norton clutch has a very strong backplate firmly attached to a large flange/clutchcenter. Not much to bend and wobble here.


The chainwheel might wobble a bit when the clutch is dissengaged because it comes loose from the backing plate but there is no power transmitted when it runs free. As soons as it starts to catch and start transmitting power it is pushed to the backing plate again and running true again. Would that really stress the sideplates of the chain that much?

Think there must be another reason/advantage to design the chainwheel out of the clutchpacket.....

Cheers, Michiel

Re: Clutch conundrum

I’m not familiar with the Norton clutch. Does the chainwheel have cork inserts like the BSA 6 spring? I have assumed that the primary function of the backing plate, on these clutches, is to support the corks in the chainwheel and prevent them being forced out, rather than add strength to the assembly.

The use of corks in the chainwheel or a friction plate that bears on the chainwheel would appear to indicate that it is desirable to use the chainwheel as part of the clutch plate stack.

Except in the 4 spring version :thinking_face: , which would also allow the chainwheel to wobble whether the cluch is engaged or not as it is not forced back onto a backing plate or the shoulder of the clutch hub.

still confused here! :confused:

email (option):

Re: Clutch conundrum

Yes, the Norton clutch has inserts in the chainwheel, just like the 6 spring BSA.

The chainwheel is part of the clutch plate stack and the backplate takes up the full springpressure, it also acts as a frictionplate and keeps the chainwheel run true.

In the Triumph 4 spring and WDM20 clutch the chainwheel/basket does not take up the springpressure and the chainwheel relies fully on the bearing assembly to run true, engaged or not. So there is much asked from these bearings and they are stacked with rollers and needs to have little play.

The clutch with the backplate has a much lighter bearing as there are only forces on this bearing when the clutch is disengaged.

BR Michiel

Nieuwe pagina 1