Switching 10" Spring Clutch to 9.5" Diaphragm

[rfs1957]

Fear I must be really dumb as I know this topic has been touched on a hundred times, but - whatever Advanced Search I do - I cannot find a thread that covers this procedure from start to finish, and certainly not the issues I'm facing.

With apologies, then, here goes.

Having used the early spring-type clutch for 15 years, after some hesitation and reading (lighter pedal ? more reliable ?) during my engine rebuild I have had the flywheel machined to accept the foot-print of the later diaphragm clutch, referred to as the 9.5" one I believe.

For those that aren't aware, this just involves making holes and fitting dowels in positions that are on a slightly smaller diameter than the original clutch.

SNGB were out of stock of the critical parts at the time (October 2018 ?) so I ordered the clutch plate, the clutch cover (which contains the diaphragm spring), and the release bearing, from Precision Clutch Components in Somerset.



Now that I have come to fit these elements, I have hit a problem that I am sure I have seen referred to on the Forum before, although perhaps - in my case - it's with a twist.

Everything assembles correctly, but now when I fit the bell-housing I discover that the release fork and bearing run out of space to operate, even before the bearing begins to push on the clutch.

Now, the new release bearing I've been supplied is about 6mm thinner than the original, which is obviously wrong.

Should it in fact be much thicker ? It would have to be about 10mm thicker than the original to optimise the available movement of the operating arm, which seems unlikely ?





The release fork is a new one, but appears to be identical to the one it replaced.

Do the two clutches use the same release fork ?

There is about 25mm of movement available at the thrust face, measured from maximum-withdrawal (where the clips on the release bearing hit the back of the bell-housing) seen here -



which equates to this, seen from the outside -



to maximum-advance (where the operating lever strikes the window in the bell-housing) seen here -



equating to this seen, from the outside -



but the thrust face is barely touching the clutch centre at this point.



So the whole thing is about 20mm wrong, a mismatch that is so massive that there must be some glaring clue somewhere in these pictures ?



Before I talk to the clutch supplier, could anyone have a stab ?

[mgcjag]

Hi Rory....you are trying to mix and match by steering away from a 3.8 set up and useing the 4.2 clutch....the first thing is that 3.8 and 4.2 bell houseings are different...i dont know if this extends to the pivot point in the casting for the clutch fork......or if the 3.8 and 4.2 fork is different....but i do know that not all of the forks(for the various jag models are the same) some have more of a curved end......also i have looked at the clutch covers in the past and the depth does vary (measurement from where release bearing touches to where cover mates to flywheel)..... cant post photos for some reason but if you look at post17 in this link the release bearing is the largest iv found and what im useing in my C build viewtopic.php?f=4&t=15560&hilit=Release+bearing

Your simplest solutions is to bend the end of the fork.....its been done many times amongst the guys im in touch with building C types...where mixing various Jag parts is very common....Hope this helps some...Steve

[tinworm]

Rory, I should be able to bolt a new clutch cover and driven plate up to either a dual drilled '64 flywheel or a 4.2 S2 flywheel tomorrow if you want to know the height of diaphagm pad to flywheel face. The bell housings 3.8. 4.2 are the same depth . The thickness of the driven plate could be a factor. The favourite is the carbon release, but you know that. I think I have both 3.8 and 4.2 versions for measurement.

cheers Barrie

[mgcjag]

Hi Rory just to add.....on the larger release bearing in the link i posted the depth is 35mm from the center of its pivot piont to the front of the carbon face.......Also a 9.5in diaphram clutch iv used was 50mm when bolted to flywheel when measured from bearing face to flywheel.....Steve

[politeperson]

Lucky you found out before you bolted the whole lot in, as I have before!

Replace the release bearing with a carbon face that sits higher or cheaper still, bend the fork as Steve suggests (with heat a ruler and a hammer).

[rfs1957]

Most helpful, all of you.

Agreeing on how we're measuring is important.



My old release bearing has a height of 32mm, which I'd guess is a worn version of Steve's 35mm.

My new one is 25mm, which is obviously a dud, so there's 10mm to be gained with just an original.

My new clutch depth is 50.5mm, so everything sounds good there.

The rough ratio of the release fork is 100/60.

So the 10mm gained on the release is worth 16mm of movement at the piston.

Still not enough to really make use of the optimum pivot positions ?

It may yet come down to the blue-tipped multi-adjusting tool.

[mgcjag]

Hi Rory...yes measurements are as per your drawings....the differences in the release bearings isnt becaus of original vs aftermarket...there are 4 different "thickness" ones that iv seen....could be others....think it will be down to part numbers...but you would have to trawl through the parts catalogue then find the part and measure it...Steve

[abowie]

..the first thing is that 3.8 and 4.2 bell houseings are different..

With the exception of early (non E type) Moss boxes I think the only difference is the diameter of the recess cut for the front mainshaft bearing. I machined a 3.8 bell to fit a 4.2 box recently; works perfectly.

The release bearing for the 3.8 is 2590 but is listed on SNGB as HD3306 which is probably a Borg and Beck part number.

The 4.2 bearing is C22478 but SNGB list C235752# as the current part number.

Assuming that your clutch supplier has actually given you the correct pressure 9.5" pressure plate (which is C22414 which supersedes to HE4562) I suspect you have just been given the wrong thrust bearing.

SNGB list the Borg and Beck clutch kit as in stock. Personally I'd just send back what you've been sold and use one of theirs.

[mgcjag]

Hi Andrew....the Starter motor holes are different between 3.8 and 4.2.......clutch fork mount position could be different unless you have measured them...Steve

[abowie]

PS. Looking at you pictures, and I could be completely wrong, the diaphragm springs do look oddly flat to my eye.

Now I don't have another one to compare with but I thought that the bearing face in the centre sticks out a bit more than that.

It is also possible to fit the friction plate backwards.... another thing worth checking.

[abowie]

Still not enough to really make use of the optimum pivot positions ?

The release fork, in the resting position, doesn't sit centrally but is slightly forward, so your 16mm might actually be perfect.

[rfs1957]

From what I can see from parts books, and with help from Simon at Precision Clutch, these ought to be the parts we're looking at.

Caveat : see post #20 from Steve3.8, what follows is at least partially inaccurate

From the XK120 to the XJ6, all manual cars used the same clutch release fork, reference C9797.

All 10" Spring-Clutch cars, until the 4.2 in 1965, used a release bearing referred to as 2590 and 48443 (in the J30 parts book) ; these are substituted for the HD3306 by SNGB, which looks like a Borg and Beck reference.

However, I think the actual B+B reference is HD3317 ; I believe this is the "lower" thrust bearing with a 25mm height as defined by the earlier drawing. (Apparently also used in MGB ?).

All the 9.5" Diaphragm-Clutch cars, including the XJ6, used the C23575 release bearing, which became C23575/1 in 1968,

https://www.terrysjag.com/E_Type_Series ... Clutch.pdf

which was slightly (??) thicker, then C23575/2 in 1970, which was the same size but different material - all of which are interchangeable. C22478 must have been in the mix at some point, too, as Andrew says.

These bearings are now Borg and Beck HD3319, and the height is about 35mm (will confirm).

(Also Healey 3000 1962-67 ?)

If anyone else can put numbers onto these please pile in, especially if you have some bearings of different heights that may suit the mix-and-match brigade.

[tinworm]

Height of new old stock release bearing (Quinton Hazell) = 34mm measured as your drawing.

Height of diaphagm pad to flywheel face using clamped Borg and Beck NOS clutch plate = 55mm

Your other variance would be the thickness of your flywheel if it has been refaced

cheers Barrie

[rfs1957]

Many thanks Barrie.

That's not very good news - your 55mm - as it's way over my 50.5mm, which is a B+B clutch too.

Steve said he got 50mm, I think, so there must be some variation here.

I decided the easiest way to see what was going on was to make a jig, makes a change from a bracket.



The clutch thrust face is 108mm from the block face.



The wooden jig replicates the thrust faces with regard to the release bearing.



Looks like this once fitted.

If I use the new fork (no wear in the claws) with the thickest of my two bearings, it's worn but equates +/- 1mm to your 34mm one, then the operating arm is only 5mm off running out of space and is about to hit the case.





So even if I swap the thin (25mm, probably 10" release) for a new thick (34mm, 9.5" release), I'm still going to have insufficient movement.

I'm beginning to think that my Medatronics 5-speed bell-housing isn't quite like the original, because the worn "thick" release bearing (which should be for a diaphragm spring clutch) had been in use for 5 years paired with my old spring clutch, and worked perfectly ............

Could Steve tell us more about bending the fork ? I had imagined it was made of cast-iron, but if it's indeed forged steel then I'll certainly have a go.

[mgcjag]

Hi Rory....i did question previously if the pivot point for the fork was different between the 3.8 bell and the 4.2 bell...didnt realise that yours was an after market bell.......you have an option to grind back the opening..i can see you have done it to the gearbox..you could go back further.......i didnt straighten my fork but i know a few who have...the guys dont have as much engineering experties as you and all i know is they heated and bent it....Steve

[tinworm]

I have a 2.8 XJ6 gearbox to hand - which has the Original Jag bell housing apart from the starter motor mounting is for an inertia type starter. For reference I can rig up an arrangement like you have set for 108mm and measure the fork/ hole clearance if you wish.

Barrie

[tinworm]

Ok , nipped up the store shed and did some more measurements for you mainly because I am intrigued. The gearbox has a worn release bearing which (without taking it off) was @32 mm from fulcrum centre to carbon facing. I rigged up the same test set up as you at 108mm to simulate your clutch/flywheel . The fork arm has a clearance of about 10mm to the slot in the bellhousing when the release bearing is extended as you require. I also measured from the centre of the fork fulcrum pin to the bellhousing front flange face and this was 152mm , in case you suspected your bellhousing .

Barrie

[mgcjag]

I would also question that all the forks are the same.....they may have the same part number but there are differences....iv see some with more curved than others.....look at the one in Barrie,s photo..not much curve and the notches on the rear of the bend....Steve

[rfs1957]

Much appreciated Barrie, this is most useful.

The funny thing is that when I went to 5 speeds, I just flipped the release fork and bearing out of the Moss 'box and fitted them to the Medatronics housing. So I must have had the "thick" carbon bearing in with the 10" spring clutch for 10 years ........... va comprendre.

Anyway, emboldened by Steve's glib encouragement, I had a go at sorting out the fork - as I have two, an old and a new, I thought I would do a dummy run with the old one to see where it needed to end up in order to maximise the geometry and get the release bearing where I wanted it.



The precision of the plane of any bend you make is, happily, not particularly critical as there is some leeway and movement at the pushrod end, but making a proper tool to apply the effort seemed important, especially to move quickly and efficiently once the heat had been got into the fork.



The locating pin for the pivot makes a convenient locating place to fasten the lever in place.



The curve of the lever is because it was an offcut from the pergola roof, the weld-marks on it came from it being turned into a chain-whip at one point to grip a bicycle cassette, and the holes in it were from when it was a Renault-5 servo holding prop. A monument to recycling ........



Firebrick offcuts are a good way of getting heat into a part like this.



In the event the bending was pretty uneventful, and after two iterations I had a shape that seemed to produce the desired result.



In the rear, the new, original shaped one ; foreground, the blue-tipped adjusted one.

Seen below, now in position ; this is at the point of contact between the thrust bearing and the clutch pad, and from this point on there is enough movement to give fully 13mm of lift at the clutch before the arm hits the casing.



The kinetics and geometry of the pivots and linkages look perfect.

Seems too good to be true - am I missing something ?



The clips that were supplied with the new thrust bearing are quite different - has anyone used these ? TBH they don't seem to have any "spring" in their steel, and had no pre-load when fitted, and then took a quite easy "set" when I tweaked them, so whilst I like the principle I don't think I'll be fitting them.



Finally, a tool about which I was about to say "don't tell anyone I made this" has worked sufficiently well to merit a spot of wire-brushing, and even get labelled - otherwise I'll look at it in 2 weeks time and have no clue what it's for.

[steve3.8]

If anyone else can put numbers onto these please pile in, especially if you have some bearings of different heights that may suit the mix-and-match brigade.

Cross reference of vehicles /bearings to add to info if it helps.