Talk Morgan Forums Morgan Sports Cars Morgan Three Wheeler & Super 3 Centa rubbers

If the Centa housing could be rotated by hand a few degrees and the nut was still tight, then it must have been bottomed out on it's thread before clamping the spline adaptor? All good feedback that should help with a fix.

I spoke to Craig at M3WServices last night, he confirmed that the ut does not bottom out on the thread.

I had to leave before Phil got the engine out. I assume the nut was tight against the Centa end of the spline, but the crank end was worn... so the nut was still tight, but the spline under the Centa end was moving radially.

Hi Laurens

You didn't say what MMC told you to torque the nut up to, a Harley specialist told me 170ft/lbs minimum more if possible.

Just to throw the cat amongst the pigeons, I have had 2 nuts come loose and I'm still on the old Harley compensator so that may be at odds with your rubber stiffness theory.

Here's mine, on both occasions the nut loosened off after extensive periods of idling, long queue of traffic, traffic lights etc. This excessive vibration with the oscillation of the compensator due to the poor fit of the splines allows it to act as a rotary hammer drill and in time loosens the nut.

The last time I did mine I used red Loctite, a pair of new Nordlock washers and 170ft/lbs of torque.

Didn't last 3000 miles.

If all the parts, mainly crank were machined to tighter/more acceptable tolerance the nut would not be needed as the Centa housing cannot go anywhere because it is held at both ends. Standard gearbox, clutch flywheel configuration springs to mind.

Food for thought.

"You didn't say what MMC told you to torque the nut up to, a Harley specialist told me 170ft/lbs minimum more if possible."

MMC Centa installation book states 237 NM = 170ft/lbs

Phil mentions 450ft/lbs?

Regarding its on a HD which has no cushion in its action - is that a similar issue with stiffer/smaller rollers - there is a lot more sudden jolts affecting the center nut than with the profiled MMC rollers in a Centa coupling?

"This excessive vibration with the oscillation of the compensator due to the poor fit of the splines allows it to act as a rotary hammer drill and in time loosens the nut" - understand but question is that with the 2500 odd Centa drives out there with lots of mileage on them, I am only aware of 1 that has had a loose nut issue - I'm sure there may be more but it doesn't seem to be a known or publicised problem.

"If all the parts, mainly crank were machined to tighter/more acceptable tolerance the nut would not be needed as the Centa housing cannot go anywhere because it is held at both ends"
Errr...no its not - the outer case is only retained in position by the nut. If no nut it could work off along the splines.

I think the point is: the Centa drive can't "fall out" as its sandwiched between the clutch and gearbox (which is not free to move back and forth due to the gearbox mounts) and the crank (not free to move due to the engine mounts). Indeed the rational for a "spline" is to allow back and forth movement!

The inner or outer rotor aren't located at either of their respective ends. The rollers are the sandwich filling between the two. If their was no nut at the outer rotor /crankshaft splined end, then there is nothing to stop the outer rotor working its way 'backwards' towards the inner rotor/gearbox end. I'm not sure how much it could move (there is at least 40mm clearance between rear of outer rotor and flywheel bearing plate, as that's the space required to remove the rollers) but I would imagine any movement in rearwards direction would make the outer rotor unbalanced and be 'floppy' and probably wreck the crankshaft splines even quicker? Okay it won't fall out but it will do a lot of damage (such as rotor/stator) trying to find its way out if there was no nut to keep it in place.

PB re-engineered the inner rotor to be supported at both ends by machining the outer rotor hub to accept his new style inner rotor at the crankshaft end - the original Centa inner rotor had no support.
In the case of the PB unit the outer rotor cannot move anywhere as the re-engineered inner rotor is keeping it in place.

Splines - your correct but the meaning relates to "prevent rotational movement between two parts" - whether this be fixed or sliding.

From the information i have been given, the nut has been found to have loosened ever so slightly - 30 degrees or so - not much but still loosened for whatever reason.

Very true, but I think the point was more that Morgan could have engineered a solution without a big nut (think spring -separated bearings etc). But it is what it is (and currently: a solution where the nut doesnt stay attached, and/or the spline wears)...

The Harley unit does have cushioning via the 2 lobed rotors that are pressed together with a bellville washer pack. That's how it works. What it dosen't have is any backlash unlike the Centa drive so when it works it works really well.
Also the Harley unit is designed to run in an oil bath which in the M3W is impossible so keeping it greased regularly is a must.

Perhaps the tolerances between the coupling and the shaft are not consistent and vary from car to car.

Once there is the slightest bit of "backlash" on the splines the game is over, as anyone with experience of splined wheel hubs will know.

The load reversals on the drivetrain will contribute to loosening the nut and perhaps a positive locking device is the answer.

Most flexible couplings are predominantly loaded in one direction only, it's less common to have such load reversals as experienced on a car transmission.