Clutch and Fluid

The clutch appears to have been designed by Josep Serra (Xiu) of XiU-rdi Engineering.  It is quite similar to the ones in my OSSAs, however, it uses 4 friction plates instead of 3.  The clutch has 3 stages of adjustable spring preload.  The clutch lever has a very light pull and the clutch works well for my level of riding.  Honestly, it makes me think the OSSA clutches could work much better than they do and that Xiu has learned a thing or two since his initial design.

My concern about the clutch requiring frequent adjustment was unfounded.  My friend Mike has more than 200 hours on his first ePure and the original clutch is still working fine.  I think part of the credit should go to the electric motor which makes life easy on a clutch compared to an ICE.  The peak versus average torque is numerically much smaller for an electric motor than for an internal combustion engine (4-strokes being worse in this regard than 2-strokes).  Although electric motors do exhibit torque ripple, it is small compared to an ICE.

I specifically mentioned the clutch working for my level of riding because it most certainly did not work for Ryan Young.  For the unfamiliar, Ryan was at one time the top trials rider in the USA and an importer for Electric Motion in the early days of the 5.7.   He now conducts a trials school that the prior owner of my bike was fortunate enough to attend.  Ryan rode the bike I now own but quickly returned it when he realized the clutch was not going to behave anywhere near his needs/expectations.  Knowing what I do about the OSSA clutch setup – that pack thickness, spring force, lever adjustment, and even oil have a huge influence on behavior – I'd be keen to try setting up an EM clutch for a rider with Ryan Young's ability.

The clutch fluid specification is 260 ml of 75 weight, changed every 40 hours.  I will be using Valvoline Dexron VI fully synthetic ATF.

Flywheel removed exposing clutch primary gear 

Flywheel Puller

A flywheel puller is required to service the clutch.  In the ePure Race, the flywheel is mounted on a tapered shaft as it would be in an ICE motorcycle.  

The required tool is a standard M35 x 1.5 right-hand external puller.  EM offers this tool for about $70, but more economical solutions can be found.  It appears Motion Pro's $37 08-0349 would work.  No-name pullers are even less expensive (and what I used).


The flywheel also incorporates the primary drive gear.  Overall dimensions are 102 mm OD, 65 mm ID, and 41 mm thick in total.  There are 2 removable weights, each 8 mm thick. 

But I can't see removing them being useful.  This bike needs more inertia, not less.

Although unlikely to ever be needed, an expensive ($283) special tool, part number TL02O-T0202-00-00 is required to separate the primary gear from the flywheel weight.  I have not found a second source for this tool but I think it is just an N07-size bearing locknut socket.

Note: According to the US distributor for King Tony branded products, their KT PRO D3442N07 has been discontinued.

Flywheel weight and primary gear shown with both removable disks installed 

Clutch Service Notes

My clutch worked flawlessly for the first 100 hours, then I began to notice some slipping at the end of the 2023 riding season.  I'll admit the clutch has not had an easy life, especially after the addition of my super-heavy flywheel weight.  (See the section Adding Flywheel Inertia under My Modifications for more information.)

Although I'm satisfied with the 100-hour life itself (it compares favorably with my other diaphragm trials clutches) the EM's clutch is far more difficult to access than any other off-road motorcycle clutch.  I'm also disappointed because I had assumed there would be two control-ring adjustment positions available to compensate for pack wear, but there is only one.  More on that later.

On the positive side, everything was easy to disassemble.  No fasteners had been over-torqued by a gorilla.  The use of a blue thread-locker was apparent.  The components are all high quality and the machined-from-billet case halves are superior to typical engine castings.

Upon disassembly, I found the control ring preload setting at the #2 position.  This is the same starting point as in EM's maintenance video.  I'm told EM intended this adjustment mechanism to allow for personal preference rather than as a way to compensate for pack wear.  If you want a softer slower-acting clutch, choose setting #1.  Conversely, if you want a more aggressive faster-acting clutch, choose setting #3.

Contrary to this, I like the standard setting and will be using position #3 to compensate for pack wear.

Upon removing the clutch pack, I immediately put a zip tie around it to ensure the plates did not rotate relative to one another.

My pack was a uniform thickness and measured 9.98mm with a digital caliper.  It was nearly 9.99mm when measured with a digital micrometer.  

For reference, the three steel plates are nominally 0.8mm thick.  The four fiber plates are nominally 1.9mm thick. 

 (3 * 0.8) + (4 * 1.9) = 10.0

I had assumed my clutch slippage was due to worn friction plates, and I would see some evidence in the form of a reduced pack thickness as I do with my OSSAs (which have a similar Xiu diaphragm clutch).  So the lack of wear came as a surprise.

There were lots of gold-colored metallic particles (which adhere to a magnet) on the pressure plate.  The photo only shows about half the quantity I found on the pressure plate.  The drain magnet also had a similar amount.

A bit of ferrous material from clutch wear is normal, but I was surprised by this amount of metal.  None of my prior oil changes had much on the drain plug's magnet - even after adding the first tungsten-copper flywheel weight.  This is the first oil change since adding the second tungsten-copper flywheel weight.  I think all the extra inertia may be taking a toll on the clutch.

I reassembled the clutch with the control ring in the #3 position.  Lever effort is not at all unreasonable - very light compared to other trials bikes.

Ferrous material found adhering to my pressure plate

Clutch Pack Access Procedure

The following are notes I made during disassembly.  I used a slightly different order than EM's video shows.  I chose to remove the flywheel while the motor was still in the chassis.  This makes disassembly easier and reduces weight for subsequent handling.  Torque specifications and socket sizes are noted.  There is no torque spec for the M20 fine flywheel nut.  Mine was easy to remove with a 1/2-inch impact wrench.  Be sure to note the locations of any shimming washers between the chassis and the gearcase.  These are not shown on the parts diagram and may vary by bike.  Mine were on the left side, one at the front motor mount screw and one at the swingarm pivot.

Reassemble in reverse order.  Use blue Loctite on chassis fasteners.  Don't forget to refill the gearbox.

XiU-rdi Clutch Pack

Although EM does not mention any of this in their clutch video, following these steps will ensure optimal performance and minimize the time needed for the clutch pack to adapt” itself.  The following information comes by way of XiU-rdi Engineering (the designer of EM's clutch) for their aftermarket OSSA clutch components. 

New friction plates are not perfectly uniform in thickness.  Slight variations can make the clutch grabby, and increase initial wear. 

Assembly Process:

Webmaster's notes: To further elaborate, let's say one of your discs is perfect with no deviation in thickness.  This disk could have any orientation relative to the other three, which would have their T marks placed 120 degrees apart.   Likewise, two perfect disks would require the remaining T marks placed 180 degrees apart.

I always soak new friction plates in gearbox fluid at least overnight, preferably longer.

Close-up of the XiU-rdi Spring Support Plate

XiU-rdi Spring Support Plate

This information also comes by way of XiU-rdi Engineering for the OSSA.  The Spring Support Plate is also called the Clutch Control Ring. 

The spring support plate has 3 mounting positions, numbered 1, 2, and 3.  The higher the number, the greater the spring preload.  Number 1 is like the original clutch support.  Each number supposedly represents a 0.1mm change in preload (however, measurements on my OSSA indicate a delta of 0.15mm). 

With new plates, you can start with position number 1 or 2, depending on your preference.  After some working time, the plates will adapt and be thinner.  Then you can choose a higher number until position 3.