When I bought my 907 Ie a few months ago it allegedly had a brand new Barnett clutch, but still rattled far more than I remembered my old Sport rattling. After being inspired by DonaldoDuck's rework I decided to take a look inside, and make a few measurements. Here is what I have:
1 AL/fiber plate slightly thinner than the rest. Looking at it it has significantly less material on one side than the other. Measures 2.75mm thick.
6 AL/fiber plates, 3.2 to 3.5mm thick.
1 wavy steel plate, 1.0mm thick.
1 2mm thick steel plate
7 1.5mm steel plates
I also took some measurements:
Clutch pack I have: 37.8mm
Distance from clutch basket bottom to pressure plate, fully compressed: 36.75
Thickness at which pressure plate disengages from clutch hub teeth: 44mm
total stroke of slave cylinder: 3mm
Fiber clutch tabs: 11.6mm
Clutch basket slots: 12.2mm
So, I only have about 1mm of spring preload pressure on my clutch - to me that sounds worn out, particularly when the tabs have over .5mm of play - that is a lot of noise! I ordered another set of Barnetts because I am cheap and so are they.
Looking at the data, here is how I plan on reassembling them:
First one in has to be a fiber plate. If you have a metal plate on the bottom, then you are compressing the entire clutch pack against the metal one, and in turn the teeth on the bottom of the hub. Set up like this there is nothing keeping the basket from rattling away, making noise and destroying the tabs (his, BTW, is how mine was assembled). It also makes sense that this plate be the one that has less material on one side, because the side that goes against the basket will never see any friction.
Next metal, then fiber, etc until I run out of fiber plates.
Then a metal,
Then either the wavy or conical one
Then the thick metal.
Then the clutch pressure plate.
The reason for the above: the conical or wavy plate acts as a preload spring, giving the clutch a more gradual engagement. But you want the load on both the face of the pressure place as well as the faces of the fiber plates to be as parallel as possible, for maximum surface area engagement. The above has the thick metal one supporting the relatively thin AL pressure plate, and a second flat metal rubbing against the friction.
As far as thickness, I think shooting for 39-40mm with new components seems to be a good target. Thicker is better, but it it's too thick then the clutch pressure plate might hang up on the splines of the hub.
Sam
clutch musings...
-
samandkimberly
- Posts: 149
- Joined: Thu Apr 03, 2008 4:38 pm
- model: 907 I.E.
- year: 1993
- Location: Boston, MA
-
tewsy
Re: clutch musings...
I bought clutch basket and match fitted barnett clutch plates from Desmo Times for $450us,and slave unit.Results are a very quiet clutch and light lever action.
-
samandkimberly
- Posts: 149
- Joined: Thu Apr 03, 2008 4:38 pm
- model: 907 I.E.
- year: 1993
- Location: Boston, MA
Re: clutch musings...
So, I received a new set of Barnett plates over the weekend. I now know why the ole plates in my bike were installed the way they were -they followed the Barnett instructions, which I believe are quite incorrect! The Barnett instructions indicate that you start by installing a metal, then a fiber, then a metal, then a fiber, then the wavy thin plate, then metal/fiber ending up with the thick metal next to the pressure plate. The wear on the clutch I removed, as well as the noise it was making shows clearly that this is wrong.
If you start with a metal plate this plate will slide all the way to the bottom of the hub, where there is a flange that keeps the plate from going against the basket. This is a good thing in that it keeps the metal plate form rubbing against the metal hub, but this flange also becomes what the pressure plate is pushing against when the clutch is engaged, rather than the basket. The matal plate is relatively thin, so it bends conically, which makes the ID of the first few friction plates wear much more than the OD. If you look at the Ducati 907 manual posted on this site, you'll see that Ducati does not do this; they start with a friction plate first. This friction plate also stops movement of the plates relative to the basket when the clutch is engaged, greatly reducing noise and tooth wear.
Next - Barnett has you put the thinner wavy plate in the middle of the clutch pack. When partially engaged, this means that only the high/low points of the wavy spring are engaged, and this showed up on the wavy plate I removed - it was blued on the high points and the fiber plates rubbing against it were discolored from the bluing. This might be OK in the long term, but I can't see why it is necessary. Sandwiching the wavy plate between two metals as the last three plates before the pressure plate eliminates the local wear and still provides spring preload to make the clutch engage smoothly.
I found that by playing with thick and thin metal plates from the old and new clutches I could make the clutch pack anywhere from 38.5mm to 41mm. 39mm felt best, so that's how I have it set up. I haven't driven far on it yet but it feels very good and is very quiet.
Just my $0.02 worth, and with the falling dollar these days that's not worth much!
Sam
If you start with a metal plate this plate will slide all the way to the bottom of the hub, where there is a flange that keeps the plate from going against the basket. This is a good thing in that it keeps the metal plate form rubbing against the metal hub, but this flange also becomes what the pressure plate is pushing against when the clutch is engaged, rather than the basket. The matal plate is relatively thin, so it bends conically, which makes the ID of the first few friction plates wear much more than the OD. If you look at the Ducati 907 manual posted on this site, you'll see that Ducati does not do this; they start with a friction plate first. This friction plate also stops movement of the plates relative to the basket when the clutch is engaged, greatly reducing noise and tooth wear.
Next - Barnett has you put the thinner wavy plate in the middle of the clutch pack. When partially engaged, this means that only the high/low points of the wavy spring are engaged, and this showed up on the wavy plate I removed - it was blued on the high points and the fiber plates rubbing against it were discolored from the bluing. This might be OK in the long term, but I can't see why it is necessary. Sandwiching the wavy plate between two metals as the last three plates before the pressure plate eliminates the local wear and still provides spring preload to make the clutch engage smoothly.
I found that by playing with thick and thin metal plates from the old and new clutches I could make the clutch pack anywhere from 38.5mm to 41mm. 39mm felt best, so that's how I have it set up. I haven't driven far on it yet but it feels very good and is very quiet.
Just my $0.02 worth, and with the falling dollar these days that's not worth much!
Sam
-
Laddie907
- Posts: 174
- Joined: Thu Nov 29, 2007 11:45 pm
- model: 907 I.E.
- year: 1991
- Location: Australia
Re: clutch musings...
I recently used a Barnett clutch plate set in my 900SS; it was rattly as all hell at idle.
the clutch plates and basket are identical to a 907ie.
[Barnett indicate different springs are used for these two models]
it was a pretty easy test to determine the problem; the clutch pack had about 3 - 5mm of free play within the basket on rotation.
The plate thicknesses were all serviceable but the problem was the friction plate outer tags that fit into the basket slots were all significantly bashed around and smaller. [at least 3 - 5 mm].
My guess is these tags were not tempered (and are too soft) as they have mushroomed, rather than chipped.
The basket slots are galled; probably from being allowed to run too long with excessive freeplay - really the basket should be replaced...but the new plates were quite tight in the slots and have shut up the noisy rattle at idle fairly well. The dry clutch "tinkle" is still there with the clutch lever depressed...
I have a new Barnett basket and more new plates to install if I ever feel that the galling is causing the tags to wear at a silly rate...
my 2 cents worth on clutches and plate order is:
I don't think it matters much where the plates (or what type) are fitted.
A distorted plate (conical / bent / wavy) will give reduced friction performance BUT all the plates are either held on the spline [or friction plates in the basket] - so if slipping does occur it's happening to ALL the plates, not just the badly contacting ones...
I think the idea of the wavy plate in the middle of the pack is to act as a spring to make the plates seperate on clutch disengagement; if it was on the end of the pack [inside or outside] it might not work so well...
I installed the plates as per Barnett recommendations.
The outer friction plate only just makes it into the basket slot under tension; if the pack was any [2mm] thicker it would not be retained.
The clutch has performed well since installing...but I haven't really given it a hard time...
the clutch plates and basket are identical to a 907ie.
[Barnett indicate different springs are used for these two models]
it was a pretty easy test to determine the problem; the clutch pack had about 3 - 5mm of free play within the basket on rotation.
The plate thicknesses were all serviceable but the problem was the friction plate outer tags that fit into the basket slots were all significantly bashed around and smaller. [at least 3 - 5 mm].
My guess is these tags were not tempered (and are too soft) as they have mushroomed, rather than chipped.
The basket slots are galled; probably from being allowed to run too long with excessive freeplay - really the basket should be replaced...but the new plates were quite tight in the slots and have shut up the noisy rattle at idle fairly well. The dry clutch "tinkle" is still there with the clutch lever depressed...
I have a new Barnett basket and more new plates to install if I ever feel that the galling is causing the tags to wear at a silly rate...
my 2 cents worth on clutches and plate order is:
I don't think it matters much where the plates (or what type) are fitted.
A distorted plate (conical / bent / wavy) will give reduced friction performance BUT all the plates are either held on the spline [or friction plates in the basket] - so if slipping does occur it's happening to ALL the plates, not just the badly contacting ones...
I think the idea of the wavy plate in the middle of the pack is to act as a spring to make the plates seperate on clutch disengagement; if it was on the end of the pack [inside or outside] it might not work so well...
I installed the plates as per Barnett recommendations.
The outer friction plate only just makes it into the basket slot under tension; if the pack was any [2mm] thicker it would not be retained.
The clutch has performed well since installing...but I haven't really given it a hard time...
-
gasfireman
- Posts: 46
- Joined: Thu Aug 09, 2007 12:00 am
- model: 906 Paso
- year: 1989
- Location: Bristol, UK
Re: clutch musings...
I have a Monster with 52k on the dry clutch. They all rattle its a dry clutch after all.
It rattles like crazy - so what!
Its been past the wear tolerances since 28k - so what it works!
The basket looks worn making it rattle worse.
I dont see any point in replacing these parts till they dont work. If it breaks then the parts that will get damaged are those I would have replaced anyway. The clutch is dry so it cant send swarf round the motor.
My 906 clutch is slightly better but being its the inside out mechanism they were never as good as the later ones but it works, so leave well alone. The later clutches rattle when the clutch is out and the bike idling, the 906 / 907 rattles when you pull it in, its just an idiosyncrasie of the design.
The biggest enemy of these clutches is dust from the plates and that can be cleaned off with brake cleaner. File the burrs off the plates and basket so the plates move freely.
Then ride it and ignoore the hell out of problems.
It rattles like crazy - so what!
Its been past the wear tolerances since 28k - so what it works!
The basket looks worn making it rattle worse.
I dont see any point in replacing these parts till they dont work. If it breaks then the parts that will get damaged are those I would have replaced anyway. The clutch is dry so it cant send swarf round the motor.
My 906 clutch is slightly better but being its the inside out mechanism they were never as good as the later ones but it works, so leave well alone. The later clutches rattle when the clutch is out and the bike idling, the 906 / 907 rattles when you pull it in, its just an idiosyncrasie of the design.
The biggest enemy of these clutches is dust from the plates and that can be cleaned off with brake cleaner. File the burrs off the plates and basket so the plates move freely.
Then ride it and ignoore the hell out of problems.
-
higgy
- paso grand pooh-bah
- Posts: 3377
- Joined: Tue Jul 29, 2008 9:50 pm
- model: 907 I.E.
- year: 1992
- Location: Hilltown,Pennsylvania
- Contact:
Re: clutch musings...
It is not a good idea to use any chemicals on any clutch plate,brake clean in particular will dry out the friction material and lead to premature failure and friction material cracking.This applies to brake pads as well. If you must spray it use water only from a spray bottle and give it time to dry. Air is also a bad idea,use a good hepa filtered vacuum.
Ducati,making mechanics out of riders since 1946
There's no problem so bad that a little fixing can't make it worse! : )
Electrocuted Birds Are Bursting Into Flames and Starting Wildfires
88 750
90 906
92 907ie
There's no problem so bad that a little fixing can't make it worse! : )
Electrocuted Birds Are Bursting Into Flames and Starting Wildfires
88 750
90 906
92 907ie
-
samandkimberly
- Posts: 149
- Joined: Thu Apr 03, 2008 4:38 pm
- model: 907 I.E.
- year: 1993
- Location: Boston, MA
Re: clutch musings...
If you have installed the clutch pack per Barnett your clutch plates are always rattling back and forth whether the clutch is engaged or not. With all that free play, and AL on steel the friction splines bound to wear very fast. Mine, for example, has worn nearly 1mm and were supposedly near new! If you reverse the installation and start with a friction plate, (per the Ducati manual) then the clutch stops the basket from rattling except when the clutch is disengaged.Laddie907 wrote: it was a pretty easy test to determine the problem; the clutch pack had about 3 - 5mm of free play within the basket on rotation.
The wavy plate isn't all that important either way; I think it can go most anywhere and the clutch will work. But I think that assembling the clutch by installing a metal plate first is a bad idea. When I have some time I hope to give Barnett a call and get their take on this.
Sam
-
du907
- Posts: 346
- Joined: Fri Jul 08, 2005 12:00 am
- model: 907 I.E.
- year: 1992
- Location: Winston-Salem, N.C. USA
Re: clutch musings...
Check the link below. It gives some information about clutch stacking.
http://www.mad-ducati.com/
Having dealt with the same problem, I have some observations you might consider.
I have tried the Barnett clutches. They don't last. Ducati OEM has two types that work and they last, one set has alumn. based fiber plates, the other has steel based fiber plates. If you still have the original clutch drum, the part with the slots that the fiber plates tangs fit into, this part is steel. You want the steel base fiber plates so that you have steel againist steel. The alumn. base fiber plates with alumn. tangs will wear out prematurely rubbing up againist steel.
This happened to me.
I replaced the clutch drum (as stated above) with a new part. However, the new part is the same that is used in most ducatis and IS DIFFERENT than the orginal paso drum. The original paso drum is flatter and machined at the back. Pasos clutches are suppose to start with a steel fiber plate that is thicker than the others and has fiber only on one side. The newer ducatis start with 2 steel driven plates and so on like the link above. This DOES'NT work in a Paso. The springer plates are thinner than the other driven plates (the plates with the teeth on the inside edge) and have a dot stamped on them. I remove one of the standard driven plates and replace with two springer plates, one at the beginning of the stack and one toward the end of the stack.
du907
http://www.mad-ducati.com/
Having dealt with the same problem, I have some observations you might consider.
I have tried the Barnett clutches. They don't last. Ducati OEM has two types that work and they last, one set has alumn. based fiber plates, the other has steel based fiber plates. If you still have the original clutch drum, the part with the slots that the fiber plates tangs fit into, this part is steel. You want the steel base fiber plates so that you have steel againist steel. The alumn. base fiber plates with alumn. tangs will wear out prematurely rubbing up againist steel.
This happened to me.
I replaced the clutch drum (as stated above) with a new part. However, the new part is the same that is used in most ducatis and IS DIFFERENT than the orginal paso drum. The original paso drum is flatter and machined at the back. Pasos clutches are suppose to start with a steel fiber plate that is thicker than the others and has fiber only on one side. The newer ducatis start with 2 steel driven plates and so on like the link above. This DOES'NT work in a Paso. The springer plates are thinner than the other driven plates (the plates with the teeth on the inside edge) and have a dot stamped on them. I remove one of the standard driven plates and replace with two springer plates, one at the beginning of the stack and one toward the end of the stack.
du907