M900 belt tension newbie questions

[spuggy]

So, at the risk of boring the cr@p out of everyone, I need some advice with my first belt tension on a M900 (sitting for 2+ years due to lockdown). The Ducati Suite link in the FAQ no worky for me, so I watched the Brad (the Bike Boy) and ExactFit/CA Cycleworks videos a few times instead.

Belts were not new when left, they'd last been done by the PO's shop. This being the guy who was riding the bike with wrong length chain fitted, and 100+mm of slack in it...

When I removed the belt covers, I was surprised/concerned how slack they were. Not to mention ticked at how tight the tensioner bolts were.

Old belts were ExactFit; they looked pristine. But I had a new set, so off they came regardless. Tensioner bearings all rotated smoothly and seem fine with no play, so the new "just in case" set can sit and wait until they're needed.


My question relates to the belt tension...

I can readily squeeze the tensioner with one hand, nip the slider allen up with the other - and get belt tension high enough that the 5mm allen is a good sliding "feeler gauge" fit on the idler bearing at TDC.

However, despite trying multiple times, I just can't seem to get it tight enough that the 6mm is a no-go. It's tighter/takes more effort, but will still go through.

Yeh, belts are very much tighter than they were before. Much happier with that. Inclined to call it good in fact.

Enter the nagging doubts stage right...

The ExactFit packaging says "90 - 110 Hz" for the 2V belts - the Gates phone app seems to show 70Hz at TDC when the driven section by the crank pulley is plucked. And Brad (who says he never uses the Hertz method for 2V belts) gets 135-140Hz when he double-checks the 5/6mm method with the phone.

It's entirely possible I don't have enough grip strength to squeeze the tensioner tight enough. On the vertical cylinder, I can use a wrench as a lever between the tensioner and the belt cover casting for more tension. On the horizontal cylinder, looks like I'd have to lever against a cooling fin. Which is pretty obviously leads to "duh, no, don't do that".

Am I over-thinking this/second guessing myself? I'm kind of leaning towards "good stiff sliding fit on the 5mm - button it up". Especially as they are obviously a lot tighter than they were before - and they never jumped a tooth...

[350TSS]

I have always used the allen key method on 750SS, M900 and S4s (maybe 7 belt changes in total and probably 70k miles in total) and a 5mm snug sliding fit has always worked fine for me.

[Mr Gazza]

Well you've already done the most important things, which is fit the new belts and check all the bearings. Next is to use the correct torque on the tensioner quadrant bolts. I know the torque in my head but I won't put anything in print without checking the book. I personally never use thread lock on those and I'm not alone.

The belt tension is over thought all to often by many I think and the frequency method is often revealed to be confusing, I've never bothered with it.. It's for musical instruments not motorbikes!!
Sounds like you've got the horizontal tension cock on to me. Even when it's right with a 5mm Allen key you can still get a 6mm through with enough will, so I would be happy with that. Over tight is a bad thing, slightly slack won't do any harm. As you say it won't slip.
Remember to use a 6mm key on the vertical belt with a 7mm being very hard to pass.

It's not a bad idea to just run it up with covers off first time and keep an eye on the tracking of the belts over the pullies.. It's fun watching the white patches of paint shooting round anyway..

[slob]

most people seem to get the belts on the tight side.
have you tried the gates app?

[Nickj]

Sounds about right to me Spuggy.
A tight fit for a 5 and a bloody hard shove to get a 6 through has worked out fine for me on a few 750 motors and my S2R, probably totalled over 100K on them with this method of tensioning.
I'd be pickier and more precise on the quatro motors, my 748R was done with a tuning method but that was a totally different beast.

[spuggy]

Okey-dokey, thanks to all who responded, much appreciated!

The 5mm is a tight/snug fit with drag, effort required etc. New belts are much tighter than the ones that came off were. I'm no expert (obviously) - but they seem in the ballpark; a lot happier with them as they are.

Sounds like that's probably fine. Which is what I thought, so thanks for the sanity check. Guess I'm gonna find out

I did take Brad's advice to turn the motor over a dozen times or so (probably more like 2-3 dozen now) on the new belts with some tension just to let them settle into where they want to run before tensioning @ TDC for each.

I think the thing that threw me the most was that I'm used to a "no-go" test being "nope, NFW does that fit" rather than "if you try harder, it'll still go", LOL.

[utopia]

The allen key method is extremely subjective .. hence the various mechanical or audio/electronic attempts to standardise the procedure.
But I've done maybe ten belt changes on 750/600 engines and always use allen keys.
Once you get the feel of it, the confidence increases.
I work on the principle that a little slack is probably fine, a bit tight could be disasterous.
I only ever move the adjusters with light hand pressure (as you describe) which is more than adequate. I would never even consider using leverage. in fact its easy to overtighten using just hands.
My preferred technique to measure the tension is to push the belt away from the pulley using two fingers, one either side of the contact point, and then test the gap with the allen keys (rather than force the key through on its own and try to judge how much effort I'm using). Then I aim for what you might call a "feeler gauge fit" with the smaller key and "elastic interference" with the larger.
Using this technique I find that I can easily discern whether I am holding both the outer surface of the belt and also the allen key accurately parallel to the pulley axis. This is not possible if the allen key is simply forced through the gap in the conventional way and this is one of the factors which give me confidence in the technique.
Hope that makes sense .. particularly the last bit. Its hard to put into words .. hence my late reply.
Its worth repeating that I aim to achieve an acceptably minimal tension while being absolutely certain to avoid overtensioning.
As a kind of double check, I also try to twist the belts in their longest run after tensioning them and aim to be able to twist them around 45 deg. However this again is extremely subjective.

[Dukedesmo]

Agree with all of the above. I use the 'go' 'no-go' (or rather 'go-but-not-so-easily') method as it's the easiest way.

I do have a belt tension gauge for the 916 but it won't fit on the 2v engine, however I like to use a couple of methods to confirm, like the 45 degree twist etc. Also tried the Gates app on my phone and all give pretty much the same tension so must be about right.

I also think that belt tension is over thought by many and that there is more leeway than you might think but (as with chains) on the loose side is better than too tight and they'll be looser after a few miles anyway.

[spuggy]

Originally Posted by Mr Gazza

Next is to use the correct torque on the tensioner quadrant bolts. I know the torque in my head but I won't put anything in print without checking the book. I personally never use thread lock on those and I'm not alone.

Job done and buttoned up, ready to tip in some fresh fuel as soon as I can make it into the garage and wheel it out without expiring from heat exhaustion in a puddle... I refused the temptation to fire it up until I confirmed the torque number I used on the tensioner bolts.

Just could not see it for looking in the PDF, but the dead-tree 1999 M900 workshop manual (anyone ever seen a 2000-2001 dead tree version? Been keeping an eye open for years and never seen one in English...) gives 26nm, or a smidge over 20 ft/lbs for the M8x1.25 tensioner bolts. Which was, after rejecting obviously stoopid outliers, what the internet had said. And what I used.

I do note the manual gives exactly the same figure for the M8x1.0 fixed tensioner stud, which later models seem to have revised downwards somewhat,... And also says I can't do the job without the Ducati tension gauge and the special tool to move the idler, LOL...

The fixed stud is spec'd for thread lock in later manuals, apparently. Probably because of the lower torque spec and the coarse thread.

[spuggy]

Originally Posted by utopia

The allen key method is extremely subjective ..
My preferred technique to measure the tension is to push the belt away from the pulley using two fingers, one either side of the contact point, and then test the gap with the allen keys (rather than force the key through on its own and try to judge how much effort I'm using). Then I aim for what you might call a "feeler gauge fit" with the smaller key and "elastic interference" with the larger.
Using this technique I find that I can easily discern whether I am holding both the outer surface of the belt and also the allen key accurately parallel to the pulley axis. This is not possible if the allen key is simply forced through the gap in the conventional way
Hope that makes sense .. particularly the last bit. Its hard to put into words ..

This helped, thanks utopia!

Yeh, if I felt the key get cocked, then I'd simply disregard that "reading" and try again.

Your suggestion made it much easier to alight the key correctly and get a good read every time - cool!

[Mr Gazza]

I've got a Ducati Manual which covers 2000 M900Sie. Quite a weighty tome! and info scattered all over the place.

It gives fixed tensioner @ 19Nm M8 x 1. Mobile tensioner @ 26NM M8 x 1.25

It also says use Loctite 243 (high strength) on both, but I never do on the mobile tensioner. Good idea on the fixed one though.

Glad I didn't post from the hip as I had 22Nm in my head for the quadrant bolts. I never assume when actually torqueing down though.

Incidentally the threads in the heads are helicoiled from new, so it's worth being careful with them. If there is a problem and they need to be replaced, I understand that Wurth inserts can be used, which apparently are over-sized from Helicoils.

[spuggy]

Originally Posted by slob

have you tried the gates app?

I actually initially had high hopes for the Gates Carbon Drive app. As least as a backup/sanity-check. Especially when I noted that the CA Cycleworks packaging gave 90-110 Hz as acceptable ranges for the M900 belts.

My first read with the Gates app, on my first attempt, gave me 70Hz. Hence my question - the belts seemed about right to me. The originals were a heck of a lot looser than that when I got the covers off...

Last time around, with the belts a smidge tighter, I got readings from the Gates app of:

Nothing (quite often). And successive readings (of the same belt adjustment without changing anything) gave 250 Hz, 175 Hz. And 25 Hz. This despite as consistent a "plucking" of the belt as possible (used to own an electric bass many years ago - never expected that to be a transferable skill).

At which point I decided the Gates app really wasn't as useful as I'd hoped it would be. There's background electric motor noise that might be making it harder for it to get accurate readings. But that's non-negotiable after 9:00 AM or so. In the desert. In July.

I was tempted to use one of my Krikit gauges on the horizontal belt. But (a) there's no spec for that and (b) access to the vertical belt kind of a challenge without dropping the motor.

[spuggy]

Originally Posted by Mr Gazza

I've got a Ducati Manual which covers 2000 M900Sie.

Colour me jealous! Been keeping an eye open for one since I bought the Dark almost 10 years ago... The up-to-1999 are plentiful (and I own one) and the 2002-on one isn't that rare. But never a 2000-2001 copy. Except for some odd-ball (Italian/Japanese maybe) that didn't have English.

Originally Posted by Mr Gazza

Quite a weighty tome! and info scattered all over the place.

Yeh, The PDF is ideal for printing out a page or 3 to get greasy thumbprints on - but I find it much easier to browse a "real" copy when looking for info that' could be a dozen pages away.. Or more.

And for some reason, I find it a lot harder to parse multiple languages in different columns on a screen into information as opposed to when its on paper...

Originally Posted by Mr Gazza

It gives fixed tensioner @ 19Nm M8 x 1. Mobile tensioner @ 26NM M8 x 1.25

It also says use Loctite 243 (high strength) on both, but I never do on the mobile tensioner. Good idea on the fixed one though.

Huh, interesting. Apparently (eg "some guy on the internet said once") the 896 shop manual gives 20Nm for the fixed, with Loctite on the fixed (only? I think).

Evolution I guess.

Originally Posted by Mr Gazza

I never assume when actually torqueing down though.

Heh - sounds like a plan... "I've got a torque wrench so I'm gonna use it" is pretty much my default position. Although it's surprising how close you often get to the spec'd torque figure just based on thread size together with common sense/experience/feel.

[Brian.]

Originally Posted by Mr Gazza

Sounds like you've got the horizontal tension cock on to me. Even when it's right with a 5mm Allen key you can still get a 6mm through with enough will, so I would be happy with that. Over tight is a bad thing, slightly slack won't do any harm. As you say it won't slip.
Remember to use a 6mm key on the vertical belt with a 7mm being very hard to pass.

Good evening Mr Gazza,

I'm in the process of doing my M750 belts & tensioner bearings (that bit was fun ) and I'd like you to know why you recommend a slightly slacker tension on the vertical belt please?

Thank you!

[Mr Gazza]

It's generally accepted that the vertical cylinder runs a bit hotter than the horizontal as it doesn't benefit from quite so much cooling air.
So the clearance is a bit more to allow for a bit more expansion.