Friday, July 24, 2015

Ducati: The result of aluminium clutch basket and steel plates


When I first saw this 2003 749S at 27,000km, I inspected the clutch and found it had a steel plate set (meaning the friction plate base material is steel, not aluminium) with an original aluminium basket.  But there wasn't a huge amount of wear, and it as still working, so it was left as was.

I saw it a couple of weeks ago at 58,000km and this was the result.  I forgot to take a photo of the plate tangs, but they they were also very worn.  If it was a bigger engine I expect this would have happened sooner.




It wasn't this bad at 42,000km, when I last had it out.  At that point I replaced the pushrod and needle roller bearing due to wear, and I had to do them again with the basket at 58,000km.  Maybe the worn basket and associated movement allowed the pushrod to move around, but it seemed rather odd that it had done the pushrod in such a short time.
.

Friday, July 10, 2015

Further comment on the Ducati ST3 crank slot incorrectly machined.

.
I thought I'd put up a photo to better explain what I was talking about with the incorrectly machined crank slots, using a photo of a crank.

The crank has two slots machined in the end of it to locate the factory locking tool, and you also use these to locate the turning tools. I've never seen these slots machined off before (not once that i can recall), but these ones certainly were. It would either be an error in or damage to the machining jig I expect. I doubt anyone made the decision to machine them 10 degrees off, as it would be a simple production process.



In the pic you can see the two slots for the tool (red arrows), the alignment slot used to align the flywheel dot (yellow circle) and the timing gear keyway (green circle). The timing and cam pulleys appeared to all be correctly aligned, but I don't recall if this bike had a timing window to see if the flywheel marks lined up with the pointer, or if it was one of the later threaded plug covers. I would expect all these slots would be machined in pretty much one pass though.

As a consequence, using the factory cam timing tools would see the crank locked in the wrong place.

The only way you would know is by doing what I did, so it's not something that would come up often. Using the tools as intended would see this result, but it would be unintended and unknown. The assumption that the tool is right is the assumption you work with when using any tool pretty much, unless anecdotal evidence shows you otherwise. This is the first anecdotal evidence of incorrectly machined slots that I have seen.

Saturday, July 4, 2015

Thin base gaskets for Ducati 996


I recently had some thin base gaskets for the 996 laser cut locally.  I sourced some 0.20 and 0.25mm shim sheet, and the laser cutter had some 0.30mm in stock, so now I have quite a few of each in stock.  Unfortunately, the laser cutting process has left quite a bit of flash on one side, so I'm a bit disappointed in them not being overly saleable as is.  Especially as I have to charge $27.50 each for them.  But I now have a source of them, so if anyone wants any, they're there.

I fitted some to an ST4S engine I was rebuilding (the catalyst for getting them cut) and the 0.20mm thickness gave 1.00mm squish.  Some will need the 0.25 or 0.30mm.

Ducati ST3 front sprocket incorrect position and other wierdness.


I had an ST3 in this week for a big service, and found a few things along the way that made me wonder if I was doing it right.

First of all, it had no opening clearance on 5 of the 6 valves.  Coincidently, it was the horizontal exhaust opener that had clearance, which is also the one I have seen with no clearance on a couple at around this many km (40,000) previously, requiring replacement of cam and rocker.  But this one was ok.

I checked the vertical first, and with all valves having no clearance, I was starting to wonder if I was doing it right frankly.  There wasn't excessive drag on the cams, such as you get when the valve is receding and the closing spring is forcing the rocker onto the cam and wearing both (as seen previously).  It was more like they'd been set that way.  The closers were all around 0.25mm, so all round it was pretty crappy.

The ST3 inlet valves have very soft closing springs, so you can easily force a 'too thick' feeler gauge in between the rocker and opening shim when checking the opening clearance without really noticing it.  I gently lever the closing rocker up with a screwdriver to make sure the valve is closed while checking the opening clearance, just to make sure.  I do the same when checking the closing clearance to get the zero setting on my dial gauge.  For a model allegedly combining the best of 4V performance and 2V maintenance cost, I find the ST3 the most annoying to do valve clearance adjustments on.  Apart from the somewhat crappy access to the vertical exhaust, the ST4/S is easier imo.

Another thing I've noticed valve clearance wise on the ST3 is that some of them (but not a consistant early vs late or the like grouping) need a total combined inlet clearance of 0.25mm (0.010", or 10 thou), or you can feel the cam binding as you rotate it.  Usually I use 0.10mm (4 thou) on both inlet opening and closing, but some of them need 0.13mm (5 thou) on both to stop the binding.  No idea why, just something I've seen.  The exhausts seem happy enough with the usual opening 0.15mm (6 thou) and closing 0.05 - 0.07mm (2 - 3 thou) though.

Given the owner had commented about the bad tune up at its last service (done somewhere else), and this was an early ST3, I thought I'd check the cam timing.  It's a fairly quick job, but it's something owners will often see no benefit in them paying for, so I don't do it as a matter of course.  But this time I sort of had a feeling I should do it.  I find it quite surprising how often my "probably should do that" urge leads to a "wtf" moment and subsequent fix.

Again things turned to weird fairly quickly.  I use a Vee Two engine turning/degree wheel holding tool, and with it TDC for either cylinder is with one of the handles either vertical or horizontal.  This is because the two slots in the end of the crank are machined in the same place on all cranks, and these slots are used by the factory cam timing tools to lock the crank.

Except for this one.  This is what I got:


The slots are 10 degrees off (retarded), something I've never seen before.  I rechecked my TDC (done using a piston stop), but no matter how hard I tried to find a mistake, I couldn't.  Even the visual "watch the piston move" test showed it was about right.  So I submitted to the reality and moved forward.  The 10 degree offset didn't really correspond to the vertical cam though, which was -7/79 (133 degree inlet c/l) at 1mm lift, or 19 degrees retarded.  I don't think the factory tools would have it 9 degrees off (allowing for the 10 degree slot error), but could be wrong.  Actually, just checking my old ST3 cam timing report, it was 123 or 9 degrees retarded on the vertical as delivered.  Anyway, it's now at 113 degrees, or 1 degree advanced.

The horizontal inlet was 9 degrees off (retarded), which is consistant with the crank slot error.  Now it too is at 113.

From there on the set up was pretty straight forward.  I redid the TPS baseline setting (was 110mV), and 2.3 degrees via the diagnostic tool was actually 337mV.  I wound one of the air bleeds in quite a bit from where it was, with the running balance ok, and the idle trimmer went from +5 to -5 for 5% CO.  Seemed to ride like it should, hopefully the owner will be of the same opinion.

The next issue is one that I have seen before.  I had a 2006 ST3 in sometime ago that needed chain and sprockets.  I removed the front (countershaft) sprocket and found the chain had been wearing on the head of the oil gallery plug (an M6 cap head screw) behind the sprocket.  Assuming this wasn't meant to be happening, I had a bit of a look-see which showed offset wear (to the inside, or RH) on the chain slider on top of the swingarm and came to the following (longwinded) conclusion.

When they introduced the 916 in 1994, the gearbox output shaft had a new part # as compared to the other big block engines (900, 906, 907, 851 and 888).  Later, when the M900ie appeared in 2000 with the "748" close ratio gearbox, it had a front sprocket that has the teeth centred pretty much, instead of being offset to the outer (LH) edge.  The offset was 5mm, which as it turns out is how much longer the "916" series output shaft is compared to the 900, 906, 907, 851 and 888.  This longer output shaft was then used in pretty much every new model, the next of which was the ST2.

When the Testastretta appeared, with the dual row ball bearing supporting the output shaft behind the front sprocket and large nut holding the front sprocket on, it too took on the longer output shaft positioning for the sprocket, as that's what the chassis required.  To achieve this, there is a spacer/seal runner behind the front sprocket that is pulled up between the bearing inner and sprocket by the nut and its 190Nm.

This spacer is the same thickness on all the Testastretta models, and when the Monster series was revised in 2001 (S4) and 2002 (900) to the ST based chassis, these big block models used the 5mm longer shaft as per the ST4 and ST2 they were based on.  In 2003, the 900 became the 1000, with the Testastretta's bottom end and therefore front sprocket location method.  For the M1000, being based on the ST, there was no sprocket alignment issue as its chassis was designed around this front sprocket location.

However, the 1000SSie chassis was still that of the 900SSie, which itself was much the same as the 900SS from 1991.  Including front and rear sprocket location.  So, with the Testastretta based bottom end, the 1000SSie required a 5mm thinner spacer/seal runner to move the sprocket in 5mm to replicate the 1991 - 2002 position.

This gives two spacer/seal runner part #, being 71312041A (1000SSie) and 71311571A (everything else, supersedes to 71313931A)    And, as usual when there's two of something, inevitably the wrong one gets fitted somewhere along the line.  For these spacers, that appears to be the ST3.  The part I'm talking about is # 13 in the diagram below.


The photo below gives the first visual clue: the length of output shaft protruding through the nut.  It's more than I would expect to see on anything other than a 1000SSie.


This bike is a 2004 model, so there's no oil gallery plug behind the sprocket for the chain to wear away.  But the offset wear on the chain slider on top of the swingarm was as expected.  If I put my little laser chain aligning tool on the side of the front sprocket and aim it at the rear sprocket (assuming the rear is in about the right place), you will see the laser dot hitting the sprocket teeth.  It should be hitting the outer plate of the chain.

To solve this, we need to move the sprocket out 5mm.  Typically, either spacer/seal runner part # is ex Italy, giving the obvious option of machining a 5mm thick spacer to go between original spacer and sprocket.  This presents a problem of suitable material at hand, as it needs to be 25 x 35 x 5 and steel.  I'm rather suss of making this sort of thing from aluminium, especially when it gets 190Nm of squash on it.  The first time I used an old nut (possibly primary drive) narrowed to the right thickness.  This time I went on the scrounge to the fencing place just up the road (often useful), and found some big arse gal nuts.  From one of these I turned up the required spacer and all was good again.  I'm going to back order the correct part for stock, just so next time i have what I need at hand.

As to whether or not all ST3 have this issue, I don't know.  I've not heard of it before (see below), and it's the sort of thing you only notice when it causes something noticeable.  The 2006 model with its oil gallery plug gave me something noticeable, but the 2004 wouldn't have made it obvious.  From now on, I'll look at every ST3 I get in.

**A quick google shows others have noticed unusual wear, but I don't think anyone else has come to this conclusion.  Sample size is what brings the anecdotal evidence, and that's easier for someone like me.