Cam Timing A Discussion Of Settings And Effects Of Changes - Written Nov ‘04
A foreword of sorts: all this is just my opinion, from my results. Yours or others may vary. My acquired status as an internet expert does not mean I am one in any way, shape or form, or that I want to be one. As Doug said once, an expert is “someone from out of town”. For many of you I’m on the other side of the world. You are under no obligation to believe me.
There are a few comments along the way talking about the feel of a result. These can be important. All the graphs shown are taken from the reports on each specific model - if you want more info that’s where to head.
And if you want to know for yourself, you can always experiment. It only costs you time and money after all. I waste quite a bit of both.
Update March ’07 a little bit about cylinder pressure, in red, before the practical examples start.
One question that I get asked a lot is “what cam timing do you recommend?” These days I tend to just tell people as it doesn’t really matter that much to me anymore to keep things a secret. Except for locally I guess, but even then I’m not really that worried. Mainly because saying “I won’t tell you, but we’ll do the job for you” just gets me whined at.
There are two main reasons for this.
This is a point I try to push because I see tuning things as the final step. Many don’t bother, which really annoys me in the short term, but just disappoints me overall. I see quite a few badly tuned bikes and knowing they could be better if the owner just let someone with the right equipment and knowledge at it kind of bugs me. The fact many of them have a box of inappropriate eproms they persist with highlights the misunderstanding of the whole tuning thing. This is something that really gets Doug wound up. But, I can’t save everyone and they’re only motorbikes anyway.
So, I’ll go through this report given the specs I’ve tried and what I have decided on using and why and why it might not work for you and fuelling changes required. But first I’ll start with a short discussion of cam timing and what does what.
The way I talk cam timing, as do most people, is in terms of centreline. The centreline of the cam is the theoretical centre point between the opening and closing points at the chosen lift related to TDC (overlap, or the end of exhaust/start of inlet stroke). The lift used to express centreline cam be important if a cam isn’t symmetrical the centreline at 0.10mm lift may be different to the centreline at 1.00mm lift. This is why there are now accepted standards for these things.
Generally, all the Ducati specs are given at 1.00mm lift. Some of the manuals late ‘80s, early ‘90s 2V mainly - list both 0.20mm and 1.00mm timing. But the later ones use 1.00mm, and it’s a figure that’s a bit easier to use I find. If you come from anything based on American V8s you’ll be used to reading “50 thou lift” (1.25mm). Using this range of lift figures gets you away from the opening ramps which may confuse the process the old std was “advertised duration” which itself varied around the 0.006” or 0.15mm lift spec and generally made comparing various cams very difficult. What spec you use is up to you I always use 1.00mm lift but I never worry about the valve clearance. So the numbers I get are always a little on the inside of the specs.
So you need to remember the centreline is a mathematical thing that exists only on paper. If you go looking for the centreline you’ll have a lot of trouble finding it. And it bears no relation to the max lift point. If the cam is symmetrical it should be the same, but many cams aren’t symmetrical. To calculate a centreline you use the following formula, set out in steps.
For a 916 with cam timing specs of 11/70 inlet, 62/18 exhaust we get
So, the centrelines are 119.5 degrees ATDC (overlap) for the inlet and 112 degrees BTDC (overlap) for the exhaust. Not max lift points, just the theoretical centreline based on the opening and closing numbers. It doesn’t tell us anything about the actual cam profile, but is very useful in expressing what’s happening where. If you don’t want to use centrelines you don’t have too.
The next bit I need to cover is what numbers we are paying attention to. The most important number is inlet closing (IC). On a single cam engine, inlet closing is what governs where the cams goes, thereby dictating IO, EO and EC as well. On a twin cam engine EO and EC can do their own thing. As IC sets IO, you can move the exhaust cam around to get the sort of overlap you want (EC and IO), but I don’t really do that so much I’ll get onto why later. I tend to ignore IC these days, looking at centreline instead, something that caught me out somewhat when playing with the 1000SS ds ie.
When you move the cams around you either advance or retard them. Advance means they open and close earlier in the cycle, retard means they open and close later. In number terms, advancing the inlet makes the centreline number smaller, whereas on the exhaust it makes the number bigger. Conversely, retarding the inlet makes the centreline number bigger and the exhaust number smaller. Advancing an inlet cam will also reduce the piston to valve clearance tightest usually from 5 to 15 degree ATDC inlet opening (overlap), whereas retarding the exhaust cam will reduce the piston to valve clearance tightest usually from 15 to 5 degrees BTDC exhaust closing (overlap). To increase piston to valve clearance you retard inlets or advance exhausts.
So if you see cam timing specs with large inlet and exhaust centrelines there may be a piston to valve clearance issue behind the settings.
Next - a generalisation. Advancing the inlets gives more midrange and less top end (therefore implying retarding the inlets gives less midrange and more top end).
This one is a real favourite. It is basically true, but assumes something that turns out be incorrect for some Ducatis. That is, the std cam duration and settings are the best possible (some people like to use the word optimum, which to me infers perfection or delusions there of. I prefer ‘desired’). They aren’t. The duration of the 4V Strada cams is too long (and they’re under valved), and the std settings can be improved upon. Add to that the fact the “as fitted” timing is not that accurate and there is obvious room for improvement.
This is changing though. The engine updates and revisions from Ducati in the last few years starting with 996SPS in ’97, 748R in ’00, S4 and ST4S in ’01 and now onto the Testastretta and 620/800/1000 2V updates in the intake (or breathing) area have all been about bigger valves (except S4 and ST4S which carry the usual 916 and 996 valves) and shorter cams. With good results. Some engines have made the same power with less cam duration (998/999 and ST4S compared to 996) or made more with the same duration and bigger valves. Compare the 748R to the 748Strada, the 998S to the 996SPS, 749 to 748 or any of the 600/750/900 to the respective 620/800/1000. All the reports for these models give specs that show what’s changed.
So the relatively easy power increases we’ve seen in the older models through cam timing changes are now often not there to be had.
Update March ’07. One thing I thought about long after I’d written this report initially (and removed and sold the 600 engine even in my Monster) was how my 600M felt a bit rough down low compared to others, and how it seemed happier on premium unleaded fuel. Being a tight arse the idea of putting premium in it all the time really upset me, but it definitely did smooth it out. Given the 600 has the 400 engine heads, the combustion chambers are rather small. To compensate for this somewhat, the 600 pistons have bowls in the crowns to lower the compression a bit, although it’s still quoted at 10.7:1, whereas the 750 and 900 were 9.2 9.5:1.
So recently I got to thinking about it (prompted by a question someone asked) and realised that the cam timing I was running would most likely have been causing it, combined with the single step ignition system these carb bikes run. The question that prompted this came from someone who had a high comp or big bore 750 that was pinging. He had been recommended by someone else to fit the longer duration 900 carb cams to reduce lower RPM cylinder pressure, thereby reducing or stopping the pinging.
So here’s another factor to think of with cam timing in very general terms what we’re doing when we advance the inlet cam timing (considering the basic air pump concept) is raising the cylinder pressure at any RPM under the torque peak RPM and lowering the cylinder pressure at any RPM above the torque peak RPM. Now if you already have too much cylinder pressure for your given chamber shape, ignition advance or fuel rating to handle, then advancing the cams is a backward step in that regard.
And the opposite is also true you can reduce cylinder pressure at any RPM under the torque peak RPM by retarding the inlet cam timing. But, keep in mind you may not always raise the cylinder pressure at any RPM above the torque peak RPM by retarding the inlet cam timing that’s also very influenced by the air flow potential of the cylinder head.
Another point. The better the heads, the less cam duration or more advance you can run before you start losing top end. Sort of inferred from the above, but better pointed out too. Good heads (not necessarily big valves) and shorter cams work better overall than bad heads and big cams (the old way). Bad heads encompasses small valve heads, not that small means bad, just bad for the application. 851SP3/4 being a good case in point. Results like this one http://www.visi.com/~moperfserv/joe_lee_872.htm of a 748 with 95mm bore (872cc) Bruce Myers built with 996SPS cams and std 33/29 valve sizes in well ported heads really prove this. Simple yet very effective.
Torque curve shape is another thing that many people don’t understand, especially how it relates to how a bike feels in use. On my 888 I ended up with a very flat torque curve (and smooth power curve) that was not overly exciting to ride, but made for good performance and acceleration from any RPM. On a 748 a flat torque curve takes away the top end rush that the model has std, something an owner might not want. It’ll be faster through the midrange, but won’t feel like it did before when it came on at the top end. Whether or not this is important is up to the owner. There’s no right or wrong in all this it just depends how you like it.
There’s a well known Australian tuner who has a trick of changing the Tacho drive ratio electronically on some of the 4V race bikes he prepares, simply to fool the rider into thinking they are using more revs than before. It’s a mental tune up on the rider, nothing more, but many will report better performance because of it. The main results are much better engine reliability / lower service requirements, simply because they aren’t revving as hard. And it’s a very good example of how sometimes you can make bikes go faster without doing anything to them.
Cam availability. Also see http://moto-one.com.au/performance/duccamspec.html for all the cam specs.
A very important issue here too. For the 2V engines from Paso to 900ie and ST2 there are 4 original fitment cams available, plus 4 from Vee Two (including an identical grind to the DP 900 cam) that I know of. So you have 8 profiles to choose from, some fairly similar. The later 620, 800 and 1000 all have unique cams; the 620 and 800 are interchangeable but differ from the older style slightly. You could probably swap them, but they are a bit shorter at the LH end and only have one bearing on the RH end not two. The 1000 cams run in the head direct and are nothing like any of the others.
All this means that the cam selection is not huge, especially when it comes to cam grinds you’d actually want to use. Add to that the cost of two of them and their uniqueness in terms of developing and grinding the corresponding opening and closing profiles and swapping cams gets pricey very quickly.
So it’s not like a small block Chev or even a GSX 1100, Honda XR250 or Harley, where you choose a grind and go forth. There’s even more Guzzi grinds out there than this. Which is the main reason I went looking for low cost mods that worked in the first place. And Ducati handed us an easy one.
Onto the practical side.
Now for some settings and dyno charts. Regardless of what a dyno chart will say, I’ve never had anyone tell me a bike with the cam timing reset has been less responsive or not as smooth. Although usually there’s some other changes made as well, we have done cam timing only jobs over the years and have never had a negative response from an owner. We did have two 750 with identical mods done around the same time at one point. One owner said it felt a bit better (he later claimed a 14 tooth front sprocket did more), the other said it wheelied out of corners in second. You can guess which owner was more enthusiastic about using the throttle. This also has a big influence on how obvious the result of the job is.
Some of these results go back quite a few years, so not all have air/fuel curves.
I also will point out at times how far you can go for a minimum piston to valve clearance of 1.5mm. This is based on checks we’ve made on std engines that we know haven’t been apart. If the seats have been cut the valves will be sunk slightly and you may get an extra degree or two. Generally, every 0.1mm of extra piston to valve clearance is another degree of available advance. As a general rule, aftermarket high compression pistons have much larger valve reliefs allowing much more advance. It is possible to go lower than 1.5mm; it’s just an accepted standard (60 thou in the old money). I have safely run engines down to 1.2mm, and would do so happily on my own engines. But that’s up to you.
I’ll start with the baby of the range, the 600M.
As you can see in the above table, the 600m cam spec is for 119 degree inlet centreline. On our bike, the “as delivered” setting was 114 degree inlet centreline, meaning the cams were already both advanced 5 degrees. This struck me as very unusual, but I’ve only checked one 600. I’ve done a few 750 (same cams) and they’ve all been around 120. When I advanced the cams I went as far as I could and still have 1.5mm piston to valve clearance in this case that was 107 degree inlet centreline, or 12 degrees over spec and 7 over as delivered. The 600 has small valves too so that certainly wouldn’t help. In the graphs below green is with cams at 114 inlet centreline, red is with cams at 107. The bike had Megacycle slip on mufflers fitted for these runs. First is power, the torque and air/fuel. Advancing the cams has leaned the mixture all through the range, as it did with std mufflers fitted. The result is a case of deciding on the compromise to suit yourself, typical of an under-valved engine.
750SS carb model. Same cams as the 600, so refer to the above table. Again under-valved as well. This bike also had a jet kit fitted so some of the result will be attributable to that. This bike had the cam timing around 120 degrees inlet centreline as delivered, which was advanced to 114, which is about the limit on a 750 for 1.5mm piston to valve clearance. Green is 120, red is 114. You can see the power peaks a little earlier with the 114 setting and then falls away quicker past the power peak. The fall away is so far past the power and torque peaks (5,500 RPM) though that it’s pretty much irrelevant.
900ss carb model. I got to try this recently on a trade in for curiosities sake, mainly because we just don’t see as many of these as we used to anymore. A ’97 900SS with Termi slip on mufflers. As delivered the timing was 114 and 116 degrees inlet centreline. This was advanced to 106 degrees, again this appears to be consistently as far as you can go with a 900SS and still have 1.5mm piston to valve clearance. So that’s 8 degrees advance on one cam and 10 on the other, 9 degrees over spec.
These long inlet engines make peak power at 7,500 RPM and that’s about all there is to it. After that the power falls away fairly quickly with this cam timing it falls away faster again. That may be a problem for some people, but it’s more the nature of the engine. If you were running short inlets and flat slides or Dellortos this could certainly be too much advance. You’d need to experiment. Heads would be a big factor too. I have seen graphs of these cams making 95 hp in a 900 with very well developed heads and short manifolds. With std heads and long manifolds low 80’s can be the wall.
The graphs below show 114/116 in green, 106 in red. The first graph is power, then torque and air/fuel. You can see the air/fuel trace go from leaner to richer about where the torque curves cross. This is expected on injected bikes, not so much on carb ones. It’s pretty rich too it may have had some wear in the needle jets (affects the lower RPM at WOT) and the 140 mains can generally be reduced to 130 on one of these with the std air box lid fitted. The third graph is acceleration on the dyno. You can see the greater peak torque of the 106 setting giving it the edge in the midrange whereas the greater top end of the as delivered 114/116 setting catches it up again. Although by that point the engine is past its best anyway.
Another 900SS we did years ago, compared to this one, has an after curve very similar to this before curve. We didn’t do a before on it.
900SS ie A model we’ve done quite a few of, we had one of these as a demo that I played with back in ’99, developing a modified ECU map in conjunction with FIM. A bike I very much enjoyed riding too.
I have no idea of what the std settings were for this bike I just fitted some Vee Two pullies and went from there. On later model bikes with adjustable pullies fitted as standard we usually find them set around 114 to 118 degree inlet centrelines. I’d assume this one was much the same. As you can see, IC figures around 65 degrees are becoming popular. Again, 106 degree inlet centreline gives about 1.5mm piston to valve clearance.
On this one I first tried 108 or 109 degrees, then after seeing the result went to 106. The graph below shows a ’99 900SS ie fitted with Staintune slip ons and an open airbox lid with K&N filter (I think, long time ago). All runs use the same DP ECU without any tuning for max power. Green is std, blue is 109, red is 106. You can see the power peak has come down from 8,200 RPM to around 7,800 RPM, but it still makes more power all the way through to the limiter.
The torque graph shows the 106 and 109 degree settings dropping the torque away a little quicker at higher RPM this is what some owners feel as less top end. It’s not less, there’s just proportionally more midrange improvement which gives the feeling of less top end. Given it’s a 2V, there’s not much you can do this simply (as a cam timing change) to help the top end. Retarding the cams might give the feeling of more top end, but it’d only be because of the diminished mid range. It’d be slower without a doubt.
906 Paso Not one I’ve done but some runs a guy named David sent me. So, thanks all to David. A nice illustration it gives too. Same cams as the 900SS carb model, but with the Weber 2 barrel downdraught carb with longish manifolds. Not the best set up, the manifold for the rear cylinder does a rather abrupt 90 degree turn just under the carb. These things really responded to 40mm Dellortos. Not so well to crashing into the rear of council works department caravans with the front wheel in the air in second, as it turns out. The fuel tanks are not kind to scrotums under such conditions.
As I understand the supplied runs graphed below, green is timing on spec, blue is 5 degrees advanced, red is 7 degrees. Being 113, 108 and 106 degree inlet centrelines respectively. No top end fall away because it’s all over by 7,000 RPM. Power first, then torque.
ST2 A 944cc fuel injected and water cooled 2V, with another cam grind that precedes the 900 ie cam by a year or so. It has a little more duration on inlet and exhaust, slightly different timing specs and the same lift.
Again, cams advanced as far as possible for 1.5mm piston to valve clearance. Again, IC around 65 degrees. I don’t remember where they were set as delivered, but it was somewhere about the spec, maybe a little retarded 114 say. I’m sure I moved them 8 or 9 degrees. I know the (lack of) result surprised me for the change made.
ST2 are funny things. They don’t make as much power as you would expect and have a long flat power peak. This one had Arrow slip on mufflers and a FIM UM161 chip. I haven’t tried one yet with an open airbox lid, although I’m currently rebuilding the engine in one I bought recently so I’ll know that soon. To ride they feel a bit doughy to me, the result of a much heavier flywheel I’d expect. It’s about 1 kg heavier than the normal flywheels. Makes them smooth and more tourer-like I guess.
Green is as delivered, red is with 105 inlet centreline. I get the feeling they’re over cammed personally. Plus the early ones at least (this one was, as is mine) looked like they had 907 heads they have the same oil leak from the chamber under the exhaust port problem 906 and 907 have which where never a high point in 2V head design.
1000SS ds ie The first of the late model 2V I’ve had a play with, this one made me have a bit of a think about what I was doing. Which is always a good thing. Compared to the 900ie and ST2 cams they have less duration and lift with bigger valves.
As you can see, the duration is the same as the 600/750 cam and spec for IC is 65 degrees. As delivered, this one was set to 116 degree inlet centrelines on both cylinders. I tried the usual ‘advance until 1.5mm piston to valve clearance’, which turned out again to be around 106 degree inlet centreline. Then I backed it off to 111 degrees and finally retarded them 6 degrees from spec to 121 degree inlet centreline just to see what would happen. Otherwise you don’t know. What I got is below.
Red is 111, green is 116, blue is 106 and yellow is 121. I’d say they got it pretty right. There’s a little more to be had through the midrange, but it has so much already there’s not a lot of value in it in my opinion. And it takes away from the top end with any advance at all. Surprisingly retarding the cams 5 degrees from as delivered dropped power everywhere, so the std timing specs are as good as it gets it would appear from these tests.
All runs are with Staintune slip on mufflers and the std ECU without any fuelling changes. The 116 and 106 runs are done with the airbox lid removed and the std paper filter sitting in place. The 111 and 121 runs are done with a cut open lid and a BMC filter. This does appear to make some difference to the air/fuel reading, but not the power.