748SP - Written 06/11
Summary: 748SP with changed cam timing, with comparison of result to other 748SP, 748 base model and 955cc with similar spec.
This is the first 748SP anyone has bought me to play with. For those not familiar with the model it’s a 748 with long duration cams the last generation of this style inlet which has G lift, but 10 degrees less duration and the A exhaust from the 851SP. The 748SPS from 1998 onwards had Ti rods and the 45/50mm exhaust system, but the earlier SP had normal 748/916 steel rods and 45mm exhausts. Chassis upgrades were single seat with aluminium subframe, cast iron front discs, Ohlins rear shock and yellow paint. As a general rule they had little midrange and needed to be revved.
This one was a dedicated track bike, so I didn’t get to ride it at all. My main aim was to try to fill out the midrange a little and see how the power curve responded overall. The last bike I did with these cams, a 955cc, big valve 748SP, had the inlets set to 97 degree centreline. This gave a power peak at 8,500 rpm, a dip at 9,500 rpm and then another peak at 10,750 rpm. Quite a wacky shape actually. It’s in the 955 report. The lesser capacity of this 748cc version raised the rpm at which the peaks and dip occurred, but the shape was much the same. Not sure why, I’ve dynod 955cc versions of the 888 SP5 and 916SP with the G/A cam combo and they have both had this dip and not had this dip. Maybe it’s exaggerated by this inlet advanced timing.
I checked the piston to valve clearance and the most I could advance the inlets was to 94 degree centreline 10 degrees advanced from spec. So we ran them at that and the exhausts at 103, which is 2 degrees retarded from spec. It had Arrow 45mm slip on mufflers, a large inlet front fairing without screens and a UM141 eprom.
Not being able to ride it there are no before dyno runs and the after runs are not directly comparable to other 748SP run at Dynobike as their new Dynojet 250i reads about 5% or so lower than the old 200. I did a rough spreadsheet comparison with 5% added so the curve shape could be compared, but as far as actual power comparisons I don’t have one.
I did some + fuel runs to see what it liked in terms of fuelling and it was the usual inlet cams advanced result, but quite exaggerated. The first graph shows the power runs, with std UM141 fuelling in blue, +5% in red, +10% in green and +15% in purple. More fuel really fixed the 5,500 rpm hole, but it also killed the top end. But it didn’t affect the dip at 10,000 rpm (neither does changing ignition advance it seems), although possibly it need -5 or -10% in that range to really make a noticeable difference. The air/fuel curve shape is very choppy, but you can get that when playing with cam timing, especially when it starts moving the tuning peaks around. Power graph first, then air/fuel.
What really surprised me was that the +5, +10 and +15% air/fuel traces converge around 11,000 rpm, indicating that it’s running out of time on its single injectors. But this is where the need for dual injectors per cylinder arises it’s not so much capacity driven as rpm driven.
A cleaner curve is shown below.
Based on the dyno runs I modified the fuel mapping and created another eprom, making it quite a bit richer in some places and quite a bit leaner in others.
You can see how the torque curve drops after 9,000 rpm, which kind of kills the power above 10,000 rpm. Even though it makes 5 more hp at 11,000 rpm than it does at 9,000 rpm, the hole at 10,000 rpm made me wonder whether it’d be worth revving it above that. So I explained all this to the owner, who took it to a track day the next week and tried it out. To add some extra to his day, it was also his first time out on the bike with a slipper clutch, new forks (reworked 749 from the look of them), magnesium wheels and slicks. I made him two eproms, both with the same revised mapping, but one with the std rev limit (11,750 or so) and one with the rev limit lowered to 10,250 rpm. His comments are below:
Wow, that made a difference! It pulled beautifully (relative term) from almost everywhere. The big dead spot was a much smaller dead spot and didn’t pose as much of a problem as previously. I could exit corners mid rev range and not have to worry about getting bogged down, it was brilliant. As the power didn’t come in with such a big rush it was easier to get onto it earlier, which gave me better pace out of corners onto straights. It was actually almost funny how much earlier I could get on the power, and I would easily make big overtakes on the exits….only to get dusted as soon as the faster bikes were upright.
I was still (over) revving it in the early sessions before it became obvious that there wasn’t much benefit in doing so, and started to change earlier. I will be fitting the lower rev limit eprom before the next track day, and see how that goes.
The slipper clutch also turned out to be great too, the bike was much more settled on the way into corners and let me get the gearing right for the exit. Having said that I don’t think I have the right gearing for Phillip island on it, maybe one or two more teeth on the back would be ideal.
Some might think that taking the highest revving engine in the range and changing its characteristics to allow 1,500 rpm or so to be taken off its limiter is a bit strange, but that’s how it worked out. Not having ridden the bike personally I can’t comment directly on how it feels to ride. And if you were paying someone to race it then there’d be no way you’d lower the limiter. Or run this cam timing to start with I know that Duane Mitchell mapped the 748SP run in the national Supersport championship by the Ducati Dealer Team with 100/102 cam timing as that was what they came up with as the best for racing. But as a road or ride day bike taking those revs off over 10,000 rpm will make the valve train much happier.
The spreadsheet I made up simply has 5% added to the runs of this bike to compare it somewhat with two other 748SP run on old Dynobike dyno. The result is the next graph. As you can see, the curve is exaggerated below the first peak and the hole deeper after it, with the peaks all moved down the rev range by 500 or so rpm. Except for the peak at 4,750 instead of 4,250.
I modified the above graph to show only this bike and ended it at 10,000 rpm. In this light, it doesn’t look that bad or unusual.
After I had dynod this SP I ran a std 748 on this dyno in preparation for it getting an 853 kit. The comparison of these two is below - power first, then torque, SP in blue, base model in red. The SP is 10 hp stronger at 9,000 rpm, something I didn’t expect. And overall is stronger in the 6,000 to 10,000 rpm range, but the delivery is not so smooth. Depends how you like it I guess. Changing the cam timing on the base model bike would improve its midrange, possibly to the point where it would be stronger under 7,500 rpm, but I don’t think it would lift the peak power much. On the 748 the cams were 120/111 and 125/105 when I checked the timing after the 853 kit went in, so there was certainly some midrange going begging on the graphs shown.
Final graph compares this bike to a 955cc 888 running the SP5 G/A cam combo (90/103 timing) and large valve heads run on this dyno. This one was limited to a bit over 10,000 rpm to help the motor, but you can see the curve is similar in shape, but with more everywhere and the tuning peaks moved down the range as expected.