996 TO 1026 Stroker with SPS Cams - Written Sept '06
This bike was originally a ’94 916 that the owner bought a 996 engine to put into. Being one of the first year single seat bikes (Strada) it was fitted with the P8 ECU, so it required a 996 wiring loom and 1.6M ECU conversion. The main differences between the ECU are that the P8 uses two pickups one on the flywheel, another on the cam gear while the 1.6M uses only one on the cam gear (with a totally different trigger pattern). Also, the P8 has external ignition drivers, whereas everything is internal in the 1.6M ECU. Meaning the wiring looms are totally different. Finally, the P8 can fire two injectors per cylinder independently (one at low pulse widths, two at higher pulse widths) as it has 4 injector drivers, whereas the 1.6M only has two injectors drivers (one per cylinder), firing any injectors for that cylinder at the same time. This is what leads to the 3,500 RPM roll on misfire the 996 Biposto models experience (for whatever reason), although with SPS cams the problem isn’t apparent.
The owner brought in the 996 engine he’d bought that allegedly had very low km on it, a 900SS crank (3 phase ’99 on 900SS), some Ferraci 996 stroker pistons and some 996SPS cams. I tried to talk him out of the whole engine splitting thing given all we were doing was fitting the 900 crank (and reshimming the cranks, usually reducing the preload a bit in the process) due to how much money it sucks up. The 900 crank gives 68mm stroke, compared to the std 996 66mm stroke, for 1026cc. Personally I don’t see the extra 30cc as anything special (3% capacity increase, so theoretically 3% more peak torque). And, in real terms, there wouldn’t be much power in it due to the std heads he didn’t want to get into. But he was very keen so off we went. I fitted the SPS cams to the heads when they were off (much easier) and set it all up when it went back together.
We ended up using the std pistons and custom 1.2mm base gaskets (although Bruce Meyers later told me he just uses two 0.6mm base gaskets) instead of the Ferraci pistons. This was because the Ferraci pistons, although being 1mm shorter in the compression height, were made for big end located rods. Given all 996 have small end located rods I’m not sure why these pistons were made that way, and I didn’t get any sense out of Ferraci when I tried contacting them about it. Plus they had very big valve cutouts you’d be able to run G or other Corsa cam grinds with them that would have dropped the compression about 0.3 points. As it was finally assembled, with the std pistons, 900 crank and 1.2mm base gaskets it had 0.90mm squish and 12.75:1 compression. Enough for every occasion I’d think. I wasn’t concerned about starting issues with that much comp as it was a ’01 engine with the revised starter drive ratio. The earlier build 996 can get a bit cranky on starting when you run SPS cams at this timing even with std compression.
The 996SPS cams were set at 107/107 centrelines as we normally use for the SPS cams in 996, mainly because the UM239 eprom we use was developed on a bike running 107/107. It’s about the most advance you can consistently get on the inlets without piston to valve clearance hassles in a 996 engine, and the exhaust timing is spec. I had played with another 996S a little while previous to this one and had changed the two top lines of the spark map to give less advance, so that’s the eprom (UM239B1) we started with here.
The owner picked up the engine and installed it himself, bringing it back after he’d done around 1,000km to have the belts adjusted, the idle mixture set and some dynoing to see how it went. He was very happy with the engine and how it went compared to his 916. He had changed the gearing from 15/38 to 14/38 on my recommendation, as the 996 engine has the 1.84:1 ratio primary drive compared to the 916’s 2:1. So the effective gearing at the rear wheel is about the same. After the belts had been done and the idle set properly I took the bike to the dyno to see how much more the crank had given it as compared to other 996’s fitted with SPS cams.
The one other difference between this bike and others of 996cc I’d done was the exhaust. I think it’s an Arrow full system, due only to the unique carbon fibre heat shield having the Arrow logo on it and the mufflers attached using Arrow look clamps, but the mufflers almost look Termi (and had Termi decals) so I really don’t know. Visually, the system comes from the head in the two small pipes then joins to one and tapers up from there to 50mm. It’s quite a distinct system to look at that I haven’t seen before.
On the dyno I did some +/- fuel runs with the Hand Held Terminal to see what it wanted fuel wise. I was a bit surprised by the peak power and the midrange, both less than I expected. Actually, I didn’t know how much peak power to expect, but I thought I might see over 130 given the SPS cams. But, it would appear std heads are good for around 130 no matter what else and that’s the end of it. The graph below shows the extremes of the +/- runs, base is green, +10% is red and 10% is blue. From these runs I went back to work and made up a custom eprom as required. Power first, then torque and air/fuel. The air/fuel curves show how much difference 10% makes.
Although the curve looked pretty smooth, the dip at 5,000 RPM surprised me a little, so I tried comparing it to a few other 996. The next graph shows this comparison, with the base run from above in green. Blue is a 996 with std cams at 108/108, Arrow 45/50 half system and squish set to 1.0mm giving 12:1 comp. Red is the original shop demo 996 with cam timing at 108/108 and Termi mufflers. Purple another 996 with SPS cams at 107/107 and Termi mufflers. Yellow a 996S (SPS cams, 50mm Termi system) with cams at 107/107, heads ported by someone else with std valves and Pistal 12.5:1 pistons. All fuelling set for max power through the range, the SPS cammed bikes all running UM239 eproms, the std cammed bikes either UM191 (way back when for the shop demo) or UM222 (the single injector one, making 126hp!).
As you can see, the 1026 really doesn’t do anything flash until 9,000 RPM, and on the road the bottom end in particular certainly feels flat. I was quite surprised when I first opened it up it really didn’t feel that strong. But if you look at the torque curve below (green line remember):
You can see the angle of the curve is quite steep from 5,000 RPM to 7,000 RPM, where it flattens and peaks. On the road that translates to lacklustre at 4,000, coming on at 6,000 and if you don’t back off you’re off by 7,000 RPM. And that’s in second gear. The 14/38 gearing helps, but I’ve ridden modified 998S with that gearing and they don’t come on like this thing does. So when it gets going it’s very strong, but in comparison to some of the company on this graph there’s a lot of money not delivering. As the only real difference is the exhaust, that’s what I suspect. Don’t know why it’d have this much impact, but there’s nothing else really different. And realistically, the top end is lacking because of the std heads. The ported heads with std valves really haven’t done anything, so bigger valves are the next place to go. Bruce Meyers says 38/32 really helps them.
So that’s the 1026. Overall this report is one of those “well, if it wasn’t for …..” ones that is very much relevant to this unique build. I don’t expect it to be indicative of all 1026 jobs, and spending money on the heads instead may yield a much better result. But it comes down to how much do you want to spend I’ve done a rough quote for someone else who has a 996 engine he wants similar things done with, plus porting and bigger valves, to put into a 916. For us to strip and reassemble, do the heads with bigger valves and porting, supply and fit SPS cams, supply eprom, tune, etc, is coming out at over $10,000. It’s a hell of a lot of money.