955 Kitted 748 and 916 Compared - Written 01/07
Summary: 748SPS and 916SP fitted with 955 kits, comparison of results between them and other models of similar capacity and/or spec.
The two bikes of this report are:
1/ a 748SPS (748SP with Ti rods, ’98 ’99 models) that had previously been fitted with a 66mm crank, 96mm piston kit, 37 inlet and 31mm exhaust valves. Running original 748SP cams.
2/ a 916SP stripped to have cracked crankcases replaced, fitted with a 96mm piston kit and Carrillo rods. Otherwise original spec.
The cam specs are shown below. Both use the same exhaust cam the original 851 kit ‘A’ grind, although they’re spec’d a little differently in the manuals. The 916SP has the ‘G’ inlet, the 748SPS the ‘748SP’ inlet. As usual, we reset the cam timing to non std specs, which are given below the std specs for each model.
The 916SP reset inlet timing of 62/62 is quite funny in number terms, and shows just how long the ‘G’ cams are. In both cases we had the inlet valves closing around the 65 degrees ATDC point.
This 748SPS had something of a turbulent history, and was one of those bikes that many of the usual suspects I spoke to knew about, but denied any involvement in. One suggested quite strongly that I tell the owner to come and get it and to refuse to work on it. Another gave me a “last time I saw that bike” story that was somewhat disturbing. The new owner didn’t really know much about it, except that it didn’t idle, although it allegedly had different cams in it and had been recently rebuilt the usual deal. I’m never interested in getting into the ‘who’s done what’ crap, but needed to know exactly what I was looking at so got into it a bit. The cams spec’d out as 748SPS and it had very light (Corsa) valve closing springs fitted, hence the non idling. The heads were done by Vee Two, running 37mm inlets and 31mm exhaust valves with opened ports and chambers. The owner had bought a new set of dual injector throttle bodies which we fitted running as dual.
One easy way to tell if a poor idle is a valve closing spring issue (assuming you have good compression and leak down) is to watch the idle vacuum. I had the idle stop wound up quite a lot to get it to idle at 2,500 rpm, where it had good vacuum. Wind the stop down to reduce the idle though and as soon as it hit 2,000 RPM the vacuum dropped to almost nothing and the engine stopped. Someone had mentioned the valve closing spring thing to the owner before it came to me, so the idea that it could be the issue was conveniently already in his head.
We pulled the heads to check the allegedly recently replaced pistons, give it a quick head job and fit original heavy valve closing springs. Heavy is a comparative term they’re still not that strong, but close with much more force than the light springs. The heavy springs were introduced with the ’91 model 851 and 851SP3, and make a huge difference to idle quality and low speed running, especially when the closing clearances open up a bit. We revisited this recently with the ’04 ST3, which had the exhaust valve closing springs replaced under a service bulletin warranty with stronger springs to fix the bad idle and low speed running issues they were having, just like my ’89 851 does.
At this point this job got more involved, for all sort of reasons that you don’t expect when you start. One of those bang head against wall and watch the bill blow out jobs that always cost us money. In the end we split the difference between the original quote and what we should have charged to arrive at a total the owner was still unhappy with. As he said “I could have bought a 996 motor for this much”. Pity he didn’t say that much earlier.
Anyway, we put it all back together and fitted an eprom based on the 748SPS with 25% more fuel overall and spark map revisions based on the 916SP and 955SPA eproms as a starting point and tuned it from there. Turned out fairly acceptable, but I was somewhat disappointed with the result power wise.
The 916SP belonged to a regular customer who only tracks it. Prior to him it was raced in New Zealand by the importers and ridden by Robert Holden. Not surprisingly, it cracked the crankcases. The crack that made us aware of the problem ran from the RH main bearing out to the horizontal cylinder base under the timing shaft. It leaked oil and we assumed it was the timing shaft seal until we looked hard. As it turns out, it also had cracks from the LH main out to the timing shaft and gearbox input shaft bearings, but they’re hidden under the alternator cover.
The new 916SP cases we used are late production pieces they had a ’02 or so date stamp and the cast in part number was 22730088C from memory, the same as the late production 888SP5 I have. They’re very much a 996 style case, with the vertical cylinder cross bolts from the RH side, bigger oil galleries and are much thicker under the timing shaft where the original 916SP cases cracked.
The owner was somewhat apprehensive about the bolts in the Pankl steel rods, which have a specified life of 3 torque cycles only. Given a set of 4 Pankl rod bolts (2 per rod) is about the same as a pair of Carrillo rods (yes, really, around au$1,200), he went for the Carrillo. He had the whole assembly balanced and we reassembled it, going through the rest of the stuff like we usually do.
I didn’t check the final compression ratio on either engine, but going on previous engines like this I’ve done I’d expect it was around 12:1, maybe a bit higher on the 748 based engine as its combustion chambers looked a bit smaller. We just set the squish on both to between 0.90 and 1.00mm and left it at that.
The 916SP we tuned at Phillip Island using the Lambda probes and data logger. It’s quite surprising how much more fuel the vertical cylinder of the 916 style bikes likes around 6,000 RPM compared to the horizontal, and how confined it is to a small RPM range.
All the graphs I’ll show in this report are generated by a spreadsheet, not the Dynojet software as they’re all made up of various +/- fuel or spark advance runs. I don’t really have any straight before and after runs for any of them.
The first graphs are for the 916SP, before and after. Power first, then torque. For an increase of 39cc (4.3%) and inlet cam timing changes it’s a pretty good result. I’m not sure how much I advanced the inlet cams, but I’m guessing this bike would have been set around original spec in its previous life, so about 9 degrees maybe. The heads and ports were untouched, and it looked to me like someone had sunk all the valves quite a bit (1 to 2 mm) at some point. The seats were certainly much lower in the heads than I would expect. Didn’t seem to affect it before though, as this bike dynod almost identically to another ‘as delivered’ 916SP we dynod long ago.
The ‘after’ power curve may have come up again after 10,500 RPM too, but I didn’t do the after runs so it wasn’t run into the limiter like I usually do. We could probably even drop the limiter in this case to save the engine a little, but this owner doesn’t use the limiter like many others do so it’s no big deal. He was certainly impressed with the power up to 10,500 RPM, and felt it was much stronger at the top end. Given the power peaks around 8,700 RPM, the longer duration cams help hold the top end feeling, as both the power and torque curves are heading down from 9,000 RPM onwards.
To the previous graphs we’ll add the 748SPS based 955 in blue. To recap: in comparison this particular engine has bigger valves (inlet +3mm, exhaust +1mm), 10 degrees less inlet cam duration with the same lift, the same exhaust cam and larger header pipe diameter (50mm or so compared to 45mm on the 916SP). Both bikes ran 50mm crossover and mufflers.
I really thought this engine would have made more than 120 Hp. Had I been allowed to play with the 748SPS a little I may have advanced the inlet cams some more to help fill in the midrange, but as mentioned it was way over budget and I was sick of the sight of the f#@king thing anyway. I guess the surprising part is the fact both bikes have an almost identical power peak at 8,700 RPM. Most of the 916cc and up 4V engines do, regardless of tune spec. The bigger valves hold the power out the top end much better, but unless you wanted to rev it out a lot that’s almost irrelevant. On the road using road RPM it certainly wasn’t that inspiring. Given the valve sizes it has I’d think it would be a much better road bike with std cams, but that’s only speculation.
I have heard from other people since who are running 955 engines with G cams. In those cases they’re running G exhausts as well, which are quite a bit bigger than the A exhausts these engines use, especially in terms of area under the curve. These engines make good power around 10,000RPM and hold it out to 12,000 or so, which appears to be a better combination. The 926cc 888 production Corsa of ’94 went to the G exhaust (as would have the factory 916 Corsa of that year), and they ran that same exhaust cam until the end of the Desmoquattro Corsa in ’01.
I’ll add another couple of curves to the previous graph, being a 916 Strada with SPS cams and 50mm exhaust in orange and a 996 Biposto with dialled std cams and 45mm slip on mufflers in pink. Just for comparison purposes.
The 916 with SPS cams does pretty well against the 748SPS and 916SP, but as expected the simple capacity advantage of the 996, combined with its std cams and good sized 36/30mm valves, easily gives the better road bike power curve.
The next graph I’ll show is a bit of a contrast. I’ve dynod a couple of 888SP5 (one of which had been 955’d), which is basically the same engine as the 916SP in a different chassis. The two implications of the chassis variation are quite simple airbox and exhaust. I know a few people have said the 888 chassis (and specifically swingarm) gives a better style exhaust system, whereas the 916 chassis has a clearly better airbox configuration. Although both exhaust use the same 45mm headers into 50mm crossover and muffler sizing. 916SP is green, 888SP5 is blue. Both bikes are original capacity.
Next up I’ll add the 955’d 916SP in red and the 955’d 888SP5 in pink. The 888SP5 was not one I’d been involved with apart from some basic mapping on the dyno, so the unknown cam timing settings may influence the shape of the curve somewhat, especially the mid range. The 888SP5 power curve has kept the std 888SP5 bottom end shape, but the top end mimics the 916SP. Which kind of shoots down my exhaust style theory.
This report is a little like the 1026 report, where the power output of an engine combo is influenced (for better or worse) in parts of the RPM range by things like exhaust, etc. I’ve never done an engine with the G exhaust cam, but maybe it’s what these big inlet cam engines really need.