olmann

Thanks for the feedback Roger. I will move forward with the S3 cams. Might be interested in the GT cams.

the flyin' scotsman

I assume the GT cams bolt right in? no need for new valve springs?

Need a Greg rebuild to go along with the cams

EMan 928

I put GT cams in my 93 GTS. Bought the cams from Greg. He also suggested different valve springs, which I did, but I believe you could still use the original springs if the car was primarily for street.

Rob Edwards

Factory 32V valve springs are all the same, '85 to '95.

EMan 928

You're right. But I think Greg sent me stiffer springs, maybe 944 type?

ptuomov

If it's the stock GT profile and if your engine redline is stock, it won't need stiffer springs. If it's a more aggressive profile ground on GT core and/or you spin it way higher than stock redline, it might need it. And if you need new springs for some reason anyway, then there's the question of cost.

jcorenman

Great summary. Of course this assumes everything else is the same: lift, duration, etc. Which it never is... except for the S3-GT cam comparison. Those are the same specs with only LSA changed, 114 deg for the S3 and 110 for the GT.

There are a number of reasons why they might have: Flappy is one, but also the wider LSA of the S3 cams reduces dynamic compression (lower combustion pressure) which reduces (and broadens) torque a little, but more importantly reduces the tendency for detonation. Remember that the S3 did not have knock detection.

The larger question is why Porsche went to smaller cams (less duration) with a tighter 106-deg LSA for the S4. It is arguably a good match for the auto, but not so much for the 5-speed. Which of course segues to the introduction of the GT two years later-- back to S3 cam-specs with a tighter 110 LSA for the 5-speeds.

Porsche says that the max torque is the same for the S4 and GT, and +10hp for the GT at a little higher RPM (316 @ 6000 vs 326 @ 6200, SAE). I don't think anyone really believes that, GTs typically dyno maybe 20hp higher than S4s but there are large variations. No one, as far as I know, has done a before/after for a S4/GT cam swap with nothing else changed. Certainly there is a difference in character, with an eagerness to pull to redline that is much less pronounced in the S4.

It is hard to argue with the comment that GT cams should stay with GTs, there were only 2000 or so made (a few hundred came to the states). But fitting GT cams to a 5-speed S4 (or GTS) would be very tempting.

There are a couple of caveats: The factory used the same springs for the GT in spite of 1mm higher lift, and they are marginal in the GT which leads to pitting. Also the newer low-zinc oils have taken their toll on cams and lifters for those owners who were unaware or made poor choices. So finding GT cams in good condition is rare, and they won't be cheap.

Which leads to other choices. GT cams are still pretty conservative as cams go, but doing anything bigger really needs to be part of a general upgrade. I don't think there is any point in fitting larger-than-GT cams without also doing headers and a less-restrictive exhaust, at least some mild porting of the intake, Sharktuning, etc.

A slippery slope, but fun!

ptuomov

There's the question of GT cam pitting. Does anyone really know whether that happens at low rpms or high rpms?

If the pitting is there because the spring loses control of the valve at high rpms, then stiffer springs will help. There should be other symptoms of loss of valvetrain control in GT at high rpms if that's the case, though. And a lightweight lifter is the simplest solution (with other benefits).

If it's basically for any other reason, higher spring load on the nose will make the pitting problem worse, not better. In particular, if it's low rpm pitting because of the oil film thickness isn't great for the GT profile (that's determined by the lobe geometry), then stiffer springs will make the pitting problem worse not better. The only thing that I'd expect to help for sure against pitting on the nose is a little higher idle speed. Lighter lifters and correspondingly lower spring loads might help as well, or at least not make anything worse.

There's also the question of higher lift you bring up. Holding the duration constant, higher lift requires more aggressive cams in terms of average velocity and likely also peak velocity and peak acceleration. But when comparing GT and S4 cams, the duration is not held constant. GT has a longer duration than S4. The digitized acceleration profile of those cams is very similar in terms of what it requires from the springs to keep the valvetrain under control. The stock springs give GT cams a little higher load on the nose than they give S4 (due to higher lift of the GT profile), which I don't think is necessary for the same rpm. But considering that GT redline being higher than the S4 redline, then it's probably spot on in terms of the load on the nose.

Then there's the question of what else needs to be done with the cam upgrade. I generally agree with you that other changes should be made at the same time. I think that the most important issue in terms of cams is the intended rpm range. I've got three points on this topic, which I believe to be true, but who knows.

First, the 928 exhaust manifold geometry is such that the 90-degree interference really starts hurting two cylinders above 6000 rpm. #1 is the first victim and #6 is the second victim a couple hundred rpms higher. In my opinion, it's a bit of a fools errand to try to cam the engine to 7000+ rpm if the stock exhaust manifold is retained. One needs headers, really any kind will do that will delay the 90-degree interference pulse from screwing up the overlap of #1 and later #6. We've really dug into this problem with the turbo exhaust manifold design. Fortunately, turbo doesn't need to be spun to 7000+ rpm to make power, so we're good.

Second, if the valve events are set for high rpm use, the later IVC will take away low-end torque. One really should increase the static compression to reclaim some of that lost low-rpm torque, which is lost because of the later IVC of the longer duration intake cam. There's relatively little downside if this is done in moderation.

Third, to make good mid-range power, one needs camshaft overlap (i.e., a tight LSA). With a plenum manifold, this is a tradeoff between idle quality and low-rpm surge at low throttle angles and WOT torque. With ITB, less so. With either ITB or plenum manifold, it's a tailpipe smog test tradeoff as well at low rpms.

Just to give you an idea about how beneficial overlap is to torque production, consider two engines.

First, 968. The 968 Variocam shuts down the LSA to a very low level and has significant overlap in the 1500-5500 rpm range. It's not exactly the same setup as the 928 otherwise, but I think it tells you a lot that the 968 advances the cam by 15 (!) degrees from 928sh settings at 1500 rpm to make more torque. The 1500-5500 LSA in the 968 engine is about 105, and that's with 224.5@1mm intake cam and 210@1mm exhaust cam.

Second, the 2013 Engine Masters winner was Kaase's 4-valve modular Ford. The scoring formula looks at torque and power in the 3000-7000rpm range, right where the 928 should be run. That engine had 230 degree intake and exhaust cams at 0.05" and 98 degree LSA installed straight up. It made a stupid amount of torque throughout the competition rev range. 11.4:1 compression and headers, though. That's another piece of evidence that if you don't need to pass tailpipe emissions test and can live with a higher idle, tight LSA is the way to go.

All just logic as theory, as always. If it makes sense, then it might be correct.

928cs

Please note that Porsche was still playing with the cams when the S4 was introduced.
They have produced a very early CS prototype in September 1986, certainly with modified S3 cams.

And what is known as GT cams are in fact CS cams. Even if the references are different, they are identical.

I think they used the same cams for the manual and automatic S4 only for a matter of cost.