Lizard928

-Only the exhaust,
-only the inside spring broke, outers were still 10% intact
-The springs hit coil bind at 0.679" (+/-.010") lift when measured. camshaft has maximum lift of .437" lots of room.
-Seat pressure was a smidge under 80#, installed height of 1.398"
-I worked with the spring manufacture to find a spring that would work well, and work with mild boost pressure. All ramp rates etc were provided to the spring manufacture.

Mike Simard

This is no surprise, not just these springs but aftermarket valve train parts lately seem to do this a lot in recent years. By "this" I mean breaking with no traditional problems like coil bind.

I don't recommend aftermarket springs with the cams I make. I designed the lift profile to not need high spring pressure too, there's a lot of possibility for valve train problems or success from small design parameters that you can't see when looking at a cam lobe. It's possible to induce bad harmonics from a poor lobe profile and a heavier aftermarket spring is itself a problem.

A well designed lobe will be able to use surprisingly little spring pressure and not float. Putting a heavier spring than needed causes high wear at idle in traffic.

ptuomov

Did you measure the coil bind yourself or did you rely on the manufacturers spec and the 1.398" installed height? Also, did you measure the coil bind for both the inner and outer spring? Regardless, there seems to be a lot of room so I'd say coil bind is unlikely.

If only one spring type breaks only on one camshaft type (inner, exhaust), then my guess would be that there's a resonance issue, possibly combined with some small manufacturing flaws. The exhaust cam profile may resonate with the inner spring at some rpms. Out of curiosity, two more questions:
1. were all broken springs on either the smaller or the larger tri-flow exhaust cam profile?
2. what's the resonant frequency of the inner spring, the spring manufacturers usually report that?

The seat pressure doesn't say much to me without the spring rate. But my off the cuff complete speculation is that the spring loads are higher than they would have to be. My personal view is that the valve springs should be pretty close to the edge and barely keep the valves from floating and bouncing. If a dual spring setup breaks one of the springs, then the piston should hit the valves. Since they didn't, from the first look it seems that the springs loads were unnecessarily high. Supporting this conjecture, doesn't Alex "Cheburator" run this profile with new stock GTS springs, or am I mistaken about the profiles? If you know the maximum negative acceleration and the positive and negative extreme jerk values of the exhaust cam profile(s) that had broken springs, we could try to do a bit of spring math.

Because I think the spring load should be right on the edge, I also think that single conical/beehive springs are better than dual springs for our kinds of engines. If the dual spring is right on the edge, then either spring breaking should kill the engine. With a single beehive, the failure probability is approximately half of the failure probability of dual spring (this is a gross generalization).

On factory vs. aftermarket springs: Factory springs are usually better because they have to warranty the whole engine and because they know the exact camshaft. That said, whether the spring is from a factory Porsche or factory Ford, the spring and then engine don't know it. Both companies have a lot of resources and high warranty costs in the case of failure, so springs don't fail with the intended camshaft profile. Different cam in a Porsche with Porsche springs or in a Porsche with Ford springs both work or fail the same way if they match or don't match the cam profile.

Many aftermarket performance companies try to sneak in whatever junk steel made in China to squeeze a bit more profit and then warranty only their component. There appears to be at least one reputable aftermarket spring manufacturer left in the US, Peterson American Corporation, whom I personally trust. But regardless of whether they are factory springs or PAC springs, the spring still has to match the cam.

Not pretending to be an expert, just trying to get a conversation going about valve springs so I could learn.

linderpat

to answer the question first posed - my view is that no, the vendor has no responsibility for your labor. The typical rule of thumb is to replace the broken part. That means the vendor would not have to replace the springs that are not broken either. Any other result would make it cost prohibitive for vendors to sell the stuff we need. Sometimes things fail. Go after the manufacturer if you can (probably in China using cheap steel).
Having said that, some vendors for purposes of good relations will step up and go further and help with your other consequential losses, like lost labor. I have no idea who the vendor is, btw.

Lizard928

Tuomo,

Danny has base stage II cams, but tri-flow is only on intake cams, not exhaust. So no difference there.

I'm afraid I didn't get the resonant frequencies.

As to coil bind, I measured it myself.

Yes Alex is running stock springs, this scares the crap out of me. Because stock springs are very weak, and hit coil bind early. I kept very nice ramp rates on my grinds, but still errored on the side of caution, and with the ability for some boost to be run without consequence.

ptuomov

Since it's a straight spring, I think one can compute the resonant frequency from the basic spring characteristics, including installed height, seated load, spring rate, and number of coil loops.

Re: Alex "Cheburator" springs and cams. Because all these computations aren't always accurate, empirical data are valuable. It's highly informative to me that Alex can run the stock GTS springs without valve float with your Stage 2 cams. If that works, then that's the type of loads one should run with those cams. Running too high loads leads to nothing but power loss and wear.

Boost doesn't require any large intake spring loads. Turbo does require slightly higher seated loads on the exhaust, but only with very restrictive hot sides.

DJF1

So my ill-informed conclusion would be that the factory S2 springs could actually be the ticket in my case??
I'm extremely hesitant to put the same type of springs back in the car and button it up without knowing if it was a wrongly manufactured batch of springs that ended up in my car.

ptuomov

First of all, don't trust anything I am saying. Think with your own brain. I believe this valve spring stuff is a problem that a hobbyist can work out with effort, but I don't know if it's true.

What I would do in your position would be the following. Send the cams to someone with a cam doctor, like speedtoys. Or someone with laser measurement system. They'll get you a data set with the two intake and one exhaust profiles. They can also check for manufacturing consistency of the cams.

Next, measure the valvetrain component weight. Lifter is hardest because it requires an assumption about oil. (If you get new lifters, hardness test on the surface is not a bad idea.)

After that, spend a little bit of money buying software that can do valvetrain dynamics. Since we don't have pushrods, software is easy. Use the software to determine what spring force is needed at each lift.

Buy springs matching that.

Just an idea.

Lizard928

I'm not at my computer so don't have my notes from when I was gathering all the data.
Can anyone confirm what installed seat height should be on the 944S2 springs, along with seat pressure, pressure per inch, and exactly what lift they see coil bind at?

BC

I don't have the numbers in front of me, but I know they are sort of high in relation to the other sets we use, and at our installed heights its best left to larger lift cams.

ptuomov

By the way, one thing that people suspect is that Porsche is changing the spring specifications. The demand for spare parts is so low that they may be now be selling the same spring under many four-valve part numbers. for example, what are the odds that the 968 valve springs are the same as what they were when the car was new and not something like 944 S2 springs?

This is pure speculation, but might explain why Colin's Stage 2 cams seem to work with new "stock" 928 springs...

Rob Edwards

Notwithstanding that 944S2 and 968 valve springs are the same part number, I'd like to believe that the Germans wouldn't do something like that, kind of smacks of fraud, no? I'd also like to think that any engine builder that matches sets of springs of a given factory part number would have noticed that the free length and/or spring pressure at a given compression had changed, and would have numerical data to back it up. Without _proof_ that they're different, what good is speculation?

I don't understand the economics of old German car parts supply very well, but I do know that several flavors of 944/944S/944S2/968 springs have tripled or quadrupled in price in the last three years. On the surface that says to me that they're selling plenty, though it could also be that it doesn't make sense to make more, so they're metering out the remaining supply with discouraging prices.

ptuomov

I think the explanation is that I just got confused. I have different spring tables for stock original 968 and 944 S2 springs. But it's probably just explained by something like different installed heights or spring-to-spring variation.

Rob Edwards

I do recall buying more than 32 sets of springs in order to find 32 that were within Greg's acceptable variation. And more than 32 lifters. That guy is a perfectionist pain in the ***.

Ducman82

any update?