Cam Sprocket Alignment

[bdonally]

I am at the stage of checking the cam sprocket alignment.
I clamped a straight edge across the case:

Then measured the offset to the sprockets and to the nose of the intermediate shaft, to calculate the offset according to the manual. Did one side at a time, measuring three times, breaking down the setup between measurements.
While set up I decided to measure to the face of the housings as a possible check that the straight edge was parallel to the case, because the reference surfaces by the intermediate shaft are quite close together.

On the left side, dimensions 1 and 4 were within 0.15mm, and 2 and 3 are within 0.05mm, and since these are less than the sprocket alignment tolerance, I am OK with it.

On the right side, though, dimensions 6 and 9 were 0.45 mm different, and dimensions 7 and 8 ( on the sprocket) are 0.20mm apart, but in the opposite direction as 6/9.

I made the measurements four times, with three setups. with consistent results.

If I hadn't checked to the housing, I would have just adjusted the sprocket shims by 1mm and carried on.

Has anyone seen this situation? Should I just ignore it and carry on?
Admittedly the cam drive housings are not a critical part of the assembly and perhaps are not made to really tight tolerance?

[bpu699]

How sure are you that your case halves are machined perfectly? I can tell you that on my case the lip around the intermediate shaft cover is slightly off. I can catch my nail on it...

My guess is, many cases may be imperfect...

That would throw readings off...

[bdonally]

At the case joint, I can feel a groove but a straightedge across the joint doesn't show any offset, certainly less than 0.001", so I think its safe to assume its OK.
I redid the setup, and still find that the right side housing is off parallel by 0.4 to 0.5mm, as mentioned in my original post.
Anybody ever checked this?
It isn't specified in the manuals.
I suppose that there is some effect on the alignment of the tensioner and the chain ramps.
Worrying about nothing?

[Trackrash]

From my experience with the stock shims you can only get within + - .25mm. Which is the tolerance. Shims are .25mm? thick. So you will need custom (thin) shims to get any closer.

It will be interesting to hear from the gurus.

EDIT: Bruce, IIRC, the shims are indeed .5mm, HOWEVER, as a side note Wayne's books states .25mm.

[Flat6pac]

The only measurements you need are 3, 5, & 7.
You’re not going to redesign the case. Too much info and it draws confusion
The spec is .25+/- the shims are .5 mm
Use factory spec, make sure your cam gears are installed correctly, add in the strait edge width and make the numbers work
Bruce

[bdonally]

Originally Posted by Flat6pac

Too much info and it draws confusion
Bruce

Thats what I eventually decided.

The sprocket spacing wound up center of tolerance after adjusting the shim count.

I can't find a spec on the cam end play. It doesn't look to be adjustable but I'd expect it to be specified.

[Walt Fricke]

No adjustment or spec for cam end play. I don't think anything puts an axial load on the cams as long as the chain gears are aligned to spec.

[bdonally]

Couldn't resist checking the end play - its 0.20 mm or so.

BTW, I find the cam timing method to be a pain - the pressure from the valve causes the lobe to turn the cam when the pin is released.
Instead, I locked a bolt into the cam using a jam nut, set the crank to Z1, then rotated the cam to the correct spot and inserted the pin. Worked really nicely.
Possibly been covered somewhere.

[Walt Fricke]

Well, there is a fair amount of play in the pin. Unless you are awfully lucky, you can't just find the hole where you can insert it, and expect things to stay the same when it comes to tightening down the bolt (you must have the bolt system, and not the big nut). To have the pin exactly place things it must be placed so the valve spring etc force - or maybe it is the direction tightening the bolt will move the gear - has it very tight (in other words, at one side of the slop)so things can't move when you torque the bolt. There are learned posts on this - how many degrees nominally the vernier system itself gives, and how many more degrees/positions there are if you jump the chain one tooth, or more. It appears you can bring the pin in so it is held tight. The pin is pretty strong - I once had a cam seize on installation, though I didn't know it. Starting the engine sheared off all the half teeth on that part of the system.

But the pin isn't intended to transmit the force to the cam - that's the job of the bolt's clamping force. It is just there so you can set the timing.

I like to use the Z1 and rotate cam to spec approach. Quicker to get in the ball park. The tricky part usually is how things move when you torque, or even just undo that double nut.

[bdonally]

I understand everything you said, and yes it moves and needs to be checked.
However it sounds like you use the same approach.

[Walt Fricke]

Wayne Dempsey explains two ways of setting cams, with one for one circumstance, and another for another. I've never understood why there should be a distinction.

It was easier to tweak the older, big nut, camshafts to get the cam to the part of the pin slop which was where you wanted to be, once you had the pin in. The cam holder does just that - holds the cam. With the bolt attachment the holder holds the gear. If you use a nut on a bolt to move the cam while holding the gear, when you remove the bolt, take off the nut, and reinstall the bolt and tighten, there is nothing to hold the cam against the slop. Some have tweaked the cam not with a bolt/nut, but with a small screw driver, with the pin in place, poking through one of the other gear holes. If you are lucky, things may stay in place while you put the washer and bolt in and snug them.

Maybe I prefer the big nut system just because of more experience with it. There are posts you can find on fine tuning these bolt attachment cams. Getting the settings you want can be done.

Me, for my next build I think I may spring for a different system - the center of things is held by the bolt which you torque, but a slotted three smaller bolt system allows aligning things and tightening those bolts - the cam/sprocket alignment doesn't change while doing this tightening.

[bpu699]

Originally Posted by bdonally

I am at the stage of checking the cam sprocket alignment.
I clamped a straight edge across the case:

Then measured the offset to the sprockets and to the nose of the intermediate shaft, to calculate the offset according to the manual. Did one side at a time, measuring three times, breaking down the setup between measurements.
While set up I decided to measure to the face of the housings as a possible check that the straight edge was parallel to the case, because the reference surfaces by the intermediate shaft are quite close together.

On the left side, dimensions 1 and 4 were within 0.15mm, and 2 and 3 are within 0.05mm, and since these are less than the sprocket alignment tolerance, I am OK with it.

On the right side, though, dimensions 6 and 9 were 0.45 mm different, and dimensions 7 and 8 ( on the sprocket) are 0.20mm apart, but in the opposite direction as 6/9.

I made the measurements four times, with three setups. with consistent results.

If I hadn't checked to the housing, I would have just adjusted the sprocket shims by 1mm and carried on.

Has anyone seen this situation? Should I just ignore it and carry on?
Admittedly the cam drive housings are not a critical part of the assembly and perhaps are not made to really tight tolerance?

Bringing this post back, spent 2 days measuring and remeasuring, and not sure what measurements to go with...

Motor was resealed, bearings changed, etc, with all original components...

Stock was 4 shims on the left and on the right.

Stock, right cam gear and idler show wear on side facing you when you are working on it. Left, shows wear on side away from you.

When I measure, I get the same issue the poster above got...

If I simply lay a straight edge against the case by the ims, and measure at the sprocket, I get 2 totally different readings on the left and right side of the sprocket. And, the measurement suggest I need to add 2 shims going to 6... which sounds way wrong as there is already wear on the face suggesting it needs LESS shims.

So, spent some time on the lathe and made a tool that holds the bar stock rigid to the cam housing, worked fabulous, pics to follow. I then aligned the ruler to be 100% parallel to the cam sprocket. This then created a .004 gap at the left case face by the ims...

But, the measurements now were correct at the original 4 shims...

The left had 4 shims, now measured to need 3...

A tiny difference of .002-.004 on the ims case halves is a huge difference at the sprocket...

Here’s the odd follow up question... should the chain, when the motor is turned by hand, now sit on the middle of the sprocket??? Meaning, I should be able to put the same size feeler, like a .005, under the chain on the inside and outside of the sprocket?

I ask, as the chain seems to ride against the face of the cam sprocket on the left, and the back of the sprocket on the right...

[bpu699]

Here are some pics of the tool I made, it works absolutely fabulous. Makes the readings completely reproducible, time after time.

Anyone trying to measure a .25mm difference by holding a ruler against the case, is likely getting very dubious readings. Just holding the calipers at a slight angle gives different readings. Applying pressure to the ruler/stock effects the readings also.

Sorry, pics loaded sideways???

The Stomski tool looks fabulous, but as posted above it very much depends on boths sides of the case being perfectly flat, which, I suspect they aren't on all cases.

My tool is made from aluminum stock, and screws onto the cam housing bolts. It is machined to accept a ruler, that is held on by vise grips.

This easily allows you to make sure the ruler is parallel to the cam sprocket...

To make fine adjustments you just unscrew the tool off the post by 1/4 turns allowing easy reproducible measurements...




Note that for the ruler to be parrallel to the sprocket, its off .004 thousands at the case...

[bpu699]

Easy to attach at the case, just use a bungee...

Holds the bar stock extremely rigid, really can be deflected. You can press against it quite a bit with no issues. Very reproducible readings.

This tool is basically 1$ in material and 1 minute on a lathe, but helped massively in terms of allowing reprodicible readings. If you don't want to allign it to the sprocket, you dont have to. Align it against the case, and lock it in with locking pliers, takes 1 minute to set up...

So, what do you guys think? I would think this would be a marketable item, as its a needed tool, and much cheaper than whats out there... Of course, folks can make it themselves too...

[Gordo2]

I sweat the alignment during my rebuild - measuring and remeasuring for nearly a full day...

Meanwhile, as I researched the subject I found a thread where someone had installed a sprocket backwards - and ran it that way for quite a few miles with no unusual wear.

Bottom line, even though I sweat the details - in the end, based on my research, I came to believe close is close enough.

Good luck.

[ChrisBennet]

I measure this with the Stomski tool:

https://www.pelicanparts.com/More_Info/PELSTOMSKISR008.htm?pn=PEL-STOMSKI-SR008&bt=Y&fs=0&SVSVSI=911M

On a 964 3.6 I rebuilt, the sprockets were not aligned according to my measurements but I heard of others encountering this "problem" so I put it back together as it came from the factory. If I recall, both sprockets had the same number of shims on both sides. It was a looong time ago and my memory isn't so good.

[Walt Fricke]

I don't dispute the fact that checks on whether or not the reference bar straight edge is exactly normal to some centerline - crank, IS shaft, case? - can differ. No question that the IS cover face is pretty narrow for this purpose, and that a small difference there leads to a much bigger one out at the cams.

The Stomski tool replicates what the factory manual shows. Its machinist grade straight edge would seem to remove any worries about how accurate a hardware store piece of aluminum might be.

The sealing surfaces of the chain housings ought to be irrelevant to gear parallelism. They play no role in that. The crush of the rather thick gasket would allow some variation in angle, and precision would not seem to matter as the housing doesn't locate the cam - and the O ring seal allows for in and out variations of the cam, so would allow for angular variation as well? The cams depend on the cam carriers for the position of the end of the cam, and the shims. Same with the cam sprockets. If the cam is parallel with the IS shaft, there should be no side to side difference with measurements to the near and far sides of the cam sprocket?

What might put the cam itself out of parallelism? Bad luck with tolerance build-up? There is a spec for variation in cylinder height. Could that induce a small angle if the tallest cylinder were on one end, the shortest on the other, and the middle one no taller than the average?

There are no axial forces on the cam, are there? It is held in place by the pressure of the oil which comes out toward the ends of the cam and encounters the plug on the flywheel side, and the thick beveled washer on the chain end? To some extent like the idler sprocket, the cam can float axially a bit depending on what the chain itself likes from the more fixed IS shaft?

Chains, it would appear, can live with some lack of alignment of their gears? I'm thinking of the five rear gears on my bicycle, whose angles to the crank gears depend on which gears are selected. Of course, far less speed and power involved there, and the gear teeth are narrower relative to the chain's width?

I wonder what Porsche's specification for tolerance relative to the machined face where the IS end plate fastens? I don't know how Porsche sets up the casting to start machining, though I'd think maybe they establish the crank centerline, machine that, and use that as a reference for everything else?

A mechanic once told me that if you got more than two shims off, you could hear the resulting misalignment. But not if less. Or something like that. He was also a machinist, so not inclined to be sloppy.

Still, very interesting measurements.

[targa72e]

I had a SC that had a replacement case. Lots of little things different than production case. examples were no provision for mounting lifting lug at back of engine (area around breather not machined) and the IS area was also not machined flat, it was left as cast. I had to come up with a way to get everything parallel. I used a spacer off the end of the IS shaft itself to a straight piece of bar stock (verified with precision straight edge) and bolted everything down using the threads at the end of the IS. I verified everything was parallel like bpu699 by making sure both sides of the gear where the same distance. This worked pretty well and was very repeatable. I have since continued the practice of checking the distance on either side of gear to verify things are parallel even when front of case is machined. The last three engines I built were all slightly off from being parallel (two mag case and one aluminum).

john