Question about cam timing nuances...

[BGCarrera32]

O.k., I've sort of harassed some of the experts here via PM to death as I've been stepping through this (sorry guys you know who you are ) so I'll post some additional questions to the board...

I am setting cam timing on a 3.2 motor using 964 cams. Overlap spec is 1.26 mm. Getting the left and right side chains tight (not gorilla stupid tight, but tight) using a vise grips and mechanical tensioner (following the Pelican Parts rebuild book) I was able to set the left cam at 1.27mm, and the right at 1.26mm with the main bolts on each cam torqued to spec at 90 ft lbs. Now for giggles, I released the tension on both chains a bit, rolled the crank around a few turns, and tightened the chains back up. No different than had I released the tension to install the pressure fed tensioners.

Just doing that caused the left cam to drift to 1.34 mm, and the right to drop to 1.21. I've got digital indicators on each bank so no errors were introduced in moving gauges side to side, etc.

Which begs the questions:

A) am I doing something wrong?
B) what ensures that overlap is correct after releasing tension and placing the pressure fed tensioners in?
C) do I have the chains too tight during setup?

And on the same note, why wouldn't I just install the pressure fed tensioners and then set the cam timing?

[BGCarrera32]

Hmmm...so here's an update. I decided to remove the vise grips on the left side and remove the solid tensioner from the right side that I was using to hold the chains tight. I put the pressure fed tensioners in their place. Both sides now read 1.34 mm overlap. Not in the middle of the spec, but they match. The chain tension is a little less than when I used the grips and the solid tensioner. Why am I reading in the Pelican Parts rebuild book that I should be setting the cam timing without the pressure fed tensioners in place? I find it hard to believe that the oil pressure sent to these tensioners will tighten up the chains that much more than the spring in each one, but perhaps I am wrong. Wouldn't just setting the timing with the pressure tensioners in place be closer to real world operation?

Thoughts?

[lindy 911]

I agree; the oil simply locks the tensioner by taking up space, not applying tension through pressure. The spring in the tensioner is what makes it tight. Some say that a new chain needs to by slightly over-tightened to make sure it's at full length which might not be possible by using the tensioner. Sounds to me like you're good to go.

[KTL]

Yeah Brian and I discussed this via PM and his results confirm what I felt- the artificial tensioning of the chains is masking what the true TDC overlap lift spec is going to be. Just put in the tensioners that are going to be used in the engine and set the spec accordingly. Be they self-contained "930" tensioners, Carrera tensioners, or mechanical tensioners (I set these with 1.5mm cold clearance btwn top of adj bolt & idler arm), the cam spec is going to be set at your desired TDC overlap lift spec. Vise grips just don't let you set the tension consistently between cyl. banks and also doesn't represent what chain tension is being applied by the actual chain tensioner itself.

[304065]

The factory gives a range for the cams, not a point value.

The reason for this is cycle-to-cycle variation.

Try this: nail the cam timing precisely on spec, to within a hundredth of a millimeter (as you list above).

Now rotate through ten revolutions of the crank and note the cycle- to-cycle variation. There's a lot of it. It's just not that precise of an assembly.

Which is not to say you shouldn't try to nail the center of the range. It's to say that once you get it nailed, button up the engine and move forward, you've achieved a tighter consistency side-to-side than the Factory specified.

[BGCarrera32]

Originally Posted by 304065

The factory gives a range for the cams, not a point value.

The reason for this is cycle-to-cycle variation.

Try this: nail the cam timing precisely on spec, to within a hundredth of a millimeter (as you list above).

Now rotate through ten revolutions of the crank and note the cycle- to-cycle variation. There's a lot of it. It's just not that precise of an assembly.

Understood. I guess my question really then is why use mechanical tensioners as the PP rebuild book advocates, to set the values, and then swap out and install the pressure fed units? Seems to me if you're close with pressure fed units installed, then you'd be good to go as that's what that engine will be actually running. I would think you could introduce more error farting around with a vise grips and mechanical tensioner and swapping out, but then what do I know...

[lindy 911]

Again, new chains need to be stretched a little to make sure they're at full length which is what a mechanical tensioner (vise grips, bolt and block, whatever) will do. Don't think the tensioner has enough spring to make that happen. It'll keep it tight once run in a little but not stretch a new chain to length.

[304065]

Chains do not "stretch." What we think is stretching is actually wear of the link pins over time, resulting in elongation. Think about it: if you could actually stretch steel with your bare hands. . .

The whole point of using the vise grip or a mechanical tensioner is to be certain that tension is held on the chain during the cam timing procedure, so the chain doesn't jump teeth on the sprockets, and to the extent that there is any play between the sprockets and the chain, that it's kept to a minimum during the procedure. You just want to positively take up the slack.

The original factory workshop manual calls for the old bucket-type tensioners to be installed, then a screwdriver used to lever the idler arm to hold tension on the chain to set the timing. This is bascially the same principle, but the tensioner is already installed. No reason you can't do this to take up the slack in the chain with a more modern tensioner installed-- the vice grip method just makes holding tension easier, but makes subsequent removal of the vice grip and installation of the tensioner more difficult.

[lindy 911]

I hate to disagree, but I disagree. When chains are new there are tolerances between pin, plates and rollers. Until pulled tight against one another the chain is not at it's full length which means it stretches slightly even if only a few thousands. Your reference to chains stretching over time is accurate.

[nocarrier]

Wouldn't keeping the chains tight help to remove any backlash between the cams and the crank? I would imagine that removing the backlash would increase the accuracy of the timing.

I've never timed a Porsche engine before and only did it over the past weekend. So I completely understand the frustrations BGCarrera32 is having.

I mad a simple tool to take the slack out of the chain while timing the engine. I seems as if I was getting much more consistent results when I removed the backlash.

any thoughts?

[Eagledriver]

Using the hydraulic chain tensioners would be ideal, but I don't see how to do it in a practical way. Everytime you wanted to move the pin you'd have to take out the tensioner and compress it again. We use the mechanical means of tension so we can do the adjustments without messing with the tensioners.

-Andy

[KTL]

I think the key is consistency. The amount of tension is somewhat irrelevant IMO. As long as you put *enough* tension on the chain to get a consisent lift reading on the dial every time you rotate the crank, I call it good. Because the artificial tension you are applying is just that- artificial. Once the tensioner is installed, that's going to be what sets the tension regardless of how much tension you think is needed on the chains.

Play around with the chain tension during the timing setup & you'll see it moves the dial indicator a significant amount. You don't have to do much to see this. Just snag the chain with your finger & pull/push on it. You can see the dial indicator move quite easily.

But in being consistent, you also want to apply an equal amount of tension on both sides. How do you measure tension on the chains when clamping a vise grip on the idler? Kinda hard to quantify. Whereas the actual chain tensioner, or the mechanical one set to the same number of turns-out on the adjuster bolt (or zero lash btwn idler arm and adj bolt, meas'd with drag method of a feeler gauge), is going to apply the same amount of force on each side of the engine. Or is it? Is the geometry substantially different on the cylinder banks? Probably not since the chains are the same for both sides of the engine.

I think we're making a lot of noise about something that's not as precise as we want it to be? As long as you get repeatable results you're good. Because once the engine is operating w/out your physical input, the lift spec is probably not as close to the dial-indicated value as you think. The hydraulic tensioners (be they the self-contained type or the Carrera pressure-fed type) vary their piston projection depending on engine load. So the chain tension is going to vary.

That's one thing that's always concerned me about running the mechanical tensioners. Sure they're fail proof for racing but they don't provide the damping effect on the chain load that a hydraulic tensioner does. Nonetheless I run them in my racecar engine because it's one less thing to fail. That mech tensioner is one of the few things that didn't get @#$%ed up when I recently killed my engine!

[ChrisBennet]

My post in this thread may be helful:
Cam timing and chain tension

[KTL]

One last thing I thought of. Be sure to check the condition of your cam hub if you're having difficult getting repeatable dial readings. The hub is the small diameter "toothed" piece that slips over the woodruff key and locks the chain sprocket to the cam via the cylindrical pin. Over time that hub can be widened-out on the tooth that the pin has been riding in for so long.