Cam Timing Question?

Do a compression test and see if perhaps an intake cam is off a cog or so. You should notice somewhat decent drop in psi from the good bank to the bad bank if that is the issue. No other way to check the cam timing other than pulling the covers.

 

I see no mention of it in the manuals...so maybe it was done ex factory.....but on this 'ere car, back of the flange where the cam pulleys mount, each cam has a notch.

Setting flywheel mark to TDC 1 will align these notches with the arrow on the cam cover and the ribbing on the belt cover. A bit rough and ready, but probably better than peering in the oil filler hole and hoping the intake cam on the rear bank is a good indication for all of them. The hard part is overcoming parallax error on the front bank - suggest climbing on top of the engine to look down rather than trying to judge from the side of the car.

 

As Frog said, the notch on the back of the cam pulleys is in the same relative location as the notch on the camshafts under the cam covers. There is a ferrari tool (AV1198) which is used to line up the external notches (on the back of the pulleys) so that you don't have to pull off the cam covers to access those notches. Without the tool it's a bit iffy but basically, those notches need to face up 90 degrees from the plan of the head.

Hope that helps!

 

As i recall, there was some debate of the factory marking to the actual proper setting. On the 308 that was here the markings were a bit off as i recall. Maybe Guiseppie has one grappa too many for lunch that day?

 

Ha! You may be right Steve, depending on how good the grappa is they could be more than a bit off!

I suspect the cam shaft and pulley markings are really just a baseline mark, in other words, good enough for government work and sufficient for the cams and crank to be in reasonably good time (perhaps +/- 1 degree, maybe two). Obviously, degree'ing the cams is the way to go if you want to be spot on, but for the average DIY guy they just want to know they're not a whole tooth or two (or more) off on the cams when doing a belt replacement.

 

I have to wonder about the timing marks under the cam covers. The cams are set when the engine is at room temperature. The timing belts are installed at room temperature. Aluminum has one of the largest coefficients of thermal expansion of any commonly used metal. The engine expands more than the belts so the tension in the timing belts changes. This will change the cam timing a small amount.

So now the question is whether this is taken into consideration when the engine was designed or is lining up those marks to the nth degree a waste of time.

 

I thought the purposes of the cam belt tension bearings was to counteract any changes in belt tension?

Phil

 

They are locked down once the tension is set so the spring contributes nothing more. They are not like some serpentine belt accessory tensioners which rely on ever present spring load to maintain tension. The purpose of the tensioner is to preset the tension in the belt and to provide a lot of wrap around for the exhaust cam. The tension in the belt pulls the belt's teeth into the intake cam pulley. Without a wrap on the exhaust cam pulley, the thing would be prone to jump cogs. When the engine is running, there is less slack on the tensioner side than on the drive side. That's another reason for the big wrap of the belt on the exhaust cam pulley.

 

There is absolutely no way the thermal expansion of aluminum is going to change/move the cams relative to the timing marks on the cam shaft caps. Sorry to be blunt but that is the dumbest thing I've heard in a long time.

 

Think of a bicycle chain with a slack side. If you do not let the pedal sproket turn but leave the driven side free and increase the distance between the sprockets, what happens?
Not to be blunt but, surely, even you can see that the driven sprocket must turn. The relative angular position of the sprockets changes. The driven sprocket advances a small amount.

 

While this is a fairly accurate statement for the Ferrari 8 cyl cam drive model, it is untrue for the Porsche 928 V8 engine. In that case, the very long timing belt is in fact timed for hot engine operation. The right bank of cylinders are advanced slightly greater than the left bank so that the long belt is allowed to expand as the engine expands, thus putting both banks in correct time.

As to the OP question, the actual answer is no, there isn't. When the cam blank is made, the notches are set on the grinding machine so that they will be ground relatively the same in relation to the cylinder bank. With the blank being ground, the important thing is getting the timing between the cylinder lobes accurate, and not necc accurate to the cam gear. That is why we can adjust the cam gear(wheel) on the camshaft to accommodate for the modest differences in the grinding process.

The best test without removal of the cam cover is a normal pulse compression test. It may show an average difference of 3-15PSI from band to bank. This would indicate that one of the cams has jumped a tooth(or more). The only way to correctly resolve this is to set the engine to TDC, remove the covers and check the indicated valve depression. Sorry...

 

In theory, you are correct. However, I strongly suspect the net effect on the cam timing would be negligible and would not even register on the degree wheel.

 

When I change the belts, I always turn the engine a few revs by hand after everything is done, to make sure the belt teeth are seated properly and that nothing is loose. I can see the tension bearing moving - slightly - but moving back and forth as the belt passes over it as the engine rotates.

But maybe I'm thinking of this action before the final tightening down, when you observe the extent of movement of the bearing and then tighten it so it is locked in....only doing this once every 3-4 years and being I'm at work, I don't have my service notes with me.

Anyway, I wouldn't trust any marks on the pulleys or anywhere else.

I always toyed with the idea about getting a spare front cam cover and adding a viewing port to it, like the rear cover's oil cap, so that you don't have to pull the front cover to check the marks.

Phil

 

Yup, that makes sense - I get it.

What that doesn't consider is that belts have a CoE, same as aluminum. So, the relevant data point is the delta between the CoE of the belts and the CoE of aluminum.

Frankly, I don't believe the independent CoE of either the belts or the aluminum would be noticeable on the degree wheel. Now, when you have both those CoEs working in common (as they do) then you're definitely not going to see anything on the degree wheel. Just my $.02.

It sounds all cute and oh-so-technical to fuss over, but there ain't anything there.....

 

The should bore anyone who reads it....

The CoE of the belts is less than that of aluminum. If the belts are carbon steel, the CoE is about 6.6d-06 in/in-deg F. If the belts are austenitic stainless, the CoE is about 9.5d-06 in/in-deg F. The CoE for aluminum is 12.3d-06 in/in-deg F. Furthermore there is less of a temperature rise in the belts than there is in the engine block. Thermal expansion is the product of CoE, the length, and the temperature change. So you have two reasons for less thermal growth of the belt. One is the material; the other is the temperature rise. Even if the temperature is the same, you still have a 28% larger CoE for the austenitic belt and about a 86% increase for the CS belt. (I should have saved my old belt and cut one open to see what material is inside.)

So let’s say the owner gets through painstakingly lining up his marks in the cam area. He starts the engine but the marks will no longer be exactly aligned due to the stretch in the belt (engine first started and at room temperature). The cams will be a little bit behind where they should be. The slack in the belt due to its spring like stretching builds up on the tensioner side - hence the big wrap so the exhaust cam does not jump a cog. Without knowing what the tension is in the belt as well as the spring constant of the belt, I cannot determine the stretch. But it is greater than zero. The people who design these engines would know this.

The engine runs and comes up to operating temperature. Let's assume the operating temperature of the block to be 200 F. Assuming the cams were set at 70 F, that's a 130 F temperature rise. Let’s guess that the operating belt temperature is 180 F.

The next step is to determine the growth of the belt as well as the growth in distance between pulley centers and compare them. I’m guessing the pulleys are 15 inches apart; I don’t feel like removing my wheel and fender liner to measure the thing.

Growth of stainless steel belt: dl = CoE * length * dT = 9.5d-06 * 15.0 * (180 – 70.0)
Growth: dl = 0.016 inch

Growth of engine between pulleys: de = 12.3d-06 * 15.0 * (200 – 70.0)
Growth: de = 0.024 inch

The difference is 0.0080 inch.

The diameter of the driven pulleys I guess is about 2.75 inches. Assuming there is sufficient slack in the tensioner side, the rotation of the driven pulleys due to engine/belt thermal expansion difference is:

(de – dl) = radius * theta (with theta being angle of rotation expressed in radians)

Plugging in the numbers you get

Theta = 0.008/1.375 = 0.0058 radians.

Expressed in degrees:

Theta = 0.0058 * 180/pi = 0.33 degree of cam rotation.

Therefore the crank rotation is 0.66 degree.

Going through the same calculation with carbon steel belts gives a final result of about 0.55 degree of cam rotation which amounts to 1.1 degrees of crank rotation.

So, as a previous poster mentioned about the Porsche 928, the cam timing falls behind but catches up as the engine warms. And as I said before, the engineers should consider this when designing an overhead cam engine. Where the timing ends up cannot be determined without knowing the load on the belt due to cam shafts as well as the resulting belt stretch. The thermal expansion of the engine can be determined quite accurately by sophisticated 3 dimensional finite element thermal and deflection models of the block. Deflection is always determined first with further processing of the deflections being used for the stress computation that is also done for engine design.

Before I end this the Porsche procedure for the timing belt on the 944 turbos (4 cylinder engine, ½ our 328 engine) is to install the belt, tensioning it as prescribed. The vehicle should then be driven for several hundred miles. The final step is a second tensioning of the belt after the initial permanent stretch has taken place.

Permanent stretch? That does provide extra slack on tensioner side to rotate the cam so it catches back up to the drive pulley.

 

Nice analysis!

Some data you're missing:
Cam gear OD(top of teeth): 3.535" [89.79 mm]
Cam gear OD(bottom of teeth: 3.375"

Cam belt strands are fiberglass. This came from DAYCO during SMG2's new belt system design. I'd always thought they were Kevlar, so my old posts were wrong.

However, I don't think this changes your analysis conclusion.

Especially given that the finest adjustment is 1/2 degree, and the WSM allows +/- 1 degree.

I posted the length of the QV's long belt section to one of smg2's new belt system threads back when the design was underway.

 

If it were me I would keep it simple and pull the cam cover off.

 

This is the very reason people degree the cams according to belt stretch or, "installed 1 degree advance".

 

Degree wheel installed, pressurize cylinder number one to about 10-20 psi. Any higher and you wont be able to turn the motor. Rotate engine manually, slowly, note when pressure is lost on degree wheel - those are valve opening events. Note when pressure rises - those are valve closing events. Repeat cylinder number 5 and compare.

 

Yup, exactly right.

With regard to the porsche 928 engine, yes, absolutely right there....that's a 7 foot long belt with thermal expansion along the axis of two banks for the one belt, so it is a consideration. Not relevant at all for a 308 engine.

 

I noticed this on eBay. Looks like a nice easy fit for a degree wheel. Has anyone tried this style before ?
Image Unavailable, Please Login