Bulvot

I'm replacing the timing belt on my 1987 928 S4, and discovered that the cam gears were worn, so I'm replacing them. Which means ensuring that the timing is right. So, I used the PK32V tool to check the existing timing before I removed the old belt and gears. This is what I found:

Driver's side: +8 degrees
Passenger side: +1 degree

I checked it at TDC and 45 degrees, rotating the engine multiple times to be sure. It measures the same every time in every position.

The car runs great. Plenty of power, no issues. The belt is in very good shape, so I don't think that it has jumped any teeth. The motor was rebuilt before I bought it with custom ground S3 cams. If it matters, it also has a x-pipe.

The question is...does that timing seem right for this motor with those cams?

My plan was to set the cams to the exact same timing with the new belt, as I have with the old belt. Is there any reason not to do so? Should I set the timing a bit different to allow for belt stretch? If so, how much should I adjust it by, given the current settings?

PorKen

The driver's/5-8 side is one tooth (7.5°) off. Best to re-string the belt but you might have enough space to adjust to 0. (Slots allow for about 2 teeth movement.)

If you installed a new regular (non-Racing) belt, adjusting both sides to zero will allow for approx. belt stretch which creates about -2 retard on the 1-4 side. Factory dial gauge specs for all versions of 32V equal 1-4: -2°, 5-8: 0°, measured cold with a 32V'r. (Hot becomes 0, 0 due to engine expansion.)


Advance feels good due to higher TQ at lower rpms but trades for HP off the top.

Bulvot

Are you certain that the timing for S3 cams in an S4 motor should be identical to S4 cams? I just replaced the cam gears and used the old belt to put them at exactly the same timing as the old ones were.

I am putting a new belt on, so I can retime with the old belt (which is still on there), or the new belt. It's all apart, so it's easy to make changes, but I want to make sure that they are the right changes.

FredR

The 32VR system works on the basis of setting cam timing relative to the notches in the cam casing back cover that are used as an index relative to the position of the cam shafts. Thus when this method is used the cam timing is the same no matter what cams are installed.
If one uses the factory method for cam timing then this is based on lift at a given degree of crank rotation and that is different for different cam sets.

Bulvot

Ok, thanks for the clarification.

It seems unlikely to me that the two cam shafts would be timed so differently on purpose. As Ken said, the driver's side is probably off by one tooth. Either the belt was installed wrong, or it jumped a tooth. In that case, that means that the current timing for both cams is supposed to read about 1 degree advanced. Whether or not that was intentional by the engine builder is another question entirely.

I will go ahead and put the new belt on, making sure to correct the one tooth off condition on the driver's side. Then recheck the timing with the new belt and see where it winds up at that point. My inclination at this point is to set them both to zero, as measured by the 32V'r.

If I'm reading the 32V'r manual correctly, if both sides read zero after rotating the engine a couple of times, then I've done it right. To help get it there quicker, if I set the passenger side to +2 initially, it should settle to 0 after a few turns of the motor.

FredR

If you are going to change the belt a few pointers:

1. Do the changeover with the engine at exactly 45 degrees BTDC.
2. If you are using a new belt then there a procedure to correct for the new condition as per the instructions in the 32VR manual.
3. Before you remove the belt, for the sprocket that has the timing out, take a look to see where the three 5mm bolts sit relative to the slot- they should sit around the centre of the slot on a correctly set motor, if they are to one extreme the adjustment is out, if they sit around the centre and timing is out by that amount then the sprocket is one tooth out and that apparently is a common installation error although ho anyone can miss such is beyond me as it is so obvious.
4. With the 32VR tool you can set the timing at the 45 BTDC position but note that to set the belt tension correctly you need to do so with the engine at TDC on number 1 firing stroke.
5. Another approach is to do the cam timing with the old [stretched] belt, then remove it and install the new belt knowing that it will stretch into the correct position after about 1000 miles or whatever. Thus one needs to know whether it is the adjustment that is out or if it is a complete tooth in error.

Bulvot

Fred,

Thanks for the input. I've got the flywheel lock tool and am doing the change at BTDC. This is my first timing belt job on a 928, but not my first timing belt. So far, this one has been much easier than other vehicles that I've worked on.

The 32V'r manual says this: "You may have to add a degree or two of advance to your desired amount to compensate for belt stretch when doing the 1-4. EG. for -2°, set to 0.. After rotating the engine the reading will be correct." Which leads me to the conclusion that the important thing is that the reading is 0 after rotating the engine a few times. It doesn't specifically say anything about a new belt. The supplemental manual essentially says the same thing. Am I misreading that? Should the 1-4 cam be 2 degrees retarded even after spinning the motor a few times?

The bolt holes on the sprocket were off toward one end of the slots. Not all the way, but definitely not centered. Something interesting about the old gears is that they were installed with red loctite (!) and seemed to be torqued way past 50 ft lbs. So, I'm really not sure what the thinking was when they were installed, or what happened with the timing.

As for tension, I'm installing the "PKT" automatic tensioner. Please, let's not comment on the merits of it in this thread. That is what is going on the motor.

My plan is to check and adjust the timing at BTDC, and to rotate the engine several times between measurements.

So, here's my real question...should I set the timing to 0 on both cams with the old stretched belt, or should I set it to 0 on both cams with the new unstretched belt, or...?

FredR

The things people do!

If you are fitting the PKT just fit the new belt and be done with it. Setting the 1/4 sprocket at 0 is the compensation for a new belt if my memory serves me correctly. For a stretched belt the setting is minus 2 degrees or so I seem to remember.

When threading the new belt, start at the crank sprocket, wind it over the oil pump and then the 5/8 sprocket. Before threading the belt set the sprocket so the 5mm bolts are centered and when you have the belt over the 5/8 sprocket teeth and everything appears aligned, tension the system anti clockwise using the 32mm washer on the cam sprocket- you may need an extra pair of hands to do this- then thread the belt onto the 1/4 sprocket.

Bulvot

Ok. So...new belt, set both cams to 0, put it back together

FredR

After about 1000 miles or so you can always go back in and check the cam timing again- by then it should show some degree of retard on the 1/4 sprocket but the 5/8 should still be zero.

DeWolf

Do you then adjust the 1-4 back to zero?

FredR

No- if there is any adjusting to do it would be to set 1/4 at minus 2 degrees assuming my memory is correct and the "initial stretch" has taken place. To be clear, one is not going to notice a slight degree of advance or retard and as Ken has already pointed out advance favours low end torque at the expense of top end power but 1 degree either way is not of much significance and one would be lucky to spot the difference on a dyno given run to run variances or so I suspect.

PorKen

S3, S4, GT, and GTS cams have differing amounts of lift at 20°, so the WSM specs vary between them, but they all measure out at -2, 0, on a cold engine, or 0, 0 on a hot engine, measured with a 32V'r.

Most likely the belt was strung incorrectly the last time. Instead of taking off the belt again, the gear was moved so when you installed the belt, correctly this time, the reading was off.

The engine does this for you each time it heats up.

Bulvot

Fred, thanks for your continued patience and explanations. I'm good to go at this point.

Except, why does the WSM not list a torque spec for the cam gear bolts? Or, maybe a better question...why am I incapable of finding it?

Bulvot

Ken, Would you believe that the +8 degree reading was taken before I removed the belt or did anything? I installed your tool on the existing cam gears with the old belt, and took the measurements.

Anyway, mysteries aside, I'm just going to put the new belt on, and set the timing to 0 degrees on both cams, and put everything back together.