JEC_31

Numbers 1 and 2 are solvable with a simple design. For example, the force of a spring-loaded design is limited to the force of the spring and it's travel would be limited by a integral and adjustable stop. The trick is finding the "range of the belt" between not enough force (insufficient tension) and too much force (snapping the belt - oops).

Number 3 - if a belt goes more slack than the limited travel can compensate for, we do indeed have a problem. What we need here is to adapt the warning system to trigger at a certain point in tensioner travel, and perhaps even allow the tensioner to travel very far to keep the poor floppy belt from falling off. I do admit that determining this warning point would be tricky and perhaps costly.

Number 4 - No, I'm just making life simpler for the person in the equation, the person who would rather not have a heart attack when a little light comes on, and then would rather replace one belt and a bearing or two - and not rebuild all 2000 pieces of the tensioner system as regular maintenance. Less chance of human error too.


The warning system isn't flawed, it's great - as you pointed out. Maybe some heavier duty wiring and hardware would help reduce the loose-connection sydrome though...

My main point in this post is that I believe that the tensioner itself is an overly complicated design, and could be simplified for the sake of robustness.

PorKen

Most of the modern timing belt tensioners I've seen (for engines with more than four cylinders) employ an internal or external spring, along with a sealed hydraulic dampener. The systems are also much more compact, the tensioner pulley arm is much shorter, and everything is replaced at timing belt changes.

All of these systems have a 60K+ change interval - except for the SAAB/Opel V6, without dampener, which has a 30K interval, and a track record for disaster similar to the 32V 928.

The sprung dampener doesn't apply increasing force. It keeps a constant tension (within the range of the spring), irregardless of engine temperature, belt stretch, or pulley wear.

Most 928 tensioners are no longer dampeners. Wear, leaks, and age have made them solid adjusters, with little compensation left in the bi-metal washers.

Here's what I see as the main issues:

~ Belt age
~ Belt tension
~ Water pump
~ Gear wear
~ Pulleys
~ Tensioner pulley arm bolt
~ Tensioner pulley arm bushings
~ Mismatched tensioner parts
~ Tensioner wear
~ Oil leaks

Less likely:

~ Cam gear hub cracks
~ Camshaft cracks

Dave Howerdel

Ken,
Wasn't the Porsche recommended belt change interval 60,000 mi. To be prudent, most of us take it to 45,000.

I agree that most 928 tensioners are no longer dampeners but this is due to poor maintenance not a failure in design.

Newer systems are different as things have advanced a bit since the 928 was designed.

but...
I still don't believe that a newer system fitted to a 928, unproven in such an application, that would not extend the replacement interval, is better than the original system when properly maintained.

Dave Howerdel

Do you really believe that the changes you propose would simplify the design?

I believe that they would certainly make things more complex as such a system would need to be designed, implemented, and proven over time (all while keeping your fingers crossed that you've accounted for everything).

As it is now, I have a high level of confidence with the system and my ability to troubleshoot/repair it. I have the manuals if my memory fails me and a wrench I can take it to if my confidence wavers.



Can anyone design a better system than what Porsche originally designed? Sure.

Will it increase the interval between belt changes?
Probably not

Is the current system flawed as to require it or are the benefits of such a redesign potentially so great as to justify it?
Not in my opinion.

John Veninger

I wonder if a better material would prevent this type of failure. I did noticed that the bushings seem a bit loose the last time I changed them. The old ones seemed tighter when I replaced them!
Does anyone know what material those $18.00!! pieces of plastic are made of?
I'm sure a set could be machined to a tighter tolerance for the same price out of a better material.

Garth S

My current bushings were turned from delrin rod** stock - with a longer throat and a grease groove .... and a dab of hi-tack synthetic grease on installation. Nice & tight after two years .....

** edit: the material used was black Nylatron, a molybdenum disulfide filled nylon. Scratch delrin, my error ...

JEC_31

I will be the first to admit that my opinion is greatly influenced by an irrational and egotistical desire to improve things.

IMO: my changes would...
> simplify and strengthen the design
> NOT increase the belt change intervals, but make them easier without tensioner rebuild difficulties,
> be admittedly difficult and dangerous (to 32 valves at least) to test and tune - although I have some ideas there too.
> reduce the overall chances of timing belt system failure due to tensioner problems.

Given the "difficult & dangerous" part, maybe I should consider chains after all.

After all my crazy schemes, the most sensible choice still is simply the usual expensive and time-consuming scrupulous system maintenance. The fact remains that rubber belts break.

John Veninger

And you didn't make enough for everyone

heinrich

Let's face it .. the whole system is fragile. it does do a very good job, MOST of the time, but there are better ways to do it. Unfortunately they are too expensive. Belts fail due to:

age (teeth lost)
abuse (overtensioning)
failed water pump due to overtensioning or age
failed or eaten drive gears 9sharp or lobsided).

Hpw to resolve? Keep an eye. Perform proper service and replace everything at that point.

The 928 belt is the strongest part of the whole setup

heinrich

Statement:
If we replaced the tensioner with a sprung one and the water pump with an electrical one we would see perhaps no belt failures ever.

EB338

Truly! I want some too!!


Perhaps some form of the originally suggested guides could be made to add another level of indication to the warning light if a belt moved off center would be a good idea, then the light could indicate both tension and misalignment, but not to try and hold a belt in line.

Something else to consider when comparing the apparent simplicity of other timing belt designs is that most of them are not V8's driving 4 camshafts over 32 valves, and the water and oil pumps. That is a lot of load for a single belt expected to maintain precise cam timing at the same time. And unless I am mistaken isn't the 928 belt the longest timing belt made? That is a lot of slack and vibration to try and manage. How many other cars have this kind of demand on this length of a belt? Then add in the hostile heat and potential oily environment the rubber must try and exist in on the engine. These cars were never really built to be the every day grocery getters many have come to be, they are high performance sports cars and German ones at that, higher levels of maintenance were always expected by their designers and are more important now as they age than ever. Frequent maintenance and inspection is just part of life with these cars.

Perhaps finding alternatives to drive the fluid pumps could reduce load enough to help and this reduce belt length again reducing complexity. Though I must admit one of my reasons for keeping away from a dry sump oil system with external pump is that in our case the oil pump becoming seized or the belt failing does stop the whole engine, rather than counting on the operator to realize the oil pressure has dropped and shut down the engine before more than just the valves are destroyed. But if we go to the point of finding alternative drives for the water and oil pumps then one might as well find a way to eliminate the belt all together and create some sort of oil bath housing and (gasp) a chain? The whole thing snowballs from there into a complete redesign of the engines primary drive system, and if we go that far why not swap in a different engine which already has no timing belt..................It seems much simpler just to accept that we need to perform very thorough frequent maintenance to the entire timing belt system.

ew928

So have any of the aftermarket 928 suppliers thought about making
CLEAR timing belt covers?

Wish I could see where the electrical faults were on a 928.
Ernest (NYC) How much can I sell my used timing belt for on Ebay?

H2

If the tensioner system were so good, why did Porsche change the disign to include a different style tensioner arm ('86.6? year on?) then a brace to the water pump at some point? Not sure if there were other improvements made. Both of those, I thought, were to help stabililize the system and help keep the belt centered on gears...so they wouldn't slip off. I don't think that plates on the cam gears would be a cure of anything...but it might pre-empt a belt slipping off for whatever reason...and, as has been mentioned, may provide an audible warning. Also...why couldn't one just regularly look through the tbelt cover hole (on top leading to the air intake tube) and take a look and see if the belt's anywhere near the new flange? I do that now.

I worry a bit because my Tbelt still rides 1/16th of an inch from the edge of the cam gear and I've replaced EVERYTHING except the cam gears and crank gear. I would feel more comfortable if the belt rode in the center of the cam gear. I have had the car on the track much of a day at between 3000 and 6000 rpm and the belt never moved a whit so I guess it's OK. And I've heard from others that this is not uncommon. I just like to be sure. Maybe I just need to buy a backup car. As Heinrich says, that whole belt system setup is a bit too fragile for comfort.

On the other hand, there are a lot of high mileage 928s running around....and an unknown number that have experienced Tbelt problems...usually with bad results.

Harvey

JEC_31

First find a clear plastic material that would A) hold up to the hell of an underhood environment, and B) be injection moldable - and then tooling up a mold (or multiple molds if a "family" mold won't work) to produce replicas of the intricate stock items is very high $$$

Although cutting view-slots into the covers and fitting them with a tough clear plastic "windows" might work. Or leave them open for cooling or for belt shavings from a failure to escape easier.


As for the used belt... I'll trade you for a large gently-used bridge (plus you pay shipping on the bridge).

heinrich

brace is not for pump - it's for that bolt that alway bends