dr bob

So, to verify the correct performance of Ken's tensioner, I would need to somehow lock the piston in place before using the factory tension tester? Can I get a "flash" reading with the Kempf tool bfore the tensioner piston is retracted by the extra 'test' tension caused by rolling the belt? Is this a perfect time to look at Paul Jaeger's twang method and the perfect musical note? This would provide the least amount to displacement of the tensioner piston.

Ken?

GregBBRD

Note that I'm at a great place to go back and forth to find out what is happening....

gb

PorKen

Gates makes a 'sonic' tester, for a no-contact measurement.

blown 87

I mean that it does not sound good for the belt tension to be that low.


Great come back Ken.

I do not have thin skin, so you are going to have to do a lot better than that to hurt my feelings.

I do have one question, did you do any engineering on this set up or did you just make a bracket to mount the Audi tensioner to the 928?

If so what methods were used to figure out the forces at work here?


This is a honest question, I have one on my car and have installed them on customers cars.

I also have a Mechanical Engineer going to give it a look over for his views on it.

We are trying to figure out a way to measure the forces while the engine is running.

I am not a expert like Greg Brown and Ken, but I am under the impression that loose is not good.

PorKen

I deleted that after I finished my coffee. Grrrr....

The latter. Then tested it on my 16 and 32V engines. The real testing was already done at Audi.

You need -enough- tension to keep the belt wrapped on the crank gear, with some damping for good measure. No more, no less.

GregBBRD

You really need to quit slinging **** and start discussing this thing, in my opinion. Your lack of data regarding the load on the tensioner combined with your attacks on me are not answering any questions...they are simply diverting attention away from the questions at hand.

I measured the tension on the belt, without allowing the tensioner to retract...of course. That being said, you have to know that the factory 9201 tool doesn't put as much "force" on the belt as the Kemph (?) tool does, when it twists the belt 90 degrees.,,,so this way not a difficult thing to do.

I do not own anything but the factory 9201 tool.

And....

I never bothered to see what happen when the engine was turned backwards....there was no need, since the belt was already way too loose. I'm not willing to bend a set of valves to prove a point...so until I can get some reasonable belt tension numbers when the engine is going the correct rotation, I'm not about to turn it backwards......

As far as the millions of cars equiped with this tensioner...I repeat...and it is starting to sound like an echo....taking a part from one application, applying it to another application with a competely different "fulcrum" is not the same thing!

This is not an Audi. This is not an Audi. This is not an Audi. The Audi has the fulcrum way way farther from the crankshaft and has completely different leverage on the tensioning pulley than you do.

Try leaving out the "Duhs" and the wise *** cracks and discuss this thing. Offer me some ideas to find out why the tensioner on this engine is so loose. I'm trying to use your product and join in with the other "followers" and become one of your flock.

GregBBRD

Did this. Held the tensioner wheel in place with an allen. 3.6- 3.8 "belt units" on the factory 9201 tool.

blown 87

Frankly that scares me, that you just made a bracket and said looks like enough tension to keep the belt on, no more no less.

Audi tested it on a engine with a shorter belt and not a 928, so I do not see where putting it on a 928 has anything to do with what Audi tested.

The factory engineers do not always get it right, just look at the revisions to the hydraulic tensioners on Volvo's.



Not trying to stir any thing up here, but rather to find out if I should be installing these on customers cars or my own car for that matter.

PorKen

If you are unconfortable using it, even though it has been shown to work properly on many 928 engines, do not use it.

If it will make you feel better, I will retract the 'duh'. But not the 'backforce' - that's still silly, IMO.

Fulcrum, smulcrum. The design works on a 928.

The belt is not 'loose' as you have measured it, it is just not overtight.

There need only be enough belt tension to keep the belt wrapped around the crank gear at all times.

PorKen

The length of the belt, and the amount of tension delivered by the tensioner is only relevant between the 1-4 cam gear and the crank gear. Between the other gears, the tension is dependent on the action of the crankshaft, and the resistance of the valve train. There's no magic in the 928 timing belt system.

The Audi belt is thicker/wider/heavier than the 928 belt. The tensioning system is understressed used on the 928.

There has only been one revision to the Audi design. After the first year an additional attachment bolt was added to the tensioner/damper for a total of 3.

mark kibort

Im not going to get into the fray here, but isnt the tension on the driver cam much greater, especially, under rapid acceleration of the engine speed? throttle blips, etc. Ive never been that impressed with steady state tests, like referencing "200mph, 6700rpm". I want to see stats or test where you are going from 4500rpm to 6700rpm in under .5 seconds.

From experience, i have lost 3 tooth positions when I was flogging the car in every way at a WC race. It measured out to be in the "4" tension range. It seems that the belt would only jump teeth on the passenger side cam, as the angle from the drive belt from the driving force is coming from the water pump pulley. the tension (I imagine) would be based how fast you reving the engine. the tension on the driver side would be determied by this as well, because the oil pump is basically just a free wheeling pulley and the main drive for the system starts at the crank pulley.

so, it seems to me that the tension is really critical on the passenger side, where if you are revign the engine quickly, and that tension on the back side of the crank gets too loose, you will skip teeth on the passenger side cam pulley.

I can understand Ken's position that you only need a certain tension, and not over tensioned as our 928 system can add, BUT, if the requirements under fast acceleration is a need for "overtension" the Audi based system might fall short.

I havent given this that much thougth, but has anyone measured the belt tension by the oil pulley to see the relationship vs the measuring point and values?

Personally, having raced a 928 engine more than most anyone over the past 10 years, Im pretty confident in the stock set ups. However I do appriciate the work done to find new and better ways to improve our cars.

carrry on.

mk

PorKen

The 1-4 gear has a lot of belt wrap, and is in little danger of coming off. The crank gear must have as many teeth enaged as possible at all times, because the oil pump gear precludes a 'wrap' pulley on the pulled side.

Acceleration is where the Audi type tensioner/damper is superior. With a quick acceleration of the crankshaft, the belt is stretched around the gears, and there is temporarily more belt on the 'loose' side. The T/D extends and retracts to keep the tension more or less constant. More importantly, the belt stays wrapped on the crank gear. With the stock detensioner, no extension is possible so the belt flaps!

High belt tension will limit the flapping to certain harmonics, but at the cost of belt, and bearing life.

dr bob

I see Ken's position is that the tensioner, with the extra stroke and quicker action available even with lower static pressure on the belt, is able to adequately take up the slack that MK describes between the crank and the pass side cams on rapid engine acceleration. This is enough to keep the belt from 'walking' over teeth on the pass side gear. Ken, does that summarize your thinking on the tension question?

In a carefully-analyzed drive, tension would be maintained at 'just enough' as you describe, with belt displacement between spans carefully limited either by tension (raises frequency and therefore lowers displacement) or by adjusting the distances between rollers (again raises frequency and lowers available displacement) or some combination. The lower the vibration frequency, the easier it is to dampen mechanically. In service though, the specific belt tension changes dynamically both with speed and a function more related to rate of accelleration of the engine itself.



From an engineering point of view, I'd be looking at supporting (but not distorting) the path of the belt as much as possible in the longest slackest sections to make sure that the oscilations are never of a magnitude greater than some arbitrary tiny percentage of the tooth height. With the HTD belt with the rounded-faced teeth on the belt, this percentage is small, such that the primary drive is ALWAYS on the section of tooth that's square, and never on the rounded portion of the tooth. Three rollers, spaced at odd prime fractions of the span (1/3, 1/5, and 1/7 of the span distance) would eliminate any harmonic relationships between the sections so there would be no reinforcing oscillation transfered between sections. I've used Delrin guide tracks with contact but virtually no preload for this duty where cogged belts need to be stabilized. [I've also added a second cogged belt, driven at the same speed but with different tension and therefore different resonance, to 'back' the primary belt in critical sections; not possible in the current drive setup and way too complicated anyway. In a past life I designed high-speed packaging machines. ] There are other methods, particularly twisting and then un-twisting the belt between supports, that are amazingly effective at eliminating oscillation prior to a drive or driven roller.

danglerb

How much effect will reduced tension have on the 5/8 timing?

Roy928tt

Wow what a great discussion, not to much mud slinging and a good bit of understanding and thinking.

I must say I tend to favour PorKens view and his upgrade, would probably never shell out the money to buy it, but I agree with the thinking behind it. I view it simply that he has 'evolved' the 928 timing belt system as Porsche may have done had the engine continued production.

Great stuff.

Cheers Roy