docmirror

You are an engineer, and so am I. If you don't like the math, I will do the experimental work for you. Rates are $470/hr, min 4 hours. I expect this job would run about 6-7 hours, payable in advance.

If you send a PM, I will reply with the mailing address. PO will be; #928-6672A.

If you don't want to believe the eng handbook, I guess that's ok.

PorKen

I believe you have read what the book says about the ideal washers in the book and mean no offense.

But, you are the one stating that this applies to the squishy copper washers installed on a 100K mile 928 engine, under tension, so it's coming out of your budget.



I have measured them, badly, and found the range sucks. Mainly, I have seen that the belt flaps, so I know objectively the stock tensioner isn't doing a good job.

docmirror

Look, I didn't mean to stir it up, but really, you can decide the moon is made of cheese. I"m ok with that. Your bold statement was the stock tensioner does nothing. Prove it! You asked me to do the work for you, and I'm gonna, once you have the check sent. Belleville washers have been around for >100 years I think. The principle is well known, albeit surely old tech. But old tech or not, it works. the math is there, the tools are readily avail, and I'm happy to do the work.

GlenL

It's in that thread you linked above. You did the work. You said that the washer stack shortens as it heats. As the block heats the Al expands and shortening the tensioner compensates for that. It's not much but the block isn't moving much, either.

PorKen

Actually, you'll note I'm usually quite careful to qualify my statements about the stock tensioner.

"The stock tensioner does adjust for engine growth" implies it has enough range to keep belt tension constant from 0-90C.

That's on you. I won't even suggest a donation for my procedural advice.


I have observed across 16 and 32V engines that the (72mm PKT/stock) tensioner pulley needs to move about 2mm cold to hot. The stack I measured moved ~0.5mm from 0C to 30C unloaded, IE. no pre-flattening of the discs under tension. Not enough.



Not that it matters, but, FWIW, I am enjoying this discussion, for once, y'all.

Rodrv6

0C to 30C is 32F to 86F, not exactly normal engine temperature range, is it???? Wouldn't a more accurate range be 0C to 100C? Or am I missing something???

PorKen

Yes, but I didn't care enough to find out. There's an opportunity there for someone!

At ~30C I was able to make a range of ~2mm... in a big vice (which is likely a destructive test to the washers). The washers should flatten further from 30-100C but I can't imagine the total would be more than 1mm.

Perhaps a new washer pack performs better. Maybe the '78-'82 design (with 8 sets of 5 washers) was designed to do 2mm when new.

GlenL

Well, the stock tensioner moves in the right direction to compensate for temperature changes. About 0.5mm over 30C means about 1.65 over 100 which is close to the 2 you report. No, I'm not going to do an experiment or the maths of how much longer the belt path is getting nor the change due to the movement of the tensioner pulley. I do expect that some engineer at Porsche did it, though, and 2 or more other engineers checked his work. And they tested it on multiple engines.

SwayBar

Is there a video available showing the flapping belt with the stock tensioner?

Likewise, is there a video of the PKT under the same temperature showing less/no flapping?

That would be interesting.

FredR

The above discussion is interesting even if it is somewhat off topic but that is a matter for the moderators.

I posted my thoughts about the stock tensioner a couple of weeks ago in another thread wherein I stated my conclusion that the Belleville pack primarily exists to help de-tension a warm system. I estimated that the engine grows by about 1mm from the crank centreline to the axis of the cams and does so for both banks so Ken's observation that his PKT changes by 2mm from cold to hot operation confirms the number I concluded.

For the operation of the Belleville pack to be understood one needs to know what Porsche intended it to do. That the stock tension alarm does not arm for 3 minutes tells us that Porsche knew that the thing might alarm when the motor is still cold. That the thing might alarm at this stage does not mean the belt is about to jump and my guess is that Porsche knew full well that not arming the system for 3 minutes is not a risk especially so if the driver is sensible during the warm up period.

I figured that the Belleville pack could shrink by about 1mm or so from the data I found. That Ken measured 0.5mm over a 30 degree differential should tell us that over a logical differential of say 85C [15C to 100C] we can expect movement of close to 1.25mm. That to my way of thinking explains what is going on and those numbers are logical. To expect the Belleville pack to travel the same as the PKT is not realistic as they are designed differently. Clearly the stock tensioner has a degree of pre-tension in it and this has been very carefully determined by Porsche. As I can tell it works reliably- whether cam sprocket wear is excessive compared to what it might be with a different system design I have no clue but I remain puzzled as to why the sprockets seemingly wear on the top flats rather than on the faces of the teeth that are intended to take the shear loading.

As to the function of the oil therein that remains less clear. The logical assumption is that it helps conduct heat to the Belleville pack more efficiently and if correct, loss of oil should result in the pack not heating correctly and thus the belt might be over tensioned if the correct amount of heat does not get to it. As to whether the oil and washer pack act as some kind of damper remains to be seen- personally I doubt it because of the spring rate of the pack but I am open minded on that one. Greg has reported that lack of oil can cause an alarm- this is interesting in that at the moment I can see no immediate logical reason as to why that should happen.

Even if the alarm were triggered during warm up this would not be a disaster as it is primarily a pre-alarm. I figured that the alarm exists to warn the driver that some intervention is needed- thus why it indicates "toothed belt service" and not "Engine about to grenade". In effect its main "raison d'etre" is to advise the driver that it is time for a re-tension which Greg advises is notionally every 15k miles or whatever. To my way of thinking it is not intended to be a "time to call a tow truck" type of alarm but somehow folklore seems to have guided many in that direction despite what the alarm says.

Engineers design things "not to fail" so to consider an alarm as a pre-cursor to immediate failure to my way of thinking does not ring true. That things can and do fail is invariably because the owner was driven by cost constraints to skimp and thus supply/install inferior kit. Thus my suspicion that the alarm is there to advise the owner it is "time to do something" but not drive him [or her] to "panic mode". I have no doubts that someone somewhere will have been saved by the alarm going off but I doubt the problem was caused by premature failure of stock Porsche items.

I have seen a couple of false alarms over some 18 years of ownership- both before I started doing my own servicing. Both times I pulled over, cancelled the alarm, kept the motor running, blipped the throttle a few times and listened for any strange noises. On both occasions the alarm did not repeat so I figured it was a false transient. Nonetheless I drove home and as per the advice on the dash had the belt tension checked- no issue either occasion.

If something serious was brewing the alarm should repeat and if it did then I would consider whether to drive it for service or call a tow truck. A genuine condition of alarm should not disappear- a genuine alarm should repeat until rectified!

On a general note I welcome good technical debate and I would encourage those participating to keep it that way.

Imo000

I agree, the light is a warning something is happening and that's all. When my belt light went off I cycled the ignition key and when it pooped up the second time (a few minutes later) I shut the engine off imitatively.

I feel the Audi tensioner is keeping the belt less tight than Porsche intended. The last one I worked on had a new belt and I was able to pry the belt off the cams gears with my finger tips. To me, that seems really loose.

Adamant1971

Same with my 87 when I went in to change the water pump this summer after 20k km's. The PKT was extended to 9mm and the belt was easy to pull off. With a new belt and pump it was down to 7mm. This winter I'm going back to a factory tensioner.

The 89 motor I just built got a stock tensioner as well.

Imo000, I'm assuming you are talking about Danny's 91. How far was the pkt extended on his with the shaved heads? Did you use the thicker head gasket? Just curious.

OBehave2.0

This is a video when I first started dyno testing the Hanson stroker car. I noticed the flapping and odd movements of the belt through the rpm range.The most troubling is the action of the belt when the throttle is closed at 7000 rpm at the end of the pull.

PorKen

That's a cool video, but it doesn't show the flapping I'm talking about, between the crank and 1-4 cam gear. It shows how the belt performs at high rpm with either tensioner. It's scary! The belt goes more and more horizontal coming out of the water pump pulley on its way to the 5-8 gear the faster the engine spins. There is a similar movement, less pronounced, from the 1-4 to the water pump.


The random video below (no affiliation, 2013) shows the belt flat between the crank and 1-4 with a PKT (on a S3). With a stock tensioner, especially cold, the belt will flutter. (Worst case, with too low tension, the flutter will ride over the 1-4 or crank gear.)


(Hope he found the source of the noise.)

PorKen

Obviously that's at the end of the PKT range, but also because the Audi tensioner/damper compresses (to dampen), the tension will lower the more you work the belt.

Sounds like you are sold on the factory unit, but you could alternately upgrade to a PKT-S bracket, which was made with shaved heads in mind.