One man's tensioner solution
#76
Inventor
Rennlist Member
Rennlist Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Greg,
I updated my post while you were writing your response. The 928 tensioner washers are bi-metallic, heat sensitive, concave washers, which flatten out when heated to compensate for engine expansion (kinda). Different beast than your always springy Bellevilles.
I updated my post while you were writing your response. The 928 tensioner washers are bi-metallic, heat sensitive, concave washers, which flatten out when heated to compensate for engine expansion (kinda). Different beast than your always springy Bellevilles.
![Stick Out Tongue](https://rennlist.com/forums/images/smilies/tongue.gif)
#77
Supercharged
Rennlist Member
Rennlist Member
Join Date: May 2002
Location: Back in Michigan - Full time!
Posts: 18,925
Likes: 0
Received 61 Likes
on
35 Posts
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Rather than pointing out the problems with Hammer's design compared to the Porsche OE design, let's flip that coin.
The Porsche OE design as Mark and other have pointed out is a positive pressure design. I don't know if anyone has bothered to check the value of this pressure, but let's assume it's X. Then the problem comes when the tension falls below a certain threshold, say X-Y.
As long as you can regularly check your belt tension, and keep it in > X-Y you should be in good shape, right? I mean, isn't that what the tensioner dummy light effectively (or ineffectively) does? It tells you when the tension has dropped below X-Y.
The key problem with the OE tensioner system is that it requires user intervention to adjust the tension. How do you get around this without adding pressure sensors and little motors that adjust the tension? Porsche added these bi-metal washers that have a similar effect, albeit quite limited.
So, now, let's look at Hammer's tensioner. It too is a positive pressure tension system as long as you set the initial tension to some modified X, let's call it X'. As the belt loosens or tightens, the spring should be able to compensate for the Y much more effectively such that X'-Y will always be greater than X-Y.
The concerns about spring rate K are valid, only to the extent if K<Y (I think). Also, the concerns about the natural frequency, should be mostly mitigated vis the oil dampner in the middle. It acts just like a coil over shock in a car. edit: When a car has a bad strut, it'll bounce all the way down the highyway. With a good strut, it'll be nice and steady.
So I don't see too much cause for concern here as long as the spring constant can accomodate the window between in-spec tension and out-of-spec tension, everything should be fine. Plus Hammer's eliminated a real stinker of an oil leak.
I applaud the efforts.
The Porsche OE design as Mark and other have pointed out is a positive pressure design. I don't know if anyone has bothered to check the value of this pressure, but let's assume it's X. Then the problem comes when the tension falls below a certain threshold, say X-Y.
As long as you can regularly check your belt tension, and keep it in > X-Y you should be in good shape, right? I mean, isn't that what the tensioner dummy light effectively (or ineffectively) does? It tells you when the tension has dropped below X-Y.
The key problem with the OE tensioner system is that it requires user intervention to adjust the tension. How do you get around this without adding pressure sensors and little motors that adjust the tension? Porsche added these bi-metal washers that have a similar effect, albeit quite limited.
So, now, let's look at Hammer's tensioner. It too is a positive pressure tension system as long as you set the initial tension to some modified X, let's call it X'. As the belt loosens or tightens, the spring should be able to compensate for the Y much more effectively such that X'-Y will always be greater than X-Y.
The concerns about spring rate K are valid, only to the extent if K<Y (I think). Also, the concerns about the natural frequency, should be mostly mitigated vis the oil dampner in the middle. It acts just like a coil over shock in a car. edit: When a car has a bad strut, it'll bounce all the way down the highyway. With a good strut, it'll be nice and steady.
So I don't see too much cause for concern here as long as the spring constant can accomodate the window between in-spec tension and out-of-spec tension, everything should be fine. Plus Hammer's eliminated a real stinker of an oil leak.
I applaud the efforts.
#78
Rest in Peace
Rennlist Member
Rennlist Member
Join Date: May 2006
Location: Bird lover in Sharpsburg
Posts: 9,903
Likes: 0
Received 2 Likes
on
2 Posts
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Every one else uses a spring also Mark.
I cant think of a engine that has a timing belt that there is not a spring in the tensioner if it is a compensating type, as most are.
Can you name one compensating type that does not have a spring?
Is the spring rate very high on the Belleville washers in the 928 tensioners, yes, but it is a spring, none the less.
I also have never seen a engine that requires as much tension as the 928 engine does on the belt.
Part of the porpuse is no doubt for thermal expansion, it also has some give to it.
There has to be a better way than what comes stock on a 928.
I think Ken and Hammer have found a better way.
again, it makes me nervous thinking about a spring, where basically a rigid shaft would be.
Mk
I cant think of a engine that has a timing belt that there is not a spring in the tensioner if it is a compensating type, as most are.
Can you name one compensating type that does not have a spring?
Is the spring rate very high on the Belleville washers in the 928 tensioners, yes, but it is a spring, none the less.
I also have never seen a engine that requires as much tension as the 928 engine does on the belt.
Part of the porpuse is no doubt for thermal expansion, it also has some give to it.
There has to be a better way than what comes stock on a 928.
I think Ken and Hammer have found a better way.
Originally Posted by mark kibort
again, it makes me nervous thinking about a spring, where basically a rigid shaft would be.
Mk
#79
Administrator - "Tyson"
Lifetime Rennlist
Member
Lifetime Rennlist
Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
---
Originally Posted by From a lurker and well known 928 guru that does not participate with the forums. This is his comment when Ken originally shared his idea. Seams to apply here
Is there really a problem with the stock tensioner if it's properly
installed and maintained? That one in the thread may work okay, but I sure
wouldn't want to be a test subject for it. It is a stock Audi part, for an
Audi engine. I doubt that the Audi has as long of a timing belt as a 928,
or that there's as much block expansion since it's probably a smaller
block. Is that tensioner off of a big Audi V-8?
The Toyota and Honda both have timing belts that are tensioned the same
way. There's a spring that pulls a roller and applies tension to the belt.
The engine is rotated several turns with the spring applying tension to the
belt with the engine cold, to get things equalized. The bolt securing the
tension roller is then torqued down so that it can not move anymore. The
spring tension is only used for setting the tension initially and nothing
else. There's no automatic tensioning using a spring while the engine is
running. Those belts are shorter than the one on a 928, and there's less
block expansion, so tension adjustment as the engines warm up isn't needed
as much. Even with a spring mechanism already in place, neither Toyota or
Honda seemed to think that using the spring to maintain tension on the
running engine was a good idea.
installed and maintained? That one in the thread may work okay, but I sure
wouldn't want to be a test subject for it. It is a stock Audi part, for an
Audi engine. I doubt that the Audi has as long of a timing belt as a 928,
or that there's as much block expansion since it's probably a smaller
block. Is that tensioner off of a big Audi V-8?
The Toyota and Honda both have timing belts that are tensioned the same
way. There's a spring that pulls a roller and applies tension to the belt.
The engine is rotated several turns with the spring applying tension to the
belt with the engine cold, to get things equalized. The bolt securing the
tension roller is then torqued down so that it can not move anymore. The
spring tension is only used for setting the tension initially and nothing
else. There's no automatic tensioning using a spring while the engine is
running. Those belts are shorter than the one on a 928, and there's less
block expansion, so tension adjustment as the engines warm up isn't needed
as much. Even with a spring mechanism already in place, neither Toyota or
Honda seemed to think that using the spring to maintain tension on the
running engine was a good idea.
#80
Rest in Peace
Rennlist Member
Rennlist Member
Join Date: May 2006
Location: Bird lover in Sharpsburg
Posts: 9,903
Likes: 0
Received 2 Likes
on
2 Posts
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Originally Posted by PorKen
Greg,
I updated my post while you were writing your response. The 928 tensioner washers are bi-metallic, heat sensitive, concave washers, which flatten out when heated to compensate for engine expansion (kinda). Different beast than your always springy Bellevilles.![Stick Out Tongue](https://rennlist.com/forums/images/smilies/tongue.gif)
I updated my post while you were writing your response. The 928 tensioner washers are bi-metallic, heat sensitive, concave washers, which flatten out when heated to compensate for engine expansion (kinda). Different beast than your always springy Bellevilles.
![Stick Out Tongue](https://rennlist.com/forums/images/smilies/tongue.gif)
![thumbup](https://rennlist.com/forums/graemlins/thumbup.gif)
Being bi-metal is what makes them tempature compensating.
![Smilie](https://rennlist.com/forums/images/smilies/smile.gif)
#81
Inventor
Rennlist Member
Rennlist Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Guru, smuru.
There is only a few mm's of travel needed for the tensioner pulley, as it only takes a slight movement in the pulley's arc to take up a large amount of belt length.
An undamped spring would be bad.
![Big Grin](https://rennlist.com/forums/images/smilies/biggrin.gif)
There is only a few mm's of travel needed for the tensioner pulley, as it only takes a slight movement in the pulley's arc to take up a large amount of belt length.
An undamped spring would be bad.
#82
Administrator - "Tyson"
Lifetime Rennlist
Member
Lifetime Rennlist
Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Hey now, just sharing data to go into the pool.
#83
Rest in Peace
Rennlist Member
Rennlist Member
Join Date: May 2006
Location: Bird lover in Sharpsburg
Posts: 9,903
Likes: 0
Received 2 Likes
on
2 Posts
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Originally Posted by PorKen
Guru, smuru. ![Big Grin](https://rennlist.com/forums/images/smilies/biggrin.gif)
![Big Grin](https://rennlist.com/forums/images/smilies/biggrin.gif)
Originally Posted by PorKen
There is only a few mm's of travel needed for the tensioner pulley, as it only takes a slight movement in the pulley's arc to take up a large amount of belt length.
An undamped spring would be bad.
An undamped spring would be bad.
A undamped spring would not be bad for long, I would give it maybe 30 seconds.
We will not get into how I know about this other than to say, look at Hackers post about the honda spring and how you should tighten the bolt BEFORE you start the engine, but I can promise you that statement is 100% correct.
![ducking](https://rennlist.com/forums/graemlins/icon107.gif)
#84
Inventor
Rennlist Member
Rennlist Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
I find your lack of faith...
...disturbing.
![](https://members.rennlist.com/porken/smiley_vader.gif)
Originally Posted by Hacker-Pschorr, guru medium
Hey now, just sharing data to go into the pool.
#85
Administrator - "Tyson"
Lifetime Rennlist
Member
Lifetime Rennlist
Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Originally Posted by PorKen
I find your lack of faith...
...disturbing.
![](https://members.rennlist.com/porken/smiley_vader.gif)
![Smilie](https://rennlist.com/forums/images/smilies/smile.gif)
#86
Rennlist Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
and thats what we we are talking about here. the oil in the tensioner provides no damping. this design we are discussing is an undamped spring and that could be bad if you understand how spring mass systems work. they can get excited by resonant frequencies and exacerbate the movement. a step motor is a spring mass system. it has running torque. however at resonance, the torque in the opposte directino can be greater, knocking it out of synchronous speed and completely stalling it. we are dealing with so many factors here, i dont even know where to begin to start.
mk
mk
Originally Posted by PorKen
Guru, smuru.
There is only a few mm's of travel needed for the tensioner pulley, as it only takes a slight movement in the pulley's arc to take up a large amount of belt length.
An undamped spring would be bad.
![Big Grin](https://rennlist.com/forums/images/smilies/biggrin.gif)
There is only a few mm's of travel needed for the tensioner pulley, as it only takes a slight movement in the pulley's arc to take up a large amount of belt length.
An undamped spring would be bad.
#87
Rennlist Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
Ill go out an measure it for you. i have one right in front of me.
trust me, its an order of magnatude more than the undamped spring propsed here.
For giggles, what is the spring rate of the spring proposed here? like a valve spring?
.
Mk
trust me, its an order of magnatude more than the undamped spring propsed here.
For giggles, what is the spring rate of the spring proposed here? like a valve spring?
.
Mk
Originally Posted by blown 87
Every one else uses a spring also Mark.
I cant think of a engine that has a timing belt that there is not a spring in the tensioner if it is a compensating type, as most are.
Can you name one compensating type that does not have a spring?
Is the spring rate very high on the Belleville washers in the 928 tensioners, yes, but it is a spring, none the less.
I also have never seen a engine that requires as much tension as the 928 engine does on the belt.
Part of the porpuse is no doubt for thermal expansion, it also has some give to it.
There has to be a better way than what comes stock on a 928.
I think Ken and Hammer have found a better way.
I cant think of a engine that has a timing belt that there is not a spring in the tensioner if it is a compensating type, as most are.
Can you name one compensating type that does not have a spring?
Is the spring rate very high on the Belleville washers in the 928 tensioners, yes, but it is a spring, none the less.
I also have never seen a engine that requires as much tension as the 928 engine does on the belt.
Part of the porpuse is no doubt for thermal expansion, it also has some give to it.
There has to be a better way than what comes stock on a 928.
I think Ken and Hammer have found a better way.
#88
Inventor
Rennlist Member
Rennlist Member
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
The tensioners with refill ports ('83-) are engineered to be oil dampened - there is a check valve and sealing o-ring to create this.
In implementation, there is little to none, unfortunately.
In implementation, there is little to none, unfortunately.
#89
Rennlist Member
Thread Starter
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
To All,
This is what happened that leads me to believe that this tensioner is within the ranges necessary. I mocked up the engine & belt on the stand with the tensioner setup torqued in place. I rigged up a test light to the circuit. I cranked up the adjusting bolt until the light came on. I tightened the adjuster up slightly until the belt tension was within the proper range on my $59 tension gauge. I spun the motor over several times checking at intervals to see if the tension changed. The tension was stabilized. At that point I saw about .060" compression in the tensioner. With a pry bar I pushed the tensioner roller swingarm into the tensioner mimmicking the belt/engine expansion when warm or at higher rpm. The tensioner compressed about .031" & immediately sprung back when I let go. It was NOT easy to compress. I'm just over 6 foot & 235 pounds & I can't compress this thing without a vice although it was easier with the prybar. But by no means sloppy. That leads me to believe that this tensioner will stay within factory limits enough for me to test on a running engine. I'm thinking of putting a pressure tester on the belt somewhere when running like they use on the Formula 1 cars. (I know someone who has one of those systems that they removed from their car a few years ago & tried to sell me but I didn't want it at that time. I know he still has it) But that will be as a last resort. I have no reason to believe that the tensioner won't work as I planned at this time. So when someone merely looks at a picture & says 'the sky is falling' maybe they should do a similar amount of testing on their mouth before criticizing the work of others.
There, ...I feel much better now!![hiha](https://rennlist.com/forums/graemlins/roflmao.gif)
Regards, Hammer
This is what happened that leads me to believe that this tensioner is within the ranges necessary. I mocked up the engine & belt on the stand with the tensioner setup torqued in place. I rigged up a test light to the circuit. I cranked up the adjusting bolt until the light came on. I tightened the adjuster up slightly until the belt tension was within the proper range on my $59 tension gauge. I spun the motor over several times checking at intervals to see if the tension changed. The tension was stabilized. At that point I saw about .060" compression in the tensioner. With a pry bar I pushed the tensioner roller swingarm into the tensioner mimmicking the belt/engine expansion when warm or at higher rpm. The tensioner compressed about .031" & immediately sprung back when I let go. It was NOT easy to compress. I'm just over 6 foot & 235 pounds & I can't compress this thing without a vice although it was easier with the prybar. But by no means sloppy. That leads me to believe that this tensioner will stay within factory limits enough for me to test on a running engine. I'm thinking of putting a pressure tester on the belt somewhere when running like they use on the Formula 1 cars. (I know someone who has one of those systems that they removed from their car a few years ago & tried to sell me but I didn't want it at that time. I know he still has it) But that will be as a last resort. I have no reason to believe that the tensioner won't work as I planned at this time. So when someone merely looks at a picture & says 'the sky is falling' maybe they should do a similar amount of testing on their mouth before criticizing the work of others.
There, ...I feel much better now!
![hiha](https://rennlist.com/forums/graemlins/roflmao.gif)
Regards, Hammer
#90
Rest in Peace
Rennlist Member
Rennlist Member
Join Date: May 2006
Location: Bird lover in Sharpsburg
Posts: 9,903
Likes: 0
Received 2 Likes
on
2 Posts
![Default](https://rennlist.com/forums/images/icons/icon1.gif)
I think that is a good idea Mark, are you sure that the dampner he is using is just a undamped spring?
I thought that there is a dampner inside the tube, I never did understand why Volvo went with a spring only setup in the later model inline engines, but they do work in that app.
The true test is going to be how does it perform.
I think that a compensating type is better that the stock setup, but I am by no means a enginer, just a man that turns wrenches for a living.
Jim is right, most 928's are not any where near where they should be as far as tension goes, at least the very few that I have seen.
I thought that there is a dampner inside the tube, I never did understand why Volvo went with a spring only setup in the later model inline engines, but they do work in that app.
The true test is going to be how does it perform.
I think that a compensating type is better that the stock setup, but I am by no means a enginer, just a man that turns wrenches for a living.
Jim is right, most 928's are not any where near where they should be as far as tension goes, at least the very few that I have seen.
Originally Posted by mark kibort
Ill go out an measure it for you. i have one right in front of me.
trust me, its an order of magnatude more than the undamped spring propsed here.
For giggles, what is the spring rate of the spring proposed here? like a valve spring?
.
Mk
trust me, its an order of magnatude more than the undamped spring propsed here.
For giggles, what is the spring rate of the spring proposed here? like a valve spring?
.
Mk