Chain Tensioner - Why Not Just a Spring?

[Jeff Higgins]

I've just experienced my second chain tensioner failure. These are the old style mechanical tensioners, not the pressure-fed Carerras. I understand the concept of how they work and what Porsche was trying to accomplish. It is starting to strike me that maybe they were gilding the lilly just a bit on this one.

Just about every other chain or belt tensioner I have ever seen is a simple fixed postition, with a shoe or a roller. They are adustable for postition, but once set, have no internal compensation at all. The other variety consists of a simple spring loaded shoe or roller.

I know some racers used fixed (solid) tensioners in 911 engines. I've seen all the warnings and pitfalls related to those. What I have never heard of or seen is a simple spring type tensioner. Has anyone ever tried fitting a spring inside an old mechanical tensioner body in place of the piston and check valve mechanism? What are some thoughts on this? It seems too obvious, so I'm sure I'm missing whatever very good reason(s) there might be for not doing this.

[HawgRyder]

I would think a spring will "give" too much at certain point around the revolution of the cam.
Depending on the number of valves being opened at any particular time, the back force on the cam will differ.
The hydraulic portion of the old tensioners slows down this lag effect (like a shock absorber) and the spring and pressure inside the tensioner takes up the slack at the first opportunity.
Bob

[randywebb]

springs bounce - you would then need a shock absorber

if you look at eh equations of motion for a harmonic oscillator you'll see what the curves look like (postition over time) - that's why we have damped harmonic osc.'s -- hence the British term for 'shocks'

you can bet P AG looked at various fixes...

[Tyson Schmidt]

Yeah what Randy said.


With just a spring, the cam timing would jump around, the chain would flail around, and the sprockets would wear out 10 times faster. (not to mention the potential for jumping time. picture the motor bouncing off the rev limiter and imagine what the chains would be doing with an undampened tensioner) You need the dampening.

The 911 engine grows quite a bit, so that needs to be compensated for. The tensioner allows this to happen slowly.

Solid tensioners work fine when set on a hot motor. But that means the chains are sloppy (and noisy) when it's cold. But the potential for either getting them too tight on a hot motor, or too loose on a cold motor, (or a little of both) is pretty high. It takes experience to know how to set them.

[Jeff Higgins]

Makes sense. Dampening chain oscillations would seem important, especially with what - about three feet of chain whirring around on each side. Granted at half the crankshaft RPM, but still moving at a pretty good clip. For a really good visual representation of an irregularly timed load / slack cycle will do to a chain, I guess I can think back to my old chain-drive Harley, and watching all the other ones I rode around with. The slack bottom side of the drive chain flailed up and down at such a rate it's amazing they ever held together. I guess the last thing we would want is a similar thing going on inside our chain boxes.

[randywebb]

got me thinkin'....

do m/c's use anything in particular to address this problem?

big chain saws?