Proper Timing Belt Tension

I too wonder about this. I don't feel the spring and tube thing is a very reliable method of setting tension. I was reading an article last night by an engineer that worked with Ford on the 2.3 motor trying to fix a harmonic noise coming from some of the cars. They found that by setting tension with a mechanical tensiometer that when they went back to check them after running, they were all over the map, from sloppy loose, to just about at the low safe limit of tension. The engineer who wrote the article claimed that they needed a readjustment after being run up to temp, and he felt annually as well. Of course this was with the newer HTD type of belt, not the stone age trapazoidal we have on a Ferrari. But I imagine the same situation would be suspect.

I have also wondered why its such a no no to rotate the motor backwards, when virtually everything else in the world has no issues with that. Again, I believe it is because the belt can be pretty sloppy and the risk of skipping a tooth becomes an issue. Perhaps the old way of testing the tension by being able to only barely twist the belt 90 degrees in its longest section by hand is still a good measure? Perhaps the best way is with these new electronic tensiometers that measure frequency and convert it to pounds. We then simply only need a number to shoot for.

 

The spring does all the work.