Chain Tension Issue

[Miztaziggy]

To measure it, just fit the sprocket and measure from the very centre of where the axle goes through to the outer tip of one of the teeth, then take a measure 90 deg round, and again, and again. All 4 should be exactly the same if it is a perfect circle.

You will find that 2 of them will be different to the other 2 and that is because the sprocket isnt circular.

If you work it out, how much bigger the radius has to be to create tightness, it is tiny.

Circumference = 2 * pie * R

Chain goes round approx half the sprocket, so distance it has to cover is = pie * R

So if R increases 1mm, then the extra length of chain required to go round the sprocket will be multiplied by pie or approx 3.14.

So for every 1mm your sprocket is out, you will feel a tightness of 3mm length lost on the chain, and if you think about how much difference to tightness moving the wheel back 3mm by using the adjusters makes, then it makes sense that a pretty small 1mm oval on the sprocket can cause a tight spot quite easily.

[Miztaziggy]

It shouldnt matter if you're directly opposite or not.

If you measure the same thing for both measurements then compare them.

Im assuming you are measuring diameters rather than radius then?

If so, measure diameter across one way, then 90 deg round. As long as both times you measure from tip to say tip on 20 teeth round anti clockwise, it doesnt matter.

[Miztaziggy]

Did you post and delete that for some reason?

[vtryorks]

Finding the centre is one of the major problems here. You'd have to do it on the bike, which introduces a higher risk of human error when using a ruler or other measuring device. Doing it off the bike is much harder, unless I'm missing something. I struggled to identify the exact centre because to do that, you require two intersecting lines, which brings me back to the issue of measuring the diameter.

The theory is perfect but how do we find the exact axle centre in practice? Tolerances are so fine, I think it might be harder than it sounds.

PS. Yes I did post and delete because I misread what ya put

[Miztaziggy]

Read that middle one. You dont have to go across the centre. Just measure near to the centre.

Choose a tooth, call it number 1.

Measure from the tip, across to tip of tooth 20.

Now measure from the tip of tooth 5 to 25, then from 10 to 20 etc

Each will be not quite across the centre, but each measure should be exactly the same, it if its an oval, they wont be

[vtryorks]

But mate, and I do see what you're saying here and sorry if I'm being thick, but you say measure from/to the center. Unless you measure the diameter first, at the two widest points, how do you know where the centre is to begin your calculations?

[vtryorks]

Can do it like this i guess:

http://www.makeitsolar.com/science-fair ... center.htm

[Miztaziggy]

Yes, sorry first I was saying measure from centre to outer tip, and do that all round the sprocket. That will tell you if the sprocket is oval / the sprocket is off centre with the drilled holes etc

To check if the sprocket is round or oval you can just measure the distances I said above. It wont tell you if the centre hole, and the bolt holes are properly in line, but it will tell you if the sprocket is round or not.

Draw a circle on a piece of paper then draw an oval (make it an obvious oval), then draw line across it where im telling you to measure, see the difference in measurements??

[tinysmall]

Read Ziggy's post again then think about what you just asked. Lol

[vtryorks]

I guess I'm not reading it right, I'll give it a go.

Failing that I think we should go to the pub.

[Miztaziggy]

Like this

[vtryorks]

I didnt read your last line. Sorry mate.

You're talking about measuring literally tip to tip on each tooth which are roughly opposite.

Or in other words, using segments of the circle.

[Miztaziggy]

Yes, as long as you measure like for like, it doesnt matter whether or not they cross the centre, just make sure you measure with the same number of teeth between them (eg 20 teeth round the sprocket).

[vtryorks]

Seems to be circular then, so only possible explanation is offset centre and/or bolt holes.

[vtryorks]

Cheers Chris, tbh I hadnt thought of using segments, neither had anyone else I'd talked to (offline) including an engineer!