I stripped a cassette lockring with my big clicky wrench last night - I kind of knew it was failing to click - backed it off to a lower setting, where it clicked happily enough - then wound it up toward 40nm in small steps, when on the way, bang - stripped. Happily it was the lockring and not my freehub.

I always thought about changing to a beam wrench because the clicky one only measures torque on normal threads - it's no good forthe left hand thread on bottom brackets.

It's not hard to guess a working cassette lockwring or BB tightness. I do like using my little torque wrench for other carbon fittings though.

That is the exact problem most people have, they do not maintain a right angle when they are torquing down a bolt. They put all the pressure at the end of the wrench and think nothing of it and with every extension and adaptor you use there is more torque lost. If you do not believe me that you lose torque with every extension you use try it out yourself your self but I would suggest you try it with a dial wrench since it is easier to see. if you use a click type you have to mark it before you check it.

yes I am

Here's a link to a serious discussion of this on an motorcycle forum. The consensus is that extensions don't make a difference if the wrench is turned at hand speed, and I agree.

Note, that if you use a hinged, or U-joint extension all bets are off.

The easiest way to think about this (with rigid straight extensions) is that Newton's laws apply, and the torque at the top of the stack has to equal the torque at the bottom, otherwise there would be a net torque within the extension which is impossible.

This would be true if there were no connecting joints, because at each joint there is a little bit of give. If you have a torque wrench that reads at a fine scale try it out for yourself. I'm sure if you're just using it on a bike there won't be enough a difference to care but in other fields it does and I just happen to carry those standards to everything I work on.

Yes, there's give and the extensions even flex, but this doesn't change the Newtonian requirement that the torque at each end of the stack must be equal and opposite.

The only time it would make a difference is if an impact wrench were used, because each connection works back and forth costing you some inertia.

The only time compensation is required is when the driver of the Torque Wrench is offset from the fastener. The most common case is when a crows foot wrench is used. On my bicycle, I need to use these at the rear wheel mount of my suspension since there is no way to tighten it on the centerline.

Most adapters and extensions are not long enough to impart a noticable torque loss. They will add backlash however, but that does not effect applied torque as long as things are still lined up on centerline.

Dan

If that were true, the same would have to apply to linear motion, since the physics involved in both cases is the same, and I'd be able to reduce the amount a set of scales read when I stand on them by wearing platform boots. If the torque into the top of the extension is not equal to the torque out, the shaft will slowly twist, at a constant rate.

Shame on you!!! You were on the right track saying the system top and bottom torques had to be equal, then you blew it. If the top torque exceeds the the bottom the shaft will rotate, at an accelerating (not constant) rate.

Ah, dammit! This is why I shouldn't post while tired!

My point still stands to an extent, the one thing the shaft won't do is remain totally rigid and do nothing, which is what the shaft of the extension on a torque wrench does.

Yes, we agree. I just couldn't resist the gotcha.

Those who think the input and output torques are different, need to take a moment to try to figure out where the extra force is epended in a slow or static arrangement. Of course, this applies only when things are in line. If they're out of line, you need to work the geometry to calculate the multiplier.