On the cx type thing, I’ve got a n/w chainring, SRAM X9 clutchy mech.

The clutchy mech seems to introduce quite a bit of drag. Probably not much, but it bugs be whenever I spin the pedals to oil the chain.

With the wheel spinning, (so no inertia from wheel, or brake drag) it takes 150 grammes hanging from the spindle before the pedals move round. 165mm cranks, I make that 0.26Nm. Plugging into here http://planetcalc.com/1908/ I get 2.45 Watts.

So I can relax and forget about it, right?

Why do you think that’s due to the clutch?

It cant be adding drag if its not moving (the clutch) which it wont be if everything is round (chainring / sprocket)

This isnt the case with my bike as it has an oval chainring – its not much but the clutch moves a teeny tiny bit on pedal turns – not enough to bother me but it will be adding drag as its a friction device

things that will cause drag…. Bottom bracket bearings, freehub bearings, jockey wheels, pedal bearings, chain

It cant be adding drag if its not moving (the clutch) which it wont be if everything is round (chainring / sprocket)

True, but since the OP is testing the “startup tranisent” (for want of a better word), there may be some change in chain tension causing it to move very slightly.

I doubt that the force the OP has measured is coming from the clutch however – what about the inertia of the cranks/chainrings/cassette and chain (breaking) friction?

cheers guys, some added detail: I took the chain off the chainring, pedals spun nice and freely. Chain back on, clutch mech locked forward, (unlike shimano, where the clutch is on or off, you pull the der. arm forward and push a button in, locking the arm forward with a no tension and a dangly chain) and there’s a bit more drag, but again, nothing I’d get worked up about. It’s only when the tension’s back on that “blimey, that’s a bit much” drag comes in again.

And it’s not breaking friction, I gave the pedals a little nudge, less the 150, it would stop, more than 150, it would keep going. And to give you an idea of accuracy, I was using a coiled/wrapped road tube with different sized allen keys stuffed in it as a weight, and weighed the one that was just heavy enough to keep it moving. So not CERN standards!

But I get your point, if the arm isn’t moving, the clutch isn’t doing anything, so it’s a combination of the tension and the ability of the jockey wheel bearings to keep things moving under load. That sound right?

True, but since the OP is testing the “startup tranisent” (for want of a better word), there may be some change in chain tension causing it to move very slightly.

If this is even measurable then wouldn’t the tension result in a slight increase in overall chain length? – causing the cage to move slightly rearwards to take up slack – i.e not working the clutch which is only engaged when moving forwards.

you pull the der. arm forward and push a button in, locking the arm forward with a no tension and a dangly chain)

Your not releasing the clutch by doing this you are releasing the tension on the chain from the spring in the rear mech. If the chain is slack the cranks will turn easier

yup, got that ndt, thought I’d covered it above.

Yeah to measure this accurately you need to have a known input and known output – the force and power losses will vary based on cadence, gear and torque/force applied.

There will be far more tension on the chain when you pedal than is applied by the rear mech, if you really wanted to test the effect of the clutch you’d need a non-clutch mech ready to swap in, and even then unless they were both new what does it mean?

if you really wanted to test the effect of the clutch you’d need a non-clutch mech ready to swap in

Or a Shimano clutch mech where the clutch is easily switched off. As before though I suspect the difference would be zero

(Oh and of course then you have to decide how important a certain loss is.)

I realise I’m not doing this accurately! 🙂

I’ve got away from the gear issue by measuring the force needed to turn the pedals when effectively freewheeling. I missed the detail that I used 90rpm in the calculator get to 2.45W.

I was spinning on the turbo last night, and at very light warm up spin, was doing 50W (according to the trainer), so 5% of my power was going straight to transmission drag.

And to be fair, I never said in the first place that it was the clutch, I said it was the mech (which is a clutchy one).

But I wasn’t clear here, sorry:

But I get your point, if the arm isn’t moving, the clutch isn’t doing anything, so it’s the drag in the transmission that I’ve noticed is a combination of the tension in the chain from the mech and the ability of the jockey wheel bearings to keep things moving under this extra load. That sound right?

So from here,

a) does my maths sound right?
b) if it is, and I decide to care about reducing the drag, would jockey wheels be the next step?

would jockey wheels be the next step

I doubt it – I think you have become fixated by the rear mech since it noticeably reduced drag when you took the tension off it. But this tension increases the amount of drag produced by all the other components. I don’t know any numbers and I don’t do power meters but if I was a betting man (which I’m not) Id be looking at the bottom bracket. If its a good one its likely to be well sealed against the elements – and seals cause drag. The question is of course – do you want a poorly sealed bottom bracket for cx

I certainly wouldn’t.

you could check that the cranks are correctly preloading the bottom bracket bearings?

I doubt it – I think you have become fixated by the rear mech since it noticeably reduced drag when you took the tension off it.

And because it wasn’t as bad with the old (old old!) XT job that was on there before.

You’re right though, tension’s going to load up all the bearings, not just the jockey wheels.

Not a bad idea to check preload anyway, ta.

old (old old!)

spring probably quite worn then ?

As someone said – when you load the drivetrain with your legs rather than a little weight – increase in chain tension will overshadow anything caused by the mech