I bought a couple of rear derailleurs - X9 Type 2.1 and a 6870 Ultegra Di2 SGS (medium to handle up to 32 cassette). Dismantled them today. First I measured the force required to move the cage 25mm.
Anyways opening up the Type 2.1 proved to be a bit more tricky since this latest iteration seems to have added a tamperproof pin. I had no way of extracting it so just had to force rotate the clutch tensioner out folding the pin into the plastic knuckle and aluminum cap/tensioner. Oh well. I suppose I could've tried drilling the pin out or see if a drill could cut into it. A screw extractor would not work since the pin was less than 2mm in diameter. This pin would not even allow any adjustment of clutch tension.
- X9 Type 2.1 - 3000g
- 6870 - 900g
Anyways opening up the Type 2.1 proved to be a bit more tricky since this latest iteration seems to have added a tamperproof pin. I had no way of extracting it so just had to force rotate the clutch tensioner out folding the pin into the plastic knuckle and aluminum cap/tensioner. Oh well. I suppose I could've tried drilling the pin out or see if a drill could cut into it. A screw extractor would not work since the pin was less than 2mm in diameter. This pin would not even allow any adjustment of clutch tension.
First noticed that the clutch was very well lubed and rides in a nylon shell. It seems like it would need a lot of push on the end to get it to not rotate freely (excuse the damaged aluminum shell from the pin)
There's a difference in cage lengths for the mediums for each brand. The top pulley is far more offset (from the pivot point) with the 6870 as well.
X9 pulley to pulley 75mm pulley to pivot 73mm
6870 87mm 85mm
X9 pulley to pulley 75mm pulley to pivot 73mm
6870 87mm 85mm
I think round one is going to be trying just more spring tension. The spring has about 1.2mm clearance from the pin without tension. There's two settings/spring pickups on the cage and it was already set on the stiffer one. I could try and cut another one on the cage or drill another hole in the knuckle itself. This can't be too radical a change because the knuckle shell is just plastic and the spring could just tear out the hole. Perhaps to not reduce the amount of plastic hanging on the spring hook it is better to drill the cage and pin bushing with a third position.
OK I drilled a third hole in the cage and pin bushing. Did it from the outside but it's hard to keep a drill straight since it hits the steel pin right at the edge and the drill wants to wander outwards rather than drilling the steel. I had to file it back inwards and have an elongated hole.
It definitely has added some tension and instead of 900g for movement it takes 1500g now. Hopefully not threatening to break the spring out of its plastic hole the other end.
Also I played with the X9 clutch without the spring in. The one-way clutch definitely works. It's pretty nice how it works and is much simpler than the Shimano variety. I also tried it with the spring and without the cap pushing on the clutch. It seems that just the tightening squeeze from attaching the cage actually puts quite a bit of pressure on the clutch. So the cap does not come into play unless you want to exceed the clutch force. Strange.
I backed off the cage attachment a tad to measure the spring-only force and it is identical to my modified spring in the Di2 (about 3.75lbs to move 1/4 turn).
This is sort of crazy since the clutch tension is set by the assembly torque on the cage bolt. Just attaching it and torquing to a reasonable hand-tightness results in a lot of clutch action - more than my X01 Type 2 on my other bike and more than the Zee even though it is a very short cage. I don't get it. I suppose I could loosely attach with threadlock?! I get a measurement of almost 10lb for the quarter turn (hard to measure since its not continuous force with the one-way).
It definitely has added some tension and instead of 900g for movement it takes 1500g now. Hopefully not threatening to break the spring out of its plastic hole the other end.
Also I played with the X9 clutch without the spring in. The one-way clutch definitely works. It's pretty nice how it works and is much simpler than the Shimano variety. I also tried it with the spring and without the cap pushing on the clutch. It seems that just the tightening squeeze from attaching the cage actually puts quite a bit of pressure on the clutch. So the cap does not come into play unless you want to exceed the clutch force. Strange.
I backed off the cage attachment a tad to measure the spring-only force and it is identical to my modified spring in the Di2 (about 3.75lbs to move 1/4 turn).
This is sort of crazy since the clutch tension is set by the assembly torque on the cage bolt. Just attaching it and torquing to a reasonable hand-tightness results in a lot of clutch action - more than my X01 Type 2 on my other bike and more than the Zee even though it is a very short cage. I don't get it. I suppose I could loosely attach with threadlock?! I get a measurement of almost 10lb for the quarter turn (hard to measure since its not continuous force with the one-way).
RSS Feed