Hiya Folks.

I've got a SunTour Cyclone mk2 GT RD that I know is a bit bent, but it sort of works.
What I can't work out is where it's bent.

Situation: Working on my Holdsworth Whirlwind..
I knew it was slightly bent - I've got another Mk2 on my other bike and that's definitely straighter - but this meant I could measure it.

I don't want to remove the new RD so here's the Mk2 in question on a spare frame (1982 Avanti):


And here's the SunTour Cyclone II GT diagram:


So my guess is that this Mk2 was in a collision and is bent somewhere, but I can't see where.
Anybody know where they bend ?

Have you tried grabbing the cage and flexing it? Maybe the pivots are worn.

I've done a full rebuild on all my SunTour RDs so I'm fairly sure it's not worn anywhere.
But I'll check in the morning ...

I think the #19 has a bend

Could very well be, yeah. (Parallax in the image posted makes it tricky to visualize from the rest of the geometry.)



Yellow is vertical here. Red is the axis of the cage axle, green is the plane of the two idler cogs. Green and yellow ideally should be parallel, but they're not. I'd expect the cage axle's axis to be perpendicular to vertical here (in yellow) but it's not; possible that the body's twisted a little?

Bottom bracket housing in the background is a tell too, along with your rear wheel's hub. Compared to how those sit I'd expect the cage axle to be parallel to both..

The bend appears to be the pulley cage (#9) clamped too tight at the upper pulley (lower pulley in the photo) pivot bolt #19. Is the pulley bushing (#11) the correct length? Does the bushing #11, plus 2x the pulley cap #10 thickness match the pulley cage (#9) spacing between inner and outer plates?

The quick-disconnect-chain design, which I personally love, has the disadvantage of only one full-length pulley cage plate, and is easily disturbed from vertical by various forces. I've seen it happen whilst re-assembling these RDs post-service--particularly when the pulleys are also replaced.

Thanks for the excellent pictures. It looks like the pulley wheel cage is bent. With the long cage, Suntour designed so that the chain could go between the upper and lower pulley. While that design makes it easier to install a chain, it also means that the cage is only a single piece in that area and therefore vulnerable to getting bent.

The two pulley should be inline with each other. These look to be off a bit. There may be other places that are bent as well. It may also be worth at least doing a visual inspection on the derailleur hanger.

Just grab that cage with two hands and pull toward you until it’s straight.

My theory is that if I work out where it's bent I can probably fix it with 'measured' violence.

I was using BBB RollerBoys replacement pulleys with shim washers to make them exactly the same width as the original SunTour pulleys.
https://www.amazon.co.uk/dp/B000VT2D...d_asin_title_1
In that photo I've nicked them and moved them to the Mk1.
So the Mk2 vs Mk1 alignment test was done with the same (RolleyBoys) pulleys.
In the photo above I've reinstated original SunTour pulleys.

Tomorrow I'll try a better photo with the camera the exact same distance from the wheel as the pulleys are.
I'll also take the cage off and check it's flat - I can also photo the cage with the shaft (#19) installed but no RD body.

Generally speaking they bend at the weakest point.

And although often this is some part of the cage, I have had an upper pivot bolt that although it allowed the body to swing freely, was either bent or had been manufactured with the fixing threads off-axis to the bolt itself.
Which was the stimulus for some pre-1066 language when discovered.

Before you take anything apart, slip the chain off, place your eye at the camera location and swing the cage as far as possible in both directions. Doing this should help you see where the issue is.

I tried it with the chain off, couldn't spot anything.
I decided to try swapping parts with my good Cyclone Mk2 GT.

The good Mk2 is on my Cyclone
left = good, but forgot spacing washer on RD mount
right = bad
Camera and bike should not have moved between shots.


I also checked the cage is flat:

I'm calling that flat.

The advantage of my new approach is I can try swapping the entire cage + axle assembly to see if it's the cage or the body that's bent.
But the TdF has hit the Alps so that's going to have to happen later

Edit: I measured the angles in Photoshop.
Good RD.
Bad RD

I did not read or catch that the frame derailleur hangar was checked for alignment with a tool such as to set up index shifting.
I would start there if applicable.
I am also assuming the wheel is the correct measure for the hub.
no excess slop in the mounting pivot.

these mechanisms are pretty stout, possible the parellogram is tweaked, not easy to do other than stuffing the cage into the spokes.


I was given a Nuovo Record mech 40 years ago, was performing poorly, had a twist, variable effect on performance, really depended on the cage orientation. had or made the tools, it was slightly twisted. risky to correct but what to lose? Time in the shop, a mill and a lathe, leverage.

Yep the drop outs were checked - by a very good frame builder.
I built the wheels - they do fit
Also rebuilt the RDs - no slop I believe.

I'm mostly trying to understand which bit is bent - can't see anything obvious with the cage so it might be the parallelogram somehow.
But I've now discovered that the bad RD is off by 3 degrees which isn't much and possibly is not something you would see just by looking at the RD.

And I swapped the cages+axles

Left = good body + bad cage (SunTour pulleys)
Right = bad body + good cage (RollerBoys pulleys)



Measured angles:
Left.
Right.
That looks to me like most of the misalignment is with the "bad body".
I'm not sure there's much I can do about fixing the parallelogram.

On the upside the good body + good cage gave the best results, so I should probably just use that.

First.. great photos to explain the issue! A great example for others to emulate!

Second, I've seen the same issue on one of my bikes. I don't have troubles with the derailleur hitting the spokes, though.
I've used the Park DAG tool (derailleur alignment gauge?) to confirm that the hangar wasn't at fault.
I haven't followed up on it, but I don't expect that it is due to bending. I'm more inclined to blame wear at the upper and lower pivots. This is just speculation, though.
The upper pivot doesn't move very often, so I'm less inclined to blame it.
The lower pivot moves a lot, and doesn't really have a good way to preserve the lube , nor a good way to add lube periodically. You have to pull that lower pivot all apart, and that's a big pain.

I did retire the Cyclone GT that's been on the bike for a while, mostly because the parallelogram spring was getting weak. I've got a Cyclone GT Mk II on it now, and the pulley cage is much more aligned.
Ideally, I'd pull out that old Cyclone GT and try to quantify the play in that lower pivot. However... other tasks have higher priorities right now.

Steve in Peoria

I pulled out the bike, got the tripod and camera, and took a shot of the Cyclone GT Mk II that's currently on the bike.
As can be seen, the pulley cage is a bit out of alignment with the cassette and wheel. The derailleur hanger has been checked and tweaked before to address this sort of visible misalignment, and was clearly not the problem.




I will note that if I grab the bottom of the pulley cage, it doesn't take much force to deflect it 5 or 10mm to the right.
I'm inclined to attribute this to wear or slop in the derailleur somewhere. Or just some flex? Not sure.

Steve in Peoria

Cheers, that's definitely worth trying, looks very similar to what I'm getting.
My good RD moves about 3 or 4mm under medium finger pressure.

But I've reassembled my RDs several times, I think it's just flex.
Some amount of flex is probably a good thing, and is why it still works 40 years later and hasn't snapped off yet.