Aardwolf

Hiya Folks,
went for a test ride on my new Holdsworth Cyclone earlier today, but one cog is having shifting issues.

The freewheel is a SunTour New Winner Ultra 6 13-28
it's 13-15-17-19-21-28 so the cogs/spacers should look like this:
..
The T17 is the one having issues.

I originally identified it as T17 when I bought it but it isn't marked, I've since got another that's actually stamped R17 and I can't see a difference.
So my questions is what's the difference between T17 and R17 ?
And would using the wrong one cause skipping ?

steelbikeguy

as far as I can tell, the difference is the flange.
Both have a flange that creates the desired spacing between the cogs.. mostly.
The flange on the R cogs have an extra task, and that is to be tall enough to push against the UTN spacer that keeps the A cogs compressed against the back (inboard) side of the freewheel.
Here's the SunTour illustration of the setup for an Ultra 6 New Winner that might help...




I've got some spare T17's, if you need some dimensions from them.

Steve in Peoria (where I've spent too many hours trying to understand the assorted cogs in my stash)

repechage

The description of the issue with the 17 is missing.
17 is a pretty common cog to use way back, 42 x 17 is a useful tooling around gear.
implication, they wear faster.

Aardwolf

Cheers, that makes a lot of sense, that's the one illustration I didn't really study

Here's some measurements on my cogs

R17 + other17UTN spacerSo the R17 flange extends about 1mm across the UTN spacer, pretty much exactly like the diagram.

If you get bored a measurement of the thickness of a T17 would be handy
Edit: just measured the T15 as 3.8mm thick.

Still not clear why T17 replaced by R17 would shift badly though ?

Aardwolf

Issue is it skips quite a lot and just doesn't run smooth.
My first thought was wear, then I spotted it might be an R17 instead of T17.
So maybe it was trying to jump to an adjacent gear if the spacing was wrong.

Here's the 17 that was giving the problem

steelbikeguy

this one tooth does grab my attention, and suggests excessive wear.




Steve in Peoria

Aardwolf

I've just spotted something else

R17 - teeth chamfered on drive side, straight on small cog side
T15 - teeth chamfered on drive side and small cog side

My R17 has less wear so I can try that, but I'm thinking the missing chamfering is the problem.

Edit: Just checked the diagram up the page and that shows the chamfering difference.

steelbikeguy

I think I've noticed that difference in chamfering before, but forgot about it.
Too bad there's no one around who could tell us why chamfering on one side is good for some cog sizes/locations but not others.

I haven't dug through my box of cogs, but I have measured thicknesses before, and both the R and T are 3.8mm thick.
Somewhere in the past, I did take pics of a R15 and T15 cog to compare them.
In both photos, the R cog is on the left and the T cog is on the right.







You'll note that these cogs don't have any machining marks (i.e. grooves) in them. I did take a photo of two R cogs.. one with the marks and one without. No idea why there is a difference.




Steve in Peoria

pastorbobnlnh 

Just a side note: Finding replacement threaded sprockets in fair to NOS condition for the New Winner model is challenging and problematic.

A work around for these sprockets could be to use Perfect/ProCompe threaded sprockets. For instance, a 17T second position Perfect/ProCompe sprocket could work for Aardwolf . It would at least offer an alternative. I'll take a look and see what I have on hand.

jethin

I feel like the case of the Suntour freewheel assembly is one C&V mystery I’m never going to solve.

Aardwolf

Cheers, I was wondering about Perfect 6 speed etc cogs.

There's a "Perfect XF 14/15/17/19/21/24" on Ebay at the moment - £10 because it's fairly crusty.
You say 'second position' so that freewheel doesn't have a suitable 17T then ?

pastorbobnlnh 

In the ebay case listed above, the 17T would be tabbed and not threaded. In most cases* on Perfect/ProCompe the first two sprockets (smallest ones) are threaded. They range in tooth count beginning with 14T and end with 18T. These fit in the middle positions on the New Winner (which only uses tabs on the two largest sprockets). The final three or four (dependent on 5 or 6-speed models) sprockets are all tabbed.

(* There is a 6-speed Perfect variant which has a 13T threaded first sprocket, followed by two threaded sprockets, and ending with three tabbed sprockets in the largest positions. These sprockets will all fit on a New Winner Body, with the possible exception being the 13T.)

pastorbobnlnh 

A few pictures 😁

Top view of a Perfect 17T next to a New Winner body.

Bottom view of 17T

17T Perfect sprocket threaded to a New Winner body.

I can't locate my digital calipers. So not exact, but the thickness with the built in spacer is slightly larger than 3.5mm.

Hope this is helpful.

steelbikeguy

That's a good point!
My eclectic collection of cogs is subdivided into silver, gold and brown cogs of each type.
I don't have "show" bikes, so I really don't care about the color. They all end up the color of "used chain lube" anyway, so it doesn't matter.

My collection also includes cogs with either no markings, or strange/wrong markings. I'm guessing they came from later generations of SunTour freewheels.
For that reason, my spreadsheet of my cog inventory includes details like whether it is splined or threaded, how many splines, how thick it is, I.D., etc.
Never a dull moment.

An example is this cog marked A21 that looks a lot more like a R or T cog.




.... or this three-spline A2-15 cog. Perhaps it isn't even from a SunTour freewheel?




Steve in Peoria

pastorbobnlnh 

The A2, 15T, three-splined is for a Regina CX body. It will also fit on some early Shimano freewheels.

The A21, 21T, is I believe, a Suntour for New Winner bodies, but could have also been for the very first-generation Winner body. IIRC, the threaded Winner sprockets went all the way to 24T ones.

dddd

The OP's skipping under heavy load would seem certain to be the result of a new chain engaging a visibly-worn sprocket.

You can usually eliminate such skipping on cogs with this level of wear by beveling (perhaps .5mm for starters) the driven-side corner of each tooth at a 40-degree angle to a radial line out from the cog center. This will eliminate the problem of the chain roller catching that corner and causing a failure of the roller to fall between teeth under heavier loading conditions (where the chain is running in a more-advanced position relative to it's original position on an un-worn sprocket.
I've repaired dozens of such precious freewheel and cassette cogs this way, using a Dremel with a 3/8" or 1/2" diameter stone (usually not even removing the wheel from the bike to do so) in perhaps two or three minutes for typically smaller cogs like this one or smaller.

The same condition can occur on a chainring, but where the much more greatly-tensed top run of the chain (where the chain first engages the teeth) forces the rollers to fall into the spaces between teeth. In such cases when a new chain is fitted, a rumble is heard and felt as the pedaling force is increased. The fix is the same, except it's the driving side of the chainring teeth instead of the driven side of the cog's teeth.

The key in both cases is not to remove too much metal on the first go. Very small cuts will make a noticeable improvement, leading up to a totally skip-free cog in one to three rounds of grinding.

Some particular cogs and freewheels are becoming quite scarce on the ground these days, so I am loath to discard worn cogs when easy replacements aren't available.

Aardwolf

To check if it was the missing bevelling I got the dremel out:



No difference, still skipping. So it's either just worn, or there's a difference I haven't spotted.
I can try it again another time.

Found a T17 on Ebay, arrived today and weeeeeeeeeeeeeeeeeeeeeeeeeee all the gears work properly

dddd

Just to be sure, your beveling is not the same beveling I was referring to in the previous post (I was referring to the driven side of each tooth, not the left of right side of each tooth).

Wear to the lower part of the driven side of each tooth is what is causing the chain to cease meshing between the teeth under heavy load, and is corrected by beveling the un-worn tip corner of the driven side of each tooth, as shown exaggerated below on the one tooth up around 12:00 position.
The bevel should also be a little more vertical than depicted, but you get the idea.

Aardwolf

Thanks for the details.
I did get your point about the wear, but I wasn't addressing it with the bevelling I did - I was fixing known differences between R17 and T17.
It may well be that the wear you describe is the actual problem.

It's just occured to me there was an easier way to test if it's wear - create a freewheel that uses that R17 cog in the correct place and see if it still skips

zukahn1

No real mystery the suntour's at the time where limited by design or maker at 13 as were other makers I have seen a lot of vintage corncobs that went 13 to 25 From a dozen makers but no 12 or 11 As a note 12 and 11 won't work good on most bikes and are not needed . On most bike's a little older to semi current you can just change the small chain ring and get negative gearing.. I have gotten vintage correct 13-32 32-52 pretty easy which should work for nearly any situation. which will give you a max low of less than 1X1 with slack slightly negative and 13 X 52 for close to for 4x which is the max chain wrap for a tor standard chain old or new.

daka

I'm not sure I understand everything you are trying to say in your post but 32:32 is exactly 1:1, not less, resulting in a 27 inch gear with 700C wheels.
Based on my own experience with a 34-52 chainwheel set up, it's hard to imagine that a 32-52 shifts very well and would likely require a double shift on anything but the most abrupt changes in terrain.
This setup will also require a RD that can wrap 39t, which limits choices.

plonz 

Here’s my Tier 1 tech support question: Have you checked hanger alignment? I’ve had bent cages and misaligned hangers give me fits on a single cog. Even if it’s not solely due to misalignment, it might be amplifying effects of a worn tooth or chain.

Aardwolf

Yep, I checked the alignment first - it was slightly off and the RD couldn't pull the chain onto the small cog: 1961 Holdsworth Cyclone build
All that was fixed before the T17 issue showed up on a 14 mile test ride.

Aardwolf

Well T17 has been working fine for the last 100ish miles, but I spotted another problem:
..
I'm now giving the Cyclone its 110 miles checkup so I did some research, tried a few things and I've just got it working:
..
They are bevelled to allow the chain to move but it's not obvious which side is bevelled and I've got an A32 which is definitely bevelled on both sides.

If it's got a mark, eg "A 32" that side goes towards the spokes. The splined sprockets were stamped and the letter is on the 'smooth' side of the stamping.
So the sharp edge side of the stamping is away from the spokes.

My A21 and A28 have no markings so of course I guessed and got it 100% wrong.
Tested with A24+A32, both marked, and found I was doing it wrong.
Just been for a short ride and all shifting is now working properly

steelbikeguy

Did we discuss bevels before?
I seem to recall commenting that it's odd that some cogs are beveled on both sides, while others are just beveled on one.
The interesting thing, sort of, is that the SunTour Winner drawing actually shows the bevels on the splined cogs, and shows the bevel on the inboard side (i.e. towards the spokes).



...although it's a bit subtle.
Here's an enlarged look, which should make it easier to notice the bevels...




Amazing how much useful info is in that little drawing!


Good to hear!

Steve in Peoria