dddd

Being able to adjust the "chain gap", the distance from the upper (guide) pulley and the freewheel teeth is re4ally important when the freewheel isn't almost too big for the particular derailer.
The idea here is that you want the smallest gap possible without the pulley and cog teeth ever touching, so that the cage movement isn't so much used up by the chain simply angling over. Hence, modern (i.e. relatively super-flexible) chain can be very problematic with certain older derailers that don't closely follow the freewheel's outer profile.
The old Shimano rear derailers, as well as Simplex and some Huret derailers had springs at both the cage pivot and at the mounting bolt pivot, so the derailer body could swing forward and back to better track the outer profile of the freewheel.
Suntour rear derailers on the other hand often had a "B-tension screw" at the mounting bolt pivot which allowed manual adjustment of the derailer body angle (and of the chain gap), but which had an insufficient range of adjustment to closely follow the outer profile of smaller freewheels. I've often removed the B-tension screw entirely and even filed some metal from the "B"-screw's threaded lug in order to allow the derailer to swing forward further (and which swung the upper pulley closer to the cogs.

Older "bushed" chains were quite stiff laterally and didn't require the top pulley to be very close to the freewheel, but different brands of freewheels performed very differently while shifting due to the shape of their teeth.

So in addition to maxing out Suntour derailer's forward swing by removing the B-tension screw and filing metal off of the screw's lug, Shimano and other "dual sprung" derailers can typically be disassembled and with additional holes drilled for the end of one of the pivot springs in order to effect some adjustment of the derailer body angle.
Note that Simplex derailers often featured a locknut where the cage pivot shaft met the outer cage plate, allowing spring tension adjustment at the cage pivot in order to accommodate a wider range of freewheel sizes (size of the largest cog).
Shown below, this Simplex derailer has had it's cage pivot spring tension increased using said locknut with an Allen key to turn and hold the cage pivot shaft during adjustment while the locknut is slackened slightly and then re-secured. This allowed me to retain super-crisp shifting using the smaller 13-24T freewheel that I wanted to use.
You can see the super-close proximity of the top pulley to the freewheel teeth, giving super-responsive shifting even using a flexible, modern, 9s chain.




Lastly, clunky or recalcitrant shifting behavior can also be somewhat the result of things like metal-to-metal contact between the cable and any guide features or cable housings, and even can be due to a poorly- or non-lubricated shift lever pivot (I always disassemble and oil up friction lever's pivoting internals with plain motor oil, which lasts for quite a while as in years). I also affix plastic "spaghetti" tubing/liner material in any metal cable guide groove or tube, securing the liner using compression-fit or with a spot of adhesive (depending on the configuration of the cable guide and also depending on the diameter of plastic tubing I'm using). A fully plastic-lined cable path hugely improves shifting even on friction-shifted bikes!

gaucho777 

Sometimes when I'm reading this site on my phone, (due to my weakening eyes) I zoom in in such a way that I only see the comments but not the bar on the left with the user names. When I read this comment:

I said to myself, this has to be dddd! Love you, man.

gaucho777 

There are a lot of variables and not a lot of specifics provided. Some vintage set-ups may indeed feel clunky, but this is not always expected or normal for all vintage bikes. Nuovo Record derailleurs with Nuovo Record or Gran Sport shifters can feel a bit sloppy and require overshifts, but my bikes with top-end derailleurs and Simplex retrofriction shifters are super-smooth across all gears using a variety of chains. Sometimes the smooth, quiet shifting of a nice friction setup is preferred to the clunk-clunk-clunk of an indexed drivetrain.

sd5782 

A Trek 412 I am working on had fairly poor shifting in back with a favorite Suntour VX-GT across a nice evenly spaced Suntour narrow spaced 14-30 six speed. I was surprised as that is not usual. I even used a vintage chain with little wear and it was the same. Taking out the B screw and a bit of filing still didn’t make it pleasant. I scratched my head and put on a vintage Shimano Exage that got the upper pulley closer. Upon further inspection and measuring, I found that the derailleur mounting location was a bit lower than most, so not allowing the combo to perform well.

dddd

I don't know if the claw-style hanger on my Nishiki Kokusai is relatively long like yours, it seems so, but I see what you mean about getting the top pulley on these older Suntour touring derailers close to the teeth of a 30T or even a 32T largest cog!
Pictured below, the B-screw is all the way out, and there is still a generous gap up from the top pulley to the teeth of a 32t largest cog.
Just look at the body angle, tipped slightly up from horizontal.
I later fitted a smaller 13-30T, 7s Alpha freewheel, and had to do a bit of metal removal (to the claw-hanger's tang this time) in order for the top pulley to approach the teeth of the 30t cog.
Shifting is, under-statedly, "pleasantly responsive" now, with the short, fully-lined cable path up to my fave Suntour "beer tap" shift levers.

Again using 9s chain here (what I often found to work best with 7s Suntour freewheels).

John E

What is your definition of "stretched" (technically, pivot-worn, because stainless steel links themselves do not actually stretch with use)? I follow Sheldon's 1/2 percent standard, i.e., 1/16" elongation across 24 half-links, or 1 part in 16*12 = 1/192 = 0.5%.

gazman22

Great input! Thanks! there is a lot to digest here and it is exactly what I am looking for to understand all the elements at play.

gazman22

Park tool CC-3.2 with .5% and .75% measurement gauges. While I think 5, 6, and 7 speed chains are good up the .75% mark, if the is elongated more than the .5% I will replace.

gazman22

Another piece in the equation that I am not clear on. When the dropouts do not have a way to locate how far back the rear wheel axle should be I tend to go all the way back. Now I wonder if more forward, as in getting the cogs clover to the upper jockey wheel is the way to go?

If the derailleur has a claw hanger (no RD mount on the drop out), should the rear axle mount tight in the claw hanger? Or, slide forward to locate cogs as close to upper jockey wheel as allowable?

Thanks!

gazman22

Want to thank everyone for their input, questions and advice. It is all very helpful.

My wife is tired of hearing about these things and frankly her answers are not as insightful as yours...

sd5782 

@dddd, look at the VXs. The upper pulley sits quite a bit higher because of how the cage plates are mounted and pivot. The S can handle 30 tooth sprockets and I think a 27 tooth wrap. I think your bike falls in that range. I love the S version. It came on the Trek I am working on, but I put a triple up front so the S can’t handle the wrap.

dddd

Typically, when a claw-hanger is used, it gets pushed back as far as the securing nut allows, and which puts the axle at least mid-way forward along the dropout slot.
One then only has to straighten the wheel to center between the chainstays when installing the wheel with a driveside claw hanger.

Having the axle fully rearward in the dropout (without any claw hanger, axle adjusters or separate axle locating stop) can be problematic in terms of chain gap (but might gain you some needed gap if using a larger freewheel size that exceeds what the derailer was designed to handle).

Adding a single axle-stop fitting to one dropout is perfectly adequate, requiring only that when installing the rear wheel, one perhaps rest a shoulder or hand on the saddle for some downward force (while steering the wheel straight with a knee) while securing the quick-release lever.

I always had trouble aligning axle stops equally when I tried installing them on both dropouts, which must be why the industry typically used either a stop on one side, or dual adjusters, or fixed-position "vertical" dropout slots.

dddd

You're right about this. And not just the VX-S, but I believe also some of the similar, later and lower-tier models.
One such mostly-steel, mid(?)-cage model came on my 1984-ish Motive Defiant, and seems made for the 13-30t 7s freewheel that I am now using.
Shown below, with bolt-on axle sitting pretty far back, and with what looks like an earlier upgraded 6s HG 14-28t freewheel on there (before I finally went with a 13-30t, Suntour Alpha-7 freewheel).

Mr. 66

The old parts can get banged around. Some derailleurs need to be retired replaced. A derailleur cage can get tweeked. I would look at light lube all the pivots, and check the cage make sure the pulleys are lined up with the freewheel cogs. I would check the rd hanger also.

Bad Lag

Yes, you do HAVE to do that. Also, you absolutely MUST let off on the pedal pressure (none, except what's needed to keep the chain moving) to get a smooth shift. It will never be silent but does not need to be noisy (grinding your gears). Hand-foot coordination is needed.

Is your frame of reference modern indexed shifting or is this relative to how they used to shift (and somehow are no longer)?

gazman22

My frame of reference is not modern indexed shifting. It is a comparison of what appear to be relatively similar generation RD's with some shifting smoothly and others not. I am trying to gain a better understanding of all the elements at play so I can diagnose and improve performance in the future.

While I have the laymen knowledge to get me there some of the time, I suspect (and see from some of the insights and comments above) there is more fo me to consider to fine tune the RD and its relationship to the cogs to improve shifting.

Please keep the suggestions and observations coming!

gazman22

When setting up the bike are you making these adjustments with the chain off? That is not what I currently do and I could see that it may be an easier way to see the path of the RD and its relationship to the cogs.

dddd

I would always be checking things out with the chain on.

Note that some rear derailers have the top pulley concentric with the cage pivot, while others have the top pulley offset from the cage pivot.

So by having the chain in place, and tested on both chainrings, one can visualize the position of the top pulley as it will be while actually riding in either chainring.

On bikes with an offset top pulley, my final "acid test" of the shifting quality and adjustments is always done while in the big chainring, because the pulley offset increases the chain gap as the chain is pulled tighter. Another test using the small chainring mostly just verifies that there is no chain sag in the smallest cog and that there is no cog-to-pulley interference on the largest cog.

Note that the old Suntour touring derailers like the V-GT had no pulley offset. They did however feature a spiral-shaped cage plate which moved away from the freewheel as the lower pulley was pulled forward, so again it was the big-ring position where I would verify shifting performance shifting toward larger cogs.

The Suntour rear derailers also featured a slanted parallelogram to better track the outer shape of the freewheel, but which really didn't much compensate for the bigger changes in chain gap that occur when shifting across wide-ratio freewheels.
The amount of slant that could be employed was limited by the vertical forces of road shock trying to auto-shift the rear derailer, too much slant and the spring had to be made stiffer and the shift lever friction increased (both are bad for shifting feel).
SRAM's recent "one-by" (single-chainring) rear derailers for off-road use have done away with any parallelogram slant for this very reason, and use a heavy amount of top pulley offset to track the steep profiles of their extremely wide-ratio cassettes. This means that these are only for single-chainring applications since chain tension variations from front shifting would play havoc with the pulley offset-caused chain gap variation.
See:
https://www.sram.com/en/sram/models/rd-x0-1-a2

dddd

Campagnolo's old Gran Turismo rear derailer is a particularly interesting study in what does and does not work well in terms of shifting.
It's large cage and huge pulley offset, combined with two sprung pivots(!) promised compatibility with wide-range gearing for sure, but it was designed well before modern bushingless/flexible chains ever saw the light of day.
So, while it could handle wide gearing and lots of chain wrap, it had an unfortunately-large chain gap which was sadly not made adjustable in any way.
I took one of these apart and drilled a new hole for one end of the cage-pivot spring, which dramatically closed down the chain gap while still handling a 13-34t, 6s cassette! The derailer would then actually index using a Shimano 7s bar-end lever, much to my surprise.
The other lesser problem with these derailers was their very stiff return spring, which introduced more in the way of cable elasticity, so begged for longer shift levers, shorter cables and even use of thicker brake cable to improve motion accuracy between the lever and derailer. The stiff return spring was kind of sad because this derailer had no slant to it and so could have used a lighter spring like the one that I ultimately fabricated for mine (not very successfully, mine was made from a spoke and so was a little light on tension).

Here you can see the massive pulley offset from eight feet away, with the chain gap very much reduced from the increased cage-pivot spring spring tension:


The derailer body had a thin rim where additional spring-tension holes can be drilled:



My new lighter spring (with added coils) proved a bit of a failure due to limited materials I had to work with:


Here's a view of how derailers can be better designed by way of positioning the linkage upside-down in order to better track the typically-convcave outer profile of the freewheel using a knee-action parallelogram. This was a revolutionary design in the late 1950's, though had to be configured for a particular, somewhat narrow range of intended freewheel size (close to 14-28t):

gazman22

Here are 2 movies, one of what I consider smooth shifting, the other not so smooth. Probably best interpreted through sound more than video. In my Dropbox here:

https://www.dropbox.com/sh/lxz1ceiac...PB-_T6s_a?dl=0

Or google drive here (which seems to be slow in loading them).

Smooth Shifting
https://drive.google.com/file/d/1RPe...usp=share_link

Not Smooth Shifting
https://drive.google.com/file/d/15HL...usp=share_link

steelbikeguy

am I the only person having trouble with the links?




Steve in Peoria

edit: maybe the link needs to be set to be available to anyone with the link, versus just those on the access list? This is how I set up the files that I shared:

gazman22

Try now or look in the Dropbox link.

steelbikeguy

The smoother shifting had some advantages.. a Shimano derailleur with two sprung pivots, a smaller freewheel (I think?), and possibly a bushingless chain (although I didn't really see what the not-smooth-shifting had).

The not-smooth shifting seemed entirely reasonable to me, to be honest. Might benefit from some chain lube?? To my casual observance, it's not much worse than my Campy Nuovo Record shifting a 13-24 five speed freewheel.

Steve in Peoria

gazman22

Spent some time looking at the "not smooth shifting".

1. the noise on close inspection is clearly all from the chain hitting the inside of the outer cage plate, which has a large diameter, visible as I shift onto the larger sprockets. This is happening where the chain leaves the upper pulley and meets the cog

2. I let out the b-tension screw.

3. also slid the wheel a little more forward in the drop out.

4. maybe collectively 2 & 3 above helped a bit.

I read up on bushing vs. bushingless chains. It seems on this "noisy" bike, if it is a bushingless chain it would be adding to the noise. Even if there is only a few links from where the chain leaves the upper pulley and meets the cog, this is where the chain is rubbing on the inside face of the outer cage plate. Wouldn't a stiffer/bushinged chain stay straighter and engage the cog sooner (less rubbing on cage)?

gazman22

As an interesting side note on the "noisy" bike. It is a Viscount, and while I don't know if it is all original it is a double up front with a long cage RD (maybe medium?) in the rear with a largest cog of 34. I had not seen this approach before or a 5 speed freewheel with 34 teeth.