Shadow MTB rear mechs Vs older design - old tucks into smaller sprockets better?

[Manc33]

On older rear mechs like this RD-M750, you could still push the rear mech closer to the sprockets by hand, it was sprung at the main fixing bolt so it could rotate...





On the newer shadow rear mechs like this RD-M772, the mech is fixed wherever you set it to and it cannot move closer to the sprockets, the b-tension is not sprung, the spring is simply not there anymore - this apparently solved the problem of "chain slap" and these days, even Dura-Ace road mechs are now using this design...




This got me thinking, if the older design can tuck into the smaller sprockets more - which they do because as they collapse more on the higher sprockets, their spring gets weaker thus allowing it to collapse more - doesn't it mean shifting is going to be better on cassettes with a large low sprocket on the older mechs?

The newer design does solve chain slap - but doesn't this sacrifice shifting performance when the mech is effectively locked in place?

I have noticed on some setups with these newer mechs if using large low sprockets like 36t and upwards, when the b-tension needs to be set to clear the low sprocket properly, the upper jockey wheel is nowhere near the teeth of the smaller sprockets, but on the older mechs it tucked in more.

Aren't we just constantly getting sold gimmicks and fads?

[pwu]

M772 (shadow)
Max. teeth large sprocket: 36T
Min. teeth large sprocket: 32T
Min. teeth small sprocket: 11T
Max. Front Difference: 22T
Total Capacity: 45T

M591 (conventional)
Low Sprocket Max: 34T
Low Sprocket Min: 28T
Top Spocket Max: 12T
Top Spocket Min: 11T
Max. Front Difference: 22T
Total Capacity: 45T

I have both, the M772 (ebay used £16.01) is a lot more accurate at shifting and responds to the tiniest amount of pressure on the shifter (Alivio M430 in my case), the barrel adjuster on the the M591 (came with bike) never seemed to work properly I'd say it shifted ok. Despite what the specs say there is a video on youtube of someone using the M772 with 11-28 and 12-23, I use 11-30, as long as the center of the guide pulley is 26mm away from the lowest gear ( https://bikebrothers.no/wp-content/uplo ... D_Tips.pdf ) it should work.

The shorter cable run and no barrel adjuster means less friction, I use a long nose ferrule made by fibrax to keep dirt out of the outer cables ( https://www.ebay.co.uk/itm/Bike-Outer-H ... 2749.l2649 ).

The M772 is so much easier to index. Undo cable, adjust barrel adjuster at shifter 8 dentents outward, manual push derailleur to second smallest sprocket and attach cable, pedal so it falls back to smallest sprocket, shift to second smallest sprocket again and push shifter just enough that its past the dead zone and about to actually pull cable and there should be a rubbing/buzzing sound because it wants to shift to the third smallest cog make sure it doesn't make the shift under that condition and the indexing is 99% done, might need a tiny adjustment on first ride after. I have zero chain noise in any sprocket.

[Manc33]

I have the M772 and it does shift well, but I wonder why they do when the chain is further away from the smaller sprockets and the mech can't rotate anymore. The best shifting I have ever seen was from a 6500 (GS). It even beat the more expensive 7700 GS which for some reason never did quite work right for me, despite the name on it. Maybe on 12-23t cassettes it does but I never tried it on such a small low sprocket, I was always on 26/27/28t when testing those out.

[scottg]

Shadow mechs don't stick out as far from the wheel as trad mechs,
so they are less likely to be snagged by brush.
So Shimano tells us.

See red line in pic, borrowed below.

[Brucey]

shadow mechs (in theory) don't work well with such a wide range of (biggest) sprocket sizes as twin-sprung RDs. But they more than make up for this by wearing far more slowly at the top knuckle, and this means that shifting performance is better for longer.

The working life of many older shimano RDs is limited by wear in the top knuckle. With a single-pivot slant mech, this is less likely to be the case.

cheers

[Manc33]

I have had one a while ago that when tightened on, wouldn't rotate properly. This would never be an issue on the shadow versions.

[nigelnightmare]

I have a Deore XT RD-M772 on my ICE Vortex fs 2011 running a custom Capreo cassette of 9-32 (so, well out of spec range) and it's shifting is fine, Better than another one with a Sram X9 on it.
Mind you I also have A Deore RD-M531 on the ICE Qnt, that since I fitted new (Sealed bearing so no float) aluminium jockey wheels it shifts even better, Weather I'm using a capreo 9-32 or a sram 11-32 cassette.

I find the 772 more noisy than either the 531 or the sram X9, But then again my head is only around 18" away from the rear derailleur so I tend to notice it more.

[Manc33]

A while back on 10s people were running those cassettes with 50t sprockets. I don't know what the shifting must be like on those with a shadow rear mech but I know the upper jockey wheel cannot be anywhere near the smaller sprockets. Maybe having a chain that's deliberately too long could help with it. I think what limits it is the chain (when chain length is the right length).

Why can't Shimano just make a rear mech that moves up the cassette in a curve as opposed to a straight line

[Brucey]

….Why can't Shimano just make a rear mech that moves up the cassette in a curve as opposed to a straight line

Actually they (and many other manufacturers) already have. Anytime you have an upper pulley that is offset w.r.t. (rather than concentric with) the A pivot, the chain gap describes a curved profile of some kind as you change gears. This design feature used to be commonplace; a problem is that it doesn't work too well when you have a wide range in the chainrings.

Currently SRAM make some single pivot mechs which don't have slant paralellograms at all, and just rely on an offset pulley to vary the chain gap. They are for 1x setups.

It used to be that a small chain gap was essential if you wanted crisp shifting. Now however, that isn't the case in quite the same way. A consistent chain gap helps (esp when everything is worn) but in 10s/11s the shifts can be made 'sweet as' over a fair range provided you have well shaped sprockets and matching chain. When the chain etc is worn the shifts can get a bit vague but until then they can be very good indeed even if the chain gap is quite large. If this were not the case then hanger extenders etc just wouldn't work .

cheers

[Manc33]

I guess it just shows how far things have come that the upper jockey wheel can now be an inch away and it still shifts acceptably.

With the old design it sort of is going up in a curve, more in line with the shape of the cassette (11-32t). I keep forgetting which way it's tensioned. If, on the smallest sprocket the older mechs are under least tension (at the b-tension spring) and on the low sprocket it's under most tension, this fits with what I am trying to explain in the first post, it tucks in exponentially more, the more collapsed the mech is - and this is what the newer shadow doesn't do, it can only ever go in a straight line up the cassette.

What gets me is people coped for a long time on the old design, making the new shadows... suspiciously gimmicky!

I can understand going to a shadow mech on a MTB because it gets rid of that loop of housing (even though someone in this thread has already shown it can still stick out) and you're going over far rougher terrain (potholes notwithstanding) but on a road bike it doesn't seem to make much sense and yet, the new road stuff is going to shadow mechs.

Also it can't cost as much to make a mech that's got the b-tension spring missing. All I keep thinking is why was it there in the first place if it could be got rid of. If mechs started out like shadow types and they added that spring later I could understand it, but it's the other way around.

[Brucey]

I guess it just shows how far things have come that the upper jockey wheel can now be an inch away and it still shifts acceptably....

yep. it doesn't last for ever, but its clever stuff.

….What gets me is people coped for a long time on the old design, making the new shadows... suspiciously gimmicky!....

ha. The 'new' shadow mechs are functionally identical to many other mechs eg many SRAM designs from the last 25 years, and even Sun Tour mechs from the 1970s; they were single sprung pivot, slant parallelogram, top pulley concentric with the A pivot too.



Shimano Shadow



SRAM X-0 mech


from Disraeli Gears

Cyclone GT allowed 32T biggest sprocket on a standard length hanger, 34T if the hanger was long pattern (32mm or more)

the Sun Tour mechs worked pretty well BITD and (by almost any standards) they work well with HG type sprockets today. Some versions of the model above (Cyclone GT) weigh about 175g, and could easily have been made even lighter if a smattering of Ti had been used instead of steel. You need spend a lot of money to get anything that lightweight these days.

cheers

[Manc33]

lol, so Shimano were being innovative adding it... but have again now taken it away. Just as the next generation of youngsters comes along to be made to think taking the spring away is an innovation. They will add it again in 30 years time.

[Brucey]

lol, so Shimano were being innovative adding it...

Nope. It certainly wasn't shimano's idea; Simplex were doing it in about 1960, and (although they had a patent of some kind) they may not have been the first. What shimano did do is they were the first to combine the slant paralellogram (for many years a SunTour patent) with the twin-sprung design, and they did so as soon as the SunTour patent expired.... SunTour in turn thought that they ought to copy shimano, and so did Campagnolo (later). SRAM have always stuck with derailleurs that have a single sprung pivot, but then they have only been making derailleurs for about 23 years.



http://www.disraeligears.co.uk/Site/Simplex_Juy_Export_61_derailleur.html

cheers

[Manc33]

Back then though cassettes typically had far smaller low sprockets, so it not being sprung at the fixing bolt wouldn't make as much of a difference to shifting performance. These days, 12-Speed cassettes are going up to a 51t (from a 10t) without the spring being there.

[Brucey]

offset pulleys, slant parallelograms and twin-sprung designs all add capacity (although they are certainly not the only things which can do so) whilst preserving some quality in the shifting. Of these the twin sprung feature adds a little more capacity to some designs, but not that much; enough to vary the range of the cassette by another four or six teeth or so.

What it does do (I think) is that it preserves shift quality even when the guide pulley is close to the sprocket, by allowing the mech to articulate away from the sprocket during the shift. If you run a single pivot derailleur that close, the shift quality deteriorates; shifts become noticeably more clunky.

However as soon as you don't need the guide pulley close to the sprocket any more, this advantage of twin-sprung mechs is of much reduced value.

BITD the simplex mech worked about as well as other mechs (which was 'not very' using chains and sprockets of the period), and had noticeably more capacity than many others. Most of the prestige models would accept a 28T sprocket. But when everything was right shifting using a twin-sprung mech was noticeably slicker than a single pivot one.

Wide range cassettes might be mistaken for a new idea. However in the 1970s Sun Tour made freewheels which went up to 36 or 38T or something, and also made (single pivot, slant parallelogram) mechs to shift them. Sound familiar? The shifting was not great mind; it took more sprockets (smaller gaps) and cunning sprockets/chains to manage that.

[edit; 34T is all I can find reference to but I'm sure larger were available (from somewhere or other)]
[edit 2; 38T sprockets are listed in SunTour's 1981 catalogue. Not sure what mech they meant to go with them but IIRC there were mechs which would manage a 38T sprocket.]

cheers