Shimergo - latest news

For discussions about bikes and equipment.
Brucey
Posts: 39824
Joined: 4 Jan 2012, 6:25pm

Re: Shimergo - latest news

Postby Brucey » 10 Jun 2020, 9:30am

from memory it is either the 17 or 19. On the parts diagram it is a unique sprocket that is only used in that cassette. IIRC the shouldering is slight; just a small (offset) variation in sprocket thickness between the centre and the part where the teeth are, I think.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

MartinC
Posts: 1872
Joined: 10 May 2007, 6:31pm
Location: Bredon

Re: Shimergo - latest news

Postby MartinC » 10 Jun 2020, 11:18am

This is so nearly a really good result. I'd be interested to see how CJ gets on longer term with this. The other problem with Shimano 10 speed road (Tiagra) is that there is only one shifter available - Tiagra STI, no bar ends or DT shifters. And these STI's only work with new Shimano brake pull. Microshift have new pull 10 BE and DT speed shifters in their but their obtainability seems to be non existent.

User avatar
CJ
Posts: 3051
Joined: 15 Jan 2007, 9:55pm

Re: Shimergo - latest news

Postby CJ » 10 Jun 2020, 11:47am

Brucey wrote:Shimano matches campag? It would be nice, except it doesn't, not quite.

If you connect a shimano 11s road mech (which is the same as 4700 10s) to a campag 10s shifter the resulting confection works, but imperfectly. I've measured and posted the actual RD movements that result about nine months or a year ago:
Brucey wrote:….. I've just done some experiments with a campag 10s shifter and a shimano 11s mech and I see what you mean about overshifting (i.e. the mech moves too far) in the lower gears.

I measured the following mech movements (low to high gears, i.e. rightwards)
1-2; 5.1mm
2-3 ; 4.4mm
3-4; 4.9mm
4-5; 4.5mm
5-6; 4.05mm
6-7; 4.3mm
7-8; 3.7mm
8-9; 4.2mm
9-10; 3.8mm

An important measurement is the 2-9 cable pull (I meant RD movement) and it should be ~28mm (7x4). I measured 29.8mm.....


I'm using a Shimano 10-speed mech. Maybe the shift ratio isn't quite the same as their 11-speed mechs. Whatever, it works very nicely with my oldish 10-speed Ergopowers and a partly worn Shimano 11-36 cassette and chain (about 0.3% elongated), which I think must be a harder test to pass than brand new stuff. And I'm not easily satisfied. By working perfectly I mean that no deliberate leaning-on-the-lever is required to execute any of the downshifts and the upshifts also happen promptly. But I'll now go out to the workshop and spend a few hours measuring what's going on - not something I really want to do now that nobody's paying me for it!

By the way, is there any authoritative source of information on the internet that maps a path through the labyrinth behind an enigma that is Campagnolo's shift evolution? None of my left-hand Ergos index the front shift - just have a load of little clicks one can adjust with the cable tension until the last click precisely aligns with big ring. This I count a good thing because I only run doubles now, sub-compact ones, and want complete freedom to use whatever FD does the job. Do they still make 'em like that? And in any event, which left-hand Ergos do and do-not have that freedom from the straitjacket of front-indexing?
Chris Juden
One lady owner, never raced or jumped.

Brucey
Posts: 39824
Joined: 4 Jan 2012, 6:25pm

Re: Shimergo - latest news

Postby Brucey » 10 Jun 2020, 12:01pm

MartinC wrote:This is so nearly a really good result...


yes. I thought 'this is so close maybe I can make it work perfectly by varying the way the cable is mounted in the pinch bolt'....? I spent quite a long time working on this approach but I didn't get anywhere with it; in order to have been worthwhile it would have to have made a difference at one end of the stroke only, more or less.

Regarding the longer-term availability of this kind of 10s stuff; I anticipate that Shimano will continue to use this shift ratio in lower 10s groups as they 'trickle down', so when Tiagra goes to 11s you can expect 10s Sora using this shift ratio, and so on.

Shimano make 10s 4700 flat bar shifters, and GRX hydro STIs using this shift ratio. I suspect that adding a cable sleeve to a Dura-Ace 11s bar-end shifter may make it work perfectly on 4700/GRX 10s. Past that there are the semi-mythical Microshift shifters and shiftmates using 11s shifters.

Looking on the bright side a shiftmate using an 11s shifter won't have much to do and (because the 'slope' of the shift ratio will be as intended) the system ought to work perfectly this way. There are also 10-from-11 cassettes and so forth, which have attractive features; for one thing the freehub body can be shorter, giving less of the evil wheel dish. For second the inevitable 11t sprocket can be ditched with no real loss too.

So yes it isn't ideal but ah, 'twas ever thus, wasn't it...?

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

Brucey
Posts: 39824
Joined: 4 Jan 2012, 6:25pm

Re: Shimergo - latest news

Postby Brucey » 10 Jun 2020, 12:54pm

CJ wrote:
I'm using a Shimano 10-speed mech. Maybe the shift ratio isn't quite the same as their 11-speed mechs.


I wondered this too, but many times I have seen a tiagra 4700 bike repaired using a 105 11s mech; it is an exact match. The geometry in the parallelogram is identical between RD-5800 (11s) and RD-4700 (10s); I think they may even use the same raw castings for the parallelogram parts. ISTR that in the GRX groupset (which exists in 10s and 11s forms) you can mix and match RDs willy-nilly. So they are the same as near as anyone (including shimano) can judge. I'll be interested to see what you find; it is possible that my ergos were worn or something, but my experience with this setup did tally with Andy A's as far as I can tell.

By the way, is there any authoritative source of information on the internet that maps a path through the labyrinth behind an enigma that is Campagnolo's shift evolution? None of my left-hand Ergos index the front shift - just have a load of little clicks one can adjust with the cable tension until the last click precisely aligns with big ring. This I count a good thing because I only run doubles now, sub-compact ones, and want complete freedom to use whatever FD does the job. Do they still make 'em like that? And in any event, which left-hand Ergos do and do-not have that freedom from the straitjacket of front-indexing?


They don't make 'em like that any more and have not done for some time. I have not got to the bottom of it by any means but IIRC they first introduced a thing called 'quickshift' which was close to being indexing for the FD in that it had fewer (unevenly spaced?) clicks, and it would only work correctly with a new generation of FDs which used a slightly longer cable pull per FD movement. Campag helpfully made a 'widget' which could convert old FDs to be compatible with new shifters. Following this they have variously declared that not all LH shifters will work a triple, and had various numbers of clicks in the LH shifter. I have rather lost track of it, and I have lost enthusiasm for it too; for the last few years the campag powershift 10s ergos have used internals which only work at all with lower cable tensions and even then they don't last long. Campag no longer sell small parts to repair them either; you have to buy a whole new mechanism. So posh campag ergos are still durable (I think) but they are expensive and similarly unrepairable too. There are isolated (and surprised) reports of 'I got XYZ to work' but I don't think anyone has made a really good stab at deconvoluting the combinations that will and won't work. Well, if they have, they have not published it that I have seen. In many respects the incentives for using campag ergos are now gone; they are not the only shifters using underbartape routing, they are not as versatile as they once were, and nor are they as repairable.

In the meantime shimano FDs for road doubles (11s and 4700/GRX doubles) now use a longer cable pull. It is comparably long to shimano MTB FD cable pull and again there are isolated reports of road STIs making MTB double FDs work (possibly not with the trim clicks doing exactly what they are meant to). However shimano's first go at the new pull ratio for FDs was not all sweetness and light; FD-5800 , FD-4700 etc use a 'long arm' which is almost vertical when the mech is positioned over the inside chainring, at which point the shift ratio is changing very quickly with small changes in chainline/FD lateral position and cable tensions required to start any movement at all can get very high. Even though they allowed for two cable routings to compensate for such changes in installation, it wasn't enough. If the cable run was funny under the BB and/or the carbon frameset used a slightly fatter seat tube, often the FD wouldn't work at all well. So the current generation of 'road' 11s FDs (currently to FD-R7000) use a pulley and an extra (adjustable) linkage in the FD to make the FD shift ratio work reproducibly despite variations in installation and cable run. The whole lot looks as if it will promptly stop working as soon as much mud infiltrates it, of course.

In the meantime Tiagra 10x3 appears to use FDs which are the same shift ratio as older shimano road triples; good for backwards compatibility I suppose but it means there is essentially no development effort going into these systems either. I suspect that 'any FD' could be used with current 10/11s shimano double STIs provided you are prepared to fiddle about and fit an extension to the FD arm so as to make the shift ratio lower in the right way.

So for tourists I expect current shimano STIs to work (eg with a shiftmate) on N-1 and N-2 systems (because the slope in the shift ratio will be a close-match) and for the whole lot to be workable as an Alpine (super-compact) double, provided the right chainset is used. Poor relations once again, mainly because the world's biggest equipment manufacturer either doesn't bother with those who might tour with dropped bars, or doesn't understand what gears they might need when they do.

At present you can cobble together a GRX-based 2x10 transmission with ST-4700 (or 2x11 with other road STIs) and that is as close to a 'normal' dropped bar touring groupset as you can get, from the big S anyway.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

User avatar
Mick F
Spambuster
Posts: 49437
Joined: 7 Jan 2007, 11:24am
Location: Tamar Valley, Cornwall

Re: Shimergo - latest news

Postby Mick F » 10 Jun 2020, 2:05pm

Brucey wrote:from memory it is either the 17 or 19. On the parts diagram it is a unique sprocket that is only used in that cassette. IIRC the shouldering is slight; just a small (offset) variation in sprocket thickness between the centre and the part where the teeth are, I think.

Looking very closely at the 19t and the 17t ................

The rear faces are identical with two teeth narrower for the indexing.
The front faces, however, 19t has two gates and 17t one.

21t is identical to 19t.

Thicknesses of 21t, 19t and 17t are all 1.62mm and flat as well. No discernible differences at all ......... other than the tooth count.
Mick F. Cornwall

Brucey
Posts: 39824
Joined: 4 Jan 2012, 6:25pm

Re: Shimergo - latest news

Postby Brucey » 10 Jun 2020, 3:04pm

the other reason long pulls are employed (as seen in shimano 9s *) is that when the number of gates on each sprocket transitions from one to two, there is usually a slightly more baulky/iffy shift. Using a slightly longer than average cable pull in that shift 'helps it', so all campag 10s shifters I have measured have one long pull in the middle of the sequence. I suppose that it is possible that campag use the longer cable pull in the middle of the 10s pattern in different ways depending on the cassette; either to match a cassette with actual uneven spacing or to better manage the iffy shift, which could be viewed as a 'virtual' uneven sprocket spacing, even if it might not be an 'actual' uneven sprocket spacing in every case. Possibly (and I'm speculating here) shimano's patents prevented them from using evenly spaced cassettes for some years....?

(*) For many years I wondered why shimano didn't (in their compatibility tables) seem to officially sanction the use of 'road' 9s STIs when combined with 9s MTB RDs and cassettes. I finally realised that 9s MTB and 9s 'road' shifters had the same average shift ratio (of course) but were actually slightly different from one another; there is one slightly longer cable pull in the sequence but it comes in a different place with each shifter type, to exactly match the intended cassettes; shimano 9s cassettes are uniformly spaced but have an 'iffy shift' in a different place depending on whether it is a wide ratio MTB cassette or a 'road' cassette with smaller sprockets. I forget the details but it might be something like the 'road' cassette iffy shift is between 4 and 5 and the MTB iffy shift is between 5 and 6.

in practice if you mismatch the parts (as so many touring setups do with 9s road shifters and 9s MTB cassette) it all seems to work perfectly when the parts are new. However it is another 'tolerance eater' and once you know where to look you will usually find an unusually prompt shift immediately adjacent to one that is a little hesitant. When everything gets a bit knackered, this is often where the rot sets in first.

Shimano 10s/11s shifting is different; the gates line up on adjacent sprockets differently and it is possible (I have not checked) that there is just a gradual progression in cable pulls to match the changing shift ratio of the RD, i.e. no anomalous 'long pull' mid sequence, and better compatibility with different sized cassettes.

Anyway when using campag 10s shifters it may be desirable to make small changes in the mid-sequence sprocket spacing of a shimano cassette to better match the cable pulls, if you want to optimise the system shift-wise. However this is a tertiary issue, behind

1) average shift ratio
2) the shift ratio 'slope'

In the case of RD-4700 with campag 10s ergo 1) is pretty close but not ideal and 2) seems problematic in the larger sprockets.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

User avatar
CJ
Posts: 3051
Joined: 15 Jan 2007, 9:55pm

Re: Shimergo - latest news

Postby CJ » 11 Jun 2020, 12:12am

I've spent the afternoon measuring, put the numbers in my old Indexing spreadsheet and produced this (9th is the zero point):

ShimergoGRX400.gif
ShimergoGRX400.gif (10.72 KiB) Viewed 290 times

Tweaking the zero position to minimise shifting errors results in a maximum alignment error on any sprocket of 1mm, which does not seem to be any problem. The overshift to bottom is not a problem because you can set the stop to, to stop it. All that is assuming I've measured everything spot on. The cable pulls I'm confident of to 0.1mm, the lateral shift not so easy to measure, maybe 0.5mm. So some of those 1mm errors could be half that. And I think gears 2 and 9 probably are off a bit, because there's a good linear trend of nicely lined-up points from 3 to 8 that looks a lot more like how the shifting behaves in practice. And I haven't actually measured the cassette, just assumed its a uniform 3.95mm pitch. Maybe it isn't - in a good way. Whatever: it shifts well, and I'm well happy.
Chris Juden
One lady owner, never raced or jumped.

tim-b
Posts: 1431
Joined: 10 Oct 2009, 8:20am

Re: Shimergo - latest news

Postby tim-b » 11 Jun 2020, 5:45am

Hi
For those of you that like to tinker GRX is the way to go. I've had a GRX-equipped bike for a couple of weeks and it's got a mix of 400 calipers, 600 levers, 800 mechs and a 600 chainset with an FSA BB!
I understand that Shimano made two 800 rear mechs, a 1x specific and a 2x specific, plus the mechanical and Di2 versions so plenty to play with
Regards
tim-b
~~~~¯\(ツ)/¯~~~~

Brucey
Posts: 39824
Joined: 4 Jan 2012, 6:25pm

Re: Shimergo - latest news

Postby Brucey » 11 Jun 2020, 8:12am

If I read Chris's measurements correctly they are pretty close to the ones I made. If CJ says it shifts OK in his setup it must do but I have to say I am surprised; IME ~0.3 to 0.5mm errors in RD position noticeably affect shifting performance in most systems and ~1mm errors are usually not at all well tolerated; this is like having a system that is perfectly matched and adjusted and then expecting it still to work when the barrel adjuster is moved three or four clicks. In a 10s system, a 1mm RD position error is over 25% of a shift.

cheers
-------------------------------------------------------------------------------
~~~~~~~~~~~~~~~~~~~~~~Brucey~~~~~~~~~~~~~~~~~~~~~~~~

sjs
Posts: 818
Joined: 24 Jan 2010, 10:08pm
Location: Hitchin

Re: Shimergo - latest news

Postby sjs » 11 Jun 2020, 9:16am

I have a mid-90s road bike which still has the original Campag 8-speed ergos and RD. Shimano and Campag 8-speed only differ by 0.2mm per rear shift, but I could not get it to work well with a Shimano wheel/cassette, even though the max error if set up for the centre of the RD travel should be only 0.5mm. Easy to fix by shimming the cassette though.

PaulS
Posts: 46
Joined: 26 Jan 2012, 6:45am
Location: East Yorkshire

Re: Shimergo - latest news

Postby PaulS » 11 Jun 2020, 10:13am

Here are my numbers from a (disappointing) play with Shimergo. I fitted new Campagnolo Potenza 11-speed levers to an old 9-speed Shimano Ultegra rear derailleur & cassette. The Shimergo table suggested that it should be close enough, but it doesn't really work. I can get about 5 sprockets to engage reliably. The others either change 2 sprockets at a time, or they jump under power.

I measured the cable pull on the lever, going both up & down the gears. Measured with a Vernier from the cable stop on the downtube to a clamp I put on the cable (rear derailleur is disengaged). Looks like there is about half a mm of positional error from the lever (I don't think that is all my measurement error), i.e. the difference in cable length going up the gears compared with going down. The Shimano mech needs a 2.5mm cable pull according to the Shimergo table. Looks like the Potenza levers pull rather more than that.

I also think the Potenza levers use a different cable pull to other Campagnolo 11-speed shifters. This is the end of my Shimergo experiment, but I'll put the numbers here in case anyone else is interested. I only took the measurements once, so take them (and especially the precision) with a pinch of salt.
Image Attachments
cable_pull.JPG

User avatar
CJ
Posts: 3051
Joined: 15 Jan 2007, 9:55pm

Re: Shimergo - latest news

Postby CJ » 11 Jun 2020, 10:46am

Brucey wrote:If I read Chris's measurements correctly they are pretty close to the ones I made. If CJ says it shifts OK in his setup it must do but I have to say I am surprised; IME ~0.3 to 0.5mm errors in RD position noticeably affect shifting performance in most systems and ~1mm errors are usually not at all well tolerated; this is like having a system that is perfectly matched and adjusted and then expecting it still to work when the barrel adjuster is moved three or four clicks. In a 10s system, a 1mm RD position error is over 25% of a shift.

I think the probable reason it works so well is the way Ergopower shifts. It was very noticeable during my measurements that there's a large over-travel on the downshift - much larger than with Shimano shifting. With this , the mech goes a good 2mm further before the shifter clicks – and then settles back. The settling back may leave it 1mm out, but the 2mm overshift has already carried the chain onto the sprocket – and apparently there are enough chain links between the guide pulley and the sprocket to accommodate the odd millimetre of residual misalignment without complaint.

The best source of Campagnolo old parts identification I've found is Disraeli Gears by Mike Sweatman (whom I remember from the Cambridge University FoE group, small world isn't it). According to this page, what I have must be the Veloce 'Ultrashift' version introduced in 2009. The 'Escape' moniker of the other version seems to indicate front indexing (like the 'escapement' mechanism in a clock I guess), which in my case I do not want.
Chris Juden
One lady owner, never raced or jumped.

NickJP
Posts: 260
Joined: 24 Sep 2018, 7:11pm

Re: Shimergo - latest news

Postby NickJP » 11 Jun 2020, 11:56am

PaulS wrote:Here are my numbers from a (disappointing) play with Shimergo. I fitted new Campagnolo Potenza 11-speed levers to an old 9-speed Shimano Ultegra rear derailleur & cassette. The Shimergo table suggested that it should be close enough, but it doesn't really work. I can get about 5 sprockets to engage reliably. The others either change 2 sprockets at a time, or they jump under power.

Campagnolo changed the cable pull of their 11-speed shifters some time around 2015, which was when they introduced the Potenza components, so I would say that your shifters use the newer cable pull that is different from that which was used to construct the Shimergo table. I have a pair of Chorus 11-speed Ultra-shift Ergos purchased in 2011 that are shifting a Shimano 9-speed MTB cassette and Shimano 9-speed MTB rear derailleur, and the shifting works perfectly well - no hesitancy or grinding or refusal to shift when changing up or down to any sprocket on the cassette.

You can distinguish the 2015+ components as they have an A in a square moulded or printed on them somewhere, like so:

Image

PaulS
Posts: 46
Joined: 26 Jan 2012, 6:45am
Location: East Yorkshire

Re: Shimergo - latest news

Postby PaulS » 11 Jun 2020, 12:12pm

Thanks Nick. And my shifters have a "B". There's the difference.
Image Attachments
P1090636.jpg