bluehills3149

Following on from the excellent post "The Definitive (But Wildly Inaccurate) Guide to Brake Cable Pull Ratios" I thought I'd do the same for gear cable but as gears are indexed (usually) it is a bit easier as the vagaries of lever size & throw are eliminated.
I started this as I’ve been putting together a 10 spd Campy shifter-Shimano geared bike (and I think I’ve come up with a simple, cheap fool-proof solution and I’ll post the info soon) so I needed to get a handle on the various shifters and how much they pull the cable.
First, there seems to a lot of misinformation and rumor about gear cable pull, not the least is the variation on the “campy cable pull is not linear but Shimano is..” see for example this (otherwise great) post
Art's Cyclery Blog » Science Behind the Magic | Drivetrain Compatibility

where it says “Campagnolo Ergo levers pull 2.5mm of cable five times, 3mm twice and 3.5mm twice” and this data seems to have been accepted as gospel and has spread far and wide.
But this bothered me. Why, if the derailleur is sweeping in a circular arc would it not be a more constant increase? And why would only Campy use non-linear movement and not Shimano, Sram et al? Surely their derailleurs all use similar geometry and would require non-linear pull or else they would end up moving the derailleur less and less with each shift.
So I measured my only Shimano shifter I have – a 9spd Durace downtube which would match any Shim 9spd shifter and if it holds for 9 spd, it would most certainly for 10 and 11 too. This I compared to my Campy 10 spd shifter. I measured Low gear (biggest cog) to high (smallest cog) last and did so by wrapping some tape around the cable and using digital calipers to measure the gap, resetting the caliper to zero after each shift.

The results are graphed below.

* Note: The first and last gear change is probably more influenced by the limit stops so can probably be ignored.

As you can see there is a clear downward trend for both levers. And, for the Campy, the average of shifts 1-9 is 2.81mm which closely matches the 2.8 given by Campagnolo. For the Shimano it’s 2.46 which is close to the advertised 2.5. This indicates my measurements are within reason. I suspect the reason the values jump around is more due to manufacturing intolerance as the detent springs sit in very precise indents and slight imperfections here could cause imprecise movement. My Campy lever was also brand new vs the Shimano which was older and on my beater bike so grit may also be an cause of its more random swings.

If anyone else has Sram or 11 spd, it would be great to see how they pull but I suspect it will be similar.

3alarmer 

...there's some stuff that might be of interest to you Here by John Allen on Dura Ace, the other Shimano stuff, and alternate cable routing to adapt one to the other.

bluehills3149

Yes- I've seen all the various sites regarding alternate cable routing to try and get Campy and Shimano to get along and the other kludge fixes. Tried them and they all suck.

DaveSSS 

Your post tells me that you've never taken one of the levers apart and don't understand how it works. Ratchet teeth can be easily be machined to produce varying amounts of cable pull. FWIW, I'm the mechanical engineer and machinist who very carefully measured these pulls on my personal bikes. It's not hard to do, using a precision machinist's rule. Wrap a piece of tape around the cable, a short distance ahead of the rear cable housing stop and measure each pull several times. If you don't get very similar readings each time, then you're technique is poor. Use a rule marked with .010 inch or similar metric increments, not a fractional rule.

Here's a picture of a very early model Campy ultrashift lever that I took apart in November of 2008. The owner of this lever was a pro racer who wanted to see if I could modify the lever to produce a heavier click to each shift.

The 6th part from the left has the detents that the small spring loaded balls in the 7th part fit into. The detents can be positioned to pull any amount of cable desired. You can also convert a 10 speed ultrashift lever to 11 speed by changing out only those two parts. I did that long ago, when I had a wreck damaged 11 speed lever with the parts to swap into a 10 speed lever.

When 11 speed came out, I changed all of my bikes to 11 speed. The shifters have the same design, just with one more detent and slightly smaller cable pulls. I'm still using a bike with a Campy 10 RD, slightly modified so it shifts properly with an 11 speed shifter. Without the mod, the RD does not have enough travel. All I did was reduce the diameter of the cable clamp bolt, right where the cable rests, to effectively shorten the lever arm length and create more travel.



Here's a repeat of something I wrote up in October of 2009:

I hooked up my 2003 Record RD to the 11 speed shifters and found that it undershifts. After all 10 shifts, the upper pulley was a little more than halfway between the cogs 10 and 11, which isn't very good alignment. I made sure that the RD travel was not restricted by the low limit screw. Setting the high limit screw a little too tight helps, but not a lot. I'd plan on buying that new 11 speed RD, unless J-tek comes up with a shiftmate to cure the problem.

I just took several sets of cable pull readings and came up with a total pull of 1.03 inch or 26.2mm. That's only a little more than the 1.02 inch or 25.9mm that I measured for 10 speed, some years ago and verified again today. The cable pulls are not all the same, just as before. I got uniform pulls of 2.5mm for the first two shifts, then only 2.2mm for the next three. The old system was consistently 2.5mm for the first five shifts. Shifts 6-9 were consistent at 2.9mm and the last was larger at 3.3mm, not restricted by the limit screw. The total of all the amounts above isn't exactly 25.9mm, but the individual shifts are rounded to one decimal place, so there's a little roundoff error. The total pull was quite repeatable. In round numbers, I get an average of 2.9mm per shift for 10 speed and 2.6mm for 11 speed, but none of the shifts is really average.

Here's a link to my very old post about the Campy 10 RD mod. I posted as C-40 on Road Bike Review, for a long time - over 10,000 posts.

https://forums.roadbikereview.com/com...in-160601.html

elladaddy

Blue Hills, your posts on cable guide have provided lots of viewing pleasure and quite a bit of practical benefit.

I just wanted to post a followup of some work I did based on your valuable measurements and graphs.

I'm running Campy Veloce 10s ergopower shifters. I chose Campy because of the MA (inspired by your brake lever work) and because I thought they looked better on my vintage Motobecane Grand Record.

However, like many others, I want to run them with a Shimano rear derailleur to get the gearing I want. The shifter/mech mix should shift a 9s cassette

However, even with with Hubbub cable routing I found shifting less than adequate. Using the work you did, I respaced my cassette to closer match what the shifters were doing. Where the Campy shifters pulled more cable, I added spacers. On the lower gears, where the Campy pulls proportionately less cable than Shimano, I reduced the spacing. The amount of changed was based on your graph of cable pull. Here's what I came up with:

1-2: 1.89mm The 10/11 free hub adapter spacer is 1.85mm. But it didn't come from Amazon in time, so I used a tiny 9s cog.
2-3: 2.29mm For this I used a single 10s cassette spacer.
3-4: 2.79mm For this I used a tiny 9s cog and a 1mm spacer readily available as part of 10/11s adaptation kits.
4-8: 2.54mm Standard 9s spacer
8-9: 2.64mm Standard 9s + 1mm spacer is perfect, but didn't have another 1mm, so I just used a standard 9s outer gear.

I used 9s cogs to maintain proper C-t-C pitch.

It shifts much better! Up and down between 2-3 is still the most balky. Perhaps adding the 1mm at the end will alleviate an accumulation of error. As will adding the 1.85mm spacer. Finally, my cogs are mixed from different cassettes, so tooth alignment isn't ideal.

I may get a standard 12-36 cassette and grind off the rivets, but I want a custom cog collection, 13-34 "mega-range", so I'll probably keep my cassette.

Maybe next time I'll have DaveSSS machine my shifters

Andrew R Stewart 

Two unrelated comments- First is that SunTour was know to have different cable pulls per shift across their early indexed systems. IIRC their goal was even cable tensions between the shifts, or something like that. See what that got them WRT market share... Second is that i'd be interested in how Shift Mates adapters relate to all this. I've used a number of them on personal bikes, including the hard to get right even with all matched stuff. tandems. Andy

HillRider

Have you used Shiftmate adapters? As Andy noted and my personal experience confirms, they work exceeding well to make Campy and Shimano get along. I've used two different models to match up:

A Shimano 9-speed derailleur and cassette with a Campy 10-speed brifter
A Shimano 10-speed derailleur and cassette with a Campy 10-speed brifter
A Campy 10-speed derailleur and cassette with a Shimano 10-speed brifter

All of these systems shifted perfectly.

Remember Shiftmates only change the input/output cable pull ratio by fixed amounts. They have no effect on the linearity or non-linearity of the shift mechanism itself.

bluehills3149

Hillrider: I have not tried shiftmate and I understand they work perfectly but I wanted to avoid having buying something else and want a simple solution.

Elladaddy - I followed up this post with this :
Campagnolo shifters & Shimano mech – easy, cheap solution

I found if you play aroud with cassette spacing the shifting is sub-optimal as the chain width is designed for a specific cog spacing and increasing it will degrade shift quality.

The post above shows how to modify the cable pulley ( shown as the grey part 2nd to left in DaveSSS's post above) which is quite easy to access inside a campy lever. By shaving it down 0.5mm mm off its radius so it pulls less cable per shift it will mate with near perfect presicion to a Shimano 11 rear der and Shimano 10 spd cassette.

I run this setup on my Trek Emonda with Campy Veloce Ultrashift 10 pd levers and Durace 11spd rear der and 10 spd cassette to give a groupo that's cheap, reliable and as light as Durace.

Recently, I did a similar mod to some Veloce 9spd levers (ultra-shift) and ground down the cable wrap-pulley so it too pulls less cable per shift and mated it to Shimano 11spd rear der and Shimano 9sd cassette. It shifts really smooth and nice - like it's factory. However these are older style levers and to remove the cable wrap pulley requires a compete tear-down of the levers and the part is hardened metal so it is harder to file down rather than the plastic pulley on later gen levers. But with Shimano 9spd cassettes at $25 and a Tiagra der at $23 I don't have to pay the Campagnolo "tax" anymore so it may be worth doing if you have some old Campy 9spd levers lying around you are dying to re-use.

And just to follow up on DaveSSS's post, he (I assume he) correctly points out HOW campy makes the lever pull different cable amounts per shift - the jagged saw-tooth ratchet (shown in his photo) has different spacing between each indent (and if you look carefully at his pic you can see this) so it pulls more cable with each shift. I stand by my original assertion that the cable pull increases with each shift so it matches the arc of a rear derailleur which is non-linear -and this is true for all brands of derailleurs.

Unfortunately I hoped the original post would spark some more comments from others who have a curiosity about the geometry of deraillleurs (which are essentilly 4-bar linkages) as there is a lot of mis-information out there that needs clarifying.

elladaddy

Thanks for the reply Blue.

I've found some chains can be forgiving. I ran a KMC 9 and Campagnolo 9 on a Shimano 8, the former giving the slightest degradation, maybe one balky shift/ride the latter with no real problem. My current project, KMC 10 on S9 isn't as reliable, but that's my campy/shimano mix so many more variables. Also, the 9-10 difference is bigger than the 8-9 difference. I'd assume downshifts would be where narrower chains would suffer, and my worst shift is actually my 2-3, which has narrower than stock spacing. And while that custom spacing, 2.29mm, is only .25mm narrower than stock S9, and my KMC10 chain is a whopping .72mm narrower than the intended S9 chain, the upper gears shift fine with 2.54mm spacing.

Perhaps it's my wariness of mauling (I tend to be careless) the insides of an expensive Campy shifter, but I feel like arithmetic can still carry the day. I did the above respacing with the assumption that as you said derailleur ratio is "non-linear ... for all brands of derailleurs". It must be if cable pull varies and spacing is constant (in the Shimano case).

Your Tiagra 1.4 find was golden. I've read that site about 62000 times and still missed that. They kind bury it but that's no excuse. My concern remains that the Campy and Shimano non-linearity don't match, as your cable pull graph clearly shows, so custom spacing would still be theoretically needed.

Given that it's a simple calculation (which I haven't done cause I'm too busy talking about it) to figure out the Shimano derailleur ratio graph from your valuable cable pull measurements. I half-assed this: inferring the non-linearity differences from the difference between the Shimano and Campagnolo cable pulls. To do it right you'd use the graph/table of the derailleur ratio combined with the shifter cable pulls.

Again, this is a theortical ideal. It sounds like you've mated a 4700 to Veloces and it sings. That's really the end of the story.

My other project, Campy 9s Barcons (pre 2001), XT derailleur and S7 cassette isn't faring so well. It worked great with a newer Racing T derailleur (which broke) and S8. Maybe this is a 4700 candidate.

bluehills3149

Actually, after I posted that I realized you do not need a Tiagra rear der but ANY 11 spd Shimano der is OK. I was thrown because (new) Tiagra is 10 spd so I initially assumed they had a special ratio rear der whereas in fact it is the shifter which has a special pull ratio unique to Tiagra and the der is geometrically identical to all the Shimano 11spd models. However, this setup will result in a very slight overshift (ie the der will move slightly too much with each shift) - normally this is not too bad as there is a bit of slack built into the designs but you need to keep adjusting them more often. Shaving the cable pulley a smidge reduces its diameter and thus the amount of cable pulled each shift and makes perfect..
And note if you don't want to butcher your precious campy lever, those models that used a pulley like in the pic above is part number EC-SR060 and they may still be available.

DaveSSS 

I decided to measure the cable pull on my very early model Campy11 speed shifters. To do that, I use a precise machinst's rule ( preferably Starret brand, not some cheap brand). Either an English or metric rule can be used, but the metric versions only have .5mm between marks, while an English model has .010 inch increments, that are significantly finer. With the wheel off the bike and the chain off the chainrings, place a piece of masking tape around the exposed cable, so it starts at some major increment, like 1 inch or 20mm, with one end of the rule resting on the cable housing stop. Go thorugh all 10 shifts and write down the position of the tape, after every shift. Repeat several times to be sure that the results are repeatable. What I found is just what I expected - the cable pulls do NOT steadily increase starting with the smallest cog shifts and progressing to the largest cogs. What I found is not much different than my previous measurements of a 10 speed system, 9 years ago. I got a consistent 3mm pull on the first shift, but that shift can be affected by the limit screw - you have to start somewhere. After that, the next six shifts were consistently 2.5mm, so the cable was at the 18mm mark at the end of those shifts and 20.5mm after the next. The pull increased to 3.5mm for the last two shifts, and the limit screw was not limiting the cable travel. The total travel was 27mm.

I also know that the 2015+ shifters pull more cable than the older models, but I have to measure the 2018 shifters that I have to find out how much. Unfortunately, my Colnago C-RS frame has no exposed cables, so I'll have to figure out some way to measure the pull on those shifters another day. I have a 2018 RD on my bike with the old shifters and found that the 27mm of cable pull was inadequate for the new RD. I tried the same trick that I did 9 years ago, to get a 10 speed RD to work with 11 speed. All I did was grind down the diameter of the clamp bolt threads to the root of the thread and file a little material off the cable clamp plate, to shorten the effective lever arn length on the RD and increase the travel a bit. It now shifts quite well with the old shifter. Before that change, if enough cable tension was applied to shift to the two largest cogs, the shifts to the smallest two cogs would hang up. More RD travel was needed.

DaveSSS 

I built a cable pull measuring fixture that includes a 2-inch travel dial indicator with .001-inch graduations. The shift cable is unclamped from the RD and connected directly to a spring tensioned slide on the fixture. A cable tension adjuster is included, to be sure that there is no slack in the cable and any movement of the finger lever starts the dial indicator moving. This device allows very small differences in each cable pull to be measured. One thing that it proved is that the cable pulls do not increase with each shift, as the RD travels from smaller to larger cogs. Both the original model and the new model include a large first cable pull and both have a significant amount of overtravel on the first shift, that returns after the finger lever is released. Both have some excess travel available on the last cable pull that shifts to the largest cog. That excess pull returns, as soon as the finger lever is released.

I made at least five runs through all 10 shifts on both models, to determine repeatability and get the best average value for each shift. Most shifts repeated with a total deviation of no more than .004 inch or .1mm. For reference, a piece of notebook paper measures about .003 inch thick.

Total travel was 27.3mm for the 2009 shifters and 29.8mm for the 2015+ shifters. That’s a difference of 2.5mm. If measurements are taken from the exposed cable on a bike, using a machinist’s rule, the first shift pull is likely to be less, since the limit screw affects the total pull and so does the initial cable tension. My measurements all began from a zero point, with the slide resting on a fixed stop. Below are the cable pulls, in millimeters.

2009 ultrashift 2015+ ultrashift
3.35 -------------- 4.0
2.5 ---------------- 2.8
2.5 ---------------- 2.5
2.2 ---------------- 2.7
2.2 ---------------- 2.7
2.6 ---------------- 2.7
2.6 ---------------- 2.7
2.9 ---------------- 3.0
2.9 ---------------- 3.4
3.5 ---------------- 3.4

elladaddy

Yes I just found out that all 4700 and S11 are the same. The 4700GS seems to have the biggest max rear cog so that's my top choice still. I'm discouraged to read about the over shift. I'll bet part of that problem is the variable cable pull. Campy has a big jump (between cogs 2 and 3) and having to adjust for that seems like it could cause over shifting elsewhere.

Thanks for the pulley PN! I haven't found one yet, but I'll keep checking.

elladaddy

Dave,

Have you happened to have measured S11? I'm still looking at comparing individual RD shift ratios with individual cable pulls to derive individual cog spacings in Campy-Shimano combos.

DaveSSS 

I only own Campy, so that's the only shifters I've measured.

fietsbob

Just 2? not that definitive .. spread sheet all of them..

DaveSSS 

All you need to do that is a precision measuring device like I have and access to every type of shifter made for more than 20 years and lots of time. My measurements of Campy 10 and 11 shifters, made many years ago were as accurate as I could do with a precision rule, but repeating the same test with a much more accurate measuring device showed some small errors.

The OP had preconceptions of the expected results and set out to prove them. I did my tests with no idea of what the results would be and made many repeated runs through all shifts and took the average of all runs.

There are many lists of shifter cable pulls out there, but most are probably not very accurate and many only list the average cable pull. The average means little.

I can remember claims of some SRAM shifters that supposedly had the same 3mm cable pull for all shifts, working with some Campy drivetrain. By my tests, the RD position would be off by a full cog after only 4 shifts. There's just very little reliable info info out there. A company like j-tek probably knows more than anybody.

I know how to make a Campy 10 RD or 2015+ RD work with pre-2015 campy 11 levers.

fietsbob

Go for it Spread sheet makers ... you can compile all that in one place..

Old 6 speed freewheels were wider spaced, and an old Campag RD moves further for less cable pull, than the newer, slant panto type ...

Geekage

Who's keeping the spreadsheet? Here are the measurements I made:

Shimano 11-speed road cable pull, as measured off a 5800-series 105 shifter:



(how do you do a monospace font? or other advanced formatting?)
This is assuming that the cog position starts from smallest to largest, so opposite from conventional "gear" where 1st gear would be the lowest gear, thus the largest cog.

Also note that my setup wasn't all that great, I measured with a digital caliper from the downtube cable stop to a piece clamped to the shift cable. The problem is that the cable stop wasn't very flat. Measurements could vary by +/-0.2mm. I'm not sure if that was my measuring technique, or just the shifter variability. I also only took a set of six measurements, shifting up and down across the range three times.


Measuring with a caliper. Yes, this is the bike I just added the disc fork to, mentioned in my other recent thread in this forum.

Here are my measurements for a 10-speed SRAM S-700 shifter. This is the Exact Actuation technology, so should be the same for 10-speed mountain or road shifters and work with 10-speed (and below) mountain or 11-speed (and below) road derailleurs.
And Shimano 9-speed Deore shifters, so also works for road and mountain 9-speed:
I've since come up with a method of measuring the shift cable travel on a bike with exposed downtube cables that doesn't require good caliper technique and can be done with no special equipment and without modifying the bike. So pretty much anyone can do it.

Tape a ruler to the downtube and a vernier scale to the shift cable, and you only need to shift, then read out the dimension on the scale. The challenge is reading the scales. I've designed scales that can be printed for measuring to 0.02mm and 0.05mm accuracy.


Vernier measurement

It does make things easier if you can mount the bike on a stand and remove the rear wheel, that way you can just shift without having to turn the cranks. Backing out the adjustment screws on the derailleur also helps give better readings for the top and bottom shifts, but they're not so important.

I left the ruler pretty long so it's easier to verify that your printer is rendering at pretty close to a 1:1 scale.

Kimmo 

Haha, nice one.

Just to complicate matters...

ian3d

I know how to make a Campy 10 RD or 2015+ RD work with pre-2015 campy 11 levers.[/QUOTE]

I would love to know how to make a pre 2015 11 spd chorus lever work with a post 2015 Chorus RD and FD. Thought I had cable friction issues until I recently discovered there was a pull ratio difference after 2015. Would be great to not have to buy a new set of levers.

DaveSSS 

The RD movement can be increased by grinding the cable clamp bolt down to the root of the thread, in the area where the cable contacts the bolt. That shortens the effective lever arm length and increases the travel. I'd probably buy a couple of new M5 stainless steel bolts from Ace Hardware to experiment with. I used a dremel cutoff disc to do the grinding. The FD is an entirely different issue. The early model left shifter has plenty of cable travel, but there are 6 or 7 clicks, so it would even operate a triple FD. The 2015 FD uses 2 or 3 clicks to shift to the little ring and has a separate click to trim the cage to the left a little, only needed if you're using the two largest sprockets. With the early model shifter, you have to figure out how many clicks will make the shift to the little ring, without going too far and dropping the chain. One more click might work for trim. If not, you just have to set the low limit screw to barely clear the chain in the little ring and largest sprocket, then hope it won't drop the chain when making the shift. A chain keeper is a wise thing to have.

ian3d

Thank you DaveSSS for your reply and information. I’ll give it a go. It will be great if I can avoid buying a new set of levers. Cheers.

70sSanO

I've decided to continue this thread since there is so much good info. And any thread that includes the word "definitive" in the title should be just that.

If the 11 speed road (10 spd 4700) actuation ratio is in fact 1.4, then the cable pull for Tiagra ST-4700 should equate to an average of 2.82mm. This means that a Tiagra ST-4700 can be used with an older 1.7 actuation RD to work with an 8 speed system spaced at 4.8mm. If this is correct, it is an upgrade over running ST-R2000/2400 Claris.

My questions have to do with the front derailleur. There are all sorts of posts that say you can't run road and mountain FD's due to differences in cable pull which I understand. And some that say you can fudge it, with a double, which also seems logical.

But I have never seen any actuation or shifter cable pull numbers for "index" shifters for a front derailleur. I have read that the front derailleur cable pull is greater with the newer 10/11 speed shifters.

Does anyone have these numbers?

As time goes on I will be moving on from my downtube shifters and going with Ergo or STI. I am running a 3x8 with XTR RD-M910 and FD-M900. The 10 spd Ergo's would be great, but my left thumb has enough arthritis that I will need to go with STI's. I would like to stick with the derailleur setup as it gives me a really wide range with Dura Ace performance.

But it would help if front derailleur shifter cable pull numbers existed to see if the longer Tiagra pull will work.

It is not life and death. I do have an FD-7703, and I could probably run an FD-4703 with my FC-6503 crank; provided it moves the chain enough to get to the outer ring. And there is always Claris.

Thanks in advance for any suggestions.

FWIW... I routinely break open cassettes and re-space them, run alternate routing, (when needed), and I'll do mix-n-match with freehub bodies so I don't mind doing an out of the box mod.

John

carhillclimb

I have a lot of numbers on Shimano & Campagnolo shifter pull lengths, many done with Alex Wetmore's tool. They are not as precise as you might like but they can get you started. I did not have the late Campagnolo 11 speed. My Campy 12 speed is internally routed and I built it before I had the Alex Wetmore tool, so all I have is the overall pull at 31-33mm which is not very accurate. It is not easy to measure the pull at the rear derailleur. I do have the Sensah 12 speed*. If there is interest I can convert from the .xls spreadsheet.

My understanding is that Campagnolo Athena 11 rear is the same cable pull as 2014 & earlier Chorus/Record/SuperRecord, with the caveat that it does not allow more than one cable release shift at a time. Can this be confirmed?
My understanding is that Campagnolo Potenza 11 rear is the same cable pull as 2015 & later Chorus/RecordSuperRecord, with the caveat that it does not allow more than one cable release shift at a time. Can this be confirmed? Also the comment is that the Potenza rear derailleur has a weaker spring so may not shift as well with the Chorus/Record/SuperRecord (may need really slick cables and housing for example)
Can anyone comment on the cable pull for Centaur 11? There is some belief that it is not the same as the other two.

As with 70sSanO, I have a lot of non-standard configurations. Most of my bikes do not have standard rear sprocket sets. All my derailleur bikes have shimano freehubs. More than half have Campagnolo shifters. I have had good success with soda can shims and thinner spacers than standard in cases where required. I also retime the sprockets as needed, and dremel cutouts as needed.

My experience is that knowing the numbers will allow you to make a start with the setup. But there will almost always be final tuning needed with sprocket spacing or something.

My experience with front brifters has been - build it and try it. For example, if you need more or less movement change the front derailleur cable clamping and see what happens. But I have Shimano RSX front derailleurs with Shimano 9 speed brifters and a half-step+granny triple**. I have Shimano Tiagra triple shifter with 105 triple derailleur shifting 24-24-47 on a Stronglight 93, and yes I know you cannot put a 34 on a Stronglight 93. But I can. I use custom "buttons" and 4mm bolts, and while I have broken 6 or so bottom bracket spindles, including a Specialized FACT, I have never broken any of my many 4mm chainring bolt setups. This also requires dremel on the derailleur because the "modern" Shimano triple needs the shift ramps and the Stronglight 93 chainrings do not have those. I have Campagnolo 11 speed chorus/record pre-2015 shifting an $8 front derailleur 24-42-52 on Shimano cranks, on a bike with Campagnolo 11 speed rear, I have Campagnolo 11 speed front brifter with a Shimano 105 10 speed front derailleur and 26-46 TA 5 pin cranks, 11 Campagnolo rear with modified SRAM Red 10 speed derailleur. I have SRAM Red front brifter shifting a Dura Ace 10 speed front modified on the bike with the Campagnolo 12 speed rear. I have di2 with 26-46 chainrings on the Specialized FACT with Lightning 94bcd spider. I have Shimano RSX triple left brifters shifting Sturmey Archer S5 hubs (where the 2 speed side is a Campagnolo 10 speed brifter operating through a TravelAgent). Similarly an RSX brifter and travel agent on a bike that can accept any of the Sturmey Archer ASC, TF or S3X. I have Campagnolo Veloce Escape triple brifters on a bike where the inner shifter position is fixed, and the outer two positions shift the front derailleur between the 2 chainrings for freewheel - yes it has 2 chains and switches between fixed and derailleur with the front brifter. For the inspiration search for bichain fixed free. I have a Campagnolo "10 speed" front brifter shifting 10 rear sprockets with an 11 speed chain and a Shimano 9 or 10 speed 105 rear derailleur, and unequal sprocket spacing, on a bike where the left side has a trials freewheel backdriving the crank to simulate a multispeed fixed; I believe 2 shifter clicks are needed for the innermost shift. The setup right now has a downhill top forward gear, and a higher than that backdrive gear, and even the trials freewheel is not sufficient to keep the backdrive lash low enough for my liking. It was relatively OK but not great when both were around 74 inches with a Suntour rear derailleur and an Atom/Regina 5 speed freewheel.

The Campagnolo 12 speed front brifter did not seem useful, so it is sitting in the box unused.

*I thought the Sensah 12 speed would solve my winter mitten shifting, but the motion of the shifter is not as nice as either Shimano or Campagnolo. There is a LOT of lever movement for the next larger rear sprocket. The shift joy is not there, so the single tooth spacing shifts do not happen as fast as desired to make the bike as fun as the mainstream. And resting the hand on the front brake lever, something I do when the hands are really cold, causes the front to shift to the small ring, which is not desirable. And it is very hard to set it up so that when on the large rear sprocket and you attempt to shift to a larger rear sprocket the bike does not shift to the next smallest rear sprocket. I had thought this last was not possible, but the last time I rode the bike that at least seemed OK.

**Half-step is not worth it. First, half your shifts will be both front and rear. Second, you will not remember which shift you did last, so you will not know how to make your next upshift or downshift without looking, and at night you cannot look. Third, the front shifts are all the tension part of the chain which is not as good as rear shifting where the untensioned chain is moving; this is important when you are upshifting at speed, but even more important when doing the double shift downshift under power for the slight grade change. Fourth, is that there are no ramps available for the front shift, and that will be non-trivial to create on your own.