Harold74

Background

I've rebuilt a friend's road bike as a wedding present. New cable & housing, new chain and cassette, and everything cleaned and lubed that can be. The bike came to me with a rather severe RD shifting problem. As a result, I don't know what "normal" is for this bike. I'm under a bit of time pressure to get this steed back on the road for the remainder of Canada's short riding season.

This is my first time setting up an FD for triple crankset. I've set up FD's for a bunch of double ring setups in the past and have been following the guidelines from Shimano (manual), Park Tool (video), and Leonard Zinn (book).

As I understand things, the available FD index and trim positions for this system are shown below.

Tiagra 3x9 system. 4500 I think. 4400 Series / 4403 FD

The Problem

Shifting from position [2] to position [3] does not cause the chain to move to the middle chainring as I initially expected. Instead, it just rubs severely. If I increase the cable tension to the extent that it starts to shift, two nasty things happen:

a) Other aspects of the shift setup deteriorate.

b) It alternates between just barely shifting to the middle ring and throwing / dropping the chain to the inside. This, even with the low limit screw properly adjusted. My friend tells me that this chain dropping has been a common occurrence for him and was, in fact, how his chain got to be mangled in the first place. The mangled chain seems to have been the proximate cause of the RD shifting issues.

Having tinkered with, seemingly, every conceivable permutation of FD height, FD angle, limit screw adjustment, and cable tension, I'm now starting to think that this behavior is just how this system is designed. Can anyone confirm or refute that?

I've discovered that a "full swing" lever push from either position [1] or position [2] will successfully move the FD to position [4]. I'm starting to wonder if the right way to shift from the small ring the middle ring is by always doing full swing shift of this sort, bypassing the problematic [3] position for the ring change. The [3] position still has utility but, perhaps, the way to access it is by shifting to [4] and then back down to [3].

So, is my hypothesis about the shifting behavior of this FD correct? Or do I need to keep tweaking the setup until I get a proper ring jump when I shift from [2] to [3]?

Harold74

Some photos of the current iteration of the setup in anticipation that may be asked for.

masi61

If this is your first triple FD I am wondering if you installed the cable like you would for a double FD and expected it to work the same?

What I’m talking about is when you installed the cable did you shift the front shifter down to the granny ring (low) position and just pull all the slack out of the cable? Because if this is what you did, you will have problems getting into the big chainring. The cable will feel way too tight.

I’m looking at your drawing and while your numbering system is unique - I can sort of understand what you are talking about.

if you could verify if the shifting behavior is like Imdesribed, then I can send a reply explaining how to get it to shift more easily and reliably. The triple Shimano road FD’s are a bit tricky to set up but you can set them up to work properly and NO they are not crap and they did not just run like crap due to some design flaw.

Harold74

Thanks for your response masi61. No, I don't believe that I made the error described above. In terms of cable tension, I set it up form the large chainring down, per the instructions that I've been following and my understanding of conventional best practice. Everything shifts as it should except, perhaps, that non-jump in rings from the [2] position to the [3] position. And even that doesn't produce the drops so long as I don't try to add cable tension to force a ring jump there.

maddog34

Verify WHICH generation of Tiagra triple FD you are dealing with... also.. Exactly WHICH Brifter is on the bike... tiagra 4700 is an oddball that Needs a VERY Specific Setup to work, and Only Works With A Tiagra Brifter.(GRX is also said to work with Tiagra 4700)

not sure if earlier tiagra needed the special setup, but it is online and MUST BE FOLLOWED to achieve correct operation... it involves setting the cable up while the brifter is shifted into position two, NOT pos. One..... TWO.
look up your particular model to confirm correct setup procedures.
my Fuji CF bike is equipped with Tiagra 4700.... the group was purchased "new take off" and had no instructions with it..i fought that silly FD for a day, then looked up/followed the setup instructions online... instant happy maddog!

Harold74

I'm a structural engineer and this is really my first "bike" diagram. I winged it 100%. If you'd care to teach me the proper nomenclature, I'd be happy to revise my sketch. I would like to know how to do it properly if there is "properly".

It certainly was not my intention to suggest that this system is "crap" masi61. I simply seek to properly understand the systems limitations and design features. Even with double ring setups, there are some things that surprised me when I first encountered them. I've stuck to indexed downtube and bar end shifters on my own bikes thus far. My only experience with brifters has been on my wife's bike and the bikes of some friends that have had me do work on their setups.

Harold74

Tiagra 4503, per the attached manual from Shimano.

Might you direct me to that special setup? I've googled this extensively and have yet to come across it myself. I'd be willing to give it a try even if it's meant for a different setup. I have time for a little experimentation.

Andrew R Stewart 

Shifting from the inner ring to the middle one is often aided by over moving the lever. The lever will trim back to the "nearest" click point in the lever. If this "auto trim point" produces chain rub then use the lever's trim click to only slightly move/center the cage.

Differing Shimano STI front triple levers had different trim features. From none (like most of the MtB STI levers are) to each ring having it's own trim click. This is one reason why we might ask for the specific version of a series as over the years of that series (like Sora) the features change with model years and trickle down evolution. One year a Sora triple ft lever might not have a large ring trim capability, the next and it might.

Shimano triple ft ders are well known for their tendency to drop a chain onto the BB shell during a poorly done down shift. This is why aftermarket chain watchers are so common.

As an example of how much a rider's technique can affect ft shifting my wife complains about ft shifting much more at the end of a long ride than at the beginning. She's getting tired and not being as on the ball as a couple of hours before. Andy

dedhed

https://www.sjscycles.com/Instruction...structions.pdf

Harold74

Andrew R Stewart thanks for your comments.

Awesome, thank you so much for that. I'll give that a go tonight.

If I follow the procedure below from my position [3], it's apparent that will mean that a shift from [2] to [3] will indeed trigger a chain jump from the small ring to the middle ring. So that pretty much answers my original question.

The proof will be in the pudding but, based on experience thus far, I suspect that setting up the FD this way will make for a larger gap between the FD outer plate and the chain in the large ring than I'm used to.

freeranger

I have a 4300 (fairly sure) FD on my wife's LeMond. Looks identical. Didn't experience any problems with it. That is, other than the spring rest eventually breaks (poor design). FD's for triples aren't common now, but if one is needed, a Microshift FD-R539-F works, and the design looks like it should last. And it shifts just fine. Did require a band adapter, but they might make one with the band integrated (my old FD already had a band adapter). Our LeMonds have 4300 brake/shifters which do have the trim feature.

KCT1986

You probably mean FD-4403. This was the first Tiagra series. It had the design with the H/L screws on the cable attachment pivot, instead of vertically on top.

freeranger

KCT1986 you are correct. It is a FD-4403 (broken one laying around in the garage). We've had 2 break at the spring rest. And in terms of design and shifting, I actually prefer the Microshift.

KCT1986

This uncommon pivot arm was a carryover from some of the earlier pre-brifter FDs. Some of those even had the design that moved the cage forward as well as outwards, what Shimano called 'Leading Action Pantagraph".

Harold74

So, after closer inspection, it turns out that I also am working with Tiagra 4403 levers and FD. So freeranger 's experience maybe particularly valuable here.

I was able to find the technical service instructions for this system, similar to what dedhed posted earlier. Those are attached. Procedurally, there seems to be no difference between the 4403 and 4503 setups.

I tried the instructions for an hour last night, until my FD cable gave up the ghost. Poor results so far.

One thing that is confusing me is the number of indexed stops involved. The first clip below indicates that the middle ring has two stops. Check. However, the shifter itself has four stops plus the null position and, therefore, five available positions total. So it seems to me that there are two rings that have two stops. Which rings are those? Is it the small ring and the middle ring? Or the large ring and the middle ring? The answer, obviously, affects how I approach the setup.

The second clip below suggest to me that it is the middle and large rings that have two stops, such that the small ring would have only one. However, two things about my hand's on experience thus far suggest to me that it may be the small ring that has two stops instead of the large ring:

1) The gap between the two large ring stops is so large that the top ring shifting becomes a mess. 3 mm gap between outer plate and chain at the top in the high position and a low position that rubs egregiously no matter the rear cog.

2) The whole setup seems to work much smoother with the two positions placed on the small ring, save the issue that I mentioned in the OP.

Anybody care to comment on where the extra index stop ought to go? That answer, alone, would reduce my strategic trial and error efforts by 50%

Andrew R Stewart 

I long ago stopped trying to memorize the newest specs. To figure out trim and primary lever positions I watch the amount of cage movement per lever click/position. The closely positioned clicks are the trim ones and the large apart clicks are the primary ones, usually (Campy Ergo is a whole 'nother thing).

I think the Shimano instruction image showing the cog/ring combos that might have some rub (for a "Shimano instructions followed set up") is only the second time this aspect has been mentioned here on BF. Back when indexing was first coming onto the market and each year it would trickle down another cost level more and more new bike buyers were unfamiliar with indexing (and shifting in general). Shimano felt the need to be somewhat proactive in their written claims and began to include the little pictorial guide of no-no cog/ring combos. I have mentioned this to, possibly, hundreds of customers over the years as a way to explain why their chain rubs. Andy

KCT1986

So you are working with a 4400 shifter, not a 4500. That changes a bit.

Based on your numbering diagram. Your 'primary' positions are 1, 4 & 5, where you should get a quiet chain run with the 7 or so sprockets in the back with the respective best chain line.
Your trims are 2 & 3. #2 for small ring and outermost sprockets. #3 for mid ring and inner sprockets. Large ring has no trim, you get whatever is possible with the correct setting for the outer cage plate while in the smallest sprocket. As noted in Shimano's diagram, noise in some combos may still not be 'trimmed' out.

This is based on Shimano's setup, cable tension set with shifter in position #3, mid chainring and inner most sprocket, with minimum clearance of inner cage.

Also, this is based on Shimano's correct crankset, (chainline, chain ring size combo & shift aids, & straight running rings, etc). Your large ring looks like it doesn't quite follow the curve of the FD cage. The rear part of the cage looks a little far from the ring, even considering the height that it is at. Length of the chainstay will also affect chain rub in some combos.

Harold74

It's a Truvativ, hopefully Shimano compatible. Based on the catalog info for the bike that I included in the OP, it appears to be OEM for the bike.

Yeah, I would expect a longer chain stay to reduce chain rub potential. At 63 cm, it's the largest frame that this bike came in. At the same time, my impression of the frame geometry is that it would promote a short chain stay. Fairly steep seat post and very tire gap at the seat post.

I tried that a few more times last night and haven't been liking the result:

1) Based on a taught cable with the FD down at the low limit position, I have a 5 mm gap to close at the #3 position that pretty much uses up all of the barrel adjuster travel. There are other ways to cheat the cable tension of course.

2) This setup produces significant chain rub when in the small ring and either of the two largest rear cogs. We live in the mountains so having noisy climbing gears is pretty annoying.

3) In the large ring, this setup produces a 3mm gap between the FD outer plate and the chain. At 3X the recommended, this would seem to invite chain drops.

Presently, I'm inclined to abandon the setup procedure based on the middle ring and return to the one that I started with and described in the OP. That seems more functional given that it would produce quiet gearing options across the entire gearing range and provide the best protection against chain drops. That, at the modest cost of fairly simple modification to shifting technique.

Without swapping out existing parts, the only other variable that I can think to play with here is a non-parallel FD angle relative to the rings. It appears that a clockwise rotation might alleviate some of the problems mentioned above. That said, I certainly don't get the impression that the system was designed in anticipation of a non-parallel FD setup.

Harold74

These are some shots of how the "official" setup procedure panned out last night. That's the procedure focusing on a tight FD gap at the low, middle ring position and the chain on the largest rear cog.


Starting from taught at the low limit position, there's a 5 mm gap to close with the barrel adjuster.

At the big ring in the smallest rear cog, there's 3mm of gap between the chain and the outer plate.

Harold74

Thanks for mentioning the microshift option. After the shifters, the FD is the part that I've been most worried about having to replace if I bung it up.

I've got my own triple build in the works and procured an IRD Alpina for that. I might have gone with the Microshift if I'd know that it existed. The Alpina was expensive for me to ship to Canada and Microshift stuff is available locally.

I note that both the Alpina and the R539 sport similar forms: both shorter cages than the tiagra on my friend's bike. Perhaps that is more amenable to triple setups with sub-53T large chainrings.


Microshift

IRD Alpina

maddog34

try... leaving slack in the cable when in the small front ring, then tighten the high limit screw to use some of that extra "3mm" when shifted into the big ring... this is how i set MOST triple FDs.
1mm is usually all it takes to eliminate chain rub.

question: what is the Height of the FD cage above the Big ring? more than 2mm CAN put the mid shift bulge in proximity to the chain when shifted to the big ring, depending on contour of a cage.....

Also... the rotation of the der on the frame gets VERY critical on narrow, high gear count drivetrains.... i Always Look Directly Downward at the cages when setting the rotation.. the outer Cage Should Be DIRECTLY ALIGNED with the outer ring WHEN SHIFTED To the outer or mid ring... the Cage Can Change orientation when in the low gear position... and DON'T let the shift bulges fool you!


and Just Yesterday, a bike came thru that had had the FD cage BENT OPEN, and it was refusing to shift up because of that bending... cages Get Bent sometimes... in that case, it was a low end bike, and a pair of small channel lock pliers had things working in a few minutes... look for any bowing (even a hint of bowing can make a difference!) in the cage walls.........

Smitty2k1

Do the chainring sizes match the specs for that FD-4403 derailleur?
I've got the Tiagra FD-4503 on my touring bike and honestly the triple setup has been difficult.