Sir_Name 

Hey all, as the title suggests, I recently purchased a set of TA Pro 5 Vis crankarms from a seller in France. They were sold as a “track” set, which I had assumed was on account of the mounted 1/8” chainring. I plan on swapping over double chainrings from my current 49d. Going by the following Compass/René Herse link, the recommended SKF BB that I’ve been wanting to upgrade to is JIS taper: https://www.renehersecycles.com/wp-c...y_20220918.pdf

So, as luck would have it, a used example of that BB in 121 mm length with JIS tapers was available on eBay at the same time as the crankarms. I bought both expecting a straightforward swap. I’m currently running a 118 mm JIS taper IRD QB-95 with Stronglight 49d crankarms. The bottom bracket is beyond shot at this point, and the NDS crankarm has never seated properly. I’m assuming it was run loose by one of the previous owners. Hence the parts search. I like 165 mm arms which narrows my window on the market, so lucky me to find compatible crankarms and BB at a time of need.

Or so I thought. The crankarms sit way outboard on the BB spindle compared to what I’m used to. Enough for me to doubt whether there’s sufficient engagement to not foul the crankarm tapers over time. Of note, I’m not a lightweight feller, and I bounce over rock and root through New England singletrack trails regularly. I need tight equipment. A few pics below.

NDS arm torqued pretty darn snug with a 6.25” ratchet:






This isn’t going on more than another mm or maybe two at full torque I reckon. Current gap from spindle face to the plane that terminates the arm taper is ~5.5 mm.

5 mm hex key showing the additional engagement from the previous owner’s crank:



This previous high water mark is more or less what I’d expect for proper engagement. A couple mm less given the 5.5 mm outboard gap of the previous pic.

After removing the arm from the spindle and using the grease marks as reference, engagement here was approximately 9 mm out of a total 15.5 mm taper length (parallel to the BB axis). Around 58% engagement of the crankarm taper length.

So, should I just go ISO taper? JIS narrow?

Or is there some weird NJS or similar track square taper that’s different from the usual JIS/ISO/French melange that might account for this seemingly less than ideal engagement? I couldn’t find anything online indicating There could be TA track specific arms with a taper and BB to match. Plus, why go thinner on a track spindle?

I do have a loose Campagnolo Record spindle that I slid the arms onto for comparison. Not much better.

This crankarm set is French threaded and does not have the ‘capped’ blind pedal threads of earlier versions, if that figures into the calculus here.

What do?

Sir_Name 

Rough measurements on the currently installed 49d and IRD JIS BB show ~2.9 mm from the spindle face to end of taper. Less than 4 mm of exposed taper between the BB shell and inside face of the NDS crankarm.





Noted that this thread seems to more or less recommend ISO which would track with wanting to sit farther up the taper. Guess I should do some trig and rough out a difference of engagement for the two tapers given my starting point.

https://www.bikeforums.net/classic-v...tory-info.html

merziac

I got nuttin but usually stick my fingers in my ears while going la, la, la, la right before I get out the bag O spindles for a round of mix and match.

That said, I've never had one I let be that far out that was going to be a rider and wouldn't be thrilled about the allen wrench spaced one either, I've probably just been lucky for the last 50 years.

Hopefully the experts will be along to straighten us out.

Sir_Name 

Yeah, that’s what’s bugging me. I’ve mixed a fair amount of vintage square taper cranks with old and new spindles and the only real concern has been chainline and keeping the Q factor more or less tight and centered. This is a first of the tapers raising concern before installing things. That’s with two bikes running Stronglight 49d crankarms on modern JIS bottom brackets. Dunno. Maybe I just need to find an ISO BB and really torque it all together on the frame to see where I’m at. I just don’t see the current arms and BB coming together to a point I’m comfortable with. Bummer.

Sir_Name 

For comparison, the Record spindle and Record crankarm loose fit versus Record spindle and TA crankarm loose fit:




(Note there’s an overhang here)




~~~
My primary assumptions on the TA crankarms and SKF JIS fitup were based on the René Herse tech sheet and that the Stronglight and TA tapers would be more or less the same wrt JIS tapers that worked for me before. Maybe I need a bigger wrench…

merziac

Exactly, maybe the real experts will come along and save us from ourselves.

I recently had the worst example of this I've had in quite awhile but its a Nervar steel square taper so all bets were off, I got it where I can live with it but its going to be what it is and being steel, I'm not going to lose any sleep over it.

El Chaba

Others have put way more research, experimentation, and mental energy into this than I, but I am of the school of thought that the whole “ jis vs. iso” taper construct is a false dichotomy….and a rather recent one. In my view each manufacturer made spindles to fit their cranks. If you want something to fit you get crank and spindle from the same of both- and from the same era-and in the size for the intended application. The TA 344 spindle and associated cups should be the perfect fit.

Moe Zhoost

Have you measured the spindle end to verify that it's JIS? The JIS standard is 12.65 mm. ISO is a bit smaller.

Charles Wahl

Sutherland's 6th indeed says that the closest taper to any of the TA cranks is JIS:


If I remember correctly, JIS is basically the "stoutest" of the taper standards, followed by Campagnolo, and ISO is the "skinniest". I seem to remember also that TA and Stronglight spindles I've got here have end dimensions that are fairly similar to JIS ones. Campy, of course, changed from their own taper to ISO sometime in the early 90s. Phil Wood didn't make a Campy taper, as I recollect, and advised people to buy their JIS-taper bottom bracket for use with the vintage Campy arms.
Sheldon Brown did not seem to have a problem, as I remember, with a crank arm overhang of up to 2-3 mm (when tightened). But you could check that out.

"Loose fit" doesn't tell you anything, and I'll wager that "pretty darn snug" with a 6.25" rachet is nowhere near the 350 in-lb (29 ft-lb) that a crank should be torqued to; it's 'pretty darn tight' with a torque wrench that's got about a 16" lever arm, for me -- but maybe I'm just weak sauce.

Charles Wahl

I don't see how the above makes sense at all. OP is trying to adapt a (supposedly) "track" set of crank arms to use as a double. The difference between a drive-side crank intended for a single chainring and a double is going to be (probably) the lack of lands for the inner ring on the backsides of the arms (no big deal), not the dimensional relationship between outer ring and the spindle end. The only reason one would use a TA 314 spindle would be to achieve having a single chainring (mounted outside the arms) situated closer to the centerline of the bike than the outer ring of a double. IF OP were looking for a replacement TA spindle to use, then the one marked 344 (see table above) would be the one to shoot for. But he's not; he wants to mount the arms on a cartridge bearing BB assembly. Were I doing that, I would look for a JIS unit (as Sutherland's suggests) with a slightly shorter spindle length than the proper TA spindle for the BB configuration that has the correct number of chainrings. BTW, I have at least 3 TA spindles marked 344, and they actually vary in length a bit: 114.5, 115 and 116.5, even though Sutherland's says 116.5 is the nominal length. I certainly wouldn't go for the 121 mm unit that OP has bought -- something more like 114 or 115 mm; unless he spends more than 75% of the time on the smaller chainring. I also wouldn't buy a used cartridge BB unit on eBay, but that's just me -- why take the risk for a bit of savings?

BTW, my impression is that one can swap TA and Stronglight arms and spindles that are nominally the same size with a fair degree of confidence that they'll work out OK -- apart from the differences between parts due to manufacturing tolerance (which, in my experience are greater for TA and Stronglight than for, say Sugino, which seem to me to have tighter tolerance) they're very close to each other in terms of taper and length, distance between nominal bearing running surfaces (fit of spindle to cups), etc. And, based on used examples that I have, they are extremely durable parts, compared to Campagnolo or Shimano; comparable to Sugino. But, if OP wants cartridge (no problem, I like those too), then all this is immaterial.

dddd

I've noticed that the bevel between each of the four flats tends to be bigger on JIS spindles than on any European crank spindles.

To me, this makes JIS spindles less suitable for use with European cranks that typically require spindles closer to ISO dimensions, i.e. smaller.
You would thus get less engagement depth AND have narrower flats using a JIS spindle, both increasing the torque-induced "camming" stress on the inside of the tapered square hole in the crankarm (all yet-worse with a bigger/stronger rider who may ride on rough surfaces with the crankarms horizontal).

If one correctly runs through the trig on engagement depth of JIS vs. ISO, using the 2-degree angle off of the spindle centerline (or 4-degree included angle), the roughly .2mm difference in spindle cross-section equates to half that, or .1mm difference between the centerline and the surface of the flats.
The tangent (Tan) is Y/X where Y is the change in the distance from the centerline of the flat, and X is the change in the engagement distance on the spindle.
Tan of 2-degrees is .035
So Y/X = .035
Substituting .1mm for Y
.1/X = .035
Solving for X
X= .1/.035 = 2.81mm

2.8mm is roughly the difference in spindle engagement between an ISO and JIS spindle or crank.

Comparing spindle cross-sections is easy if you use the thickness of the caliper's jaws (roughly 1/8") as the reference point for locating the measurement. This way, any chamfer at the end of the spindles won't complicate the measurement of either spindle being compared.
The measured difference between two spindles will be equal to 2Y in the above example.

Showing spindle thickness or cross-section 1/8" (the jaw's material thickness) from the end of the spindle, BELOW:
JIS spindle shown.
A TA spindle measures more like 12.75mm iir as compared to this one's 12.9mm, and an ISO spindle is more like 12.68mm, old Campy being close to the TA dimension...
Cheap calipers can be checked using a new ball bearing which is highly accurate. Only spindle thickness differences are important here, not the exact numerical dimension shown.
Calipers can also be set (locked) to a fixed dimension, and it's engagement depth along two different spindles can be instantly compared. Even an adjustable wrench thus might allow a decent comparison!

Charles Wahl

I still maintain that unless OP is installing arms to torque that's correct, the amount of overhang observed is inconclusive re "problem or no problem". And that a 120-something-wide spindle is almost certainly too wide for a double, especially if that spindle is JIS.

TMI follows:
I'm no expert, but I'm pretty sure there was no ISO (6695) standard until 1991, well after the age of square spindles; before that it was just Campagnolo, TA, Stronglight, Sugino, and the other Japanese manufacturers doing their respective things. If you look at Sutherland's, you'll see that even Shimano sometimes used Campy taper, sometimes JIS apparently, for different product levels. I presume that JIS must have existed as a standard well before this, since their standard nomenclature for spindle markings seems to have been used on a ton of product.

I don't have any documentation about the JIS standard, but the ISO standard for 1991 shows the width of the spindle, measured at 1.5 mm from the bolt end, to be 12.6 mm, and the "standard" overhang of the crank arm at torque to be "min. 1.5 mm". The 6695 standard was apparently revised in 2015 (well beyond the time when manufacturers had moved into all the various goofy proprietary configurations used in the most modern equipment) to change the narrow end dimension, measured the same way, to 12.73 mm +0.02 or -0.05 mm, with the overhang revised to min. 1.5 max 2.0 mm. One wonders what difference that actually makes, or has made, to manufacturers and poor users like us, and why ISO even bothered.

The measurement 1.5 mm from end of spindle I can't really figure out: it's impossible to measure directly with the dial calipers I have, all of which have jaws over 3 mm thick. So the only way to measure that would be to find a 1.5 mm shim, cut a hole in it for the spindle, and measure with the caliper face up, using the tapered/sharp ends of the jaws, with the caliper set flush on the shim. That isn't how dddd is measuring in the last photo above, BTW; though I freely grant his point that it doesn't much matter what the dimension is, so much as the difference in dimensions of two different spindles.

Sorry for nattering on, but I have nothing else to do, being currently subject to an extended, enforced period of limited activity, and especially no bike riding; plus the weather has been inclement.

dddd

The interesting about the ISO standard is that it is identical to Sugino Mighty and to the Sugino-made Suntour Superbe and Sprint.

That's right, Mighty was close to old Campy, but identical to the later ISO Campagnolo cartridge bb taper.

The older Stronglight spindles I have measured are closest to ISO/Mighty, i.e. the smallest common taper.

Old Campy bb spindle tapers are closer to ISO dimension than to JIS.

I believe that Stronglight was the first big producer of square-taper spindles, and that Sugino (with their Mighty) copied their spindle dimensions rather than Campagnolo's.

I'd guess that Sugino was at the forefront of JIS development, possibly a taper developed to work better with the softer melt-forged Maxy cranks, by increasing the length of engagement. This same strategy was later used by shimano when they developed their 2nd (and longer) version of their Octalink spindle taper, which then brought Octalink down to much lower price points.

52telecaster

Just a data point from me of probably no significance but, on my grand record that came with a stronglight double I put a campy with triplizer. The campy crank went on the stronglight axle but not quite as far. Making it so there was enough room for a triple. Honestly the spacing is cheating to the narrow end of things a bit but that's what I prefer. At any rate it seems to work fine.

merziac

I didn't know there would be math's, math's are hard, I hate math's.

And just for fun since I agree with Mr. Wahl, much of this is inconclusive without being fully, properly torqued, seated, ridden and retorqued.

Also, in a case like this where another mm or 0.5 may get you to where you might be able to sleep at night, a smudge of grease that I generally avoid to possibly overcome any stiction, may just get you there and save you from yourself.

oneclick 

1. measure the exact thickness of your jaws;
2. measure the end of the spindle - the very end;
3. measure the spindle with your jaws' outer faces flush with the spindle end;
4. do the math i.e.

width at 1.5mm = (width at end) + ((1.5/jaw thickness) * (width at jaw-flush - width at end))

Charles Wahl

There generally is no "very end" of the spindle to measure; the end is typically rounded over somewhat, and in any event, even if milled to a square edge, that's impossible to measure without cheating up the taper a little. Look at OP's photos, obviously there's a roundover or chamfer on the very end of the spindle.

The ISO approach is right: standard dimension at some distance from the end, but 1.5 mm is not a practicable number, IMO. The only way would be to measure the taper at whatever thickness your caliper jaws are, then "do the math" from that point to 1.5 mm from the end, assuming that the taper is indeed 2 degrees.

oneclick 

It does take a certain degree of physical dexterity to measure most things accurately; I can supply the procedure and the maths, but holding the bits properly is up to you.

Charles Wahl

Dexterity doesn't help when there's no actual material inline with the taper at the "very end".

79pmooney

This sounds completely believable in my experience.

Yes. My guess is that Campy was not completely consistent with their BB-crank fits. (Were they with anything else? Units fit together beautifully - the famous Campy quality but from one unit to the next? Not so much. By contrast, the good Japanese manufacturers were boring, they were so consistent. In the '70s, Sugino made the Mighty to be Campy and race-situation compatible. (I suspect those Mightys varied a bit on chainline when set on various "identical" Campy BB spindles, but generally close enough to work.) For the people choosing the standards for the new ISO, picking the rock-solid consistent Sugino Mighty would have been a easy decision.)

Slightly off-topic but: Shimano and Phil Wood's JIS spindles are very close. I have set cranksets on long Shimano BBs, measured how much further in I want each crank (knees love Qs as small as possible), calc'd the width and asymmetry and ordered from Phil. Yes, expensive. And yes, flat out beautiful! (And when you get the Phil, it has plenty of wiggle room for dialing clearances and chainline.)

oneclick 

In such a case it will help if you try to measure the end of the taper, not the end of the spindle.

Sir_Name 

Hey all, I’m back…busy week. There’s a lot to address that I’ll get to later. I kinda buried the lede. First question is whether or not there’s some obscure French track taper that differs from the norm. I assume no. It’s just odd that my Stronglight and TA arms sit so differently on the same taper. That’s my starting point.

Noted that I should just install the bb and really crank down on the crank arm fixing bolts to full torque to see where I land. I’ve been wrong plenty of times, so hopeful for a pleasant surprise I’ll say.

dddd

I didn't mention the spindle dimensions of the old TA taper in my earlier post, but it's relatively big/thick for an old European taper.
It's still at least 0.05mm smaller than a JIS taper, but it's at least 0.05mm thicker than an old Campagnolo taper and so is about 0.15mm thicker than old Stronglight or ISO taper.

So I am assuming that it must be your Stronglight arms which don't engage the spindle as far as your TA arms(??).

santa fe 2926 

I’ve had TA cranks and spindles for 40+ years, the TA 344 are 114.5 and 116.5. You think that was the change from 120 to 126 5 speed to 6 speed?

scarlson 

To add to the confusion, in my experience (~4 single cranksets and one tandem) I have come across two different types of TA taper. Broadly, I would classify them into "old" and "new". I don't know when the change happened, but this is my hypothesis. Because at some point TA DID start making things ISO. The Zephyr for sure was, along with most (all?) of their sealed BBs.

What I'm calling the "old" TA tapers are like JIS, as Sutherland's says. They are bigger than Campy and ISO. The "new" tapers are basically ISO.

The TA Pro 5 Vis has been in production a long time. Sometimes you can tell an old one from a new one, for example the really old ones come with covered pedal eyes, and the newest ones were sometimes anodized.

Here's a thread where I investigate this a little bit. I was surprised with my tandem crankset sitting off the JIS and old TA spindles I had. The cranks I had barely went halfway on (9mm engagement out of a total 18mm long taper)! I put a Campy spindle on the front and ISO spindle on the back of this tandem, and it appears to work.
https://www.bikeforums.net/classic-v...so-tandem.html

Here's what I think is a "new" crank on an "old" TA spindle. 9mm of free taper.


Here's the same on an ISO spindle. Sits 2mm farther inboard. This is probably acceptable.


Here's an old TA crank (old enough to have the capped pedal eye) on an old TA cup-and-cone spindle. 5.5mm of free taper. I'd consider this acceptable.