dddd

That must be some very good lighting in your shop, for you to discern tolerance-level differences in 2-degree tapers using your eyes.

Keeping a few worn/spent JIS cartridge bottom brackets around instead of discarding them is perhaps the simplest, most reliable way to ascertain (by a quick test-fit) what length JIS spindle is going to be the best one to buy.

Comparing the relative depth of engagement of different 2-degree tapers in any given crankarm is perhaps most easily done by locking a caliper to a 13mm opening width, and then simply comparing how deeply that any two different spindle ends will pass between the jaws.
You will still have to also compare at least the driveside lengths that the different spindles protrude past the face of the bb shell (unless both were known to be perfectly symmetric) to get the chainline right.

Yes, tapers appear (and are) different, but it's their thickness that needs to be compensated, not any nonexistent difference in their taper angles.
If the taper angles were different, the installation would not likely be reliable and the chainring rotation axis might not stay true under power.
Not that the reduced engagement depth of a JIS spindle taper isn't less than perfect, but does at least have a good track record of working properly, and I've seen OEM spec of a JIS bottom bracket with a Campagnolo Record crankset.

Taper-length differences are sometimes also cited, but since the taper dimensions reference off of the end of the spindle, the length difference rarely has any effect on the contacting surfaces, and exists only inboard (toward the center of the bb shell) of the hole in the crankarm so not touching anything but air.

Does stating something that is known true, versus not true, make someone an "expert" in any field? I mean did any of us here design any of these parts?
Should we then be praising your "expertise" for every one of the statements that you post?

bulgie 

Like Carl Sagan said, extraordinary claims require extraordinary evidence. Since everyone else says all tapers are 2°, I think you need better evidence than "I held two spindles up next to each other". I am highly skeptical that you can see taper angle differences by eye. Good on ya if you can, but please don't be insulted if I doubt it. I've been mixing spindles and cranks since the '60s, was in the bike biz for ~30 years and have remained an active hobbyist in the 25 years since then, but I don't consider myself an expert. I know for sure I can't discern the difference in angle between say Campy and Stronglight (or any other brands I've tried) by eye, because I've tried. I have always gone on the assumption they are the same angle, and have never found any evidence to the contrary. Other than Lambert, I have never seen any spindle with any angle that wasn't compatible with Campy/Sugino/Stronglight/TA/etc. Yes, spindles and arms can be incompatible for other reasons, but never for a difference in angle. Other brands with other angles (Ofmega?) are rumored, but I've never seen one with my eyes, in the half-century I've been doing this.

The only differences I have seen are the size at the small end, and the length of the taper before it ends, usually at some sort of shelf. The only things that can go wrong are:

1. going on too far, to where the back of the crank hits the shelf
2. not going on far enough, due to the taper being too big, so the small end of the taper is too far recessed into the arm, inadequate engagement.

#1 can be fixed with shims, a PITA to install but usually reliable enough. #2 is fatal unless you are a skilled machinist.

Chain-line is another issue, which can be dealt with a number of ways. I don't consider that a taper incompatibility per se.

dddd

Wait, I confess to having done that (bottoming a crank bolt washer against an ~ISO spindle going into a JIS crankarm), but the fix seemed easy, I just used a grinding wheel mounted in my drill chuck to grind a rough millimeter off the ends of the spindle. And, predictably enough, the four flat's shoulders did then bottom on the inside face of the crankarm, but with apparently not too much interference, as the crank bolt didn't yield excessively to my subsequent modest re-torquing following a weekend of riding.

For racing-grade setups, I guess maybe we shouldn't be mis-matching spindles and cranks at all, and perhaps not for stronger/heavier recreational riders either.
But (quoting Chas C here) "if it fits, it works" does seem to hold true in my limited experience of perhaps a half-dozen such "batard" installations.

One more detail of using a JIS spindle in a French, Campy or ISO crankarm has to do with the four visibly-bigger radii on the JIS square taper that effectively makes the contacting flats narrower. This is a bit sub-optimal imo, but hasn't caused me any problems as a spirited lightweight lightweight rider in the foothills. I am imagining that the narrower flats and reduced engagement depth will increase the outward force and thus pressure exerted along the edges of the flats via "camming" forces against the inside of the square hole, possibly leading to increased wear of the mating surfaces. But I have also noticed that the chainrings run perfectly true on French Stronglight cranks and ISO-sized Sugino Mighty cranks that I've mounted on JIS spindles.

Kontact

You don't look at the angle of tapers, you hold the tapers together and look at whether the spindles remain parallel.

Kilroy1988 

Sorry, but anyone who thinks they can measure within a degree of difference by eye across parallel lines interrupted by bearing races and different shades of metallic gray over the space of a few inches is full of it.

zukahn1 

If your bike was made for a true English Raleigh older style BB the 5mm or so drive side offset is basically correct.

Kontact

You should be sorry. It is really simple and easy to see.

One spindle is horizontal and on the left. The other spindle is horizontal and on the right, just above the left spindle. You press the top taper on the right side of the left spindle to the bottom taper of the right spindle on the left side and hold them with firm pressure or a clamp. It is relatively easy to see whether that one 6" spindle is parallel or not with the other 6" spindle.

If you were unable to see that, you would also be unable to perform basic tasks like hanging a picture straight or aligning a disc caliper.


You are aware that 1 degree of offset over 6" is a tenth of an inch, right? Do you think you can see a tenth of an inch?

Kilroy1988 

Firstly, a 120mm spindle is 4.72 inches long. If you overlap the tapers in the center, you lose approximately half an inch in the middle. You cannot look out to the far ends of either because a half an inch on either end is also tapered and not straight. So you're looking at about 8" distance between the far ends that are straight (i.e. inside the tapers), and you have the bearing races completely distorting the view of those far ends anyway. Finally, and critically, because you are pressing them together in a central point, the offset of the 1° difference (assuming it is there) is halved again in either direction - it's a 1° difference from the center to either end, not end to end. So, you're really dealing with about 3.5" of offset at 1° at best.

What you actually have to compare the angles are only the absolutely straight sections between either bearing race, which are themselves less than 3" long and are separated from one another by about an inch and half that's being taken up by the overlapping tapers in the middle and the inside races. You can't look straight down anything, and you can't look across any unbroken space to see the two lines converging because of the fact that your spindles are offset (one on top of the other) and that you have the bearing races in the way.

That being said, across less than 8" with four different bearing races in the way (not to mention either your hand or a clamp), you are expecting me to believe that you can see with your naked eye an approximately 1.5mm offset that resembles a slight kink in the middle of a totally obscured line? It's no wonder you think the taper on different spindles are so dramatically different with such flights of fancy.

It's not comparable at all to hanging a picture frame or adjusting disc brakes or doing any other task that requires you to eyeball 1/10th of an inch on a clean horizon.

-Gregory

Kontact

You don't "sight down" the spindles. You hold them up level if front of you. If you make one level you'll be able to see if the other is. Like hanging pictures.

If that is too challenging, hold them up in front or some other horizontal line, like a window frame. The angle of the second spindle to the window frame will be obvious for the same reason that you were able to align the first one to the frame.

People can see very small differences in angle between parallels, and we can see deviations from horizontal and vertical very easily as well. Your instance to the contrary does not change these facts.


And 1mm is 1/25th of an inch. 1.5mm is a 1/17th of an inch. Pick one.

Kilroy1988 

To the first point, I disagree that you can reliably discern the difference that way. There are too many variables breaking that line including the bearing races and the fact that on one half the background is up and on the other half the background is down, because the spindles are not being held in a straight line this way.

To the second point, that's not just using your naked eye and I never said you couldn't discern a difference if you were to create some frame of reference using a tool like a window frame. That's got nothing to do with your claim.

-Gregory

Kilroy1988 

And yeah, people can see very small differences in things. I used to play a game on my computer where all you have to do is try to create specific angles or parallel lines as part of a challenge. Sometimes I could be dead on and other times I could be a hair off. Just using your eyes without a frame of reference (like the straight edge of a ruler or a windowpane) is not a reliable way to say that you're discerning slight differences in the taper of a bottom bracket spindle. Sorry. There are way too many variables at stake when you have two small, unevenly shaped objects in hand such as that. If you think you can do it again, I call BS because you could just as easily be a degree off by thinking it's not straight.

JohnDThompson 

Perhaps the definitive test would be to measure the width of the taper at two locations separated by a known distance, and apply a little trigonometry and algebra. I suspect this would require more precise instruments than I have at hand.