what the deal with all these different modern BB standards is?
In the beginning, there was a 68mm wide BB shell threaded 1.370"×24tpi. Except for French, Italian, Swiss and Raleigh bikes, that is, but they were all there or thereabouts. This worked fine for solid steel axles of about ⅝" diameter with loose ball bearings.
Next, some clever bloke decided that 22mm hollow axles were better (and they are) so we got ISIS, Octalink and Powerspline. Unfortunately, he wasn't quite clever enough, and the thin bearings lasted about a week.
A slightly cleverer bloke figured out that you could make cantilevered cups to put proper sized bearings outside the conventional shell, and increase the axle even more to 24-25mm, which is big enough to make reliable aluminium axles saving even more weight, and we got HollowTech2, GXP, UltraTorque and MegaExo, plus Zipp pushing the envelope with a 30mm axle version. The new problem was that the BB shell had effectively grown to 90mm wide, which didn't leave room for both a decent crank thickness and the riders' heels unless you increased Q. Alongside this, we got Shimano's BB86/90/92, which simply integrated the bearing receptacle into a wider plain-bored shell.
Cannondale looked at all the problems with their engineer's eye and worked out that the good things were; 30mm axle (light, stiff), pressed in bearings (cheap to mass produce and assemble) and 68mm shell width (room for both cranks and ankles) and designed BB30, and published it as an open standard.
All was well with the world for about ten minutes, until people decided that the cost of making frames accurately was too much for them, so we got PF30 where the BB shell accepts the bearings via a mushy shim, which avoids the need for a high precision bore in the shell at the cost of putting some cheap plastic in the load path from cranks to frame. Also, some people (Vroomen&White, THIndustries) thought they were too good for a mere common standard, and started fucking unnecessarily with the bearing spacing, so we got BBRight and BB386Evo
In the beginning, there was a 68mm wide BB shell threaded 1.370"×24tpi. Except for French, Italian, Swiss and Raleigh bikes, that is, but they were all there or thereabouts. This worked fine for solid steel axles of about ⅝" diameter with loose ball bearings.
Next, some clever bloke decided that 22mm hollow axles were better (and they are) so we got ISIS, Octalink and Powerspline. Unfortunately, he wasn't quite clever enough, and the thin bearings lasted about a week.
A slightly cleverer bloke figured out that you could make cantilevered cups to put proper sized bearings outside the conventional shell, and increase the axle even more to 24-25mm, which is big enough to make reliable aluminium axles saving even more weight, and we got HollowTech2, GXP, UltraTorque and MegaExo, plus Zipp pushing the envelope with a 30mm axle version. The new problem was that the BB shell had effectively grown to 90mm wide, which didn't leave room for both a decent crank thickness and the riders' heels unless you increased Q. Alongside this, we got Shimano's BB86/90/92, which simply integrated the bearing receptacle into a wider plain-bored shell.
Cannondale looked at all the problems with their engineer's eye and worked out that the good things were; 30mm axle (light, stiff), pressed in bearings (cheap to mass produce and assemble) and 68mm shell width (room for both cranks and ankles) and designed BB30, and published it as an open standard.
All was well with the world for about ten minutes, until people decided that the cost of making frames accurately was too much for them, so we got PF30 where the BB shell accepts the bearings via a mushy shim, which avoids the need for a high precision bore in the shell at the cost of putting some cheap plastic in the load path from cranks to frame. Also, some people (Vroomen&White, THIndustries) thought they were too good for a mere common standard, and started fucking unnecessarily with the bearing spacing, so we got BBRight and BB386Evo