Sram Force Crankset 0.26mm Play at NDS Axle-Bearing

[aerobikewheels]

As SRAM europe is not willing to forward me to someone who might help to solve this problem. And the telephone lady was not willing to be cooperative in solving the problem, altough i asked very nicely.

And i am still waiting for a response of my dealer.

I will post it here and hoping for some input.

Recently i have ordered a SRAM groupset, All works fine, except the GXP crankset.

At the bearing to Axle contact point on the non drive side there is a difference of 0.26 Milimeter (0.01 Inch) This is so much that there is a lot of play in the axle.

At the Drive Side, this is a healthy 0.01mm.

Attached foto.

I assume this is a production error, But if not, that should be a very high margin.
Interested in your feedback.

[0psi]

Out of curiosity, what are the measurements on the BB in your hand?

[goodboyr]

If the crank is torqued down against the shoulder it should be locked enough to not move. The picture where you are miking the diameter of the shaft, you are not using the tips of the micrometer. I'm suggesting the shaft is actually OK, and instead the problem is either a spacer issue or lack of torque.

[aerobikewheels]

Out of curiosity, what are the measurements on the BB in your hand?

Bottom bracket in my hand is 22.19. I also measured this to see whether the problem lies within the crank or the BB

If the crank is torqued down against the shoulder it should be locked enough to not move. The picture where you are miking the diameter of the shaft, you are not using the tips of the micrometer. I'm suggesting the shaft is actually OK, and instead the problem is either a spacer issue or lack of torque.

Crank is installed with a calibrated tensionmeter to sram's exact specifications.

Tips of the micrometer is not making that much difference, Micrometer is an expensive and very accurate Mitutoyo. I did measure all sizes multiple times. Also i just checked with the tips as you suggest, same reading.

Readings with tips in shaft gave 2 hundreds extra of shaft diameter. Wich means difference of 0.03 at driveside (still good) and 0.28 at NDS (still not good)

0.26mm is a lot, you can feel the play freely, If you put torque on the crank and squeze it together you do not have the play as long as you do not apply force, However with little force, The play remains. Also when pedaling.

I checked four times if installation was correct, all spacers and parts are there. so we can rule out this.

Also frame (bb86) is checked and is exactly up to specification. So it is not play between the cups and the frame.

[goodboyr]

Deleted

[goodboyr]

The sram instructions say if you have play, then remove , regrease and retorque. I've had to do this on gxp installations a number of times, so that might get you more squeeze to stabilize things. Barring that, I would suspect the bb more than the crank. Never seen a crank out of spec.

[kauphy]

The BB in your hand has a dust-cap that the installed BB seems to be missing. I had play in my GXP crankset because I missed installing it once. All was well when I put it back.

[breezerboy]

The original SRAM BB's used a 22mm internal bearing that fitted the axle nicely. A few years back, presumably as a cost saving exercise, they changed to the more readily available 7/8" (22.2mm) bearing. They now rely purely on the two faces of the bearing being clamped solidly between the axle step and crank arm to stop movement - p!$$ poor engineering!!!
Best bet is junk the SRAM BB and fit a Wheels Manufacturing (or similar) that uses the 22mm bearing click here

[bikerjulio]

To add to @breezer comments - the original GXP design was by TruVativ and used a sleeve pressed in the NDS bearing to take it from 24mm to 22mm. If you can find a NOS TruVativ BB you will see what I mean. The NDS measures 22.1mm.

[jano]

Exactly, this is all down to SRAM producing a rubbish BB. Plus the GXP standard being a complicated solution to a non-existent problem.

[aerobikewheels]

The original SRAM BB's used a 22mm internal bearing that fitted the axle nicely. A few years back, presumably as a cost saving exercise, they changed to the more readily available 7/8" (22.2mm) bearing. They now rely purely on the two faces of the bearing being clamped solidly between the axle step and crank arm to stop movement - p!$$ poor engineering!!!
Best bet is junk the SRAM BB and fit a Wheels Manufacturing (or similar) that uses the 22mm bearing click here

To add to @breezer comments - the original GXP design was by TruVativ and used a sleeve pressed in the NDS bearing to take it from 24mm to 22mm. If you can find a NOS TruVativ BB you will see what I mean. The NDS measures 22.1mm.

Exactly, this is all down to SRAM producing a rubbish BB. Plus the GXP standard being a complicated solution to a non-existent problem.

Thanks, Now i know why i got this problem, Poor engineering for sure, I bet the professional riders all use BB30 or alternative bottom brackets. This would something to be expected of a 300USD supermarket bike. Not for a 750USD groupset. Little disappointed in SRAM, Have installed several bikes with RED groupset with BB30, never had any problems.

What would savings be? 50cents for a bearing max, And even if they are so keen to make these savings, Why not produce the bottom bracket to 22.2mm?

Rider of the bike is around 100KG with a slow cadance, which the sram crankset clearly can't handle. I really don't get this, it is so simpel to solve. No weight limit stated at SRAM's product.

(owner of the bike also produces high precision metal parts, so he was suprised, to say atleast, how the construction was made as it did not make sense. He also noted that 22.2mm was an English bearing size, So therefore we tought Crankset was too small)

The sram instructions say if you have play, then remove , regrease and retorque. I've had to do this on gxp installations a number of times, so that might get you more squeeze to stabilize things. Barring that, I would suspect the bb more than the crank. Never seen a crank out of spec.

i suspected the BB also, but luckily i made a mistake and ordered BSA first instead of BB86, So therefore i could compare the BB's, they were both 22.2
If SRAM gets complaints that play comes back every now and then, why note it in the instructions, Just Fix the problem.

The BB in your hand has a dust-cap that the installed BB seems to be missing. I had play in my GXP crankset because I missed installing it once. All was well when I put it back.

Yes, Put a lot of grease between everything, Might stop the sound temporary, And squeezing the crankset tight might solve the play for a while, However this is not something that will last. Maybe you can get it working for a year if you have a smooth pedal stroke and are not too heavy, But seriously, Just making the damn thing fit and it would last a lifetime. Also far less stress on the Drive Side bearing.

Dust cap is removed to make measurements by the way

[CerveloP56Gman]

Echo: Those numbers are way off. There should be no more than a 0.05mm difference between the journal on the crank and the inner race of that bearing, the journal being the smaller of the two, obviously. It is a 7/8" hole, not 22mm. It's some kind of weird vestigial inch standard on an otherwise metric crank design. It's a damn good design though. No pre-load on the bearings makes for a great spin.

[Marin]

The left hand side / NDS bearing is clamped between the left crank arm and the stepped axle, so there can't be any play once properly installed - end of story.

The axle is not clamped or compressed against the right / DS bearing, so this interface has tigher tolerances.

Note how you don't have to 1st set preload and afterwards clamp (á la Shimano) on GXP BBs - the NDS bearing takes all the lateral loads, as the axle is not fixed laterally on the DS.