Here is a one-piece crank and bottom bracket for a child's bicycle made by Roadmaster.

  

This appears to be an ordinary OPC bottom bracket shell, but is about 43mm ID and about 65mm wide, where an ordinary OPC is 51.4mm ID and 68mm wide.

Which begs the question: how does this mini-OPC improve on the standard size? There are no obvious clearance issues where the extra 9mm diameter of an American OPC would be problematic, likewise the extra 3mm of width. And while there is slightly more material cost in the larger shell and bearing races, a usual rule of thumb is doubling manufacturing/sales volume drops cost by 5%. And the volume of American OPC shells, bearings, etc., is huge compared to the volume of this. So despite the higher material cost for American OPC, the cost is likely lower.

The great thing about standards is there's so many of them.

— Unknown

Shoot me now! Shoot me now!

— Daffy Duck

... now offered in all popular thread formats including British, French, Italian, Swiss, Chater-Lea and Stripped.

— Andrew Muzi, circa 1979

We don't need a new 'oversize' bottom bracket standard. We already have an 'oversize' bottom bracket standard, it is called Ashtabula/OPC.

— Chalo Colina

strandard: when your bike uses some wacko-size "proprietary standard" part, it breaks, and (being a wacko size) you cannot get a replacement so you are stranded.

— Pardo

Here are some bottom bracket shell sizes. American Astabula/OPC and ISO are by far the most common, with Mid increasingly common on BMX bicycles.

• “A × B” means thread diameter A and pitch or TPI B.
• “shell C” means shell width of C.
• “D bore” means a smooth bore of nominal diameter D. This is wrong (too simple), because bores also need a tolerance. For example, D±T. In other words, there may be two standards listed here as D but different tolerances mean parts cannot be interchanged.
  • Threads also need specified tolerances, but bike part tolerances usually follow industrial thread standards. When two standards are listed with the same thread A×B, parts can usually be interchanged.
• Many sizes are available in variants, often listed in the same row, for example “road” or “mountain”. Parts may have mixed compatibility. For example, “road” and “mountain” might both be offered in 123 mm width, and may use the same cups and bearings but different axles/spindles.
• Threads may be listed as left-threaded or right-threaded. Threads without “left-” or “right-” are right-threaded.

But wait, there's more...

A few makers have occasionally offered oversize cups to allow repair of damaged bottom brackets. For example, tap with an oversize tap then screw in an oversize cup. Oversize parts of standard dimensions is common practice in the automotive and engine industries, but is rare in bicycles.

Rare, but it happens — here is an oversize cup (photo: Andrew Muzi):

  

Although the above is clearly marked, there are some bottom bracket designs where the threaded parts are of a shape where it is hard to put any clear mark on the (oversize) threaded part.

The ability to fit oversize cups is generally a good thing, but does slightly complicate your life when servicing bottom brackets.

There's also makers who just have poor manufacturing tolerances, but that's not a standard, it's just (too) common.

American/Ashtabula/OPC probably deserves some extra discussion, because it was already wide-spread at the time many of the other sizes above first appeared.

More specifically, American OPC is only rough guidance. It is common that cheap frames with American OPC have a BB shell which is substantially larger or smaller than nominal, or which is badly out-of-round. Sloppy works well enough with heavy “bash to fit” bottom bracket cups and where the bearings can be adjusted to compensate for excessively tight bearing fit.

Sloppy tolerances would work badly for a BB which requires a precise fit. However, you can imagine using a refined American OPC, in which the general dimensions are the same, but specified to tight enough tolerances so light bottom brackets can be used reliably.

Other than self-gratification, why invent a new size, when there is already an existing size?

• One common argument against American OPC is that it is used widely on cheap bicycles, so a “compatible” design that relies on precise sizes will not reliably fit frames made sloppily.

  That is, a precise-fit BB installed in a sloppy-size frame may be damaged in installation, or may go in but make noises or fail after little use, etc.

  Thus, one argument was that using American OPC for “good” bottom brackets would lead to a rash of problems which would be blamed on the bottom bracket but actually caused by “bad” (sloppy-sized) frames. In turn (the argument goes), a new size is needed so that it is impossible to fit a “good” BB in a “bad” frame.

  This seems like a silly argument since:

  • It is easy/cheap to measure fine tolerances during service. For example, using go/no-go gauges. So it is easy enough to figure out if a given frame will work well or poorly with a precise-fit BB.

  • Precise-fit BBs cost more, and people with frames of "bash to fit" quality are unlikely to install expen$ive bottom brackets on their cheap frames. This makes the basic problem much less likely to arise in the real world.

• A more serious consideration is weight, though this probably should be "weight and cost".

  An argument is good cost and weight can be achieved with something larger than ISO's 409sh mm diameter and smaller than OPC's 51-ish mm diamter.

  The overall weight should be considered: a larger shell is heavier, but overall bicycle weight may go down if a heavier frame but lighter crank spindle can be used. A lighter spindle may be possible. For example, an aluminum spindle cannot be used safely in an ISO bottom bracket shell with inboard bearings, but can be used safely where there is space for a larger spindle.

  In some cases a yet-larger bore allows yet-thinner spindle wall, increasing stiffness without hurting weight. Weight, stiffness, and bearing durability are three of the main concerns commonly given as motivations for use of "oversize" bottom brackets.

  While “larger than ISO” may be lighter overall, eventually a shell is so large that no further BB weight savings are possible. An argument against OPC might be that it is too large.

  However, many “oversize” sizes are close to OPC, and there is little more weight for going all the way to OPC.

  Further, weights were easy enough to compute at the time new sizes were being introduce. Thus, there was no need to determine this by experiments done at great cost to bike buyers.

• Reduced weight motivates large size jumps, but many of the standards listed above are nearly the same size -- is there really much difference between 35, 36, and 37 mm sizes? Between 46, 48, and 50 mm sizes?

  An instructive comparison may be the weight differnce between "Mid" (41 mm) and "Spanish" (37 mm) bottom brackets, which are similar except for dimensions. The weight difference is sometimes cited as 100 grams with both using the same spindle diameter, but taking in to account the difference in both bearing and shell weight.

  100 grams may be "a lot", but the larger bearing (123 g) is also more durable than the smaller bearing (79 g). This is 44 grams of the notional weight difference. Meaning the difference in shell weight is 56 grams.

  A pair of spacers of aluminum with 37 mm bore and 41 mm OD and 9 mm wide would weigh 10 grams (both spacers). So if everybody just used Mid shells, then folks wanting lower weight could use smaller bearings and spacers, folks wanting more durability could use bigger bearings and no spacer. A Mid shell with Spanish bearings would weigh more than a Spanish shell and Spanish bearings, but would still save some weight while offering riders more real choices than being locked in to a small fast-wearing Spanish bearing size.

  Also, it is not clear the real weight difference is 100 g. “My expensive ultra-light Spanish shell weighs less than your cheap Mid shell” is a bad argument that Spanish is “lots” lighter.

  Shells vary, so care is needed to ensure shell comparisons are fair. In one comparison the small shell is 112 grams. With a 68 mm-wide shell and steel at 7.85 g/cm³, a 37 mm bore gives roughly 40.5 mm OD thus 3.5 mm wall thickness. For 41 mm bore and the same 3.5 mm wall thickenss, shell weight would be about 125 g or 13 g heavier. This is obviously a simplification as real shells have holes and bearing shoulders, but the comparison lists the difference in shell weights as 54 g while it appears here that similar-quality shells would differ by only about 13 g.

  In turn, a fair total weight comparison between "Mid" and "Spanish" looks like 25 g difference, rather than 100 g. With a titanium, aluminum or carbon-fiber shell, the difference should be even less. Similar considerations apply when looking at American/OPC vs. other shells. For example, a superficial comparison of American OPC and PF30 suggests total weight difference around 20 grams when both use similar bearings.

  Also, using the larger Mid shell with large-bore bearings might allow an aluminum spindle for light riders, leading to lower total weight than the "lighter" Spanish shell.

• Another way to look at this: suppose an ideal BB standard is XYZ mm, but instead we use OPC, which is “too big” and, therefore, “too heavy”. What is the penalty?

  Consider a light ISO BB in a steel shell and then place it in a steel OPC shell using aluminum spacers. This should be about the biggest weight penalty, as you are using the heaviest material and doing nothing to take advantage of weight savings a larger spindle.

  In one comparison, OPC including shell and spacer is about 65 g heavier than ISO. Since ISO is the smallest commonly-used size and OPC is the largest (except for eccentric bottom brackets), it seems fair to conclude that the intrinsic weight penalty for any "too big" shell is at most 65 g.

  In other words, the penalty for using OPC instead of some other oversize strandard should be less than 65 g, often much less.

  This line of reasoning is a bit of a simplification. For example, a wider shell can increase the penalty.

  However, it does suggest where a “slightly too big” standard already exists, the weight peanlty for using it is quite small. And so why invent some new standard?

• As manufacturing sales/volume goes up, cost usually comes down. One way to use this is a “heavy” standard which is made light through better engineering; then made cheap through higher volume.

  In other words, folks building cranks, bottom brackets, and frames have dozens(!) of bottom brackets to consider. That makes it hard to get high volume. For a given cost, the bike will be heavier.

  If we all used either ISO or OPC, then every maker could have higher volumes, lower costs, and could make up some or all of what is lost by OPC being not the “ideal” for their specific market.

  It seems likely that “innovation” in bottom bracket sizes made our bikes worse for all of cost, durability, and weight.

• A “too large” shell makes it possible to fit a range of bearings, which is useful as different riders have different needs.

  Light bearings may give long service for some riders and “long enough” service for racing-oriented cyclists.

  Conversely, some heavy, vigorous, or all-weather cyclists get consistently poor service life from light-weight bottom bracket bearings. Bearing service life often rises much faster than bearing weight, so being able to fit a larger/heavier bearing can be a big win for many cyclists.

  A “too large” shell allows both a small bearing with spacer and a large bearing without spacer.

• It is desirable to fit triple chainring cranks so the smallest sprocket overlaps the shell. That is, the inside bore of the smallest sprocket needs to fit over the shell's outside diameter. A larger shell limits the smallest sprocket which can be fitted.

  With a conventional ISO BB, the shell outside diameter is on the order of 42 mm, allowing sprockets down to 16T (e.g., Mountain Tamer™).

  With an OPC shell, sprockets to down to 24T can be fitted, maybe smaller.

  There are various assumptions here about how sprockets are attached, tolerable clearances between the shell and sprocket, etc. The more general point is allowing very small sprockets can be useful, and a variety of BB shell sizes limit the smallest sprocket size.

  However, most “oversize” sizes have a minimum sprocket size close to the OPC minimum size.

So it appears a lot of standards which differ by a few mm have no particular strength/weight/etc. advantage.

At the same time, a few years retrospective says most “oversize” bottom brackets have been plagued with a mix of noise, looseness, and/or bearings that bind.

• One reason for these problems is loose manufacturing tolerances to save money.

• One common problem is designs which form press-fit bearing bores separately. Even if the bearing bores are the right diameter, making them separately means it is hard to get them aligned, and especially hard to get them aligned cheaply.

• Better is to make both bearing bores using a single through pass, or put the bores in a separate carrier which holds the bearings precisely even when installed in an imprecise shell.

Since the original complaint against OPC was it was too cheap and inaccurate, it seems sad consumers paid for lots of new and problematic ”standards“, only for many consumers to wind up with problems these not-an-OPC designs were supposed to avoid.

A summary, then, is that we probably should have gone straight from ISO to American OPC and only “backed off” to intermediate sizes once the spindle and bearing sizes and alignment and other tolerances were well-established.

More HERE.

https://xkcd.com/927 as of 2019/06:

  

¹ ``Press Fit Bottom Bracket Compatibility Update'', Full Speed Ahead, from http://www.qbp.com/diagrams/TechInfo/FSA/externalbbfitchart.pdf as of 2011/07.

² American OPC/Ashtabula deserves special note because it is made in very high volume at very low cost, and as a result shells and cups tend to be made in a range of sizes due to manufacturing "slop" — often quite far from the nominal size, and also often quite out of round. However, shells and cups are somewhat stretchy, and unlike cartridge bearings used with many press-fit shells, many cheap American OPC use adjustable bearings, allowing compensation for other errors. Finally, typical cheap units use large-diameter ball bearings that are relatively insensitive to adjustment, contamination, and damage. Thus, in practice they give good service despite low cost.

³ Sutherland's Handboodk for Bicycle Mechanics, 6th Edition.

⁴ Crank bearings carry a "precessing" load that tends to twist the cup under pedaling load. Left cups are right-threaded so they self-tighten. Most right cups are left-threaded so they self-tighten. Right-threaded right cups tend to self-unscrew. French-standard and Italian-standard right cups are the most common example of right-hand threaded right cups, and have a reputation for unscrewing. Right-threaded right cups are often installed very tight to resist unscrewing, which can make them very hard to remove for service. Press-in cups are usually specified to a tight-enough press fit so twisting is not a problem. Bearings with a loose press fit are prone to move and thus make noises. Bearing movement can wear/damage the frame's bearing holes, making them looser.

⁵ See http://www.bike-eu.com/news/3532/thun-offers-bb30-for-low-cost-mtbs.html as of 2011/07.

⁶ The Mavic/Stronglight conical seat requires a conical seat in the frame. A special milling cutter can be used to retrofit many existing frames, including those with ISO and Klein bottom brackets. Retrofit is sometimes used to deal with frame damage that prevents use of of a conventional threaded or press-in bottom bracket.

⁷ It is unclear why Italian uses mixed inch/metric dimensions. Chalo's hypothesis (which seems quite likely) is the dimensions were designed for metric diameters and manufactured using older inch-dimension lathes, which were not capable of cutting metric-pitch threads.

⁸ Italian is about 1 mm larger diameter than ISO and has the same thread pitch (24 TPI). Thus, stripped ISO left cups bottom bracket shells sometimes be repaired by tapping them oversize to Italian then fitting an Italian-thread left cup. However, the right cups are not directly compatible because ISO is left-threaded while Italian is right-threaded, so more metal must be removed. However, given a stripped frame, such over-tapping may be an acceptable risk. Note that using an Italian right cup introduces the problem of self-unscrewing, which may lead to further frame damage.

⁹ Note "brand name" and "size" are not the same thing. For example, Chater-Lea and Raleigh both made bottom brackets in their "named" sizes and also other standard sizes. Conversely, many companies have made Raleigh-size bottom brackets. In other words, your "Raleigh" bicycle may have an Italian-thread bottom bracket or your non-Raleigh frame may take a Raleigh-size bottom bracket; and a Raleigh-brand bottom bracket in your Raleigh-brand frame may use ISO threading.

¹⁰ Listed HERE as of 2014/08 as "Custom SRAM GXP, 84.5mm". Listed HERE as might be possible to fit a triple crank in place of the factory double using Shimano SM-BB91-42 adapter, but according to Specialized "I would suggest checking with Shimano and Wheels Manufacturing. Sorry, we don't have that adapter." Specialized "COMPATIBILITY GUIDE SPECIALIZED OSBB / CRANK COMPATIBILITY - MOUNTAIN FRAMES" (document CG0308 Rev.C February 2011; HERE as of 2014/08) states Shimano Hollowtech II cranks may be fitted using the above adapter, SRAM/Truvative GXP cranks using "SRAM BB GXP press-in threaded cups (#00.6415.033.040)", or Specialized carbon cranks using "circlips". It further states that the alloy OSBB shell was used for 2010-11 Stumpjumper FSR carbon models, 2009-1011 Era carbon models, 2009-1020 Epic carbon models, and 2009-2010 stumpjumper HT carbon models. Presumably "carbon model" means the frame is mainly carbon fiber. Unclear what are "press-in threaded cups". Document lists bearings as "6808 series", but standard 6808 bearings are 40 mm ID and 52 mm OD, this is probably an error and "6806 series" is intended, with 30 mm ID, 42 mm OD, and 7 mm width.

¹¹ BBright drawings HERE as of 2014/08.

¹² "T47 Bottom Bracket", http://www.paragonmachineworks.com/images/Drawings/T47_BB_SHEET_A1-2.pdf as of 2016/02/29.

¹³ The complete guide to bottom bracket standards, BikeRadar, 2019/03/14. https://www.bikeradar.com/advice/buyers-guides/the-complete-guide-to-bottom-bracket-standards as of 2019/06.

¹⁴ SRAM DUB Will Replace 2 Current Crank Standards With A New One, Wil Barret, Singletrack, 2018/01/016. https://singletrackworld.com/2018/01/sram-dub-replaces-2-current-bb-standards-with-a-new-one as of 2019/06.

¹⁵ Trek confirms use of T47 threaded bottom brackets — but with a twist, James Huang; 2019/06/06. https://cyclingtips.com/2019/06/trek-t47-threaded-bottom-brackets as of 2021/09.

¹⁶ Original Thompson/Thun uses a threaded axle and one cone that adjusts. Some 2020-vintage Thompson/Thun replacements use a plain axle with press-on cartridge bearings, then shouldered plastic cups to reach from the bearing O.D. to the shell I.D. To install, remove one cup, press in the assembly, then press in the other cup — during pressing, it slides over both the bearing O.D. and cup I.D.

¹⁷ “French Backwards” is not a common/standard name. It seems people who work on those bikes are familiar with the threading, but so far, there is no common name for it.

¹⁸ [https://restoringvintagebicycles.com/tag/stronglight-crankset as of 2024/01/04]. “Older French bicycles sometimes featured reverse threading for all the BB components, as on my 1947 Camille Daudon.”. “1947 Stronglight crankset with Rosa rings and reverse threading on the non drive side crank bolt, lockring, and BB cup.” The wording suggests French-Backwards was used on several brands, not just Camille Daudon. French-Backwards requires parts from at least one bottom bracket maker. In this case, Stronglight, as there were not other square taper crank or bottom bracket makers selling parts at the time. It would also require tool makers to produce special left-hand 35.0×1.0 thread taps — the size 35.0×1.0 was rare outside the bicycle industry, and Swiss-standard threading had not yet been introduced, so there were not yet left-hand taps for Swiss-standard..

¹⁹ [https://velobase.com/ViewComponent.aspx?ID=A11E7A88-5140-40F9-B45F-DBB1967AEF68&Enum=119&AbsPos=6 as of 2024-01-05]. oldschoolfool Vintage User on 08/09/11: “I have one threaded 1.37 x 24F. The 1.37 x 24 does not interchange because although they are the same pitch and diameter they have different thread flank angles.”.

BBright, Campagnolo, Cannondale, Fat Chance, FSA, Gary Fisher, Klein, ISIS, Merlin, Mavic, Phil Wood, Raleigh, Ritchey, Roadmaster, Shimano, Specialized, SRAM, Stronglight, Thompson, Thun, Trek, and any other trade names are trademarks owned by their respective owners.