Cycling UK Forum

Jezrant wrote: ↑12 May 2024, 3:11pm If you’re worried about fitting a new headset on a vintage Italian steel fork, you can run a die down the steerer threads

Jezrant wrote: ↑12 May 2024, 3:11pm it’s now virtually impossible to find vintage Record headsets in good condition

cyclop wrote: ↑12 May 2024, 6:51am Is this an intellectual query or are you looking at buying something?

cyclop wrote: ↑12 May 2024, 6:51am For what it,s worth,amonst my old steel bikes is an Italian framed machine which does have Italian threads at the bottom bracket.

.  _____________ iso_60
. / ____________ bsc_60
. |/ ___________ ita_55
. ||/
. |||    _______ iso_60
. |||   / ______ bsc_60
. |||   |/ _____ ita_55
. |||   ||/
. |||   |||
. 001   001
. 001   010
. 001   100
. 010   010
. 010   001
. 010   100
. 100   100
. 100   001
. 100   010

 fork__ | cups__ | match
--------------------------------------------------
 ita_55 | ita_55 | perfect
 ita_55 | bsc_60 | a bit loosely
 ita_55 | iso_60 | ???
--------------------------------------------------
 bsc_60 | bsc_60 | perfect
 bsc_60 | ita_55 | too tight to easily thread on
 bsc_60 | iso_60 | perfect
--------------------------------------------------
 iso_60 | iso_60 | perfect
 iso_60 | ita_55 | just fine
 iso_60 | bsc_60 | perfect
--------------------------------------------------

• Campagnolo 1"/French <------ I don't consider it
• Campagnolo 1"/Italian: { 25.4mm 24TPI, 55 thread degree, cup diameter 30.2mm, crown race 26.4mm }
• Campagnolo 1"/English: { 1" 24TPI, 60 thread degree, cup diameter 30.2mm, crown race 26.4mm }

• ISO 1": { 1" 24TPI, 60 degree thread-angle, cup diameter 30.0mm, crown race 26.4mm }
• Campagnolo (legacy) 1"/English: { 1" 24TPI, 60 degree thread-angle, cup diameter 30.2mm, crown race 26.4mm }

Brucey wrote: ↑10 May 2024, 8:51pm FWIW stronglight A9, stronglight delta, red 'S' headsets etc (and a few others) use the same bearing innards and as a result they are (IMHO) all crap in exactly the same way. That is, the cone angle the rollers are set to is very low. This means it is usual for at least one element in the lower bearing to be forced off-centre eg. by braking forces, but it is unusual for those elements to return to the correct position afterwards.

Brucey wrote: ↑10 May 2024, 8:51pm Some roller bearing headsets use a more intelligent roller angle but because of this, the rollers are often a bit shorter, which makes for a lower load rating, defeating to whole objective. I think the current crop from tange are like this.

Brucey wrote: ↑10 May 2024, 8:51pm Whilst I am sure that roller bearing bicycle headsets have some happy (and frankly somewhat deluded) users the truth is that most them are badly done, and furthermore they are not really necessary anyway. Show me a damaged ball-bearing headset and I'll show you a bike that was never set up correctly, no exceptions.

I am no great lover of cartridge bearings per se, but pragmatically speaking, that is what most bikes will come with these days. Also, it is always possible to rebuild a standard hoffmann type bearing as a full complement type , which is usually nearly doubles the load rating and makes them much better suited to cycle use, having (amongst other things) a far more graceful failure mode. In the case of headset bearings, it is simply a case of largely removing the second (upper) lip on the inner race before rebuilding the bearing with loose balls and no clip. Other bearings are best converted into DFC spec bearings, which is readily done using a dremel tool or similar.

• has sealed bearings, which I don't know... I don't really appreciate
• has a strange mechanism, with a "collar" that screws onto the fork

Brucey wrote: ↑10 May 2024, 6:01pm however, ISO headsets are usually 30.0 mm where they go into the frame whereas campag headsets are usually 30.2 mm. The usual thing is to have a 0.2 mm interference fit. This means that whilst it is usually possible to install ISO cups into a frame prepped for campagnolo it usually goes a lot better if a little adhesive eg 'bearing fit' is also used.

Brucey wrote: ↑10 May 2024, 6:01pm I suspect the 'zinc plated finish' is just a typo or something. This is not how Al is usually finished.

531colin wrote: ↑10 May 2024, 8:20pm I’ve used A9s, I’ve used one or two Ritchey 1 1/8 headsets which were rollers bottom and balls top, and an “orbit” headset 1” threaded which was rollers top and bottom

• IRD RollerDrive, balls top, rollers bottom, see here
• IRD Double RollerDrive, rollers top, rollers bottom, see here

531colin wrote: ↑10 May 2024, 8:02pm I think if you have got one of these headsets, fit it and see if you like it. I wouldn’t run screaming away from one, but I wouldn’t jump through hoops to get one, either.

• from a shop located in the UK, ~90UKP
• from a shop located in Switzerland, ~100 CHS

rjb wrote: ↑10 May 2024, 5:44pm Good luck with trying to find a rudelli headset.

• Stronglight A9
• Stronglight Delta
• Tange/IRD RollerDrive
• Rudelli
• ...

"roller" bearing is generally the preferred term, so I'll use it from here on. They aren't gone, but they aren't very common -- actually, they never really were that common to begin with. Nowadays, with so many threadless headsets out there, and excellent sealed units available, they are pretty much forgotten. They aren't the perfect headset in any way, but they are serviceable, last a long, long time and I like them.

The thing about ball bearings as they are used in headsets is that they are excellent at reducing the friction of the fork rotating in the frame, but not as great at handling loads. The fork of a bike (as I've described in an earlier post) is like a "front line of assault." Every bump, large or small, travels directly up the fork, transmitting shock through the headset. Ball bearings have very little contact area with their races (which is why they are so good at reducing friction), but that small contact area also means those forces are concentrated on a much smaller surface area. Over time, the bearing races can become dimpled or pitted, which leads to "notchy" steering.

Roller bearings have a larger contact area, which makes them better at handling loads. On the other hand, critics of roller bearing headsets will point out that they have more friction (I'll explain why later in the post. Read on.) -- they don't allow the fork to turn quite as freely as ball bearings. Then again, in actual riding, forks don't really move that much -- bikes turn or change direction more from leaning than by "steering." In the small amount that forks actually rotate in the frame, roller bearings work fine.

Another benefit of roller bearing headsets is they are believed to add a bit of stability to some bikes. Occasionally, some bicycles will experience something referred to as "shimmy." There are different explanations (and some disagreements) about exactly what causes shimmy, but it is essentially a rhythmic, almost resonant shaking of a bicycle that one can feel through the steering. It seems to come from a combination of the gyroscopic action of the wheels and perhaps an oscillation or twisting in the frame. The steering action twists the frame one direction, then that energy is momentarily stored then released, twisting the frame back the other direction, then back again. In some cases, it can also be exacerbated by loading a bike with luggage/bags. Some bikes that feel fine with hands on the bars can shimmy badly when riding no-hands. I should point out that if a bike that never had a shimmy problem suddenly develops one, it could be a sign that something is actually wrong -- loose spokes, loose headset, cracked frame or fork -- so if it happens, definitely investigate, but be aware that sometimes it just happens.

Long life, and low friction
Roller bearing headsets don't move quite as freely as ball bearing designs. This has to do with the fact that roller bearings are like tiny cylinders. When they roll, they want to roll in a straight line. That makes them work really well in bottom brackets and pedals or other places where they can follow a straight-line path around a more or less straight or cylindrical axle, yet as ideal as they are for those applications, their use is still fairly uncommon except in more expensive examples. But when put into use in a headset, they have to follow a curved path.

One way to help that is by setting the bearings up at an angle. The bearing races and the bearing retainer are set up in almost a conical arrangement around the bicycle's head-tube and fork steerer. That helps, but the bearings still have to travel in a bit of a curved line, and they have a certain amount of resistance to it. Given that a headset doesn't rotate very much in actual use, as I pointed out already, it isn't really that bad of a trade-off: Slightly more friction (in a component that isn't hurt so much by a bit of friction), in exchange for really long bearing life. One thing that would probably make them work even better would be to make the roller bearings tapered -- thicker at one end than the other -- but making precision tapered roller bearings is pretty expensive, so the added cost would be hard to justify for the tiny decrease in friction.

Roller bearing headsets aren't perfect or revolutionary in any way. If they did completely disappear, I'm sure we'd all survive. Many riders would never even notice. But I'll still hang on to the ones I have, and keep some spare parts handy in the unlikely event that one of them ever needs to be rebuilt. Who knows, that 12-year-old Stronglight Delta I have might need a rebuild some 15 years from now. I want to be ready.

• roller bearings on the lower part
• ball bearings on the upper part.

Brucey wrote: ↑9 May 2024, 8:48pm FWIW the ISO 60 deg thread is normally considered to be identical to the campagnolo 60 deg thread.

• ISO standard
• crown race = 26.4
• Stack height = 35.9mm
• Weight = 110g

• Campagnolo Italian threading, 25.4mm 24TPI, 55 degree thread-angle
• Campagnolo English threading, 1" 24TPI, 60 degree thread-angle

• ISO 1 inc threading, 1" 24TPI, 60 degree thread-angle
• Campagnolo English threading, 1" 24TPI, 60 degree thread-angle

• it's exactly the same 1" threading just called with two different names?
• there are differences, e.g. the internal diameter is it slightly different?