TallRider

I've been working on two 15-20 year old steel frames with threaded 1" headsets (yes, my builds are never-ending right now).
And I've noticed that there's more variability than I'd like to see in the diameter of headset pieces that get pressed into frames and forks.

I've got a crown race that slides into place with very little pressure (I can seat it all the way down by hand). It doesn't have play, but it goes into place too easily for me to be confident that play won't develop. Two other crown races of the same official dimension seat with a more normal level of force required.

I've also got one headset cup (the top race) that slides in nearly all the way by hand, although again I have to exert some force - just not nearly as much as I'd expect. It also goes into the bottom of the head tube just as easily, so I'm confident the difference is between the two cups, not between top and bottom of the head tube. The bottom race from the same headset goes in with normal level of force.

First, how normal is this unacceptable level of variability in machined tolerances of press-fit headset parts?

Secondly, what can I do about it? I've heard of some form of loctite which are designed to make press-fit interfaces tighter. Is that the best option here?

DannoXYZ 

1) My guess are these parts were forced in at an angle some time in the past, thus squeezing the cups down and expanding the crown race and headtube.

2) There are knurling tools you can use to increase the interference fit...

TallRider

Thanks Danno.

1) both problematic pieces are from new headsets
2) probably cheaper to just buy a new headset than use a knurling tool here

well biked

In marginal situations, the Park site recommends Loctite #RC609. I don't know if you saw the discussion about a headset problem in the C and V forum recently where an Italian frame was being built up and it was found that the headset that was being used was a JIS headset, which meant the crown race was too large (ID) for the fork and the cups were too small for the headtube (ID), the cups could be pressed in by hand. In that case the consensus was to use an ISO headset. In your situations, if you're confident you're using headsets of the correct standard (ISO, JIS, etc) then you could try the Loctite and see if it holds the parts securely. There can be variation from one "ISO headtube" and another, for example, a range of a couple tenths of a millimeter that's acceptable. It may be that you're experiencing a sort of "worst case scenario" within the various ranges-

TallRider

I'll try to get ahold of that loctite. I'm pretty sure the variability here is in the headset cup from one headset and the crown race from another; all of the other pieces for all of the other headsets I've got sitting around in the same standard (J.I.S.) all seem as if they fit properly.

DannoXYZ 

I suspect that your frame may be out of spec. Previous headset installations probably expanded the headtube. Multiple uses of the crown-race facing tool may have removed a little too much from the fork as well. Measure the frame, fork and headset parts and compare to the specs given by Park Tool's Headset Standards Summary. You might find that your frame's dimensions are on the large-clearance end of the tolerance ranges.

It appears in your case that there's a snug fit with no free-play; just not a tight interference fit. So some Loctite should be fine.

operator

I just took a beer can and shimmed it. I wouldn't do this on a expensive headset though.

TallRider

The reason that I think this is not the case (my frame being out of spec) is because all of the other JIS headset cups that I have seem to fit properly, there's only one cup out of three JIS headsets (so, one of six cups) that slides right in. Only one crown race slides right on, too - the other two seem as if they fit normally. That said, it certainly seems from a manufacturing standpoint that variability is more likely to come from the head tube and fork crown than from the headset parts. But the headset parts themselves clearly show variability.

Wouldn't work here - that's way more thickness than I'd need to make up in this case.

well biked

One other possibility. It could be that you're assuming that a headset you pulled off the bike has JIS headset cups because it uses a 27.0mm (JIS) crown race. It's fairly common for an otherwise ISO headset to use a 27.0mm crown race ID I believe, I learned that during an earlier discussion on the forum. Same for the frame/fork of course, a 27.0mm crown race seat can be used for the fork while the headtube takes the ISO size cups. So it might be that what you're assuming to be a JIS headset, in regard to the original headset on the bike, is really an ISO headset that uses a 27.0mm crown race and you bought the new JIS headset under the assumption that the headset had to be JIS-sized for the cups, too. Just a thought.........Of course, if you're saying you have more than one new JIS headset (that you're certain is JIS) and some fit fine and only one doesn't, then this wouldn't make sense-

edit: An example of an '80's Japanese bike using ISO cups is my '83 Centurion Pro Tour, the Tange headset cups are stamped "30.2" which is the ISO size. As for the crown race ID on that bike, I don't know if it's 27.0mm or 26.4mm because it's still on the fork and I really don't want to remove it, but I wouldn't be at all surprised if it's 27.0mm. I'm just bringing this up to point out that just because it's an '80's Japanese bike does not necessarily mean it uses JIS headset cups-

Philatio

I'm not sure exactly what the deal with the "shimming loctite" is, but you might be able to use JB-Weld? It would kind of make it permanent, but I suspect the loctite thing would be too.

TallRider

well biked: all of the cups in question are officially 30.0mm (JIS) and only one of the six (from three headsets) slides in too easily, which implies it's smaller than spec - so no chance it was the wrong spec (since ISO is 30.2, which would be too tight.
Also, the crown races are all 27.0, and the difference between 26.4 and 27.0 is pretty easy to tell. (Now that I know it's a difference; a year ago I learned the hard way when I tried to install a 26.4mm crown race on a 27.0mm fork crown, and it didn't go. Not even close. That's when I learned there were two major standards for 1" threaded headsets.)

Philatio: I agree that any form of epoxy rated for use on metal would work here for shimming action. I'll probably just end up using that.

well biked

I see now, you've got six known JIS cups and only one cup is a problem. One of the conditions for my above "theory" to have any merit was that you were dealing with an original headset and you didn't know for sure which size the cups were. Oh well-

TallRider

was worth asking