Bruce27

Hello,
I'm cleaning up and re-greasing a recent acquisition. It's a bike from the late 1970's. When cleaning the headset I noticed that the Crown-Race on the fork is a little loose.
Most headsets I've re-greased in the past had a Crown-Race that basically had a tight friction fit on the top of the fork crown. This one is slightly loose and can be moved from side to side. I think there's about 0.003"-0.005" of side to side movement.
Is this ok or do I need a tighter fitting Crown-Race for the fork before I reassemble?
Thanks

_ForceD_

Some have said that the race has to fit tightly onto the steer tube. But I recently overhauled an old head set and had the same issue. I cleaned and reassembled it, and it’s working fine. It just seems to me that as tight as the assembly is that it gets held in place sufficiently. But, maybe I’m wrong. I was also wondering if that space could be eliminate with something like JB Weld (and then sufficiently cleaned off the bearing interface).

Edited to add: The above reassembled head set doesn’t seem sloppy, and there’s no ‘clunking’ when riding the bike. Actually, the steering is smooth and improved after all the adjustments.

Dan

Bruce27

Thanks Dan. That's kinda what I was hoping for. If everything just tightens up, maybe it will be fine.
I was debating whether to have a shop install a new headset. Not sure what the long term effect would be with it being loose.

'02 nrs

Iride01 

Maybe some of the frame builders or better experienced mechanics in all things bicycle here should weigh in with their knowledge, but I don't think the loose race is a issue when the headset is all put together and properly adjusted. Though you might need to check things that might have it sitting askew or such. But if it's just the fraction of a ten thousandth of a inch less than what a more proper interference fit should have been, then I wouldn't be worried

Bruce27

I have a backup headset that I tried and checked the fit. The Crown-Race with this new headset would need to be pressed on...so it would be tight. I've only encountered tight pressed fits in the past.

Evidently the fella in the above video thought it was something that needed to be corrected.

Hopefully more will chime-in if a loose Crown-Race is ok.

smd4

We used to dimple the crown race sleeve with a punch at 90 degrees. Does the same thing as that knurler in the video.

I can't comment on whether it's OK for it to be loose or not. My gut says that it being loose defeats the purpose. I prefer a tight fit, personally.

Hondo6

If I recall correctly, crown races for threaded headsets are designed to be a press fit, requiring a driver to install and a tool to remove (though the latter can be done by other means as well).

I kinda wonder if you don't have a JIS crown race on an ISO fork. The former is 27mm ID; the latter, 26.4mm - a difference of 0.6mm, or approx 0.0236".

I'd beg/borrow/purchase a set of calipers (digital ones are both cheap and generally quite accurate) and check the seat and crown race IDs.

Andrew R Stewart 

Minor headset slop, whether from loosely adjusted bearing preload or a sliding fit of the crown race will have very little effects on riding (the fact that this bike had been used for so long before with this sloppy race fit is proof) other than a bit of knock over bumps.

I've clean and greased more than a few headsets that have had a sloppy fitting race with no issues the riders later noted, or I did on the after service test ride. Is this "best practice"? No but it is a "workable practice". The usual case of this have been on American made bikes that used Wald type headsets (and what we usually refer to as "BMX" dimensional standards). What I have had issues with is using soda can shims when the gaps were large. The Al shims can erode and mush about over the miles and fall out of place in time. If the slop is small (and at .003-.005" of gap it is small) using a retaining compound (Locktite makes their 600 series of retaining compounds in differing grades for various amounts of slop) or dimpling/knurling the race seat's shoulder surface to expand it should be good. Andy

Bruce27

You are right Hondo6 . I double checked and my estimate of a 0.003"-0.005" gap was wrong. I checked and the fork seat is 27mm and the ID of the crown race is 26.4mm.

Bruce27

Thanks Andy. Turns out I have a lot more slop than my original estimate. Slop is more like 0.0236". With this new info, do you think a new crown race is in order?

Hondo6

(chuckling) Then it looks like we were both wrong this morning, amigo. Check out my "error correction" above.

Bruce27

Thanks Hondo6 . but not 100% sure I understand. I thought you were right in that the slop is about 0.0236" or 0.6mm. Way more than my original estimate.

You've helped me narrow this down to the question... is 0.0236" (0.6mm) slop considered too much slop in the crown race to fork seat assembly?

Hondo6

I misread my mm-to-inches calculation of 0.6/25.4 as 0.00236" - not 0.0236" - thus leading to my original comment above. When I read Andy's later comment, I checked and realized I'd made the error. Hence my later edit and strikeout correction to my first comment above.

In this case, by sheer coincidence our two errors "cancelled each other out" exactly and gave the correct result.

Regarding roughly 0.0236" or 0.6mm being too much slop . . . I'll defer to Andrew on that. He has far more practical experience in this area than I do.

Best of luck with the bike. If you decide to replace the headset and your existing spare won't fit because of excess stack height, Tange still makes some nice and long-lasting 1" headsets, including some with low (<34mm) stack height.

Andrew R Stewart 

Has anyone measured the crown seat diameter of the fork? It's beginning to seem that you might have a 27.0mm race fitted to a 26.4mm crown seat diameter. If so then the obvious would be to either replace the crown race with one that is a 26.4 dimension AND works with the rest of the bearing (lower cup and balls) OR replace the complete headset with one designed for that 26.4 fork (and also fits the head tube and stack heights).

Ideally no slop should exist between bearing races/cups and their seats in/on the frame/fork. My earlier post was more to there being some bikes that are made to such a poor tolerance control that best practice fixes are impossible without basically replacing the bike. In these cases knowing what's "good enough", is safe and is agreed to by the rider is key if your a shop.

There's a lot of missing info about this bike/fork/headset that I would want a "hands on" before making any specific suggestions as THE solution. We don't even know the brand, country of manufacturing or headset dimensional standards this bike has (as some examples). Andy

grumpus

You don't need a knurler, you can just raise the surface with lots of taps with a centre punch or small cold chisel. But in this situation I'd reach for the Loctite 620 retaining compound.

Bruce27

Thanks Andy. Makes sense. Plus I've added confusion by transposing a couple of critical details.

I'll try again....

The bike is a late 70's Trek with Ishiwata 022 tubing
The crown race has an ID of 27mm
The crown seat OD on the fork is 26.4mm
So there's 0.6mm (0.0236") of slop trying to use the existing headset parts.

Somewhere along the way it appears the wrong crown race was added to this bike. It also looks like it has a different finish than the rest of the headset. It has a more stain grainy finish while the rest of the headset is a high gloss chrome.
The backup headset I have is the Tange shown below.

With the great input I've received, I'm thinking the best way to resolve this is to have the new headset installed and be done with it....

Andrew R Stewart 

"'m thinking the best way to resolve this is to have the new headset installed and be done with it...." Bruce

YES! Undoubtedly the fork or headset was replaced with other then spec at some point. However crown races being nearly totally hidden from view often don't get the same surface finishing that the cups or threaded top race do. That .6mm is a huge gap to span with retaining compounds, loose shims and/or dimpling. Andy

Garthr

If the rest of the headset bearing surfaces are still good you can get Tange Levin CDS 26.4 crown races for around $5. https://www.benscycle.com/tange-seik...-2421-hd1204/p

Can you mix brands of crown races ? Possibly yes, "it depends" on the particulars, but these crown races made in Asia/Japan and such are not items that vary much in terms of dimensions, regardless of the era they're from. It's likely the Rockhopper came with a Tange manufactured headset.

To seat the crown rack you can use a steel pipe or even pvc pipe and a mallet, just wide enough to slide over the steerer and rest on inside/top/middle of the crown race. Grease the fork lightly. Just make sure it's not resting on the bearing surface. Sure they make special tools for that but I've never used on in many decades of replacing them.

Bruce27

Thanks Garthr . That's good to know. I have another bike with an issue that could possibly use that part. I do have a few tools but no headset tools, correct size PVC or pipes so I'm going to play this safe and have a shop do the job. Hopefully the cost will be under $50.
Ah...the fun & surprises acquiring old stuff.

choddo

The bit you struck through was correct though? Or did you correct the decimal place error AND strike it through?

Hondo6

I left the original text as it was originally written and struck it through. Here's a step-by-step, best as I can reconstruct it from memory and the comments above.

1. I noted the OP's original estimate of 0.003 to 0.005" in headset slop. For some reason, I decided to see if that corresponded with a gap of 0.6mm (ISO/JIS crown race mismatch difference).

2. I converted 0.6mm to inches. The actual number I got, to 3 significant digits, was 0.0236" - which is what I wrote in the comment.

3. Unfortunately, I mentally misread that calculated value as 0.00236" - 10x smaller than the calculated value. I'd guess that I focused on the 3 most significant digits vice noting the decimal place correctly.

4. I compared this mentally-misread erroneous value (0.00236") to the OP's initial estimate of 0.003" to 0.005" slop. The numbers seemed close, so I then wrote my initial comment conjecturing that the issue might be a JIS/ISO mismatch - but I wrote the correct value (0.0236") in that comment vice my mentally-misread value. My analysis was thus "erroneous" due to my mental misreading of the original mm-to-inches calculation as 0.00236" (rounded to 3 significant digits) vice the correct calculated value (0.0236", also rounded to 3 significant digits).

5. Andrew's comment a bit later about Red Loctite for some reason prompted me to re-check my numbers. (I thought I remembered that Red Loctite would only fill gaps considerably smaller than 0.6mm.) This in turn caused me to (a) realize my mental decimal point placement error, which invalidated my original analysis of the problem; (b) do the strikethru edit on my original comment, and (c) identify the reason for the edit.

6. However, it turns out that apparently the OP made a similar error in his original measurement or estimate of his headset slop, understating the magnitude of the gap by close to 10x (also I'd guess due to a decimal point error, but maybe not). It appears that he later more accurately measured the gap and found that it did indeed appear to be a JIS/ISO headset crown race mismatch, and that the actual gap was around 0.6mm (0.0236"). Our errors thus cancelled each other out, and what I initially wrote in the comment above (0.0236") turned out to be correct for the approximate gap when it was more accurately measured later.

Pure luck/coincidence? You betcha! But in this case, apparently that bit of pure luck/coincidence ended up helping ID the problem.

"All's well that ends well", I guess.

choddo

Thank you for your comprehensive analysis. The auditors have now approved this one.