Wrench report: replacing steering head bearings in a '94 VFR

This is mostly a description, but I've thrown in some commentary at
the end:

To start: my VFR was exhibiting some poor handling, primarily a bit
of headshake at around 35mph. Other symptoms showed up as well
(feeling like it was flopping into corners, vague front end, bike
felt like it was hinged in the middle), but only several thousand
miles down the road.

I narrowed my suspects list to the steering head bearings, front
wheel bearings, or possibly the rear shock. For various reasons the
steering head bearings were top on my list, so I elected to replace
them.

To that end, I purchased:
upper and lower dust seals
OEM upper bearing set (retained ball)
OEM lower beaerings set (retained ball)
Honda's steering stem socket
Honda's bearing driver tool
2 Armstrong pin spanners

Note that I elected to use Honda's ball bearings rather than the
tapered rollers available from All Balls Racing. I'll discuss this
decision a bit at the end. I purchased the socket and bearing driver
because I'm without a grinder, so I can't make tools. The critical
dimensions of the steering stem socket will be included at the end of
this report for those who would rather make their own.

Generally, I followed the procedures laid out in the Honda FSM and
CSM. This report is not meant to substitute for those documents.

"A Day at the Races"

Part I: remove your bearings and step away from the VFR!

I hoisted my VFR's front end by strapping it to the rafters in my
garage. The straps were anchored to the frame directly behind the
front engine-mount bolts. This, of course, required the removal of
the left and right fairings. When I dropped the front wheel, I
checked the bearings for axial and radial play, and found none. I
removed the fork legs, and let the grips lay where they were.

The top nut is 30mm, and I broke that loose. The top bridge contains
the ignition switch, and is tethered to the bike unless you remove
it. I don't have a Torx bit that large, so I left it hanging as well.
As you might expect, it's pretty crowded around the steering stem by
now.

bend the tabs down on the locking washer, and remove the locknut and
washer. Before removing the adjuster nut, disconnect your horn wires!
Remove the adjuster nut and lower the steering stem out of the bike.

to remove the bearing race from the steering stem takes patience and
a bit of care. I scarred up the seating surface on the cbr600
steering stem [1] that I practiced on. Best practice, so far as I
know, uses chisels in varying thicknesses to slowly wedge the race
off, then careful work with a punch. I scarred the lower seat
slightly, so ground the scars flat with a hand stone before continuing.

Punching the bearing races out of the frame is similar, but Honda
machined two slots so that you can get to two sides of the race with
a long punch. Plenty of tapping, alternating sides and varying
location, popped both outer races free. Save these for later, you'll
need them.

With all races removed, I re-assembled the bearing sets and worked
them. While my upper bearing was quite smooth, the lower bearing was
obviously notchy. I had not noticed _any_ notchiness in the steering
of the VFR, and I presume that this is because when the front wheel
got off the ground it unloaded the lower bearing. This is
speculation, however.

Part II: Driving races

I installed the lower inner race (goes onto the steering stem) first.
The dust seal goes down onto the stem (clean the seat first), then
the inner race gets driven onto it. Honda's steering stem driver
makes this job a snap. No need to grind the inside of your old race
and use it as a driver, simply place bearing race on the stem, put
the driver over the top, and hammer it down until it's fully seated
all around. That was easy.

then I installed the upper outer race into the frame. Again, this is
straightforward with Honda's driver, but this time the old outer race
is used as an adapter on the driver (which otherwise would be bearing
on the bearing surface itself). Again, driving the race is a breeze.
However, when the new race is fully seated, the old race will be
stuck in the hole. I reached in from the other end with a rubber-
coated steel rod (though a chunk of dowel would probably do just as
well) and tapped it once with the hammer, and it popped right out.

The lower outer race is a bit more work. In part this is due to it's
location; it's awkward to reach with the bike on the ground. [2] It's
also partly due to the fact that Honda's driver stem is too small to
engage the bearing race. The OD of this race is 55mm. I ended up
using a combination of an improvised driver adapter (a nylon roller)
and a steel punch to get this race in. I used the old race as a
driver surface. Again, when the new race was fully seated, the old
race required removal by tapping it lightly with a rubber-coated rod.

Part III: Putting it all together.

Here's where it all came apart for me. So far, everything had gone
like a dream. I greased up the bearings, inserted the lower bearing
set on the inner race, and re-installed the lower triple. Then I
greased the upper bearing set, slid it over the triple, then
installed the upper race, then the dust seal. It's time now for the
adjuster nut and lock nut.

Honda's CSM tells you to torque the steering stem to 25nm and then to
rotate the steering stem a few times to make sure everything moves
freely. At 25nm however, the steering stem was _quite_ difficult to
rotate at all. This was not helped by the fact that the adjuster nut
wanted to tighten more when I rotated the stem to the left, and
loosen itself when I rotated the stem to the right.

I thought that this behavior might be stopped by installing the
locknut, so I installed the locknut as described in the CSM, using
the two Armstrong pin spanners. These tools are quite irritating, and
given a second chance I wouldn't have bought them at all.[3] Still, I
managed to get the locknut on as the manual indicated. No luck
though, since the whole unit turned instead of just the adjuster. I
then thought that I needed to jam the locknut on harder. That didn't
work either. then I thought that maybe it was some grease on the
threads, so I removed it and cleaned off the threads. No luck again.

At this point, I thought that I'd just torque up the bearings to 25nm
and install everything, then check it. So, I installed the top bridge
and the fork legs, and the front wheel. At this point, it's easier to
install the brake calipers than to leave them hanging, so I put them
back also. I torqued down the steering stem nut and...I can't turn
the bars. It's _way_ too tight. I repeat this a couple of times, each
time checking to make sure I've got everything in and in the correct
order. I also start backing off the torque, from 25 to 20 to 15nm.
It's always _way_ too tight.

One last try this night (for it has taken me about 8 hours to get
this far): I tighten up the adjuster nut by hand, put the locknut on
and tighten it up, then bolt everything up. It's stiff, but maybe it
will work itself out. Heck, maybe it's loose enough to be ok. So I
button everything up and try it out. It's immediately obvious that
this setting is still far too tight, as reactions to my steering
inputs are delayed by a fraction of a second. Halfway down the block
I encounter a weave (apparently a common symptom of too-tight
bearings), so I get back home and put the thing away for the night.

Part IV: The Second Day:

Having replaced the side fairings, I was reluctant to remove them
again, so I simply jacked up the front end with a stack of wood
underneath the headers.

I was also reluctant to remove everything from the front end, so went
about trying to remove the top bridge and bars without removing
anything else. Removing the clutch and brake levers from the bars
allowed everything to come off and hang to the sides, and then the
top bridge comes off as well. So far so good, and this is much easier
than having to tighten everything up each time to try out my latest
setting.

Now then, this time I'm trying out a theory proposed by Tony
Materazzi (find the discussion here: https://lists.cs.wisc.edu/
archive/vfr/2004-March/msg01165.shtml). So, I re-set the bearing by
hand, then re-installed the lock washer and locknut. However, this
time instead of tightening the lock washer all the way by hand and
then tightening further with the wrenches, I tightened it entirely by
hand, then _backed off_ until I could only just barely feel the nut
contact the washer when I pressed down on it. At that point, I
tightened the washer only enough more to align the washer tabs, and
then buttoned everything up.

This time the bearing force _did_not_increase_ when I tightened the
top bridge. That's a good thing. So, I re-installed everything and
went for a ride. Everything tracked straight and true. I'll omit the
details, but it seems that I got it.

[1]While the cbr600 lower triple is similar to that of the VFR, it is
not identical. The distance to the fork legs and the forward offset
from the steering stem are both different, as is the location of the
steering stops. the steering stem itself is identical, and uses the
same upper nut and bearing set.

[2] I wouldn't have been comfortable doing this step with the bike
supported with a jack underneath the headers. I'd be too afraid of
bounding it off of the jack.

[3] American Kowa Seiki is the company that is recommended for this
tool, but their online ordering was not working when I was looking
for the wrenches. Rather than attempt to phone them and get the parts
ordered (which would have required waiting until the next business
day), I simply ordered these wrenches instead. Don't do this.

Part V: discussion and commentary.

First, the decision to go with Honda's ball bearings. Most people on
this list (currently and in the archives) assert that tapered roller
bearings are better than ball bearings, and that Honda only uses ball
bearings for cost reasons. This may be the case, I don't know. I've
seen commentary that indicates that the issue is much more complex:
https://lists.cs.wisc.edu/archive/vf...msg01100.shtml

given that discussion, I decided to go with Honda's bearings. This
removed one source of uncertainty in my work because I could simply
follow the manual instead of modifying the procedure to put tapered
rollers in. Such modifications don't appear anywhere I've seen, so
I'd be on my own.

Perhaps once I'm more experienced I'll try tapered rollers. I intend
to own this bike long enough that I'll probably be doing this job again.

Second, techniques and "if I could do it over" thoughts.

I'd remove the front end of the bike. I'd just drop the entire front
fairing, possibly removing the front bracket as well. That would
allow me to remove the brakes entirely and set them aside. It would
also ease access to the relevant parts. While removing the front
fairing seems like a lot of work, easing access, protecting paint,
and eliminating crowding seem worth it to me.

Having removed the front end, I'd think about those straps to the
rafters. The got in the way a bit, but they might not be so much in
the way once I've removed the front fairing? I don't know. Going with
the iron pipe frame instead of the rafters won't help things. A frame
stand might do, but I don't own one and don't have the means to
fabricate one.

I might remove the gas tank next time. I protected it this time with
paper and cardboard, but still put one chip in the front. At that
rate, I'd be looking at having the bike down for at least 2, maybe
more like 3 days, and would probably do the next major maintenance at
the same time.

I'd not even bother with the pair of pin spanners. Certainly the
Armstrong design is too thick to allow both to be placed one over the
other. In addition, the lock washer blocks two of the holes on the
adjustment nut, further limiting placement. They (as I'll discuss
below) just don't appear to be necessary. I'd get the Kowa Seiki
models if I were to get them.

I'd buy a new lock washer...or two. Mine didn't break, but I'd hate
to be without one, and I've bent those tabs a fair bit.

Ok, bearing preload and the "locknut." I disagree with those who
believe that the locknut locks the adjuster by jamming against it.
Certainly in my case there was no jamming action; the whole works
spun. Secondly, if the locknut were to act as a jam nut, there would
be no reason for using the washer to lock the two together. Thirdly,
there was pressure exerted by the top bridge on the bearing (again,
on my bike).

However, if Tony Materazzi is correct, the lock washer makes sense.
The locknut is "locked" to the top bridge by the pressure exerted by
the steering stem nut. The locknut also serves to keep the pressure
off the adjuster nut. The lock washer keeps the adjuster nut from
turning independently of the triple. This would also explain why
Honda uses the sort of locknut they do: large diameter so it gets
plenty of contact (i.e. friction) with the top bridge 8 slots so you
can keep the adjuster as close as possible to the desired setting.

There is one conundrum however: What about the bearing preload spec?
Why did 25nm lock my bike up solid? I really don't know, and that
bothers me, especially since it seems that I've gotten things set
correctly (ore perhaps even a bit too tight) as is. If anyone's got
ideas on this one I'm all ears.

later,
Johnj
_______________________________________________
vfr mailing list
vfr@xxxxxx
For subscription and delivery options:
https://lists.cs.wisc.edu/mailman/listinfo/vfr

Oh Wise and Knowing Listers,


I have the opportunity to buy a complete F4 front end in very good
condition. Legs, upper and lower clamps, clip ons, calipers, wheel,
rotors, axle, spacers, miscelaneous hardware. The idea would be to swap in
onto my beloved 3rd gen. For this purpose at least, I think it's safe to
assume that 3rd gen and 4th gen are so close as to be the same. So those
with 4th gen experience, please chime in.

>From a engineering point of view, is it a reasonable project? Stem length,
offset, steering stops, distance from center, etc., What say ye?

Richard Ronay


_______________________________________________
vfr mailing list
vfr@xxxxxx
For subscription and delivery options:
https://lists.cs.wisc.edu/mailman/listinfo/vfr

On Friday 29Sep, 2006, at 8:46 PM, J Richard Ronay wrote:

>
> Oh Wise and Knowing Listers,
>
>
> I have the opportunity to buy a complete F4 front end in very good
> condition. Legs, upper and lower clamps, clip ons, calipers, wheel,
> rotors, axle, spacers, miscelaneous hardware. The idea would be
> to swap in
> onto my beloved 3rd gen. For this purpose at least, I think it's
> safe to
> assume that 3rd gen and 4th gen are so close as to be the same. So
> those
> with 4th gen experience, please chime in.
>
>> From a engineering point of view, is it a reasonable project?
>> Stem length,
> offset, steering stops, distance from center, etc., What say ye?
>
Well, when I did my steering head bearings, I took some measurements
on my VFR ('94) steering stem and lower triple:

Lower bearing surface diameter (so, ID of your lower inner steering
head bearing): 30mm
Upper bearing surface diameter (ID of your upper inner SH bearing): 26mm
Height of steering stem: 22.38cm
center to center distance between fork legs: 23.336cm
fork axis offset (i.e. distance between the line going through the
center of the fork legs and the steering stem...I think that I
measured to the _front_ of the stem, not the center, but I don't
recall exactly now): 53mm

While I've got an f3 steering stem here, I'd hesitate before assuming
that they're the same, since the f4 is substantially different.

HTH,
Johnj

_______________________________________________
vfr mailing list
vfr@xxxxxx
For subscription and delivery options:
https://lists.cs.wisc.edu/mailman/listinfo/vfr

JohnJ,
I read all your steps and I think it is possible that you have two separate
problems.

First (and most critical), if you installed caged ball bearings, it is
possible that you installed one or both of the bearings with the cage on the
wrong side. This is easy to do and I have seen shop mechanics that should
know better do it. This would force the cage into the race when even a
light torque was applied and cause the difficult steering, as well as
possibly causing the nut back off or get tighter when the stem is rotated.
If that has happened, (and to me it seems likely, given that you described
every other step as being done in the right order and should not have caused
the steering to be difficult otherwise) the bike is not safe to ride and you
need to get back in there and replace any damaged parts.

Second, you indicated that you could not get the locknut to lock. It is
possible that the threads are slightly stripped, either in the locknut or at
that part of the threads on the stem. This can happen when a mechanic or
prior owner tries to adjust the stem bearing the lazy way, just putting a
punch under the top triple and tapping the preload nut and locknut around
together as one unit. Note that it is possible for them to still feel like
they are threading on the stem, but have the threads too worn off too much
to allow them to tighten properly. However, being able to tighten the
preload nut to 25nm seems to indicate that this is not the case, so maybe
the only reason that you could not get them to lock is that the spanners did
not fit properly. ??

As for suggestions for the next time, I made a gas tank prop rod out of a
piece of 1/4" aluminum rod from Home Depot so I could hold the tank at it's
highest position. Then I fed a soft tie around each frame spar, as close to
the steering head casting as I could get, being careful not to pinch any
wires and hoses. I connected tie down straps from the soft ties to the
large eyebolts I have attached to the rafters, then I jacked the bike up and
tightened the straps. The bike was stable (also used a rear stand) so that
I could take off the entire front end, which made installing the tapered
roller bearings much easier.
John Harrison
Birmingham



----- Original Message -----
From: "John Johnson"
To: "vfr800 vfr800"
Sent: Friday, September 29, 2006 6:11 PM
Subject: Wrench report: replacing steering head bearings in a '94 VFR


> Part III: Putting it all together.
>
> Here's where it all came apart for me. So far, everything had gone
> like a dream. I greased up the bearings, inserted the lower bearing
> set on the inner race, and re-installed the lower triple. Then I
> greased the upper bearing set, slid it over the triple, then
> installed the upper race, then the dust seal. It's time now for the
> adjuster nut and lock nut.
>
> Honda's CSM tells you to torque the steering stem to 25nm and then to
> rotate the steering stem a few times to make sure everything moves
> freely. At 25nm however, the steering stem was _quite_ difficult to
> rotate at all. This was not helped by the fact that the adjuster nut
> wanted to tighten more when I rotated the stem to the left, and
> loosen itself when I rotated the stem to the right.
>
> I thought that this behavior might be stopped by installing the
> locknut, so I installed the locknut as described in the CSM, using
> the two Armstrong pin spanners. These tools are quite irritating, and
> given a second chance I wouldn't have bought them at all.[3] Still, I
> managed to get the locknut on as the manual indicated. No luck
> though, since the whole unit turned instead of just the adjuster. I
> then thought that I needed to jam the locknut on harder. That didn't
> work either. then I thought that maybe it was some grease on the
> threads, so I removed it and cleaned off the threads. No luck again.
>
> At this point, I thought that I'd just torque up the bearings to 25nm
> and install everything, then check it. So, I installed the top bridge
> and the fork legs, and the front wheel. At this point, it's easier to
> install the brake calipers than to leave them hanging, so I put them
> back also. I torqued down the steering stem nut and...I can't turn
> the bars. It's _way_ too tight. I repeat this a couple of times, each
> time checking to make sure I've got everything in and in the correct
> order. I also start backing off the torque, from 25 to 20 to 15nm.
> It's always _way_ too tight.
>

_______________________________________________
vfr mailing list
vfr@xxxxxx
For subscription and delivery options:
https://lists.cs.wisc.edu/mailman/listinfo/vfr

On Friday 29Sep, 2006, at 10:52 PM, John Harrison wrote:

> JohnJ,
> I read all your steps and I think it is possible that you have two
> separate problems.
>
> First (and most critical), if you installed caged ball bearings,
> it is possible that you installed one or both of the bearings with
> the cage on the wrong side.

Ok, this seems quite possible. How would I tell? If this is the case,
how do I know whether or not the cages (plastic in this case) are
damaged? Or is this a case of "you've fucked them up, buy new ones."?
>
> Second, you indicated that you could not get the locknut to lock.
> It is possible that the threads are slightly stripped, either in
> the locknut or at that part of the threads on the stem. This can
> happen when a mechanic or prior owner tries to adjust the stem
> bearing the lazy way, just putting a punch under the top triple and
> tapping the preload nut and locknut around together as one unit.
> Note that it is possible for them to still feel like they are
> threading on the stem, but have the threads too worn off too much
> to allow them to tighten properly. However, being able to tighten
> the preload nut to 25nm seems to indicate that this is not the
> case, so maybe the only reason that you could not get them to lock
> is that the spanners did not fit properly. ??

Threads are absolutely not stripped. You cannot adjust the stem
bearings the lazy way on this bike, as the top bridge completely
covers the locknut and adjuster You literally cannot get a punch in
there to adjust it without taking off the bridge. While irritating,
my spanners will spin the locknut while holding the adjuster steady.
The issue is that the entire assembly then moves as a unit. I could
maybe crank the whole works down tight enough to jam it on, but that
would be way more than 1/4 turn on the locknut, never mind finger
tight plus just enough to align a groove.

crap.
Johnj
_______________________________________________
vfr mailing list
vfr@xxxxxx
For subscription and delivery options:
https://lists.cs.wisc.edu/mailman/listinfo/vfr

JohnJ,
If you installed the top one backwards (typically the top one "faces"
opposite the bottom one), pretty sure that you will be able to see the
damage from the inner race on the cage when you look at it. If the cage was
metal, yes, you would need a new bearing AND new races. Since it is
plastic, if it is not cracked and if the cage will still hold the bearings
at the correct spacing, you may not need a new bearing. Personally, if I
found that this was the problem, I would take a VERY close look at the races
(with a magnifying glass) to see if they were damaged and I would definitely
replace the bearing/cage.
John Harrison
Birmingham



----- Original Message -----
From: "John Johnson"
>> First (and most critical), if you installed caged ball bearings, it is
>> possible that you installed one or both of the bearings with the cage on
>> the wrong side.
>
> Ok, this seems quite possible. How would I tell? If this is the case, how
> do I know whether or not the cages (plastic in this case) are damaged? Or
> is this a case of "you've fucked them up, buy new ones."?

_______________________________________________
vfr mailing list
vfr@xxxxxx
For subscription and delivery options:
https://lists.cs.wisc.edu/mailman/listinfo/vfr

On Friday 29Sep, 2006, at 10:52 PM, John Harrison wrote:

> JohnJ,
> I read all your steps and I think it is possible that you have two
> separate problems.
>
> First (and most critical), if you installed caged ball bearings,
> it is possible that you installed one or both of the bearings with
> the cage on the wrong side.

I just checked my install. Both cages were originally installed in
the proper orientation.
>
> Second, you indicated that you could not get the locknut to lock.
> It is possible that the threads are slightly stripped, either in
> the locknut or at that part of the threads on the stem.

I replied to JohnH privately, but the theads are _not_ stripped.

Ok, AFAIK I've done everything right. The bike feels ok, in general
(though another test ride is now required, I hope things go back
together well...), so without any basis for comparison I'm going to
call this one good. I will be watching my bearings like a hawk, though.

later,
Johnj
_______________________________________________
vfr mailing list
vfr@xxxxxx
For subscription and delivery options:
https://lists.cs.wisc.edu/mailman/listinfo/vfr

I found it.

I had installed the upper dust seal _inside_ the bearing! I removed
it and, seeing no damage to the bearing, re-installed it with the old
(undamaged) dust seal. Once I had the dust seal in the right place,
everything torqued up exactly as the manual stated.

I'm not sure what to think about Tony Materazzi's theory about the
locknut. In the correct installation, the adjuster won't move under
normal steering head movement, so the lock-nut is a backup. Whether
or not it acts as a jam nut or as a "shelf" to lock to the top
bridge, I don't know.

I'm ordering a new upper bearing set (yes, caged balls) and will
install it as soon as I get a chance. I'm not interested in hammering
the lower inner race off the steering stem again, particularly if
this set lasts another 30k miles. If anyone wants to send me a 4GVFR
steering stem with a brand new tapered roller race on it, I'll
happily trade, though.

Heck, 32k maintenance is coming up, so I'll probably just do that at
the same time. That way the gas tank and front end only have to come
off once. Man, what an expensive year for the bike!

big thanks to all those who helped with this, especially JohnH and
Chris Stumpf.

later,
Johnj
_______________________________________________
vfr mailing list
vfr@xxxxxx
For subscription and delivery options:
https://lists.cs.wisc.edu/mailman/listinfo/vfr