Steering Bearing and Head Shake

Yamaha FJR Motorcycle Forum

Help Support Yamaha FJR Motorcycle Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ionbeam

2 FUN
Joined
Jun 24, 2005
Messages
8,817
Reaction score
1,785
Location
Sandown, NH
I saw this via a Honda group I ride with. I guess Friday is a good day to launch this bit of mental masturbation. I first saw this before the internet was born. Thoughts?


Sooner or later, all the larger late model Hondas and Yamahas develop a problem with the steering bearings that results in a head shake when decelerating rapidly from 50 mph and less. For want of a better term, we'll call it a "deceleration wobble."

There is a rule in motorcycle suspension technology which says that problems in handling that occur under 45 mph are due to defects in front of the steering head, while those occurring at higher speeds are found in causes aft of the steering. It's a tried and true rule of thumb, and a decel wobble obeys the pattern. Many things can cause weaves and wobbles, whether on acceleration or deceleration -- tire wear is especially critical. But decel wobbles have their own special causes, and if the front tire isn't excessively worn or the wheel badly out of balance, the cause is almost always the steering bearings. But we're not talking looseness. We're talking about something that is not addressed in any service manuals, factory or aftermarket. Whether because the frame is made of soft material or what, the steering bearing races in many large Japanese bikes "walk" in the frame. That is, they shift in their recesses and become out of square with the steering stem, and out of parallel with each other.
The cause of the classic decel wobble is bearing races that have "floated."

bear_race.jpg


The resulting non-parallelness sets up torque forces in the steering which manifest themselves as attempts by the fork to correct itself, with the result: shimmy, shimmy. Again, the problem isn't looseness. Mere tightening often fails to correct the problem.

The following procedure is one circulated by American Honda's District Service Reps back in the day. Though never published, this is based on the above premise, and rooted in a procedure found in Honda Service Letter #126, a document explaining the need to fully seat replacement bearings. Though the technique originated for replacement bearings, it is valid for all originally-installed steering bearings as well. Three tools are needed: a torque wrench, the special factory steering bearing nut socket (Honda's is part # 07916-3710100), and a good quality tubular 0-10 lb. spring scale. A floor jack or something similar to jack the front of the motorcycle off the floor will be handy, too.

Most Japaneses bikes have both a tensioning nut and a locknut. The early GL1000 is an exception.

bear_tens_nut.jpg


Follow your manual's instructions for removing the top triple clamp (Honda calls this a "bridge"), so that the pair of special castlelated nuts becomes accessible. The upper one is just a locknut. Remove it and set it aside, along with the special washer. Jack the front end up off the floor, and feel the bearings as you turn the bars each side from center. If the bearings are notchy or the front end has a self-centering action, the bearings need to be replaced, no second-guessing here. After replacing them if necessary, continue. Get the front end off the floor. Turn the fork to full right lock, and with the torque wrench and special socket, tighten the bearings to 40-50 ft-lbs. The fork will be very stiff. Don't panic. It's only temporary. Now turn the fork lock-to-lock, repeatedly, at least twenty times. You will probably notice something interesting: that ridiculously high tension will loosen up; the bearings will get looser, indicating that they have squared up and settled into the frame. In some cases, you won't be able to tell, but even if you don't notice the bearings loosening up, proceed. Turn the fork to full left lock now and loosen the nut until it's just finger tight, then turn the fork to the right lock again and tighten it to 7-10 ft/lbs.

The spring scale method is far and away the best, and for some time has been recommended by Honda.

Attach your spring scale onto one fork tube, using a piece of shoestring or something similarly soft so as not to scratch the tube. With the fork assembly pointed straight ahead and the tire off the floor, slowly pull the spring scale straight ahead until the tip of the fender arcs about one inch. Note the poundage. You're looking for a 5-7 lb. pull. Five for motorcycles under 600 lbs., more for heavier machines and those with fairings. Tighten the tensioning nut as needed, a little at a time, and check with the spring scale.

bear_spring_scale.jpg


After adjusting, drop the special washer back into place, and screw on the locknut, but don't tighten it. Though you probably found the locknut jammed against the tensioning nut, that's not the correct way to install it. It should be close to the tensioning nut, but not jammed against it. Leave a little space -- about 0.020". Then bend the locktabs into the locknut to keep the two interlocked. The locknut's job is to isolate the torque of the bridge nut from the steering bearings. Reassemble the rest of the fork per the manual. If a test ride reveals that the bike sways side to side like a rowboat the bearings are too tight, readjust to a lower spec as needed. Ride safely.

Postscript 2013: The special Honda socket has long been unavailable now. It is not really feasible to do this procedure without the socket, as very few hook wrenches permit the amount of force required. But you might succeed at that. A few have reported that the socket is being reproduced on the aftermarket, and I have seen on on eBay. There may still be a few stock ones at dealers too.


It would sure be interesting to see what the before/after shoe string/pound readings are when someone changes from OEM steering bearing to roller bearings.

 
Last edited by a moderator:
Where was this info 37 years ago??? I had a KZ 650 that shook like Andy at the sight of a bookcase on decell.

Dealer said it was the nature of the beast.

 
Good info, especially the testing method, but not sure how much it will apply. The metalurgy on these bikes is so much better than what those Honda's were back then that I wouldn't think this would be much of an issue anymore, but I'm no expert either.

 
I don't buy the original premise he makes. We are supposed to accept the author's word because it is "tried and true" whatever that means.

I still say that the steering bearing can't cause the shakes. The bearing can make the bike's steering under damped, (when there is insufficient preload on the bearing) and so become more susceptible to shaking when something else is out of sorts, but it doesn't cause it. And I sure don't see how the bearing cocking, walking out of its square position would make the steering more susceptible to shaking. Must be a special Honda thing?

Setting the bearing preload via a strain gauge is actually a far more accurate way to do it than just the raw, bearing torque method that Yamaha prescribes for the FJR. The final torque value for the first castellated nut (which sets the preload) after initially seating the bearings with an initial torque of 37 ft-lb and then loosening fully, is only 13 ft-lb. Many folks would be hard pressed to even measure such a low torque value accurately, and even if they could the amount of mechanical damping can vary depending on the condition of that bearing.

But it may be worth noting that we do not have the same situation on the second castellated "lock nut", which is supposed to only be finger tight against the rubber o-ring (not run up snug to the metal), and then the locking tab of the washer slipped through the two nuts before installing the upper triple clamp and big nut on the top, torqued to a much higher value (85 ft-lbs). The idea is you don't want that 85 ft-lb nut causing the bearing preload to increase.

 
Last edited by a moderator:
"There is a rule in motorcycle suspension technology which says that problems in handling that occur under 45 mph are due to defects in front of the steering head, while those occurring at higher speeds are found in causes aft of the steering. It's a tried and true rule of thumb, and a decel wobble obeys the pattern."

This is not entirely true. By now, you're familiar with my 'affection' for the Metzler Roadtec Z6's that were standard on my '10 FJR. The headshake @ steady state & decel up until 35-40mph was a direct result of the center wave tread groove on that POS tire. The handlebars & the rest of the bike followed this wave groove. Solution was to send the Z6's to the landfill & go w/PR3's. Headshake COMPLETELY eliminated.

 
Mike Nixon is a Master Mechanic and really knows his stuff. His website is chock full of great information. Though he specializes in Hondas much of his information is universal.

I replaced my steering head bearings with tapered roller bearings when I was in the middle of a 24K complete service.(The bike was new to me so I checked everything from front to back). I followed Mike Nixons "Shucking the Shakes" procedure. I have used the same technique for several of my bikes and it WORKS! None of the bikes I have used the procedure on have any front end wobble problems...Period. They all ride straight and true.

By the way for any of you interested in knowing with my tapered rollers were torqued to 22Ft/Lbs and the pull reading was about 6 lbs which is about the pull required on these bikes due to their weight.

His website is:

www.motorcycleproject.com

Z

PS. As a side note I only replaced the steering head bearings because the original ball bearings had indentations in the races. The original ball bearing setup works fine and I subscribe to the addage... if it aint broke don't fix it!

 
+1 on the "cause" of headshake.

I did all, and I mean all of the adjusting, tightening, replacing bearings and such with my ST1100.

In the end, changing from an Avon Storm to a Michelin tire solved the problem.

 
Mike Nixon is a Master Mechanic and really knows his stuff. His website is chock full of great information. Though he specializes in Hondas much of his information is universal.
I replaced my steering head bearings with tapered roller bearings when I was in the middle of a 24K complete service.(The bike was new to me so I checked everything from front to back). I followed Mike Nixons "Shucking the Shakes" procedure. I have used the same technique for several of my bikes and it WORKS! None of the bikes I have used the procedure on have any front end wobble problems...Period. They all ride straight and true.

By the way for any of you interested in knowing with my tapered rollers were torqued to 22Ft/Lbs and the pull reading was about 6 lbs which is about the pull required on these bikes due to their weight.

His website is:

www.motorcycleproject.com

Z

PS. As a side note I only replaced the steering head bearings because the original ball bearings had indentations in the races. The original ball bearing setup works fine and I subscribe to the addage... if it aint broke don't fix it!
Cool. Another reference to this procedure. Let's bump this zombie....

This procedure worked for my FJR when nothing else did.

Decel wobble banished.

Old tires not a factor.

Full load, no problem.

 
Last edited by a moderator:
Head shake is just worn tires in general, not one tire versus another. (Probably mostly.
smile.png
) As the tires wear, the centerline of the contact patch shifts unevenly, resulting in an application of caster. Sometimes the tire contact is left of the centerline, sometimes right. That offset turns the steering if it's not held by your hands. Braking makes the effect stronger.

I submit that nearly everyone who said they switched tire brands to solve it, actually solved it just by mounting a new tire, no matter what brand. If one brand of tire wears more evenly, keeping a good centerline, then yay for you. It might be wearing more evenly because after the experience, you're keeping a closer eye on tire pressure.

I'm not saying head bearings have nothing to do with it, but "loose" bearings make it easier to shake. Still caused by uneven centerline of the front tire.

 
Last edited by a moderator:
What the W said above.. But then again, after changing to tapered bearings, no worn out sorry tire has ever given me head shake again. Just one data point, but it's factual and must be after 10 or so tire sets... Michelin PR3's and 4's.

 
What the W said above.. But then again, after changing to tapered bearings, no worn out sorry tire has ever given me head shake again. Just one data point, but it's factual and must be after 10 or so tire sets.
^This^. A new tire made my wobble go away for a little while but it always returned. I tried Diablo Strada, PR2, Angels, PP2CT and something else on the front, a Shinko I think. The head shake always came back. Multiple repack/retorque of bearings did nothing to change anything. And of course I was given "The Talk" by two dealerships how "it's normal" and admonished about riding without both hands on the bars at all times.

Tapered roller bearings is the only thing that made my problem go away permanently.

 
Top