Testing for wobbles


Doing any work or testing for wobbles or instability of a motorcycle is dangerous.  A wobble often is due to a combination of factors.  Some factors may not be spelled out here, or factors are explained here but may not be understood.  Improperly adjusted assemblies, poorly torqued hardware, over-stressed parts, or any other factors can result in a serious wobble, loss of control of the motorcycle, and possibly even an accident that can result in personal injury or death.  In this document, I merely report what I learned through my wobble experience with BMW motorcycles of a particular era.

A fully trained expert should examine wobbles or any other instability.   If you decide to test for instability on your motorcycle, you have been warned that it is exceedingly dangerous.  

You agree that the author is free from any liability by continuing to read this information.  If you do not take 100% responsibility for the results of reading and possibly applying what you read here, you may not read further.  Reading this article or using any of the hazardous testing procedures in this article signifies your release of the author from liability.

I posted this article in the /5 and later series because the wobble issue was mainly about the SWB /5 bikes.

Tests for BMW motorcycle high-speed wobble or weave

by Duane Ausherman

This photo shows a surefire treatment to cure a wobble.

First, please read this Wikipedia information on high-speed wobbles.

I avoided writing this article for years due to the seriousness and knowing how difficult it is to describe the testing.  Many BMW riders that knew of the reputation of my BMW franchised motorcycle business, BMW of Marin, encouraged me to write this.  I write this today, but still with great trepidation.  A couple of “war stories” are included, but that is in an attempt to communicate just how serious this issue is.

I trained my four mechanics to do some basic testing for low-speed instability during all test rides.  The others would ride it for the experience when something new was noticed.  We needed to be on the same page and speak the same language and vocabulary when dealing with this serious issue.

I only allowed my mechanics to test to a particular level.  If the motorcycle showed tendencies that needed more dangerous testing, they would return and inform me.  Since the business was mine, I was ultimately responsible.  I did all the more serious tests because it would be tough to live with myself if an employee got hurt or worse.

I hope that I have your full attention now.

This article also applies to all models of BMW in that era.  I divide wobbles into two groups, high speed, and low speed.  Low-speed wobbles are in the range of 25-45 mph.  Once a wobble has started and the bike is slowing down, the wobble can continue down to 15 mph.  A high-speed wobble occurs at 55 mph and up.  The speeds between 45- 55 mph are sort of a gray area.  Typically, not much happens, so one can ignore that range for this article.

The early models, the /3 and /2

By the time I was in the business, very few /3 models were still being ridden.  I don’t recall anything but a few low-speed wobbles being found, and they were easily fixed.

The /2 models made from late 1955 up through 1969 that had Earles forks were quite stable.  We often found low-speed wobbles, but they were easily corrected.  The Earles forks were so stable that they tolerated a handlebar fairing quite well.  The few high-speed wobbles were so rare that we easily found mechanical issues without even riding them.  Once the mechanical issues were fixed, the high-speed wobbles were gone.

One common error was finding the fork swing arm in the sidecar position.  There is a reason that BMW made a solo position and a sidecar position.  Just because the rider had never had a wobble, or noticed any instability, doesn’t mean that it was stable or safe.  Many riders have no sensitivity to such things.  Riders can ride many miles and never notice something we would spot instantly.  They had just been lucky.  Skill is good, but good luck is better.

In 1968-69 BMW came out with bikes that we call “US” models.  That model was a test bed for the forks that would be used on the soon-to-be-announced /5.  They have the same telescopic fork as the later /5 thru /7 motorcycles.  A few ancillary parts are different, but from a construction standpoint, they are identical forks.  They, too, will all wobble if a handlebar fairing is mounted.  Many riders have assured me that their bike with a handlebar fairing does not wobble.  On all that I have been allowed to test, I have ridden it in front of the owner and demonstrated a low-speed wobble.  No exceptions on any telescopic fork BMW from the first ones in 68 up through the SWB models of the /5.

My approach to a wobble is to carefully examine the motorcycle.  This included many bikes, of which the owner had never reported any issue.  A visual inspection can often find the most likely causes.  Correction of these is to be done before riding it again.  A special test ride regimen was followed to reduce the chance of an accident.  First, we would find and correct the low-speed wobbles and then maybe test for a high-speed wobble.

If all of the standard mechanical possibilities have been checked, it’s time for the low-speed test.  Testing for low-speed wobbles isn’t very dangerous.  I have not had any scary experiences with them.  I look for them first, as they can show the likelihood of a high-speed problem.  Our usual test ride was much more inclusive than I will describe.  Here we are only discussing wobbles.  First, I test for straight tracking.  If the bike doesn’t track straight, I want to find out what’s crooked.  My test is to get up to about 20-25 mph and let go of the bars to see if it goes straight.  My measure is the amount of lean I must use to keep it straight.  I will lean my body however much it takes to go straight.  I estimate how far off-center my head is on the bike.  If it takes 6″ of lean, that’s too much.  I want to find out why.  We often found a 2″ or less lean on brand-new bikes.  Some weren’t perfect when new.  Some lean isn’t going to ensure a wobble, but it needs investigating.  I have seen a 6″ to 8″ lean that handles very well and safely at all speeds.  I actually think that a bike with a “lean” is less likely to wobble, but I have no statistical proof.  However, after several hours of riding with a lean, the rider’s shoulders and back will get sore from keeping it straight.

We weren’t even concerned with a lean related to wobbles.  The discovery of the lean just meant that something was amiss that we hadn’t discovered during our work on the motorcycle.  What else had we missed?

Yamaha motorcycle wobble story

My cousin’s wife rode a Yamaha 350 and had felt a bit of a twitch, but nothing really scary.  My cousin had ridden it, and he had found nothing.  They were both experienced riders.  However, different riders, especially amateurs, find different results.  At 65 mph, the bike went into such a wobble that the bars were immediately wrenched out of her hands.  The speed and force of the bars were such that the knuckles on both hands were broken by the bars swinging back and forth.  She had no time to get her hands out of the way, it was instant.  Then the bike flipped over.  When someone tells you that you should speed up or slow down, remember there may be no time.  Her injuries were not limited to broken fingers.

BMW motorcycle wobble story

I asked Jim for permission to use his story, and his answer is copied directly.

Duane, after all the help I have received from your site, you are welcome to anything you need from me.  As I said, I’m no brainchild when it comes to the fine details you guys get into on that forum, but I have had an Earles fork under me since around 1972.  I went down twice over the years for the same reason.  It can be passed off as some anomaly, bearings, etc.  It comes down to proper maintenance on the rider’s part.  A false sense of security because I have this “tank” under me just doesn’t cut it.  This is my first ground-up restoration, and I have not much to go on except a parts catalog, a Clymer manual, Slabon’s book, and the most helpful so far, you guys.  I ask some dumb questions at times, but I do learn from the answers.  Thanks again for all the help.  Jim

His story

Not being the engineer type, I can only say you beat the grim reaper this time.  I agree with Duane, you never want to experience the excitement of flight twice.  Last year a high-speed wobble broke my bike and my bones.  I have since replaced every bike’s bearing, wheel, spoke, and tire.  It is still not done, but getting back on it will be a slow process.  Not being able to determine the cause of these events is the scariest part of the deal.  I have also changed out my US bars for lower Euro bars.  They are the longer sidecar bars.  I have no idea how this will affect the handling, but we will see.  I regularly rode my bike at 70 to 75 mph.  All I remember was an uneven bit of road between lanes.  As I crossed that ridge, all hell broke loose.  Things happen so fast that I only had time to realize that I was a goner.   The bike spent an eternity in the air, hitting all four corners before coming to rest right behind me.  The brunt of the damage to the bike was sustained by the crash bars, handlebars, front fender and headlight bucket, rear solo seat, and hard bags.  I have also installed new shocks, the rears being nitrogen-filled replacements by Works Performance in California.  I figured I had to start at the beginning anyway, so why not replace everything that could have gone wrong?  Keeping everything in proper adjustment is next.  You can get to a point where you feel these old Beemers are indestructible and snicker at hog riders who have to wrench their bikes daily to keep all of their parts aboard.  I have spent nearly 12,000 bucks to repair the bike.  the hospital bills were much higher.  Consider yourself ahead of the game if no doctors were involved.


Testing for wobbles

Some caution first.  I am describing these tests to acquaint the reader with our methods.   I can’t overemphasize the danger.  A wobble can maim or kill.

The most likely condition for a wobble is in slowing down while in a downhill gentle curve with hands off of the handlebars with no weight on the seat, sort of like posting while riding a horse.

This testing is progressive because I perform a test under unlikely conditions for a wobble.  If that passes, I make it more likely to wobble.  From experience, I know when to stop increasing the risk and quit.  This is not for your average owner/rider.

I take it up to 30 mph and slow down gently and feel for any sign of instability.  Then up to 40 mph and slow down.  Then I take it up to 35-40 mph, shut the throttle off, and let go of the bars.  I actually try to induce a wobble by hitting the bars.  This one will get a lot of failures.  Maybe 25 % will wobble on this one.  Then up to 40 mph, let go of the bars, bang them quite hard, and see if the shake dampens out quickly, slowly, or not at all.   At any time, the wobble can be controlled by grabbing the bars.  The last and strictest test is to “post” or stand up a bit, take the weight off the seat, and put it on the pegs.  My legs aren’t gripping the tank but sort of bowlegged to clear the tank.  I want to reduce the biomass of my body that will dampen a wobble.  All the while, I am slowing down from 40 mph with my hands off the bars, and I give the bars a “hit” with my hand.  This position shifts my body forward as I must balance on my feet.  It is very hard to shift the weight forward with a fairing but easy with low bars.  No post-1967 with telescopic forks that  I have ever tested with a handlebar fairing, will pass this last test.

Many owners have told me, “My BMW with a handlebar fairing doesn’t wobble.”  If allowed, I have gotten on the “stable bike,” riding down the street, turned around, and come back in front of the owner at 35 mph with the bike in a wobble.  I have probably done this 100 times.  Point made.

Possible causes of the low-speed wobble.  A low-speed wobble has few aerodynamic factors.  Some of the factors are hard, or impossible, to isolate to only one thing.  It is usually a combination of several things.  Often one can get a false sense of security.  The owner will remove or correct something, and the bike now shows no wobble.  Is the bike fixed?  Maybe not.  What if the wobble goes away with some other aspect removed?  What if the wobble will go away with any of several things changed?  I contend that just removing one of the factors is usually only a start.

A good example is the handlebar fairing or windshield.  Is the problem one of raising the center of gravity or of the pivoting front end?  Is the greater weight on the forks compressing the fork enough to change the trail?  A handlebar fairing will always alter results.  Remove it for the tests.  Here are some things to check.

1.  Correct tires and air pressure.  Twenty-five years ago, this meant Conti or Metzler at about 30-34 lbs.  Today it may be higher.

2.  Tires balanced correctly.  Are you sure you have really balanced them?  Can your repair shop really do it?  Read about balancing a BMW motorcycle wheel.

3.  Are the tires in very good condition, especially the front?  We found that when the tread is down to 1/2, replace it, balance it, and many wobbles will disappear.

4.  Tight wheel, swingarm, and steering bearings?

5.  Good shocks and mounting bushings?

6.  Saddlebags?  Remove them for testing.

7.  A top box is just about the worst thing on a bike.  The problem is both aerodynamic and weight.  Even empty, it can cause a wobble.  Remove it for the tests.

The buyer of our very first R90S came in to complain about a low-speed wobble.  I took a look, and he had mounted a top box.  I proclaimed that he caused the wobble with the box.   He was incredulous, as not only was it empty, but made of some space-age material and only weighed 2 lbs.  I told him to remove it and give it a test ride.  The wobble was gone.

8.  A /5 BMW (and others, too) must be neutral steering.  To test this, I prefer to ride through a 50-60 mile sweeping curve.  Let go of the bars, and it should stay in the same curve.  If it wants to turn tighter (fall down), it is too low in front or too high in the rear.  Is the bike high enough?  Does it have sagged-out springs or a front tire with a low profile?  It can also have forks bent straight back towards the engine.  It may track straight but not turn properly.  If it wants to stand up and go straight in the curve, maybe the front tire is too big, the rear is too small, or the rake is too great.  When the dealer got the BMW, it was neutral steering.  You may not care about neutral steering, but you better be concerned about finding out what has changed.

9.  The US telescopic forks (1968-69) and the /5 and later have a reputation for wobbling.  It is only partly true.  If the forks are properly aligned, they won’t have stiction.  This can be a very important factor in wobbles, especially high-speed wobbles.  The forks may need to go through the alignment procedure to work correctly.

Testing for high-speed wobbles

This requires a lot of experience, paid-up life insurance, and a bit of insanity….  maybe more than a bit.  I will describe my worst experience.

Chris Blum, my parts man at BMW of Marin, bought a new 1972 R75/5 in Curry color.  The only modification was mounting European low bars.  By 2000 miles, he complained about a wobble and was afraid to take it up in speed.  Everything mechanical was perfect, and we could find no error in our inspection.  The tires were in balance and only took the minimal weight.  The rims were quite true.  All bearings were tight.  After all, the bike was new.

I was sort of doubting that it had a problem worth mentioning.  Maybe Chris was super sensitive and was complaining about a quiver, not a full wobble.  As the owner of the shop, I had the most experience with wobbles.  I put my riding stuff on and prepared to take it out for a ride.  Chris’ last words were, “Watch it.  This thing is dangerous.”

I took it out on a seldom-traveled road to perform the tests.  It had a minor, low-speed wobble but dampened out easily by putting my hands on the bars.  Many bikes had worse low-speed wobbles and no high-speed instability.

I started the high-speed tests at 50-55 mph and showed no instability.  Of course, I was posting during each test in order to increase the chance of a wobble.  I repeated it at 60, and it was as solid as a rock.  I did the same at 65 mph, and the forks instantly went into a full wobble.  There was no time to do anything.  Had my hands been on the bars, my fingers would have been crushed by the wildly swinging bars from fork stop to fork stop.  The front tire was skipping and airborne.  It would come down at the end of the travel and make a huge squeal.  The front wheel would become airborne again and head for the other fork stop.  This happened about 10-12 times in maybe one or two seconds.  I wasn’t paying close attention to my watch.  My only reaction was to try to sit down again.  Due to the added biomass, or some unknown factor, the bike snapped out of the wobble as fast as it started.

I immediately stopped and got off.  I was so scared and weak-kneed that I could barely put it on the center stand.  I just fell to the ground and was in complete shock.  Finally, I recovered enough to get up and ride the bike back to the shop.  We had an all-new problem on our hands.

We got into the truck and drove back to the scene.  This was an hour later, and I was sort of OK by this time.  We examined the tire tracks.  Every 15-20 feet, there was a long black mark, and each was about 90 degrees from the previous one.  The marks weren’t straight but oddly shaped.  You could see where the tire smashed back down on the pavement with a wide part, and then the traction and trail forced the wheel to correct.  This correction was with enough force to whip the bike back into the air with the tire again off of the ground.  If one drew a line through all of the “left” parts of tracks and another through the “right” tracks, those lines were about 3-4 feet apart.

The only thing we found wrong with this new motorcycle was that the forks were not aligned.  It showed very little stiction.  The forks were .006″ out of alignment in both planes.  The BMW spec was .004,” so it was very close.  We had been aligning forks for about a year to .002″, but we had never suspected that such a small error could cause such a wild wobble.  We reduced our previous spec to .001″ or better.  After all, what is wrong with perfect?  The alignment procedure fixed the bike, and it never gave a problem again.

That caused me to recall all of the new bikes we had sold and align the forks.  This was at my expense and took an average of 3,5 to 4 hours each.

Earles forks on the /2

BMW introduced the Earles forks in 1956 to appeal to the sidecar driver.  They have the characteristic of being very soft and comfortable.  This is for the sidecar to allow braking while maintaining a constant wheelbase length, a good thing for sidecars.  Earles forks are heavy and swing slowly.  The unsprung weight is, however, very light.  For solo sports riding, the Earles forks are a poor choice.

This configuration has unfairly been criticized for being susceptible to wobbles.  That is simply untrue.  Any motorcycle can wobble.  It is the nature of the beast.  There is nothing wrong with the Earles fork or /2 frames design to make it wobble.  At this point, only a bare bike, as delivered by the factory, is being discussed.  If one decides that a wobble is dangerous and undesirable, then it follows that it is a fault that should be corrected.  Since they didn’t wobble when new, the ones that do wobble must have something different about them.  It is logical to say that that particular bike has a very dangerous problem.  We should find that “difference” and correct it.  Altering the design is not the solution.  Finding and correcting the fault is.

It is possible for Earles forks to be bent backward and cause a wobble.  It will certainly increase its tendency to wobble.  Can you see black paint on the front engine cover?  If not, the cover might have been replaced.  The older frames didn’t have strengthening gussets at the headstock and bent easily.  There is no “proper” distance between the engine and the front cover.  It is never less than 3/4″ and usually more.  Two types of covers and three types of cross braces were used, so it takes lots of experience to know if it is correct.  I used to be able to judge it by sticking so many fingers in that space as my measure.

Do not ever put the Earles forks swing arm in the sidecar position for solo use.

The manufacturing tolerances of the /2 were not very close in some respects.  One can easily find small variations in the forks and frame.  I know of one case where a frame was out of spec and didn’t track straight.  It was a European delivery, and the owner, my roommate, took it back to the factory.  They straightened the frame.  It was returned, and it tracked perfectly.  That is an unusual case but serves to show that these bikes were made by humans.

Standard repair practices usually fixed the few that wobbled.

My sample wasn’t large enough to see real problems.  In 1966 I cut up a perfectly good 1962 R60/2 and installed a VW engine.  It required cutting the frame in half and lengthening it by about 2.”  Knowing nothing about frame geometry, I just welded it up in alignment.  It was as stable as it should be, as it was now longer.  It also had neutral steering, but that was an accident.

I opened up a BMW motorcycle repair shop in 1967 in San Francisco and started learning just how little I knew about them.  Wobbles fascinated me, and I did all I could to learn more about them.  Here are a few factors and how to check them.

Front wheel, swingarm, and steering bearings

Testing the front wheel and swingarm bearings.  Grab the front wheel in one hand and the fender brace in the other.  Shake the wheel sideways.  If any “play” is felt, it must be in the wheel or swing arm bearings.  One can see which is loose by looking at the gap at the wheel or the swingarm pivot.  Grab the front fender with one hand and pick it up till the wheel is off of the ground.  Spin the wheel with your other hand and “feel” any vibration through your fender hand.  It should feel completely smooth.  If you feel a vibration, it is from a failed wheel bearing.  Replace them.  Read my page about BMW wheel bearings.

If the swingarm has some play, it must be adjusted to fix it.  It is more likely that the bearings need to be replaced.  If no play is felt in the swingarm, that doesn’t mean that the bearings are good, only that they haven’t yet been proven to be bad.  To make the next check: block up the front end and remove the wheel and shocks.  The swingarm is now loose and will swing up and down freely.  If it is more than a few years old, hasn’t been lubricated regularly, or both, it probably has notched bearings.  Does it swing smoothly or with some “tight spots?”  Replace the bad bearings.  Before 1965 the front swing arm couldn’t be lubricated.  It is easy to add a grease zerk for lubing the bearings.  Read about adjusting the /2 front swing arm.

With the front wheel in the air, the steering damper off, gently and very slowly swing the forks back and forth through the straight-ahead position.  Does it swing completely smoothly, or with a notch in the dead ahead position?  The rider can feel a notch while riding at low speeds, noticing that the bike won’t go straight.  It wants to curve back and forth constantly.  Notched bearings must be replaced.  Loose steering bearings can be checked in a few ways.

1.  When applying the front brakes lightly at low speeds, did you ever feel uneven braking or a kind of resistance that was related to wheel speed?  This could be an out-of-round brake drum or loose steering bearings.

2.  With the engine off, push the bike and apply a bit of front brake.  Do you feel a “click” or movement in the front end?

3.  With the front wheel off of the ground, grab the front end by the lower shock legs, just above the swingarm, and pull gently.  Don’t pull it off of the center stand.  A bit of “play” or a “click” can be felt if the bearings are loose.  Tighten the steering bearings as needed, and retest it for a notch or proper tightness.  It is common for one to tighten up the steering bearings and only then be able to feel notched bearings.

Updated 30 March 2023