First of all, this is from an speedometer from my Sanglas bike. Nevertheless, I think it could very well apply to the Honda's and many other bikes.
My first Sanglas bike has a speedo needle that bounces violently. I tried greasing the drive, the cable, but nothing. I just thought it should be the instrument.
Bought another Sanglas bike and the needle moves smoothly. So I thought "no wonder, it has 4 times less km's on it". But I found something odd. The speedo cable comes from behind, and then goes in front of the lower triple tree, and up to the speedometer. It should go just backwards and up, so I re-routed it.
First ride with it and the needle bounces violently everytime I brake or accelerate!!! No problem, I will change it at the next gas station. Good thing I've found the problem in my first Sanglas bike.
But I didn't have time to stop. Suddenly, the needle dies -but not the odometer- Damn!!!! It has broken inside.
So I opened it and found the spring broken at the point where it is attached to the armature. No problem, I will remove the pin and off we go.
The spring is there to absorb the rapid changes in speed. While the worm gears operate the odometer, there are some magnets that move a bell-shaped metal attached to the needle. The spring not only damps the vibration and rapid changes, but also limits the amount of needle travel. That is, the needle will move until the spring force equals the bell-shaped turning force. So I knew it had to be out of calibration. But would it be more speed than indicated, or less speed than indicated?
Common sense says that if the spring is shorter, it will take less degrees of rotation to achieve the same spring force, so it would mean less indicated speed than real speed. But after my first test drive yesterday, I tend to believe that is the other way round. It seems to indicate more speed than real speed, which is better than the opposite as that way I know that keeping into legal speeds I will always go slower than the maximum permitted.
So, how can that be? How can it be that with a shorter spring, the indicated speed is higher than the real speed? Obviously, what I did was to remove the needle, and once the spring re-attached, with the stationary speedometer, attach the needle at the rest position.
I guess there should be some way to calibrate the speedometer. Is it easy? Probably it involves adjusting the distance from the bell to the magnets. In any case, it will probably be "trial-and-error" if you don't have a bench setup so, anyway, I will probably leave it this way. It was harder to crack it open than the Honda instruments. I also thought about buying a second-hand unit, but I wanted to keep the recorded mileage.
