The tach on my 86 bounces around quite a bit.  It's generally accurate though and doesn't bounce around like it would if there were an intermittent connection. It just seems as though some damping part of the circuit has left the building.

I did swap in some newer, low mile 1991 coils when I got the bike but I would think that any damping of the signal takes place in the cluster, not at the coils.

I never really look at the tach while riding though so it's probably not something I'll bother to ever do anything about, just curious what the cause might be.

Is it possible there is an electrical connection issue at the connection to the coils, it takes its signal from the coils, I think it's the 2/3 coil. I also did a clean up of the plug and connectors into the clocks when I had them off for a swap.

This was a PM to Ringfad who lent me two amplifier boards when my speedo (not tacho) went bouncy, but some of it may apply. OTOH, it could be completely useless for this question...

Long story short....

Changed the amplifier - bouncy needle
Check the main board, found a loose pin on the power feed to the speedo, soldered it up - bouncy needle
Changed the main board for yours, then took it out again before bolting everything up
Modified your main board from post-86 standard (one fuel light) to two fuel lights
Having made it useless for its original purpose, I modified it further to dim the fuel lights as per my original
Put it back in again - bouncy needle
Pulled out a spare speedo, found that I couldn't clock this one to the current mileage because the construction is slightly different and you can't get the pin out that holds the small white drive cogs
Decided to swap the small circuit boards and drive coils over. Found mechanical differences between the two speedos which meant that the drive coil from one type won't mount on the other variant
Found that there are two variants of that circuit board too but they have the same part number! One has two green and one yellow wire, the other has two yellow and one brown wire
Got another spare speedo out of the cupboard. Cross-checked the wire colours and drive board layout to draw the above conclusion
Clocked it forward and fitted it

Problem solved.

Conclusion: the small circuit board on the side of the speedo was stuffed. You're welcome anytime to pick up your amplifiers - they're going nowhere.

Hmmm... maybe I will fix it. The same tach is used on all K75 and K100 models and it shows up as its own part in the parts catalog as 62112305263. I was curious how much effort it took to replace the whole tach subassembly so I pulled out an old junk cluster to try.  Surprisingly it only takes one screw and is incredibly easy to do. (A lot easier than just replacing the gear LCD which I've done before.)

I plan to swap in a K11 front end on the bike in the fall so at that point in time I'll try some other clusters to make sure I've got a good working tach and swap the tach from it into the current cluster. (I could just swap in another whole cluster but if I just swap the tach in then the odomoeter will still be the factory one with the right miles on it.)

(It's handy to have a bunch of old clusters on the shelf to loot from.)

duck wrote:(It's handy to have a bunch of old clusters on the shelf to loot from.)

Isn't it just!

Follow up:

So I decided to fix this issue this morning.

From looking at the K100 wiring diagram the signal from the coils to the instrument cluster utilizes a black/blue wire from the 1&4 coil to the cluster so the first thing I did was check that that wire looks well connected.  The answer to that question was affirmative.

So the next possible culprit was the cluster itself.  I pulled the old cluster and tried three different square pin clusters (the patient is an 86) that I had on hand.  The tach needle behaved on all three of those which meant that the circuitry in the original tach assembly had gone haywire for some reason.

So I swapped in one of the tach assemblies that worked and all is fine now.