Still sorting and cleaning and organising the second K100RS - 1984, beautiful condition but.

Symptoms:

Tried to start as though battery is flat - lights dimmed, groaned, not quite turned over enough. Immediate slight panic, since I've only just cleaned the starter relay contacts.

Pulled battery out - voltmeter read mostly full. I put it on the smart charger anyway, until it read 12.8V, put it back in.

Clock works until I turn on ignition - then there is nothing working anywhere - clock, lights, starter, anything. Turn ignition off - clock turns back on. After much thought and swearing, I test for a dirty starter motor by putting it in second, rolling it backwards a few times, then turning on - all the lights come on. Ignition off, ignition on - dead again. Put in 2nd, roll backwards, doesn't work this time. 

Fuses are fine and there are no damaged wires visible without removing the tank.

I leave it for fifteen minutes, go back - everything works, it starts and idles happily, but the speedometer and odometer are picking up a signal and start working while at idle. I put it in gear on the centre stand and spin the back wheel (carefully!) and get speed and distance working smoothly. Back in neutral - slight signal making the speedo needle jump and the trip meter start ticking over.

I suspect there is weird electrical interference and/or shorting created by a dirty starter motor. While I'm finding time to pull that out, what are your thoughts?

buggered battery
or bad contacts to any of the battery terminals

quite regularly we see bad earth from the negative lead to the chassis earth bolt
lastly ....could be a bad ignition switch and or the terminals in the loom that do that connection (have seen some shockers )

a bad battery will often recover some after charging to read a good voltage ...without load on it .....the real test is to measure the battery with the main beam on at a minimum ....measure the voltage for more than a few minutes ...if it falls below 12 ....get a new one

if you have a regular charger that delivers 4 amps for arguments sake  the general rule of thumb is that it will take about 8 hours on charge to change the electrolyte to something that is useful from a low state of charge .

quite often internally the plates of the battery shed plating of the electrodes ...which then accumulates on the bottom of the cell ...often partially shorting them ..( reducing capacity for charge and discharge )

fingers crossed but a new battery might be the best option ...if you don't investigate any other options
but that doesn't make sense ,,,so do the  first and most obvious cleanup of terminals etc

Suspect the battery. It may show good voltage when idle but see what it shows when under load from the starter.

Well, the battery was recording well above 12 while the ignition was on and nothing operating. I've met dead batteries where the voltage drops under load, but in this case the battery read 12 and nothing was reaching the clock, let alone any of the instrument lights, chassis lights or starter. After the bike began working again I left it not running with high beam on, and voltage dropped about .2 to 12.3.

The contacts to the battery are clean. I'll check the chassis earth. It's an AGM battery, so I'd be surprised if the plates had shed anything.

I've also seen a suggestion that the ignition switch could be an issue. Since I had the instrument panel on and off a few times, I could have knocked a connection and then jolted it back. I'll check that, as well.

could be worth having a good look at that instruments connector and clean up ....we love "deoxit " for this ... good news about the voltage at the battery ...but might be time to pull the tank and have a good inspection
ignition on ...means lots of things are powered ......after or during  start button is pressed more things start drawing more power ....fuel pump, ignition , injection etc

some slightly longer fuel hoses will help with this,  even of its only a temporary nature

might pay to measure the connectors before making any changes ,,,,, especially on the feed back from the ignition switch ..and the kill switch .
then do the fuses we have seen them do the usual corrosion things at irregular intervals ...measure the little pins that on the external side of the  fuse itself ...but doesn't give feedback for the output-load  side ...but can point to something before them ...for this things need to be turned on to see the loss
remember you are trying to prove the fault before claiming a repair

have you considered the earth connection to the frame (UNDER TANK ) for all the "other" electrics, eye terminals often found to be corroded ... some times using a different earth reference can tell a lot ...

just some suggestions

good luck !!!

Take the battery to an auto store that thinks you will buy a new one and get it load tested. One of the internal connections may have broken and is just touching without a good contact.
Because the starter cant get enough power to turn the engine you aren't going to find the problem at the earth point under the tank, it's going to be in the high current area of the start circuit. Battery, earth point on the gearbox, cables, relay contacts, starter motor or the battery itself are the places to look.

Check cluster for bad ground....have you tried to pull the clutch at the starting attempts?

I had similar issues, running a good thick wire from the collection of earth wires under the tank to the battery negative terminal fixed it. Either that or a dying battery.

Sounds like the battery only has a surface charge.   I'd suggest a load test at an auto parts store and then clean the grounding points and connections on the cluster and ecu.  Might want to pull the speedo sensor and give it a cleaning too.