The problem is intermittent. If it happens the bike can't be turned off unless you pull the battery terminal. Checked the starter relay and it's ok.
Looking at the electrical diagram you see that the start control circuit (controls the timing for start and decompression) is energized regardless of ignition key or start button. I put a new battery in and think that there can be more than 15V for a moment which makes the control circuit fail turning on the start relay. Is my theory correct? Will pull the YB wire at the starter relay to test next time it happens. Now that's a darn poor and UNSAVE design of the start control circuit if i'm right.
The rectifier /regulator could also be the culprit. Anyone can recommend a good, strong regulator that will handle a poor absorption of the battery which then won't accept much current and the voltage goes up?

Rubber side down is strictly for savage related issues and the cafe for all others.

Sailor wrote on 01/01/22 at 05:26:01:


The savage has the same design.

OK,
let's start with a failing starter clutch. In that case there's no power/current from the battery flowing to or from the starter motor, because the starter relay is open. That means the bike should turn off!
Secondly, a wiring short. I can not complete rule that out. The wiring in in good shape though, but i will measure to see whether there is any voltage, even small, on the YB wire which energizes the starter relay. Also there aren't many hot wires on the bike with the ignition off where it could short to, only hot to starter relay and to the start control circuit. Does anyone have a detailed circuit diagram of the start control?
Last comment: I disagree that the voltage on the lead acid battery can not be raised above 15V even with a small current. Actually, this is used to recondition the battery, revers a bit of sulfation. The bike regulators are primitive and dump excessive power by shorting a half cyle in the AC voltage of the stator. This may not be enough. The problem i have seen was after a trickle charger was on the battery for a while. So it was fully charged and the battery was new (unknown absorption characteristic).
Currently there is an aftermarket regulator on the bike. Would like to replace it with a known strong regulator able to dump 100-150W. Any ideas?

Agreed. The problem is when charging. If there is overvoltage, then it takes awhile even with a strong regulator to settle it, say at 13.8-14.5 volt charging voltage. Reason is the battery is a huge "capacitor".
The question becomes whether Suzuki has designed its electronic to handle overvoltage or whether their timing circuit for starting trips unintentionally at overvoltage. Understood, there is a starting clutch, but imagine going 80mph and the starter motor engages all of a sudden.

Don't think the problem I described has anything to do with aging, other than replacing the old battery.
I will get to the bottom of it by pulling the control relay when the starter is acting up again. If the starter motor still runs, I may have a short of the YB wire to a hot red wire, although unlikely. Pulling the YB wire at the starter relay will then proof that it's not a hung relay. All very logical.
Money pit - as with anything else repairs will cost you a lot and may be improper - so DIY! Also, there isn't enough innovation justifying buying a new bike/car/boat every couple of years.  
BTW, stay on subject - i.e. the technical stuff.

a switch, yes thought about that. Here's the problem with it. assume your bike stalls in the middle of an intersection. You want to restart quickly and not have to throw a second switch.
Fixing it is best. If the problem lies in the start control circuit, I rather bypass it because it's adding just a 0.7 second delay (and Suzuki has to answer some questions). Not sure the starter button is spec'd to drive the starter relay directly - must check.

Thanks to everyone for their inputs - and a happy new year.

To follow up: so far i can not replicate the problem - and i have ridden for several hours.
Here's the new theory: the only difference at the time the problem occurred was that the bike had been on a trickle charger for a day or so. The post 97 vs1400 have the rectifier/regulator (RR) directly and always connected (through a fuse) to the battery. Did not disconnect the battery from the bike when it was on the trickle (1A) charger. The charger goes to 14-14.5V and most likely the IC in RR wakes up and wants to regulate - but the loop is broken because there's no power from the stator and the SCRs  don't do a thing. So the IC of the RR probably rails causing the RR not to work for awhile after the bike is started, which in turn can lead to overvoltage causing havoc, like the starter motor to turn on.
Any comments? Does the bike has to be disconnected from the battery when on trickle charger?