I find it strange that in item 1, it cranked but did not start. I assume you used the same procedure. I wouldn't think that the ignition system would be that sensitive to 12.6 vs 12.3, but it could be possible I suppose.
Usually when a battery cell fails, it will not crank the starter. The static battery voltage will test out pretty good, until you hit it with a load. I don't think static voltage is really any indication of the status of the battery quality, you have to do a load test.

From my experience, with a CV carb the biggest problem is the throttle. If you can get the bike to crank.

 

You are partially right that a battery can show 12.6V and then fail a load test. A battery with a truly "bad cell" will test well under 12V, one with degraded internal chemistry can be 12.6V, but drop quickly when a load is applied. A voltage drop test is essentially a load test using the starter as the load, rather than pulling it and taking it to a shop. Static voltage is absolutely an indicator of battery charge state from which you can make some estimates of it's health. Especially on more modern bikes with more parasitic draw while off that will more quickly draw it off full charge on a tired battery.

The way to think about it is how many watt hours (a unit of energy) does it take to start a cold engine and can the battery supply that (including running the headlight and all other things that use or sap power). I'm not saying it's the problem here, just saying it could be and it's easy to check with a $10 meter.

 

I'll not agree with this statement, see it false too many times.

 

It's physics (electromotive potential), but you're free to disagree. Somedays I disagree with gravity, somedays it disagrees with me.

 

Nice discussion. My experience with lead/acid batteries is that the best indicator of its health is a load test.
Can the battery supply sufficient current at a voltage level that allows the electronics to operate?
As a battery degrades, it may have enough plate surface exposed to the acid to make 12.5V when there isn't a load but if a load is applied the reduced viable plate surface isn't capable of producing the required current and circuit voltage drops.
That's why it's a REALLY good idea to ck voltage WHILE cranking to see what your battery is doing.

 

I've never used a Vmeter on the system while starting & with a good battery. Why?
But it would be interesting to see what it is.
Not really interested in a big argument, if you want to believe the moon is cheese, it's ok with me.

 

I said in my original post (#5), "Check voltage with the key off and while cranking."

I'll explain this because it's important to people troubleshooting that don't have experience with electrical systems.

A load is just resistance in a circuit. So "a load test" at your local garage or parts store is just a box with some resistors, that simulate typical loads in a system like your headlight and starter.

A fully charged lead acid battery should be at 12.6V or higher. Less and it's either not fully charged or not able to get fully charged, but 12.6V alone doesn't mean its good.

If you check voltage while cranking, the voltage will drop, this is normal because a big load is being applied, just like a "load test". If the battery is good, it may drop to 10.5V or so wether the engine starts or not. If you keep cranking, obviously you'll run the battery down. But, that initial cranking tells you the battery can flow current and maintain a reasonable voltage and the resulting power (V x A) is okay.

An aged battery might start at 12.6V, but under the starting load will drop quickly and probably go below 9.5V. Depending on how hard the bike starts normal this may not be enough to get it started.

So, to summarize based on my original "Check voltage with the key off and while cranking." If the original poster said with the key off I'm at 12.2V, then the battery needs a charge or isn't capable of taking a full charge. If he said, I'm at 12.6V with the key off, when cranking it falls to 8V, then it's time for a new battery. Like I said previously some people are good mechanically, but have never understood electrical systems. Knowing a few 5 minute tests can tell you a lot before you go assuming it's something else. Sometimes the battery is good, but corrosion on the lugs (at the battery, starter, ground, etc.) cause resistance that add load, drop voltage, and cause starting issues. This is diagnosed with a slightly different voltage drop test that again takes 5 minutes and there should be plenty of how to videos for both tests on YouTube.

 

I didn't say it but some people interact normally, others don't. FYI, I'm an Elec Eng, majored in physics/chemistry, so get off it. ... I'm going to consider myself in the 'other' group.

 

Next time it does that lean the bike over 45 deg ish till you see fuel dribble out the overflow hose, pick it back up and it'llstart right up. One guy installed a primer bulb in the fuel line. Only happens to me when bike has been sittin unused for a long time.

 

On injected klx250 it happens at least quite often. Turns over but not starting usually being low battery. Give it some jump juice or time on the tender and fires straight up. I think the heat around here combined with short runs just doesnt charge it enough.