A battery fresh off the charge will hold a surcharge for a short time. After a two hour rest i should go down to 12.6 - 12.8V.
Of course, any load on the battery will rapidly bleed off any surcharge.
The fact that the charging system raised it back up to 13.2V says the charging system is working.
If you hold the throttle long enough, it should reach 14.5V eventually. That's where the Vreg should tell the alternator to "back Off". And this is where measuring gets tricky again.
Let's say the bike load was 120 Watts. The Alternator at full output makes 150W, with 30 watts left over to recharge the battery. But, when the battery is full the Vreg cuts the output of the alternator by about Half, or 75Watts. This isn't enough to run the bike's 120W load, so it takes some power out of the battery, which lowers the voltage. The Vreg senses this and turns the alternator back on. This off-on-off-on cycle repeats (oscillates) at a random time period, so that meters will sometimes only see an average between 14.5 and 12.8 V. 13.65V is the exact median between these, but spending a little more off time than on time can skew the voltage read average. However, let's say you had LESS than a 120 W load, such as when headlight off, then the vreg might spend more time making the alternator do 75W rather than 150W, and the average system voltage would be lower than the median 13.65V. There is yet another dependency, relating to the test instrument sampling methodology. Being a digital instrument, the device takes samples of the voltage in a time duration and sampling rate that the designers choose. Further, that sample is only passed to the display at yet another time interval.
If you have ever seen video of a turning propeller you've seen the effects of sampling rates with an un-synchronized event. Sometimes, the propeller appears to turn backwards in video, when it most certainly never does. This is called "aliasing" in the digital measurement world, where the display of the instrument does not accurately represent what is being measured. Additionally, digital instruments can also "average" several measurements and display only the average of what was measured. If there are "ripples" riding on a DC voltage potential, you may not know of the actual peak and valleys present at the point of measurement, without more sophisticated measuring equipment. Some meters have "peak hold" button or switch. Oscilloscopes, can display the waveform riding on a DC level.
From your report, I'd say that there is nothing wrong with your VReg. With some more data, there may be a slight adjustment needed. But, it is certainly trying to do it's job.
Were I you, I would next compare the voltage at the battery terminals with the the actual voltage arriving at the Black and Green terminals of the VREG. Surely, you can't expect any Vreg to do it's job correctly if the bike's wiring, connectors, and switches lose voltage before distributing to the VREG.
If you are quick you can simply compare the voltages measured at the two points mentioned and find the differential. Another way is to place one probe on the Vreg black terminal and the other on Battery POS terminal. The meter will display the voltage lost in that pathway. Record and report.
You will also make another measurement between the green wire at the Vreg and the Battery NEG terminal. The meter will display the voltage lost in that pathway. Record and report. Then you add the two measurements together to obtain the difference voltage the regulator "sees" from actual battery voltage.
Another reason to know have this measurement, is that when the alternator needs to make the full 150 Watts, it will need every bit of battery voltage to achieve full power. The VReg passes the black wire voltage to the alternator, when full power is required. But, if the VREG receives 2-3V less than what the Battery actually has, then the alternator cannot make the full 150 Watts when required.
The VREG's most important job is to keep the battery from overcharging above 14.5V. Anything lower and it will let the alternator make as much power as the RPM and load will allow.
Cranking the bike with the electric starter takes some power out of the battery. So, it was no longer full after that event. Even if the starter was only drawing 25 AMPS, that is still 250-300 watts per second being drained. Recall that the charging system will only charge the battery at a 30-40 Watt per second rate, and ONLY when the engine is revved above 2500-3000 RPM. So, you will have to rev the engine 10 times longer than the time the starter button was pushed to restore the battery to full.
Cheers,
