The tight regulation is good for a long battery life. Never draining, and never over charging... a nice constant voltage.
I agree about never overcharging for better battery life, and that is precisely what the stock mechanical regulator can do and has since its introduction. However, the draining aspect is a function of alternator capacity and system loads, rather than Vreg capability. I haven't characterized the 650 alternator. But, the other SOHC4 alternators make about 1/3 of their peak capacity at idle RPM. For the 750, 1/3 of 210 watts is 70 Watts. For the 550 and other smaller bikes, 1/3 of 150 watts is 50 watts. With lighting on, I've measured the system drain apart from battery charging to be about 120 watts (10 amps). Neither bike, using alternators within spec, can maintain a battery at full voltage when the bike uses more power than is available from the alternator, unless the bike's standard load is reduced below what the alternator can supply at idle RPM.
"Tighter regulation" is simply impossible when such deficiencies are encountered. The Battery is still going to drain, even if the vreg tells the alternator to make all the power it can. The stock regulator gives all the voltage available to the alternator field when the battery is in depletion mode, supplying hthe difference between bike demand and alternator output. Any electrical junction in the path between battery and Field coil is going to reduce that voltage by it's junction drop characteristics. Schottky devices will be at a lower differential than standard silicon, but still higher than what can achieved by a direct mechanical metal to metal contact.
I don't believe it is about tighter regulation, but rather improved efficiency in between the extreme ends of the charging system operating envelope. However, I have no data to quantify how much efficiency improvement is gained by the experimental vreg operation over the stock. There should be some. But, I don't know if it is 10% or more 0.1%. Have you calculated or measured these differences? It does make sense that more power can gleaned from the alternator at a lower RPM than the stock device. I'd like to know what the RPM break point shift is with the selected Vreg device where the alternator takes over the system load and begins recharging the battery.
In the end I have trouble accepting that, in operation, the proposed vreg will actually provide a "a nice constant voltage" to the battery of the standard SOHC4, regardless of battery technology selected.
The "tiny" battery has a 230 cold cranking amp rating, 340 amps peak. Weighing in at 2.2 lbs., it's light, and small, but quite large in it's capabilities.
The LiFePO4 battery technology may indeed need the "Tighter regulation" aspect. Particularly if you can adjust the regulator to a peak voltage of 14.4 V rather than the 14.5 V to 14.9 V the stock regulator has for a tolerance range. Lead acid battery tech is far more tolerant to input voltage variations and simply absorbs power and ignores any regulation variance with it's low impedance characteristics.
I believe it was Scottly that reported that the LiFePO4 battery impedance raised significantly when the battery achieved full capacity and was of no further use to quell or control voltage variations. Of course, the low impedance aspect would return when any load was placed upon the battery, moving it off its "full" status. But, when such a battery in the SOHC4 system was employed, "tighter regulation" may be of some help to the battery. However, battery damage incurred by over voltage excursions have not been characterized, as to service life detriment quantification, or just what the amplitude and frequency is required to make an impact on such damage for this battery technology. I question the assumption that less than "tight regulation" damages the battery in any way.
The L9911 VR IC is pretty immune to spikes/ripple on the power line. The spec calls that out in the data sheet. I can't imagine a car starter would present less spikes/noise.
Much depends on
where the connections are made as to the devices exposure to EMF spikes. The stock lead acid battery acts as a filter, whose low impedance absorbs spikes present on the distributed buss. The effect is much more pronounced in closer proximity the battery terminals. Before computers and low voltage electronics were added to the vehicle equipment list, much less attention was paid to power distribution, and the SOHC4 shows this by having a power distribution connection for the bike's electrical buss placed in line between battery and starter load. The convenient buss connection was atop the stock solenoid rather than at the battery terminals.
In summary, you can't keep the stock battery from draining, unless you change the alternator capability or the bike's system loading.
"Tight regulation" simply isn't needed for a lead acid battery. They can be charged and maintained nicely by a half wave bridge. In fact, many chargers tout "pulsing" as a means to limit/reduce sulfation effects. (Though there is some doubt of it's effectiveness upon nested cells.)
It may be helpful for a LiFEPO4 battery, but this is unquantified.
I guess, I just don't see how the "experimentation" is "done". Or, how there is a significant operational benefit to its mode of operation other than an unquantified efficiency improvement.
It does appear to be "Cheap" and "Modern", though. Perhaps it is for those who can't or have no desire to make the stock bits function as they have for 40 years.
Have fun with your experiment!
Cheers,
