> Lead acid batteries need constant current until they have a nominal voltage. As soon as a voltage is nominal they are charged. If to continue charging their electrolyte may boil out and plates destroyed.
It won't boil (100C/212F) with "normal" or even large charging current.
It will bubble. This used to be considered normal. We knew the battery was charged. I think the H2O was breaking down into H and O, making explosive gas; also the water "went away" and had to be replaced from time to time.
Now we have sealed batteries, we can't add water, we can't see if they bubble, and battery makers will NOT advise charging to the point that explosive gasses come out.
Still, I think you can trickle 1A onto a 50A-H car battery all winter no harm. For the smaller 5A-H batteries, maybe 0.1A. This can conveniently be a 15VDC supply and a 22 ohm resistor. When the battery is very dead, 10V, it will get around 0.25A charge, falling to 0.1A at 12.6V and 0.05A at 14V. (And yes, we used to volt-limit car generators at 14.4V, though we did check and add water often.)
AFAIK, Ni-Cad batteries can be charged similarly. Of course if you leave them idle for a month, they go flat; if there is any load on a series-string then an extended deep discharge will cause the weakest cell to internally short.
> keeping the part count as low as possible
You do not need individual R-C networks for the L and R mike capsules. One 1K+47uFd will easily power two capsules, and the crosstalk will be "low enough" unless you put the mikes in different studios and analyze the recordings. Hmmmm... -20dB at 14Hz, -40dB at 140Hz, -60dB at 1.4KHz where it would matter.
There is MUCH to be said for 9V disposable. You would have to get many many cycles out of an exotic-metal battery for its ecological impact to be less than fairly benign carbon-zinc trash. When the rig runs down mid-session, you change the batts in 2 minutes instead of taking a 4-hour re-charge break. When you travel, you don't care about the local wall voltage, you go to the bazaar and buy radio batteries. At home, a 25-pack of 9V batts is fairly low-price, hopefully less per hour than the value of the work you do. And of course two-9V eliminates that kick-up scheme. An LED, 6V Zener, 100 ohm resistor, and push-button is a simple battery check with zero standby loss. You check it before any critical take.
> for consumer-level output, the OPA2134 needn't be supplied with +/-18V.
Even on "Pro" gear (not counting 16-track tape decks built to hard-wire to a console), you never need +4dBm. I have not seen a true 600 ohm input in decades, and any deck with a Rec Level control seems to be happy with -10dBV if you crank the knob.
So the level-set is: crank the recorder to "7" or "10" and leave it there. Crank the preamp for nice meter readings. Now the output of the preamp is 0.3V to 0.6V max, and a real clever 3V-power preamp is ample.
Note that the actual audio power into the recorder is under a tenth of a milliWatt, more like 0.036mW. If a 100W speaker-amp draws 200W of raw juice, why is our 0.036mW load requiring 360mW of power? Over-engineering is fine when power is very cheap, as when you can't buy a wall-wart smaller than 2W. In battery operation, such a difference between raw power input and audio power output suggests a sharper pencil.
> How big of an output resistance is ideal?
What are you driving?
If you have a mile of telephone line with true 600 ohm load, that's one thing. 150,000pFd is a heavy load. You want a low-Z output to drive it without treble loss; OTOH the op-amp will be hurting bad with 0.15uFd load and be unstable unless you have enough isolation that the op-amp hardly notices the load. This gets tough. Fortunately, this is rare.
Oh, and I remember when you shorted a chip-amp (shorts happen), you hadda bike out to the electronics shop and buy another $12 chip. A resistor between the chip and the more likely short-points reduces disasters. I guess this is less relevant since the 741 appeared.
I drive 100-foot runs to few-K loads with 470 ohm series resistors. I may lose a dB midband and at 300 feet another dB at the top of the band. Not a problem in live recording. There's several-dB "errors" between sound-check and show, between one mike location and another, we fix it in the mix.
An advantage of 470 ohms is that a beefy chip-amp working around -10dBV can feed multiple recorders, be shorted on two or three 470-isolated outputs, and still put normal not-dirted signal on any un-shorted output. I have fed 7 recorders at once. So far I've had one-short situation twice (once a cable-pinch, once a misconnected output back-fed the splitter), but there's a sub-rule of Murphy's Law which suggests that up to a third of your outputs may get shorted on a bad day.