This pack seems to charge ok with the original charger. With Nickel Hydride batteries you need to be careful not to overcharge - so don't leave it on longer than necessary. If you don't have a charger, research the subject on the net. There are several approaches to a charger solution - an RC car battery charger, a DC power supply with a series current limiting resistor, or an intelligent voltage sensing unit.
It turns out that the D3 fits nicely in a "pocket" camera soft case. Mine is a LTX series by Tamrac. It has a pocket on the top that accepts the power and mic connection box. I installed left, right, and stereo input jacks as well as an in-line style stereo jack. I found that the panel-mount jacks are junk - they don't make reliable connections. So I mainly use the in-line stereo jack. I use a double pole double throw switch for activating both mic power and D3 power. I use an alkaline 9V battery (in the box too) to power the stereo mics via a series resistor to each and then AC-coupling the output of each to the D3. You will need to experiment with the value of the resistor you use for the particular mic you use, as this resistor also sets the mic gain. The value of the AC coupling capacitor affects the low frequency response. Mine are probably larger than they need to be, but I would rather EQ the lows out than not have them there to begin with. It would be useful to have a switch to change this if desired.
For the regulator, I used an LM317. Nothing abnormal about the use of this - it is right out of the datasheets. I set a trimpot to get 9V out, then measured the value of the trimpot and replaced it with resistor(s) of that value. I stuck a little heat sink on the LM317 to help dissipate the heat. You need a minimum of 3 volts more on the input than desired for the output voltage. So a 12V battery works out well.
I typically use 12V 7AH lead acid batteries for the main power source. I place one of these in a belt pack and can usually get at least 2 days of use at air races I record.
Electrets require a resistor-fed power source. The output signal is AC coupled to the recorder input by a series capacitor. The value of the resistor sets the gain of the electret as well. You might want to start with 2.2K for the resistor and 1uf for the capacitor.
There are mods individuals perform on the electrets to achieve wider dynamic range or better frequency response; like modifying the front hole diameter, or sealing the rear of the module with epoxy. I haven't tried any of these. I'm more interested in comparing various models and understanding the differences between each. I haven't been able to see these differences as accurately with a software package as I can hear with my own ears.
So, anyway, be on the watch for stereo headphones or a stereo extension cable that goes for under $1 or so at a swap meet. These make perfect wiring for your in-ear mics. Wire them as necessary for correct polarity, maybe color code them so you know which mic is which channel. Then you have several options as to how to use them:
If inserting them in your ears, you want to use mics that will fit ok, plus you don't want to insert them too far. You also don't want to cut your ear canal in any way as this is a great area for infection to occur. There are swimming earplugs that you can modify to surround the mic and protect against infection etc.
I purchase a couple prosthetic ears from an audiologist supply house (Westone Laboratories, Inc. 1-800-552-7203) for around $30 or so. These are intended to act as "dummy" ears for learning to make ear impressions (part #157-A). These seem to work great for "remote" ears for recording purposes! Really fun to play with. Purists will tell you there is more to the binaural effect than just the ears but these sure get you close.
Many say that binaural recordings have a bunch of phase distortions in normal speaker playback that become objectionable. They recommend mics placed near the ears to reduce this. I prefer to use headphones to get the full effect of the in-ear recordings so I continue to record that way.