Last night and this AM I did some testing with the ARGOS PTT and
current integrator.  Results:

1) At 8-10 VDC, transmitting every 110 seconds for 8 hrs/day will
   take about 40.4 mAh/day.  Double that for sending via 2 PTT numbers
   (i.e., we'll xmit every 55 seconds), and add 3.6 mAh/day for 150 uA
   quiescent current, and we get 84.4 mAh/day, or 30.8 Ah/year.

2) If we use strings of 7 batteries, we can only use the battery down
   to 1.05 V/cell:  7 * 1.05V = 7.35V, minus a diode drop (approx .35V).
   Eyeballing the Duracell graphs, that looks to yield between 10 and 11
   Ah/cell at 21 degrees C.  To be safe, I'll assume 10.

3) So, FOR THIS PTT, it looks like 3 battery packs will give a year's use.
   But that's nominal, and other PTTs may use more.  I think 4 strings of
   7 cells would be safe, though.

4) Again eyeballing the Duracell curves, if we were to use strings of 8 cells,
   and conditioned the batteries by running the pack down to 11 VDC, it looks
   like we could get between 11.5 and 13 Ah/cell.  I don't think it's
   worth the trouble -- I'd still want 4 strings.

5) I ran tests at 10, 8, 7, and every 0.5 V down to 4.5 VDC.  The PTT
   kept running down to 5.0 VDC, although the calculated xmit current
   multiplied by battery voltage would indicate that transmitted power
   might be kinda low.  Xmit current drops off after 6.0 VDC.  So, I'd
   guess that we might actually get useable transmit power down to 6 VDC,
   which is 0.9 V/cell.  This gets us back into the regime of 12-13 Ah/pack,
   although the end of life is at reduced transmit power.

6) Here's my results.  It was run at 10.0 VDC overnight, for 512 transmissions.
   The other voltages were run for 20 xmits each, and results calculated.
   It was still working at 5.0 VDC, and the receiver (across the room)
   got all the messages.  At 4.5 VDC, it acted like it was still trying,
   but the xmit current was down by more than half and nothing got through.
   I'd have expected the logic voltage to fail -- I wonder how logic 
   is powered..... 

7) Xmit current was calculated assuming transmissions are 0.92 seconds
   in length.  The column for mAh/day is for xmit current only, and for
   one PTT number.  Double it and add 3.6 for real useage.


                              /--Calculated--\
Battery   Rcvd xmits   Integ    xmit     mAh
Voltage    	        cnts   current  /day
-------   ----------   -----   -------  ----
10.0        512/512    3752	594 mA  40.2
 8.0         20/20      149     603     40.4
 7.0         20/20      156     632     42.3
 6.5         19/20      150     608     40.7
 6.0         20/20      154	624	41.8
 5.5         20/20      138	559	37.4
 5.0         20/20      111	450     30.1
 4.5               Crapped Out