Referring Figure 1 below, this is the design with a small PNP biasing the power TransQ(Q4). Rp was adjusted from 200k to near 0 and the Power Q would still not fully turn on. Could only get at best, 1.6V (Vce), and typically Vce=2.1V. This is not low enough to allow >2.4A thru the system. This was all fine and good if all I wanted to do was a simple battery recharge. However, I also want to power a load (3A) and recharge a battery (@1A). Under this design, the load pulled the last .6A out of the battery - not cool!
So to take this another step further, Rp was adjusted from 24k down to nearly 0R to no effect on Vce(on). The next step is to change the biasing on Q4.
Figure 1
The short-term solution was to brute-force the biasing on power trans Q4, as shown in Figure 2.
Figure 2 (could not shrink it!)
Although, far from elegant, Fig 2 shows that when Q4 is sufficiently biased, it will pass high currents. I was able to get Vce(sat) down to .5V, allowing for the recharge of the battery (.75A) and powering the load(3A).
An interesting "problem" from this configuration is that I lost the status outputs (pins 9 and 10).
So, the question remains, why won't the BQ24450(pin 15) switch to gnd to allow turn-on of Q3?