PD or Iota for my upgrade?

Salvo wrote:
The source impedance is not equal to output voltage divided by current. Source impedance is much lower than your calculated 0.203 ohms. If you studied feedback control theory then you would know the error amplifier located inside the Unitrode chip reduces output resistance by the magnitude of the amp gain. The amp has very high gain. Resistance is quite low; as it should be.

We'll have to agree to disagree. We disagree because we're talking about different things. You're talking about feedback theory, which involves changing voltages, while I'm talking about steady state conditions. You are right that the load impedance differs under non-steady state conditions. I thought I mentioned that in my post. It's just that I think the feedback is likely to converge to the steady state quickly under any reasonable battery impedance. AFAICT, the PD is designed according to the recommendations for the UC3846 Unitrode chip.

I wish you luck with your PD measurements. Make sure the error amp reference voltage remains constant when there's high current demand. PD's issues probably center around the error amp mechanization.

Thank you. I agree, the error amp voltage could be a problem, in theory, but it should show up under a resistive load as well as a battery load. The reference voltage is only five volts, and the current draw from that reference is very low. It's pretty much independent of load - so even under high current I think it's unlikely that reference voltage would vary much. It's built into the Unitrode chip and it's designed to be independent of load. If there is a problem under high current, I think your earlier comment about the input stage is a more likely candidate than the Unitrode chip itself.