Originally Posted by
OldSpark
It could be your isolator. The specs say it has a drop out voltage of 12.6V which is under the voltage of a fully charged battery (in theory 12.67V) and certainly less than a recently charged battery (which may be as high as ~13.6V but certainly over 12.7V, and an open circuit battery can take a day to lose that surface charge, or several minutes with lights on.)
So if your batteries are connected and the isolation delay is too long (maybe 4 seconds from what you suggest; usually they are 15-60 seconds), both batteries suffer the cranking dip. (Unlike diode isolators, your MOSFET isolator probably conducts both ways.)
If only some genius could come up with an isolator that is isolated when the car stops charging and doesn't reconnect until after cranking when charging has recommenced. Surely that'd be worth a Nobel Prizeor two?