What I find odd is the huge temp leaps (~13 degrees) only occur during the initial heat-up phase.  After it passes the target and starts to cool, the jmps are much more modest (max of ~ 5 degrees).

That's not odd, that's characteristic of an underdamped dynamic system. It means you could use a little less P or a little more D, assuming there's anti-windup in that firmware for the integrator term (there is in current versions of Marlin, not sure about back then)

That's the one!

I tried using P/I/D of 13.2, .71, 1.66 which I found in another post. Interestingly, it will hold 195 within 0.2 degrees with those values, but alas that is too cool for abs.
What I find odd is the huge temp leaps (~13 degrees) only occur during the initial heat-up phase.  After it passes the target and starts to cool, the jmps are much more modest (max of ~ 5 degrees).

I'll try dropping the max heater pwm (that the 255 value, right? I'll find it...) and report back later.

Thx elmoret.

You can either limit max heater PWM in firmware, or dial back the P and I values initially so the algorithm can find them correctly. Currently the controller is too aggressive, resulting in overshoot.

Is this normal? Is there something else to change in the firmware to fix this?

I did a search, and found a few other PID settings to try. I will post back if any of them work....