No problem on the short shrift...

I'll took a quick look at your kjet link... even though I have the background to follow your details of circuit operation, I'm not going to take the time now to do so...

In general, your analysis looks sound.

I have a question about the correlation of the duty cycle and dwell angle. Typically duty cycle refers to the percentage of time that "a device in on". Correct? And dwell angle in automotive terms usually means "the time the points are closed" expressed in degrees of rotation scaled by the number of cylinders. So, higher dwell == higher duty cycle. Correct??

I'm an EE, so I'd have to dig through some texts to recall the relation between pressure and flow volume. But, it seems counter-intuitive that decreased fuel pressure (control pressure) results in more flow, i.e., richer mixture.

In my experience, I wouldn't assume that a 50% duty cycle is optimum for emissions or loop stability. I'm heavily influences by the "preset/fail mode" value at 54deg. Consider that the "do no major harm" setting, i.e., engine should run well enough to safely operate the vehicle given no other component failures/misadjustments. Further, it does seem that mine wants to "lock in" on a value pretty close to 54 deg.

However, after painstakingly working with various settings last night and getting reproducible results. I can't reproduce them today. The outside air temp is less than 10 degrees different. Last night, I tried the "slightly less than 49-59 deg" with O2 sensor connected and it worked fine. No hesitation at pullout, no very high duty cycle. Set at 41-43 deg midpoint, hesitation & high duty cycle. Set at about 62 deg midpoint (original setting), same thing. set it back to the "slightly less than 49-59" from the greenbook, tan like a charm.

Anyway, yes, I did spring for a new Hall sensor. I've got a relative in the biz who was willing to get if for me at cost...

So, given the excessively lean mixture that the O2 system cannot correct for, it looks pretty certainly like a pressure issue somewhere in the k-jet. Would you concur?