[weblacky]

Okay,
I found time to pull the DS1780 off the bad PIMM and replace it.  Results are good, but testing was odd...

So everything came off cleanly...no problems
    .

After I attached a new chip, both PIMMs' in-circuit scans came back very similar to basically identical for the sensor regions.  Also I did a quick scan of the bare PCB before installing the new chip...what I found is that ALL the sensor legs do have a 1kOhm protective resistor (somewhere) and all the pads (with chip out) showed 1K readings...so protection resistors are good and values are the same all the way to PCB ground (consistent...that's good).

But when I probed the old (now removed) chip...the internal impedance doesn't show up.  But the IC protection diodes are GOOD and look EXACTLY like other DS1780 ICs (even have nearly identical voltage drops).  So out of circuit, I'd have told you this was good!  

However I cannot reasonably explain why in-circuit measurements showed internal resistance changes, a new/installed IC has the CORRECT resistances now, but the old IC shows good protection diodes and no issues during outside probing?!?!?  It is possible I was dealing with a loose leg or some kind of base soldering issue before?  I was probing using the test pads on the PCB) not the IC legs.  But to be connected (for having readings) but be missing like 1K of additional resistance (while installed) doesn't make any sense to me.  The protection diode graphs do no indicate leaking either...else I'd say leaking was going on to change resistance.  Nope, clean diode graphics on the old DS1780!

however it's very possible that some unique combination of leg connections would reveal the issue out of circuit, as when it's installed in-circuit you can have several paths going and maybe the route taken shows something that the direct ground to sensor leg route does not.  I also tried VCC to sensor legs, same results with working DS1780 ICs.

Again, out of circuit, this old DS1780 looks fine!  But new installed DS1780 no longer has observed resistance differences on sensor legs for 12V and 5V monitoring!

So you'd have to say the old chip is damaged, but I cannot confirm it with other tools...which is frustrating.