I tried out the circuit. It was based on if not the working circuit I uploaded. (It used the 324 not the 347.)
The circuit was so close to working, so close. The first run simulated. What I mean by close is an odd voltage came up, so I set up valid pins, and then the voltage changed, just as it should have. However, I'd thought I had 1v for an input differential, so I checked the gain, but it seemed reasonable. So then I checked the input differential and found it was 4v. Hence, I zoomed in and changed it to 1v. But then the next attempt at simulation produced the familiar errors.
Conclusion: It wasn't the right circuit (need 347), and it wouldn't simulate more than once.
But I really thought it was going to work after that initial run.
😞Now the question is what's common between that circuit and the original one I sent in? The initial circuit used the 347 and the multimeter, but this one used the 324 and the probe. For instance, are there shared functions between the multimeter and the probe? This could be the cause. Does the floating nature of the unwired mux's cause random starts, starts that sometimes work and sometimes don't? If so, that'd mean repeated simulation attempts would eventually cause some successes without any changes being made.
You've inspired me to make my successful circuit eventually fail by repeatedly starting and stopping it, so as to rule out floats, as this disturbs me that you're showing my successful circuit isn't as successful as it seems thus far. True, you did tweak it a little, but you've got me concerned that this one is also a time bomb (Multisim-wise).
