Our replies have gotten out of sync, and occasionally not 100% correct.? But indeed, as John (more or less) noted, David's 'D' unmodified schematic:
-- runs without errors, but
-- it shows little evidence of .STEPping (the .STEP command is not missing, but the K command controlled by it is missing), and
-- it shows no resonance splitting because the inductors are not coupled to one another.
When the K command is added back, then the splitting is evident when K = 0.01, but not for K = 0.0001.
Ignore the comments about the error with that schematic.? That was a mistake.? It was caused by a typo, when adding the K command back to the 'D' schematic.? The last uploaded iteration of the schematic ("DesplitterDJMW.asc") has no errors, and it shows resonance splitting at both V(A) and V(B).
Here are a few other points to note, though they have nothing to do with your objectives with this simulation:
-- You can specify more than 65535 frequency steps with the .AC command, but LTspice reduces the number to 65535.? Therefore, if you need the frequency steps to be closer together, or if you thought you would get more than 65535, you ought to be aware that it ultimately limits you to that number. and quietly tells you in the .log file.
-- The "SINE(0 1 1000000)" attached to V1 has no effect on the .AC simulation.? That is for .TRAN simulations only.? (Well, the DC value of the SINE function does affect it; but the frequency and amplitude have no effect.)
-- I strongly urge you to use multipliers, rather than numbers like 1000000 or 0.000001.? You can write them as 1MEG or 1e6, and 1u or 1e-6, so you don't have to count the digits, and eventually get it wrong.? We are only human.
-- Inductors in LTspice can have a built-in series resistance.? Right-click on the inductor symbol and add a number for "Series Resistance".? By default it doesn't show on the schematic, but you can change that.? The simulation runs a little faster if you do it that way instead of adding a separate resistor, but you might not notice the difference in this case.? And you might prefer the separate resistor anyway.
Andy
