I agree, Evan, using a linking value of "1" the model performs well.? I'm glad to have a model that closely matches real world measurement.? One in a row! Ha!
?
? I've read that behind the scenes, LTspice turns "1" in the K statement into four or five 9s. Behavior seems to change when I start with five nines, especially when I add measured resistance and capacitance from toroids.info.
?
? ?I just realized the data in Diz's site is for a single winding, not the actual bifilar winding. Still not sure how that affects the model.? Regardless, using "1" rather than "99999" seems to activate some decent assumptions in the model.
?
I'll have to go back and look at Diz's 2N5109/2SC5551 amp, and Q90.? I'd like to get to the point that I can model 2 or three stages in the uBITX RF chain with reasonable fidelity.
?
Thanks again
Andy
?
