tramp wrote:R is the magnetic reluctance.
Description of the modelling is here: https://www.vutbr.cz/www_base/zav_prace ... e_id=60090
tramp wrote:Yes, one have to keep in mind that we use idealised models, without been bounded to real physics.
I see that the Koren model in use in the dkbuilder doesn't' modelize the grid current (I can be wrong on this, because i just took a fast look on it). Which imply a saturated push-pull will work with greater dynamics than the real stuff. It is the main issue with spice for simulation of guitar amps, and even the Koren model that incorporate a diode to simulate the grid current is a crude approximation, because it doesn't modelize the influence of that grid current on Ia and Ig2 (it is better for triodes but very bad in that regard for pentodes). The Ayumi model is better, but it can be really difficult to make it to converge. This show the limit of all tube models based on the constant voltage curves, because of the huge changes of the tubes parameters on these curves, they become too complicated if you want to correctly modilize the tubes when they are used in extreme conditions like the ones into a guitar amp.