fine/coarse interaction
The attached patchfile pertains to changes I proposed in the use of rfield.trans_coef_zone[] and rfield.trans_coef_total[], which I believe are incorrectly applied and only at the output stage.
Related to this, the FractionThisLine code in pressure.h is also fundamentally flawed. The attached input deck (based on a model from Nick) begins strong oscillations around zone 168 (of the second iteration, but the root problem can be seen at the same point of the first iteration) due to the NII 1084A wings initially overlapping with HeII 1085, and then not overlapping at all (as temperature decreases). This cause a huge (incorrect) increase in HeII radiation pressure, which is the root of the oscillations.
Gary and Robin have described these and related issues in draft manuscripts elsewhere.