In 1particle1mode the 1p1mconf=3 case is incorrect because it does not take into account the interference between the pump and the cavity mode. In the general case, when the wave numbers or these two are different, this term cannot be represented by a Tridiagonal. The same technique needs to be used as with the ternary version of ParticleTwoModes.
Both problems can be easily solved with the new, more general quantumoperator architecture.
Wolfgang Niedenzu has a patch for this, which has never been incorporated.
Attached note shows the necessary cases of mode-particle interaction in moving-particle cavity QED.
A clause class could be defined which could store two (or more) Tridiagonals for successive application.
Fixed on commit #89989543