Equivalent bonds different exchanges?

[mfechner-cmd]

Dear all,

I have some issue with computing J's in Sr2IrO4. I am using VASP in a non-relativistic setting so no L.S coupling. What I obtain is that nearest neighbour exchange somehow become unequal see pasted below although I would expect that they are all the same. Any thoughts here? Suggestion? My wannier90 and VASP computations are well converged, I can also provide input files if needed.

this is the wanns2j output Ir1 and Ir7 are in the same plane just different direction but J the 6th row provide Jiso and it changes between the bonds?

Ir1 Ir7 0 -1 0 0.0010 2.741 -2.742 0.000 3.877
Ir1 Ir7 -1 -1 0 0.0635 -2.741 -2.742 0.000 3.877
Ir1 Ir7 0 0 0 -1.3427 2.741 2.742 0.000 3.877
Ir1 Ir7 -1 0 0 -0.0012 -2.741 2.742 0.000 3.877
Ir1 Ir1 -1 0 0 -0.2231 -5.482 0.000 0.000 5.482
Ir1 Ir1 1 0 0 -0.2231 5.482 0.000 0.000 5.482
Ir1 Ir1 0 -1 0 -1.5134 0.000 -5.484 0.000 5.484
Ir1 Ir1 0 1 0 -1.5134 0.000 5.484 0.000 5.484

thanks for any input here