Convergence problems when adding magnetic field

Hello SPIS community,

I am doing hybrid PIC simulations of a simple object in a LEO plasma. The simulations produce good results when I set the magnetic field equal to zero (results make sense, agree with theory, etc.) However, when I add in a typical LEO plasma (3e-5 T), the simulation runs, but the results are strange. I still see the wake structure, but some of the cells have the expected density and adjacent cells have zero density (i.e., non-physical results, non-smooth). Like a checkerboard pattern.

I also tried setting a very small magnetic field (1e-9 T) and this also produced the same non-physical effect.

Any suggestions on including magnetic field in a LEO plasma wake simulation?

Thank you.