Instability in Acoustic material model

[labcorrea]

I'm pretty new to Specfem and I'm currently trying to run an acoustic model in a "plate-like" geometry. I've attached the files with all the details, including a force with a custom amplitude. (SpecfemAcoustic.tar.gz)

The snag I've hit is this weird instability issue that's giving me a hard time. The frequency spectrum of pressure from all the stations is all over the place, as you can see in the attached figure. The source is center at 150kHz - curve in the left.

I've played around with different time steps (even went as low as 1e-12s) and tried out various meshes, from smaller elements (240x240x6) to larger ones (30x30x2), but no luck so far.

I was wondering if someone could lend a hand and share any tips or tricks to get around this instability in my model. Any help would be awesome.