This is what Starlink Mini sends back to the DC power line when actively operating. The "Fundamental" frequency is 250 KHz from the main DC-DC, crazily modulated with VPP up to 1.4V. It is already known that this might cause interference in some cars.

I managed to clean the line partially. Filter will be included in the upcoming project (s).

1.4v up to 1A is about > 1W peak of RF noise. That's a lot of RF noise down the cables!!! As a Ham radio operator this is a big red flag, specially on low HF bands. Can you test Reception of AM and low HF (2-7 MHz) with a handheld radio in close range of the Starlink?

Yes. I will run more tests tomorrow.

What does the ground loop look like on that scope probe? It's really hard to tell what comes in thru the airwaves vs what's on the wires. This might be one of the spots where a $20K scope outperforms a Rigol, not sure. If real, should be easy to filter. Perhaps bulky though.

There shouldn't be ground loops since the power bus and the Starlink are isolated. Sure, there is some extra noise due to imperfections of the test, but the whole thing is real. It's easy to see the DC-DC switching freq and other components that depend on the terminal operation.

250khz causing interference with 250kbps CAN bus ?? - Seems very noisy (BUC converter?)

...and here we go

Nothing will happen, actually. And the PoE board is not even used here.

true there's a noise during active data transfer in the car

Are you performing tests from a power supply alone or is there a car battery in parallel?

It's connected to PSU in this test. I know about two cases in cars. It may cause trouble with parking sensors on some Audi cars and repeater cameras on Tesla Model 3 AP 2.5. I decided to investigate it deeply.