nrf54h20 usb D+ and D- routing

[Samno]

Currently designing around the nrf54h20 and wondering if Nordic has any recommendations for properly routing the USB D+ and D- traces. 

D+ and D- are supposed to be routed to pins B7 and B8, respectively. However, pin A7 blocks access to B7 on the top layer. This seems to suggest that Nordic requires us to bring D+ and D- in on an inner layer and connect using vias. However many white papers online recommend routing USB High-Speed on the top layer, avoiding the use of vias entirely. It seems this is not possible to do with the H20. If B8 we moved to A8, and B7 and A7 were switched, then D+ and D- would be on pins along the outside of the chip and accessible on the top layer.

Does Nordic have any guidance or white papers for best practices on connecting USB D+ and D- on the nrf54H20?

[ketiljo]

Like so:

Else, follow general USB high speed design rules, like equal track length on D+ and D-, impedance control etc. 

[Samno]

How are you achieving the 90 ohm impedance matching with such thin traces? I have 3 mils track to track spacing and 10 mils dielectric thickness with RF dielectric (3.48 Dk) and 1 oz copper and it's still telling me track thickness should be around 16 mils

[ketiljo]

Here's the full stackup for the DK PCB:

[Samno]

Thanks do you have the actual materials used for PP-010 and Core-042 so I can see if my pcb manufacturer can get them?

[ketiljo]

PP is polypropylene and the core is FR4, I don't have more information than that.

[Samno]

Interesting I have never seen polypropylene used as prepreg before. So an RF substrate was not necessary for the 2.4 GHz radio?

[ketiljo]

Ordinary FR4 is fine for most designs as traces and transmission lines are normally very short.