GPS module does not lock (signal issues) about board

[G10h4ck]

The GPS module has a ferrite ( PZ2012D800-3R0TF) in series for filtering.

If I remember well an inductance in series is good to filter out high frequency noise, the lower frequency it start to be effective filtering depends on hysteresis characteristic, and inductance value AFAIR, do we have more info about that inductance so we could at least guess what frequencies of conducted noise should get filtered?

[CodeFetch]

Check if the power fluctuates with an oscilloscope. Maybe it just needs a decoupling capacitor.

[spiccinini]

Check if the power fluctuates with an oscilloscope. Maybe it just needs a decoupling capacitor.
Thanks @CodeFetch for the sugestion. I already mesured, DC power is good, and the GPS board already has a DC decoupling capacitor and also a ferrite bead in series but it seems it is not enaugh.

[spiccinini]

The culprit of this issue is that the PCIe is generating the interference. The power sources are not the problem as we bypassed them and the problem persisted. When we disable PCIe directly from the registers of the SoC then the problem dissapears. Also the noise is not present when the board is booting until the PCIe root complex starts. We did some measurements of the noise with a b200 USRP and the noise is in a wide range of the spectrum.

We are working at #54 improving the impedance of the PCIe lines in the mega board with some hope.

[CodeFetch]

Some time ago I was told by an electronics engineer that the PCIe lines are far too long and should not move along the whole board. He said even with shielding it would unlikely pass CE conformity tests. I'll kindly ask him to have a look for a solution.