GPS interference with Navio2 + RPi

Circling back to Navio2 for a bit.

How many of the Navio2+PX4 users on here have encountered GPS issues/GPS interference from the hardware? This is an issue that has essentially grounded my copter for the past few months.

In the field, the setup takes forever to acquire satellites, and when it does, it rarely obtains a lock so no RTL. Got a few crashes because of those GPS issues.

Tested this with a secondary u-blox M8N GPS mounted on the quad using a 12cm tall mount, connected to u-center (Windows) using USB. The GPS would not lock, barely acquiring 2-3 satellites for minutes. Unplugged the RC receiver from Navio2, same. Shut down RPi, no change. When I unplugged the RPi from the power source, the number of acquired satellites IMMEDIATELY jumped up and the PPS LED started blinking; I had a fix.

Clearly, when power is applied to the combo, it creates interference bad enough to prevent normal GPS unit operation.

When launched in the same location my PX4 with BN-880 (same frame, similar GPS mount height) obtains GPS lock almost instantaneously and flies beautifully.

Before I start thinking about replacing components, upgrading the RPi (I doubt a product like the RPi is creating such interference) or try to wrap this thing in aluminum foil, I’d like to hear from other users on here about their experiences. Who here has experienced GPS issues/interferrence with Navio2? What did you do to alleviate/fix?

My thought would be noise on the power supply from the RPi or whatever.
Look up isolation and filtering. I really doubt it is EMF - tin foil hats rarely work.

The RPi was connected through its microUSB port to a power brick from a reputable company, located further away.

I wouldn’t put it past Navio2 itself, I have first-hand knowledge of how bad their manufacturing quality is.

Perhaps in this case the interference could be from the flat flex cable connecting the RPi camera. I vaguely remember seeing something online about interference caused by FFC. I’ll try shielding it and see what happens.

The camera cable shielding was mentioned in https://github.com/ArduPilot/ardupilot/issues/8437

The GPS interference was caused by the WiFi dongle (RPi 2). One would assume that those operate on different frequencies, but after unplugging the dongle I’m getting much much better GPS signal quality. I’m keeping my RPi 3s and RPi 4s for other stuff, but I would be curious how the built-in wireless affects GPS in those setups, since the instructions don’t call for disabling wlan.

GPS jamming indicator reported in the log has not changed, still ~200 - perhaps it’s unreliable. But I got much quicker satellite fix and RTL finally worked, so I consider the issue fixed. I can finally move to trying dual GPS setup with navio2.

As a side note the primitive camera cable shielding I’d made with foil helped to remove interference in the camera imagery, possibly coming from the motors and/or ESCs - the horizontal lines which were seen in frames when the motors were spinning up are now gone from the video feed.