Inconsistent Sat count on GPS during cruise flight

Hello All

I’m facing a problem in Here 3 GPS. Here is a snapshot of the log.
image

Setup

  • Cube Orange+
  • 2 x Here 3 (Mounted symmetrically)
  • One GPS is connected directly to the CAN-1 port on Pixhwak.
  • Another GPS is connected in CAN-2 via CAN splitter board, and powered by a separate 5v bec ( Castle 2.0 BEC.)

GPS signal wires are shielded and grounded.

Initially, it was observed on the cruise flight, but after keeping on the ground under open sky, still we got the same issue.

Can someone help me with troubleshooting this?

Thanks in advance.

here3 gps ardupilot

Can you switch power supply of the two GPS modules to test what happens? Clean and stable power is especially important for GPS.

Hi

Yes I have swapped the power supply and it was exactly same. Even I’ve changed the UBEC.

This behaviour is same.

Is the second GPS mounted closer to any source of interference?

Also, are they configured differently at firmware level? I have no experience with HERE GPS modules, but I think they use u-blox chips? There might be a bandwidth issue if it’s configured for multiple GNS systems and getting lots of connections.

But Yuri’s explanation is more likely.

A serial bandwidth issue would cause sat count to occasionally drop completely and probably cause an unhealthy GPS warning. If sat count is reported, the UBX message containing it is making it through intact.

Still a good idea to check for consistent firmware and config.

Hi,

Yes previously we had RFD antennas near the GPS, it’s now moved other side. But, still, we are facing the same issue.
We have kept the GPS module cable as it is, not modified or done anything to it, using the same JST-GH connector that comes with the module.

Hi

I’ve checked the Ardupilot messages and there are no “Unhealthy GPS” messages.
I’ll also check for the firmware and hardware versions and try to check with the different firmware versions.

Thanks!

Are both HERE3 new and out of the box, or did you purchase one or both of them used? Have you previously changed their firmware configuration via u-center? Also please compare their configuration in the DroneCAN/UAVCAN module in Mission Planner.

Hi

Both are newly purchased ones. And, while initial testing it was working fine. Below are the snapshots of log graphs in ATOL and manual hovers.

Any way for you to bench-test them outside, without power connected to the ESCs/RFD/anything else that might cause interferences?