The AGNSS board in the left arm is very sensitive to wind and vibration.
So be sure to remove the vent and install the foam AND use some form of adhesive (rubbery/silicon type best) to secure the board.
Mine was one of the first AGNSS boards. Back then the instructions where for the Emlid board, and I dont recall getting foam or seeing the vent removed as it is now suggested (else I would have.)
I was alerted to the problem only after not getting 100% Fixes in the PPKbot.
Konstantin was straight onto the problem..."send me a photo of the install - the arm" he noticed the chrome vent was still in place.!
This is what the RTKPlot of the converted UBX file looked like - Very bad!
A week earlier, a flight looked like this. Very nice

After correcting my install it looked like this. (still not super clean) notice the cycle slips that cut through all sats.
But I should also point out that my old P4P had developed a vibration in the Fan. Which lead me to think if AGNSS is that sensative, the fan vibes wouldnt help.
So I inspected the fan assembly and noticed a few things.
- the four plastic weld rivets that join the top and bottom halves had some movement in it.
- the dji foam between the fan and the magnesium (grey) case had squashed and points touched the case.
- the black wire when bumped would kill power to fan, wires had fatigued.
Once fixed the rtkPlot looks much better.

Please note this P4P has done hundreds of km surveys (per Teo install).
The satellites listed are southern hemisphere and may look different to those in the nth.
The test flights were very short (5mins or less), the takeoff area was surrounded by trees and building with only small opening to satellites. This may not be representative of a full flight in constant fwd motion on a windy day.
Hope it helps some one.
The AGNSS board in the left arm is very sensitive to wind and vibration.
So be sure to remove the vent and install the foam AND use some form of adhesive (rubbery/silicon type best) to secure the board.
Mine was one of the first AGNSS boards. Back then the instructions where for the Emlid board, and I dont recall getting foam or seeing the vent removed as it is now suggested (else I would have.)
I was alerted to the problem only after not getting 100% Fixes in the PPKbot.
Konstantin was straight onto the problem..."send me a photo of the install - the arm" he noticed the chrome vent was still in place.!
This is what the RTKPlot of the converted UBX file looked like - Very bad!

A week earlier, a flight looked like this. Very nice

After correcting my install it looked like this. (still not super clean) notice the cycle slips that cut through all sats.

But I should also point out that my old P4P had developed a vibration in the Fan. Which lead me to think if AGNSS is that sensative, the fan vibes wouldnt help.
So I inspected the fan assembly and noticed a few things.
- the four plastic weld rivets that join the top and bottom halves had some movement in it.
- the dji foam between the fan and the magnesium (grey) case had squashed and points touched the case.
- the black wire when bumped would kill power to fan, wires had fatigued.

Once fixed the rtkPlot looks much better.

Please note this P4P has done hundreds of km surveys (per Teo install).
The satellites listed are southern hemisphere and may look different to those in the nth.
The test flights were very short (5mins or less), the takeoff area was surrounded by trees and building with only small opening to satellites. This may not be representative of a full flight in constant fwd motion on a windy day.
Hope it helps some one.
edited Apr 14 '21 at 11:44 am