looks good
The mtz8302’s dual GPS almost seems to work, but it still has a problem.
Use 2x ZED-F9P with FW 1.12 without XBEE and last config from Github. AOG is version 4.3.10 and is configured with an NTRIP client (own NTRIP server) to UDP.
Everything neatly mounted on an aluminum bar with a distance of 101cm between the GPS antennas.
Get the following debug:
[]
got NTRIP data
got UBX1 PVT lat: lon:
poor quality of GPS signal, or only 1 Antenna: NO heading / roll calc
Antenna distance set: 101.00 Ant. dist from GPS: 507
UBX RelPosNED found
UBXRelPosNED flag heading: NOT valid, checksum OK; sentence NOT used
[]
GPS does not seem accurate enough. When I connect a separate USB cable to an F9P, it shows the mode: 3D / DGNSS / FLOAT.
seems like your heading/RelPosNED F9P doesn’t get the corrections from the other one. Is the cable postion F9P Tx2 to heading F9P RX2 good? If your are not sure, connect both TX2 to RX2 (one board to the other)
Thanks for your fast response. Looks you’re right, I made a mistake switching the tx2 and rx2.
When the kids sleeping, I fix the wiring. Can’t wait. 
Made the wires cross. TX2->RX2 and RX2->TX2.
Now I get after a couple of minutes, messages like:
UBX RelPosNED found
got RelPosNED. Heading: 48.91 down vector (cm): 1743.95
got UBX1 PVT lat: x lon: x
poor quality of GPS signal, or only 1 Antenna: NO heading/roll calc
Antenna distance set: 102.00 Ant. dist from GPS: 1864
Is this proof that the two F9P units are talking? Is there another problem for the distance/quality?
After 10-15 minutes no RelPosNED values are set but the message:
UBXRelPosNED flag heading: NOT valid, checksum OK; sentence NOT used
Have you connected to each F9P with Ucenter once running to confirm both units have RTK fix?
Have quietly looked at the different parts. And found something.
Have the setup run without XBEE on top and run AOG to pass the NTRIP data via UDP to the F9Ps.
If I view the top F9P (left receiver) via U-Center, this looks good, goes to 3D / GGNS / FIXED in a short time. The bottom receiver (right) remains at 3D. So it seems that it does not receive the NTRIP data.
If I activate the Ntrip client with this receiver for the test in U-Center, it immediately goes to 3D / GGNS / FIXED. In the debug of the ESP32 there are immediately messages that the calculation in terms of heading are being carried out.
Hopefully we can fix this?
If RTCM isnt going across, there must be a reason, are UART2 TX /RX connected? Do you have both UART2 baud rates set the same (full speed 460), and are you outputting RTCM3 messages on UART2?
1 Like
Now starts to get a little crazy. Settings checked and these are good.
Had two wires running:
TX2 (right F9P) → RX2 (left F9P) (according to documentation) +
Rx2 (right F9P) → TX2 (left F9P)
With these wires I get a FIX on the F9P on the left and not on the right.
If I remove the “Rx2 (right F9P) → TX2 (left F9P)” because it is not necessary, I don’t have a fix anywhere.
For your information, on the F9P on the right, the GPS-> XBEE LED flashes and on the F9P on the left, XBEE-> GPS flashes
Configurations reloaded again in the right and left, but situation remains the same.
Beware point of confusion: TX RX on Ardusimple docs and pcb lables relate to xbee point of view not the f9p point of view, hence are reversed. Latest commit changes labels on the notes at top but not the code AFAICS but does not mention which point of view, f9p or Xbee Ardusimple’s.
1 Like
Latest commit corrects bug in relPOSHPlength, relPosHPD, relPosHPN relPosHPE. I have not checked this commit yet but the previous one gave much more reliable lengths and roll values once the code bug was corrected. Antenna distance typically within ± 5mm and rel down within ±1cm typically.
Setup for testing purposes of virtual antenna position with and without filtering using inverted pendulum:-
NB using a ground plane as shown significantly improved fix quality and rel length and roll.
Antenna spacing is 97 cm and height above pivot point ( virtual antenna position) 86cm
1 Like
I think the position antenna should report as fixed but the rel pos antenna will not be fixed as it is moving base which is slightly different.
I really don’t understand why they did this. Confusing if you don’t know.
I would like to install INO for ESP32 but when compiling I have this error:
Arduino: 1.8.13 (Windows Store 1.8.39.0) (Windows 10), Board: “Arduino Uno WiFi”
AOG_GPS_ESP32: 272: 10: fatal error: Update.h: No such file or directory
#include <Update.h>
^ ~~~~~~~~~
compilation terminated.
exit status 1
Update.h: No such file or directory
This report should contain more information if in
File → Preferences the “Show.” Option will be enabled
detailed information during compilation "
Prześlij opinię
Historia
Zapisane
Społeczność
Do you have selected the correct board type? Some kind of ESP32.
1 Like
Problem solved! I missed the ESP32 repository in the board manager.
1 Like
i have setup the dual antenna system for testing. I get heading from dual antenna and it is showing the right heading, but there is too much lag. It seems that esp is wotking only when its led is blinking.
New config files for the F9P on GitHub. Try them, might fix it
Do you mean the 1.13 config file for updated f9p?
1 Like
Hi, I have updated the f9p, uploaded the 1.13 config and now there is no lag. I am still testing on my table.
Now AOG is telling me I have RTK fix, but the two ardusimple board have the “no RTK” led always ON, is it normal?