Building my ardsimple stack tonight , got me thinking…….should the ioref really come from one source (probably the esp32, as its communicating with both boards) to comply with good electronic practice?
Have anyone thought about to connect the esp32 dev Module via ethernet to AOG with the encj2860.
Are there any expierences?
Regards, Peter
No code has this yet…I am also waiting for this, Matt / Franz have been concentrating on getting the main calculation / code working stable first, but it is on the to do list at some point…Unless anyone else fancies having a go, I would love to get mine onto ethernet and off wifi asap!
Re RTK loss, may be worth considering the F9R with built in IMU should work to some extent with complete loss of GNSS or at least stop an abrupt step in position. Also if you are useing SNIP I believe it has a “near” function whereby it uses the nearest NTRIP base to the rover. Thus if one base station goes down it should revert to the second nearest.
I think that on the course from IMU he would lose the line after a few meters. For the accuracy we require for field work, an exact position is necessary.
I think the problem I have been having is more AOG locking up while ntrip client is busy trying to re-connect, than actually loosing RTK, I have noticed today things work fine, until AOG tries reconnecting, then it hangs the entire program up / causes delays etc, have posted in the bugs thread but may be something specfic to my setup.
On the whole, while everything is working, Dual Antenna is working fantastic for me now… Only problem really now is when i am in bad areas for RTK / signal comes and goes every 15-30 seconds, and then I have real problems.
Or maybe try to solve the problem with GSM coverage. You can use a router with an external strong antenna placed on the roof of the tractor.
Yeah have been looking into boosting the signal, I have setup strsvr to act as ntrip client / push the corrections to UDP via a virtual com port now, which means I can disable the AOG ntrip client, will try tomorrow, have a feeling this may well fix my problems.
I made the test especially for you!
After disconnecting internet access for AG
Results
- 50 seconds after turning off the internet:
AG displays a message about the loss of NTRIP (GPS status still RTK fix), the tractor is steer well. - after 4 minutes from switching off the internet, a single GPS appears, there is a gradual loss of accuracy and a significant drift of the A-B line
Within 4 minutes of losing my internet connection, I didn’t notice any signs of suspension or loss of steering accuracy.
Sounds too good to be true. Why would we update RTK correction data every second if 4 minutes would be good enough. How do you define “no signs of loss of steering accuracy”?
I conducted the test while sowing corn with the following set:
I assessed the lack of accuracy loss by observing the distance of the extreme row visually. I did not measure, so it was possible that there were some deviations but they could not be more than 10-15cm.
1 Like
Do you use USB or UDP to connect to autosteer board and GPS reciever?
I have a feeling this is possibly only an issue when using UDP, hence why not many are seeing it? I use UDP both for autosteer PCBV2 and also ESP32 dualgps…
I use USB and a NATEC hub. Often when turning on the main relay PCB boards disconnect but later when using the AS switch is good.
The tablet is Dell venue 11 pro.
Latest strsvr now has udp output, at least the demoB31 from RTKlibexplorer does.
Yeah I tried using that but couldnt seem to get the data into the ESP using that, not sure why / what I was doing wrong, using a virtual com port sending to UDP worked straight away though. Will test tomorrow and see how it behaves then. Im pretty sure it will work just fine, as it only seems to be the ntrip client causing things to hang up for some reason, without that, it should work just fine as it does when the signal is solid.
1 Like
On the video you have RTK fix. The tractor is likely driving at an accuracy of an inch. Perhaps the tool wanders a bit on top of that. I was suspicious about 10-15 cm accuracy on RTK float. For me float is sometimes accurate (the correct integer solution, the receiver is just not convinced) but often off a metre.
While trying to compile the code for AOG_GPS_ESP32 & Beta2 from mtz8302, my Arduino IDE throws a lot of “/*” within comment errors. They’re all in zAOG_EEPROM. At first I fixed the block comments but now I just delete all the code that’s commented out. Is my IDE too old or strict?
Often there is a code error somewhere “in front” of the code it is claiming as wrong.
Missing semicolon, " ) or something like this.
New version of the ESP32 GPS code on GitHub. I kicked out some of the unused code. The used libarys should all be part of the standard ESP32 package. Give a try.
I started to complete parts for the dual antenna and I’m not sure what esp32 board I should use. I found something similar to yours at local auctions. Will it work with your code?
https://allegro.pl/oferta/oryginalny-wemos-d1-r32-uno-esp32-wifi-bluetooth-8708090649