I don’t see, what’s wrong on your setup… Can’t help you. I think I will order 2 receivers and antennas tomorrow so I can do some tests…
Good. Please post your findings. I would love to get it working. For now I’m going to build Mathias’s esp32 version
I got some answer on NavSpark forum about data output:
At the moment we could only get the rtk position on one receiver and the necessary baseline data for heading and roll on the other receiver. I posted a feature request to give all the output on one receiver. They answered this is already planned. They want to add another PSTI, 0XX,… message on moving base (B) which outputs baseline data between moving base and rover (B - C). If this is released I only need to change some lines in the AOG code belonging the parsing of PSTI message and everything should be fine, I hope 
1 Like
It works!!!
I just tried my test setup some meters in front of the shed. Here are some pictures:
I got some prerelease firmware from NavSpark with a new PSTI 035 output on Precisely Moving Base receiver B and updated the parsing code in AOG on my laptop from 032 to 035.
I gave the RTCM stream (my own base) to receiver B through NTRIP and Bluetooth over my phone (laptop doesn’t have internet outside) and took the NMEA stream over USB serial to my laptop (black USB cable, white is not connected only for setup of the receiver, power for second receiver and connection between the modules over the jumper cables) .
The distance between antennas is 123.7cm in my setup. I simply glued the antennas left and right on the cobin roof with some transparent silicon. In AOG I have to put a Antenna offset of about 62 cm to the right as I get position from the right (precisely moving base receiver B) antenna and not the left one (rover receiver moving base mode C).
AOG got RTK position (max 5Hz), heading and roll. While driving little circles the heading was pointing the right way - as far as I know where North is - roll was to the side the tractor was actually “rolling” to. Both values changes very smooth.
For a real on field test I have to wait for the tablet to arrive tomorrow or the day after and for the rain to stop finally.
I will contact Brian to do the little change in code so it gets into the released version. Also I will give feedback to NavSpark that the prerelease version is working at least for my needs.
1 Like
Thats great news. can you put that parsing code in v4.1.0, and can I get ahold of that navspark update ?
Also, now that all of the tx and rx pins, namely RX2, are used to communicate between the two board, where will the correction from radio for rtk connect ?
I didnt get an error free version of AOG out of my Visual Studio… but I already asked Brian to do the necessary changes in the code.
Firmware: ADVANCED MOVING BASE MODE - NavSpark Community
No, they partly communicate over SDA and SCL:
You only have to communicate with the Precisely Moving Base Receiver (Receiver B in the picture). You get the NMEA stream over USB, TXD pin or Bluetooth and you give RTCM (Base correction stream) to the same receiver over RXD2 pin, USB or bluetooth, but dont forget to set the jumper right!
→ If you use an external wireless transmitter you have to put the jumper on position “A” on both boards. (“B” for giving RTCM over USB port 2 from RTKLib or AOG Ntrip client; “C” if using your smartphone to provide RTCM stream → But always position “A” on Receiver C)
The PSTI,035 message is missing in the picture… it gives the baseline data between Receiver C and Receiver B (On Receiver B TXD). (For understanding: You get the same baseline data named PSTI,032 on Receiver C TXD. [032 → baseline to previous receiver; 035 → baseline to following receiver])
Thanks. I downloaded a copy. I assume it is to load on moving base only ? I will try it with 4.3.10 to test as autosteer valve will not work past 4.1. Also the diagram you sent has external time stamp trigger, which I figure is a speed signal pulse, correct me if thats wrong. I looked for that diagram on the user guide, but the diagram is replaced again with the NMEA output from receiver C instead of B. RTK is still input to B though.
I loaded it on both receivers…
Sorry, don’t know what external time stamp trigger does…
As far as I know (NavSpark forum) you don’t get rtk position from receiver C, but didnt tested it…
To get data from receiver B worked good for me earlier.
Did you use this User Guide? http://navspark.mybigcommerce.com/content/NS-HP-GN2-User-Guide_rev0.4.pdf 20200821
Okay, I load it on both. I tested it when we spoke last and you are correct. No rtk on rover. I did look at that page again and it is rev0.3. Thanks again, I’ll post my results as well
The last days I got my tablet, built a tabletholder in front of the steering wheel and started some tests. I can report, that most things seem to work very fine:
- Position rtk-fix takes only a few seconds after powering the receivers plus some extra seconds for a working heading and roll (second receiver); all together it alway took less than one minute
- Heading and Roll looked very stable without big “jumps”… In NavSpark forum, they give these accuracy for heading:
2 meter separation: 0.07 degree
1 meter separation: 0.13 degree
50 cm separation: 0.27 degree - A big problem (I think its a know bug?) is the use of a large antenna offset which I need as the antennas are placed left and right on the cabin roof (124 cm seperated). I get big jumps in position in the view of AOG (I think it jumps exectly the configured 62cm in antenna offset). Below is a short video (I know there are some graphical bugs in my self built version of AOG) were you see this and one older post about this problem:
Tomorrow I will try my setup on seeding some grass. To avoid the problem with antenna offset I will try to use implement offset instead. I think that should work OK for now.
For long time use I will think about reposition one of the antennas (to the center of tractor). Then I will only have halve of the accuracy in heading and roll…
OK… I repositioned one of the antennas so the position send to AOG is now the middle of tractor. I also managed to build a AOG version for my needs without any graphical bugs or something. The values shown by AOG for heading and roll are still really stable and smooth - a lot better than I thought with my antenna separation of only 62cm now.
If only using position and heading everything works fine. But there is a problem with the calculation of antenna offset depending on tractor roll in AOG. If I turn roll from GPS on, the value and the animation on the top of AOG view look good, but the tractor position on the screen jumps left/right all the time… I pretty sure these jumps are exactly between antenna position and corrected position. Is about 5cm per 1 degree of roll wich fits with the correction at an antenna hight of almost 3m.
Below is a short video showing the screen while not moving the tractor but rolling the cabin from 0 to 1 degree. You see the jumps of about 4-5cm at the end of the video. On field I got jumps of 20cm and more with roll of 4…5…6 degree…
I read a lot of the code belonging these correction calculation and I think I understood most of it, but I can’t see where my problems come from… Is there anyone out there using roll from AVR (PTNL message, I think it’s from trimble)? This uses the same calculation in AOG which I also want to use…
1 Like
Is there any practical experience with SkyTraQ PX1172RH Dual heading receiver?
I just ordered one, so I’ll post how it works when it arrives.
1 Like
Well I can say this much so far. Shipping is very expensive. Fedex shipping to Canada from Skytraq is $65 USD (ships from Taiwan), and then on top of that Fedex is dinging me for another $40 CAD for “duties and fees.” Which is odd because I’m pretty sure electronics like this are exempt from duties. Ahh well. Fedex and UPS are some of the worst gougers.
Within the EU we have to pay a least import tax and fedex disbursement flat fee of 15€ flat (up to 600€) and if you don’t have an account you pay also collect on delivery fee €12.5!
Anyway, It would be very interesting if it’s comparable to two F9Ps or dual Mosaic H receiver?
Yes even with high shipping and fees–which I also have to pay for ArduSimple, but not as much for the Sparkfun version–it’s still cheaper than the double Ardusimple boards without the radio modules.
On the itemized invoice I see that the extra money Fedex asked for was about GST actually, not duties in this case, plus a convenient extra $10 fee that goes into their pocket. I guess getting paid once wasn’t enough for them. What a racket!
1 Like
Hi I am first timer trying to get my setup working properly. I’m using a v2 board and a SkyTraQ PX1172RH. The problem I"m having I can’t get AGOPEN to process the $PSTI035 message to calculate heading and roll. If I click on the antenna I can see the SPSTI035 message but if I go to system data I only see GGA and VTG no HDT information. Is there some settings I am missing to get AOG to parse the heading message?
After looking through AgIO source code I found NMEA.Designer.cs it parses $PSTI032 which is the wrong baseline info. (tractor to base station) I need to parse the $PSTI035 base line of the 2 gps antennas on the tractor. Could a simple change to line 211 fix this problem. I don’t know how to make the change and then compile to try it.
Both messages have same sentence structure.
