0

Ultimate GPS v3 altitude correction?
Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.

Ultimate GPS v3 altitude correction?

by dimenticato on Thu Sep 16, 2021 6:23 pm

Ultimate GPS v3 breakout board, PA6H module.
ESP32 WROOM Devkit 1
Arduino IDE for programming
Problem: My location is 2m above sea level on the Central California coast (36 degrees latitude).
The GPS reports the altitude as 16.70m with 8 satellites. It has connected to as many as 10 satellites.
I purchased a second GPS board to compare output, but no significant difference.
My project is for a high altitude balloon mission and altitude is critical.
Has anyone else had this experience? Can you point me to a calibration algorithm?
Any ideas appreciated. Thanks in advance...

dimenticato
 
Posts: 2
Joined: Sun Aug 29, 2021 1:41 am

Re: Ultimate GPS v3 altitude correction?

by ZapBlaster4000 on Thu Sep 16, 2021 7:14 pm

How high? Are you calibrating to the sea as you see it or to a USGS/NGS vertical benchmark?

Sea Level can be an ambiguous term, with MSL geoid and elipsoid differing by many meters and often times pretty far from the actual level of the sea.

Getting good vertical accuracy using GNSS can be tricky and you might need a more expensive board. I use uBlox ZED-F9P boards with both patch and survey grad antennas and typically can get within about 3m vertical accuracy tracking 24 to 32 satellites.

If you really need require better GNSS accuracy at any cost you could use a pair of ZED-F9P boards in RTK mode. There's a bit of a learning curve and you will need either unidirectional communication or logging/post processing for correction data. I have a realtime RTK pair running for over 100 days now with vertical accuracy right around +/- 4cm.

Anyway, I don't have much experience very high up but I suspect GNSS might not be the most cost effective or mass effective technology. Maybe a good barometer? I've used pairs of BMP390L with good differential results.

ZapBlaster4000
 
Posts: 11
Joined: Wed Feb 17, 2021 12:38 am

Re: Ultimate GPS v3 altitude correction?

by dimenticato on Thu Sep 16, 2021 8:13 pm

Nothing I've seen in the documentation leads to calibration. I started out with a BMP388 but could not get agreement with the GPS and neither was agreeing with sea level two blocks from my desk. At different times the altitude reported negative to the geoid but the longer it ran the more it stays too high. Thanks for the info on the ZED blox. good to have options :-)

dimenticato
 
Posts: 2
Joined: Sun Aug 29, 2021 1:41 am

Re: Ultimate GPS v3 altitude correction?

by ZapBlaster4000 on Thu Sep 16, 2021 9:01 pm

Sure thing, here's some graphs to give you an idea of the stability of the F9P w/RTK, notice the graph vertical scale is .05ft or about 15mm per division:
Screenshot 2021-09-16 175813.png
Screenshot 2021-09-16 175813.png (333.61 KiB) Viewed 36 times


And here's a two week interval with 1ft per division:
Screenshot 2021-09-16 180028.png
Screenshot 2021-09-16 180028.png (329.91 KiB) Viewed 36 times


I am using ESP32 to transmit correction data over WiFi/IP

ZapBlaster4000
 
Posts: 11
Joined: Wed Feb 17, 2021 12:38 am

Please be positive and constructive with your questions and comments.