GPS progress....
I built a GPS front end using a MAX2769... alas it does not work and I do not know why. I have it connected to an FPGA (PK70 FPGA blade board) . I spent all day Saturday and Sunday trying to find sats in the data record. So this evening after work I took a working GPS, An old Novatel Superstar II, and wired on some jumpers between the front end and the correlator. When I record that data stream and change my constants for a different Fsample and Fcenter I get Sats! I can ! even compute a positionSo all my basic hardware chain is working, just not the front end that is an exact implementation of a commercial IC reference design. Not the part of the design I expected to be bad. Progress.... now to decide if I want to just shell out 2K for a small
Non COCOM GPS that can do what I want or if I should spend a month or so finishing what I started, and have something to offer the community.
Non COCOM GPS that can do what I want or if I should spend a month or so finishing what I started, and have something to offer the community.
posted by Paul Breed at 9:16 PM
11 Comments:
RTK GPS using a pair of max2769's & software would definitely pay off, but put uBlox out of business. The cost of RTK GPS is like mp3 encoders before LAME came out. It's way above where it should be for such a basic & dated technology.
I wonder what will happen if/when someone writes and open source high dynamic GPS stack? Will it be the cryptowars again?
Can't wait, keep up the good work...
It would be a great boon for the community. But I also understand the value of time.
Also, once you have something flyable, I'd be happy to contribute to flight testing on my 4" rocket. If the device is small enough, by May I should have a couple of 38mm minimum diameter dual-deploy rockets ready to go, as well.
Maxim has abstracts on using a laptop for software GPS but no source code. RTKLIB has been around for years but required discontinued $300 uBlox modules.
Whatever state you reach, any source code that converts the output of the MAX2769 into useful input for RTKLIB would be huge.
You can overwrite your old NovAtel SSII with open source firmware called GPL-GPS: http://gps.psas.pdx.edu/
I still think the best thing to do would be to record the raw GPS pulse data and do all the processing on the ground, when you have all the time in the world -- rather than try to do it in real time on the rocket. Flash memory is cheap and rugged. Put an antenna on the nose and the tail of the rocket, and you can get some attitude information too.
Why not save that project until you have more tme than money? Or when some obstacle will block you from actually building and testing rockets? Or when you just need a break from hot flamey stuff! :)
Can that really happen? :)
In my view everyone have to glance at it.
RTKLIB has been around for years but required discontinued $300 uBlox modules.
If you do it I'll have to start calling you "Batman" lol. I'd settle for a frontend logger. I just cant find the time for the curve.
Post a Comment
<< Home