Moderator Comment by Chris Anderson on November 12, 2009 at 1:56pm

RTKGPS= Real Time Kinematic GPS, ie, high accuracy. More @ Wikipedia

Moderator Comment by Jack Crossfire on November 12, 2009 at 3:27pm

He seems to have made progress since the last blog post. Actually have a uBlox-5 & uBlox-4 that's not being used. If some of you can take routes other than 680, maybe we can get the commute short enough to do this project.

Developer Comment by William Premerlani on November 12, 2009 at 5:16pm

@Jack,
Are you suggesting it may be feasible to use a uBlox to implement Takasu's method? Can that be done by simply reprogramming the uBlox?
The math certainly looks interesting, and the measured performance is impressive.
Bill

Developer Comment by Doug Weibel on November 12, 2009 at 5:46pm

Is RTK the system that allows you to get attitude from gps? I saw some info from a vendor somewhere offering such a system. It consisted of three gps receivers that used a common clock and were physically separated by some distance (think full size airframe). I emailed them for additional info but got no response. Looked terribly expensive at this point, but something that will probably be in the DIY realm in the future.

Moderator Comment by Jack Crossfire on November 12, 2009 at 6:27pm

The supported GPS modules are given on

http://gpspp.sakura.ne.jp/rtklib/rtklib_releasenote.htm

uBlox 4 & 5 are supported. No Locosys or Sirfstar of course. He said he got 1cm accuracy in 3 dimensions. He also overclocked the uBlox 4 to 10Hz by processing the raw measurements in software.

Developer Comment by William Premerlani on November 12, 2009 at 6:28pm

Doug,
The method that Takasu has developed uses a single GPS to determine very accurate location by combining both modulation and Doppler shift information. It does not get attitude.
Bill

Comment by Mike on November 13, 2009 at 3:15am

Doug,
two RTK GPS units are commonly used in the marine industry as compasses for multi-beam bathymetric surveys and for DP (dynamic positioning) for dredgers and other vessels - similar to position hold. I haven't heard of them being used for attitude calculation but would imagine that they would have to be on a fairly long base line to minimise errors.

Mike
Mike

Comment by Marty on November 13, 2009 at 11:38pm

Keep in mind that one of the major outcomes of this experiment was replacing the standard antennas with geodetic grade ones. So even if the equations can be applied successfully to our application we will still need to find a suitable antenna.

Moderator Comment by Jack Crossfire on November 16, 2009 at 10:41pm

This was too good to be true. The only uBlox chip which generated the required raw measurement data was the uBlox 4 specifically in the LEA-4T package. The uBlox 5 has that disabled & my uBlox 4 is a NEO-4s. Obviously they disabled raw measurement specifically because of open source RTK libraries like this.

The LEA-4T is discontinued & cost a lot more than the consumer ones back in the day. All the other supported modules are in the $300 range. It's not even a hardware limitation but a software limitation.

Comment by Stephen Jack on November 17, 2009 at 1:54pm

Jack count to ten breath in and out slowly.

(Paparazzi wiki) “Sourcing from u-blox
u-blox keeps tight control over the distribution of their products. They must be obtained DIRECTLY from their own reseller offices. These offices may not be available in your area, for example Canada does not have a reseller. Sample quantities can be obtained from uBlox but overnight or 2 day shipping is required which drives the cost up considerably. While it is a large hassle obtaining these devices, it is undoubtedly worth it.
Talking with ublox sale for two years, Confirmed, that Order is possiable Directly from ublox, by knowing what project it was for & how was it to be use. After long reply waiting time, the answer was: - YES, but at least 2K-3K in volume, otherwise they're not interested. Like to share the order ? It is 500pcs/Roll.”
Other Source
http://www.rfdesign.co.za/pages/5645456/Products/GPS-Products/Antar...

The LEA4-T is still available.

The LEA5-T is a disturbing issue on several levels. Besides it having a bug which makes it useless for RTK the raw data operates at 2Hz where as the LEA4-T operates at 10Hz.

U-blox Data Sheet

“1.14.3 Raw data
The LEA-5T supports raw data output at an update rate of 2 Hz. The UBX-RXM-RAW message includes carrier phase with half-cycle ambiguity resolved, code phase and Doppler measurements, which can be used in external applications that offer precision positioning, real-time kinematics (RTK) and attitude sensing.”

I wondered whether Hexagon had bought U-Blox to start degrading the T series, so I did a search tonight. What I found came as a complete shock, Hexagon acquired Novatel in late 2007 if that does not mean anything then read through the list of receivers tested by T.Takasu and A.Yasuda.
http://gpspp.sakura.ne.jp/anteva/anteva.htm

If you don't know who Hexagon is they are a Swedish company that is very anti-competative, they keep prices artificially high in the machine guidance market by literally eating the competition. Leica was Swiss now is part of Hexagon, U-blox is Swiss; say no more (for now).

The LEA6-T is set to come out in early 2010
The Novatel replacement to the Superstar II is released this month, the OEMStar

Taskasu and Yasuda have made available the software tools they used to test the antennas and receivers in their research. It will be interesting to see how the LEA6-T and the OEMStar perform, and if Open Source is under attack.

Comment