GPS-DO Project Planning

[cdev]

I hope this is helpful then!

S1315F-Raw is not a new designation. Older chips were Venus 6. I have a stash of many old Skytraq datasheets, as well as firmware for the Venus 6 devices I have on an older computer. I will have to look for them.

I think their software-based SPARC core gives them a flexibility other vendors don't have.

Many areas have free GPS data services. Where I live I can get free correction data in real time from both a nearby state that is much larger and my own state. The baselines are reasonably short.  Also my town has done surveys and calculated the exact locations of maybe a hundred of its fire hydrants and luckily there is one very close to me.  So I have access to high accuracy data to let me do a sanity check on my position fixes if I want to set up my own base station too.  I might want to do that in the future.

The S1315F's that I have are all the Venus 6 chipset-the current S1315F-RAW is a Venus 8 chipset.

I can't hardly find anything on these older ones, but they were cheap as chips so I have a handful. I'm still looking at clocking the ATmega32u4 at 10mhz with the TXCO, as suggested, that should save a bit of code and speed up the timing lock by double.

I wasn't able to find a method of using a 1hz phase lock that didn't take more room and complexity than using the ATmega to do frequency-locking, so I am still planning on the frequency-locked method.

You can also find a bit more info on the Skytraq chip on Michele Bavaro's blog at http://michelebavaro.blogspot.com/2011/04/skytraq-s1315f-raw-under-mangnifying.html

[cdev]

How cheap were they and are they still available there?

[Xnke]

They are currently available at Sparkfun for 11$/each, I bought 32 at 4$ each from a buddy of mine locally.

[cdev]

wow, thats really cheap. Well, you got a fabulous deal. You can use them for timing as well as make a super accurate GPS out of them.

By super accurate I mean accurate to a few cm, with a decent antenna and a clear sky view. You can set up your own base station and do RTK using RTKlib. They are extremely sensitive.

What firmware revision are they, can you launch GPS Viewer and extract the firmware and hardware patch levels?

[hamster_nz]

I did a little bit of playing over the weekend, and hooked a UBLOX NEO-7M to an FPGA board, and count the number of 100MHz cycles between the PPS signal. Using a DSC1033 for the 100MHz source, and logged the count for ~10 hours. Graphs of data attached.

It is worthwhile noting that the GPS module can be programmed to go into 'I'm in a fixed location' mode, which may improve the quality of the data a bit.

Posting it here in case anybody finds it of interest.

[cdev]

There is a plug-in for a wireshark-like program 'comms_champion' that promises to help document the ublox communications protocol - its on github.

[iMo]

It is worthwhile noting that the GPS module can be programmed to go into 'I'm in a fixed location' mode, which may improve the quality of the data a bit.

Does this work with NEO-7M too?

[cdev]

Definitely a Neo 7 T could. Its possible that a 7 M could, but its possible it doesn't. I have an 8M and I do not remember seeing a Kalman filter setting for stationary. To be honest with you I don't remember. Look in U-center or the Ublox developer information.

Look under "Kalman Filter" settings.

I really doubt if any non-T GPSs have TRAIM but even the cheapest $6 Skytraq navspark Mini has something thats functionally very similar where you can tell it that its stationary and it will maintain the 1 PPS as long as it has at least one satellite. When you consider how sensitive they are (so they almost always can receive a few sats) the difference is not so important to having a reliable 1 pps. (however as these cheap GPSs only have a TCXO, no ovenized oscillator they are jittery within that approximately 25 ns window.