[TangerineSDR] New Dual-Freq GPS Module
101science at gmail.com
Thu Apr 23 11:51:48 EDT 2020
Thanks! I suspected those numbers also were perhaps not actual accuracy. When I first powered it up it was showing .005, then .004 and after several days now .002 us but I agree it is probably suspect. I used 162stable32 to graph the data from u-Center and I also see the 4 nanosecond jitter over time. This is a mind boggling level of accuracy to me. Thanks for the tips John appreciate it very much.
Larry , K4LED
On 4/23/2020 11:40:57 AM, John Ackermann N8UR <jra at febo.com> wrote:
Hi Larry --
I think the offset and frequency values you're seeing are sort of
artificial -- the F9P is probably steering its internal clock to keep it
in sync with the GPS. It will always report close to dead on.
[ The Trimble NetRS has an option called "Clock Steering" that does
this, and they warn you to turn it off if you are using an external
reference. With an external reference, the processing results will tell
you the performance of the reference vs. the GPS system clock.
Unfortunately, the ZED-F9 doesn't provide for an external reference;
that would be really cool! ]
The jitter on the ZED-F9P or -F9T PPS output is about +/- 4 nanoseconds,
which is less than half what other receivers provide. So the raw PPS
output is quite a bit better than any other unit I've looked at.
It is remarkable how well the software "quantization error" message
predicts that jitter, and if you can apply the error message as a
correction value in software, you can reduce the noise by another half
order of magnitude, at least.
My strong suspicion is that using RTK correction will *not* improve the
hardware PPS output, because the PPS jitter is mainly caused by the
receiver hardware and there's just not anything inside the module that
cam improve that. An interesting question, though, is whether adding
RTK to the mix would improve the performance of the software
quantization error message. That's something I need to test.
On 4/23/20 11:12 AM, Larry Dodd wrote:
> John, All, FYI
> I have two SparkFun RK2 F9P boards for testing and learning. My original
> idea was to use one for RTCM correction data but using the internet
> NTRIP client alone is giving good results. Spectacular in fact. UBX -
> NAV - Clock status reporting .002 us and Frequency .000324 ppm. I
> received a TICC and some other boards from TAPR yesterday to learn more.
> I would like to eventually use the F9P 1PPS to correct my computer clock
> but not sure how to go about that - possibly using the RS232 interface?
> Thanks for all your professional work.
> Larry, K4LED
>> On 4/23/2020 10:05:07 AM, John Ackermann N8UR via TangerineSDR
>> Thanks, Gerry!
>> From the testing I've done on the ZED-F9P and ZED-F9T, the timing
>> performance is quite spectacular -- using software sawtooth correction,
>> an order of magnitude plus better than what are used to as "GPS PPS"
>> I just got my SkyTraq board yesterday and hope to fire it up soon.
>> On 4/23/20 9:51 AM, Gerry Creager - NOAA Affiliate wrote:
>> > I chatted with Gary Miller, the primary developer active these days on
>> > gpsd, about the uBlox Zed-f9 and the SkyTraq PX1122R dual frequency
>> > products. Gary's used more of the OEM boards than anyone else I know,
>> > and because of the ongoing work with gpsd, has a lot of hands-on time
>> > making code work.
>> > Gary's very impressed with the SkyTraq engineering, but said,
>> > essentially, that post-sales support and quality control were not good.
>> > Assuming this is a family forum, I'll not be literal.
>> > His impression of the Zed-F9 was a little muted, as he's not yet laid
>> > hands on one, but has extensive uBlox experience. He's of the impression
>> > that this system is likely both of higher quality and more readily
>> > supported than the SkyTraq.
>> > 73
>> > Gerry N5JXS
>> > On Fri, Apr 17, 2020 at 1:38 PM John Ackermann N8UR via TangerineSDR
>> > > wrote:
>> > Just FYI --
>> > I just discovered that SkyTraq has a new GPS module, PX1122R, that seems
>> > very comparable to the uBlox ZED-F9P, with a couple of exceptions noted
>> > below.
>> > It's dual frequency and has a built-in RTK processing engine -- that
>> > means you can feed it correction data and get corrected results in
>> > real-time. The claimed accuracy is 1.5m CEP autonomous, or 1 cm plus 1
>> > ppm when in RTK mode.
>> > There is a PPS signal with claimed 12 ns accuracy.
>> > The good news is that the module is $99 quantity 1 (they also have a
>> > breakout board for $125 and an evaluation board for $150).
>> > The bad news is that, at least from the data sheet, it's unclear what
>> > data outputs it provides -- there are a limited number of NMEA sentences
>> > listed, but it also supports RTCM 3.1 and "Skytraq raw data binary".
>> > I *think* but am not sure that the RTCM message output provides the raw
>> > data needed to create RINEX files and generate TEC (I'm happy for input
>> > on that point). And I don't know whether the binary protocol provides a
>> > sawtooth or quantization error correction message to improve the PPS
>> > quality.
>> > I ordered one of the evaluation boards to test, but depending on the
>> > data output capabilities this device may not be well suited for
>> > timekeeping or the TangerineSDR.
>> > Links:
>> > Data sheet: http://navspark.mybigcommerce.com/content/PX1122R_DS.pdf
>> > Module:
>> > Breakout board:
>> > Eval Board:
>> > John
>> > --
>> > TangerineSDR mailing list
>> > TangerineSDR at lists.tapr.org
>> > http://lists.tapr.org/mailman/listinfo/tangerinesdr_lists.tapr.org
>> > --
>> > Gerry Creager
>> > NSSL/CIMMS
>> > 405.325.6371
>> > ++++++++++++++++++++++
>> > /The way to get started is to quit talking and begin doing./
>> > / Walt Disney/
>> TangerineSDR mailing list
>> TangerineSDR at lists.tapr.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the TangerineSDR