[time-nuts] Three-cornered hat on timelab? -> GPS
kb8tq at n1k.org
Tue Apr 18 08:03:15 EDT 2017
There are a number of survey grade receivers that pop up on eBay from time to time.
They rarely have “timing” features like an external reference and PPS output that both
work properly. There’s also a bit of risk buying them and finding out they are non-functional.
All that said, if you shop for a while (= years) you can get them pretty cheap. The latest
NetRS is coming in for $100.
SDR is still a bit of a confused arena. There are a lot of “hey I almost got this working”
sort of projects by guys in basements. There are a few papers by universities and labs
that got things working on L1. The “big guys” often use fairly expensive hardware and not
the < $500 sort of thing the basement guys do.
So far the number of SDR’s that are capable of multi-frequency operation are few and
far between. For about $10K you *can* get a RF device that would do the job. You then
run into the need to move the samples over to your computer. 10Gb ethernet (or multiple
10Gb’s) look like the leading approach for that. Based on various papers, your PC likely
needs > 20 cores to handle three bands / two systems on a parallel basis.
The “unexplored” alternative to an all in one multi band approach would be a multi
board approach. More or less, use one board each for L1, L2, and L5. You might
stuff in a PC per band / per GNSS system. How to split that up and get it all working
is not really clear. There’s also the “unload to an FPGA” approach that at least on
the surface makes a lot of sense. You likely would have an FPGA board per band and
per GNSS system. No matter how you slice it, its a lot of gear.
The guys at Swift seem to have a very new board out that does L1/L2. It’s reasonably
priced ($600 for the board, $2K for the eval kit). It is sort of open source. It is abundantly
unclear how they handle the clocking into and out of the board. At this point it is GPS
only with all sorts of things coming “real soon now”. It is SDR so upgrades via firmware
are a practical thing.
Past that there are a number of OEM cards that do this and that. Most that I have found are
a “fill out a nine page form for a quote” sort of thing. In general there is never a reply
after you fill out the form. The little I have found is that for about $6K or so you can
get a board that will handle L1/L2 with L5 coming real soon. GPS and Glonass are
supported with other systems “real soon”.
Since this is TimeNuts and not SurveyNuts, we are not the target audience for any
of these guys. That makes the whole digging process even more fun. Simply finding
out *if* the device has a working external reference or PPS may be nearly impossible
(= wait 3 months for an answer). The existing open source software is not (yet) targeted
at a timing application. That could easily change. A lot of labs are interested in this
same sort of thing.
All that said, right now LimeSDR looks like a pretty good option. One has to trust that
they *will* ship real soon now. Boards are in the wild and appear to work. The $300
cost is rational for a single band board. They should do what a BladeRF will do for
half the cost. The HackRF One appears to simply have too many issues. The Ettus
B200 / B210 would do the trick single band for around $1000. Their X310 with two
dual rx boards would do the whole thing (for $10K).
Lots of choices, lots of work. No clear winner in the race at this point. Anybody who
wants to send me the full X310 setup, contact me off list. I’ll pay shipping :) I already
have the computers….
> On Apr 17, 2017, at 9:33 PM, Hal Murray <hmurray at megapathdsl.net> wrote:
> magnus at rubidium.dyndns.org said:
>> You can fair better if you have a double-frequency GPS setup, as it can
>> first-degree measure and compensate the ionospheric shifts, which allows
>> for a benefit over L1 CA only receiver.
> Are any dual-frequency receivers available at hobbyist prices?
> What is the status of open-source SDR for GPS? Is this a good excuse to jump
> on that bandwagon?
> These are my opinions. I hate spam.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts