[time-nuts] Re: cheap GPS sim Re: Re: Another leap second problem

[Jim Lux]

And for example, testing spacecraft that will be in orbit, zipping along at some km/s. 

And therein lies some traps - Some simulators use “nominal” GNSS satellite orbits, which are different than “actual orbits”. Some have better or worse models for “receiver orbit”.  Some simulators have different models for ionospheric effects. This all makes a difference when you’re hoping to have 1 meter position uncertainty in your processed output. 

Actually playing back the signal isn’t the hard part <grin>.. 


On Sat, 15 Nov 2025 10:39:31 -0000, "john.haine--- via time-nuts" <time-nuts at lists.febo.com> wrote:

I used to work on a joint project with the firm that made what became the Spirent boxes (joint project was nothing to do with GPS). Chatting to the GPSsim product manager I wondered out loud why they were needed when you could just point an antenna at he sky (being deliberately naïve). He pointed out that (some of) their customers needed to generate an RF signal that would be received on a mach 2 jet fighter in a dogfight, rolling and looping like crazy, as well as meeting all the RF parameters. When I worked for a GPS chip company (not doing defence stuff) they had racks of the things though probably not the same level of simulation. Our customers were mainly automotive, IoT, some UAVs, asset tracking, timing etc.

-----Original Message-----
From: Bob Camp via time-nuts 
Sent: 12 November 2025 14:33
To: Discussion of precise time and frequency measurement 
Cc: Bob Camp 
Subject: [time-nuts] Re: cheap GPS sim Re: Re: Another leap second problem

Hi

Things to dig into on the Spirent boxes:

1) Some are very simple devices. You get very little control over things and a limited number of sat’s.

2) They have old firmware in them that may or may not be happy with the “modern era”. It certainly will not do the enhancements that have been added over the last decade or so.

3) They can be used for “roll back the clock” stuff. Is doing that with “legacy” signals an issue? Do you turn up bugs that aren’t really bugs? Who knows ….

4) There are fancier boxes in their lineup. They can do a lot of neat stuff. Some can do multi-band or multi-GNSS. Some versions do show up on eBay.

5) The fancier stuff appears to need a license for their PC based software to run the boxes. When I checked, that license was not at all cheap. I am not aware of any open source stuff that will do that job.

This is all from the last time I dove into them. That was about a decade ago. Things may have changed a bit.

Fun !!!

Bob

> On Nov 12, 2025, at 7:47 AM, Ed Marciniak via time-nuts  wrote:
>
> I’ve seen some Spirent GPS boxes show up that are affordable at perhaps $100 or so last I looked.
>
> I think they only generated a single L1C satellite signal on the one happen to have.
>
> I had no idea what the software interface is like on them, but it had a GPIB interface.
>
> They’re cheap enough that you could use several of them, and only 2RU height and not real deep or heavy.
>
> Depending on what your intentions are, it ought to be possible to receive a GPS signal from an antenna, run it through an isolator or two and apply the simulated signal via a directional coupler used in an unconventional way to the live signal if it helps reduce number of signals you need to generate. As a bonus, the native antenna signal could be applied to a timing receiver to generate highly accurate local clocks to drive the rest of the chain.
>
>
> ________________________________
> From: Jim Lux via time-nuts 
> Sent: Tuesday, November 11, 2025 12:59:17 PM
> To: time-nuts at lists.febo.com 
> Cc: Jim Lux 
> Subject: [time-nuts] Re: cheap GPS sim Re: Re: Another leap second
> problem
>
>
>
>
>
>
> Oh yeah, we have a bunch of "dumb playback" devices that basically have a 1 or 2 bit DAC for each band - you use the fancy sim to generate a signal, record it with the playback device and go to town.
>
>
> On Mon, 10 Nov 2025 22:55:53 +0100, Magnus Danielson  wrote:
>
> Hi Jim,
>
> Many of the GNSS issues that was suggested to be tested doesn't really
> need a very clean RF chain, and a HackRF One with some decent source
> and some open source code will have you generate signals. Testing all
> the nitty bits of the signals needs more code to fill in and fiddle
> with those bits, but it is not stellar development really.
>
> However, the bitrate of the data on the satellite carrier is 50 bps as
> modulated onto the ranging code, and that in a very strict structure,
> so many of these changes does not happen quick. That is why I proposed
> to inject the trouble after the actual GPS/GNSS receiver as an
> alternative to have that part tested quicker.
>
> The good RF chain becomes important when you want to calibrate
> receivers and work way down in the nitty gritty. Some of us do, but
> these simpler devices can clean the pipe.
>
> One GNSS-simulator vendor had a simpler device for L1 only signals
> originally, and didn't think it would sell to the heavier design houses.
> How wrong they where, they where very popular if you only dared to
> ask, as they could have a bunch of them spread out in the lab and let
> the big expensive simulator do more of the hard work while trivial
> things could cheaply be tested separately by each designer.
>
> Anyway, do not dismiss the simpler solutions, they may be just as
> efficient for many problems, even if they currently may not have all
> the tools in place to do that. I've intended to develop something like
> that myself, but you know, a few lines of code, how hard can it be,
> and spare time did not suffice.
>
> Cheers,
> Magnus
>
> Den 2025-11-10 kl. 19:26, skrev Jim Lux via time-nuts:
>>
>>
>>
>>
>> Interesting - one would need an SDR with very good frequency accuracy and stability, if you're doing anything other than just checking GPS software.
>> That's part of what makes the commercial simulators expensive - in theory, they have good quality synthesis chains.
>>
>> I'm not sure that the AD936x (used in most inexpensive SDRs as up/down converter) is of "time-nuts" quality in terms of the LO synthesis. Certainly one can feed in a good quality reference clock (derived from your H maser ensemble, disciplined by your set of Cs fountains, of course).
>>
>> At work, I have been fooling with a variety of cheap SDRs (real cheap - as in RTL-SDRs for $30-50) and they have all sorts of interesting internal compromises or settings in the synthesis chains which make no difference for their design purpose of receiving Over the Air TV broadcasts, but DO make a difference for other uses.
>> (I note as an aside, that there are people who claim to make
>> multichannel receiver setups for direction finding, who seem unaware
>> of the details of achieving phase coherence, even with common 28.8
>> MHz clocks)
>>
>> On Sat, 8 Nov 2025 19:04:23 -0600, Steven Sommars via time-nuts wrote:
>>
>>
>>
>> GPS simulator:
>> https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_osqzs
>> s_gps-2Dsdr-2Dsim&d=DwIGaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfi
>> iMM&r=JsDsKeR7cZC8wbZhIlxxBQ&m=5hIWzm0hDjmFgHok2NnqLq2_Po_LNQy3cXnrhE
>> zuFAu8Yo7ftk71UVUaBfq35Wpd&s=K8-VckBmmdzOZgDXCtqpkyd0LKFRdrN9V19POALJ
>> jTw&e= Thanks for the pointer. [I see used Spectracom simulators on
>> eBay for $1000-$2000. This is above my hobbyist budget.]
>>
>>
>>
>>
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com To unsubscribe
>> send an email to time-nuts-leave at lists.febo.com
>
>
>
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com To unsubscribe send
> an email to time-nuts-leave at lists.febo.com
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com To unsubscribe send
> an email to time-nuts-leave at lists.febo.com
_______________________________________________
time-nuts mailing list -- time-nuts at lists.febo.com To unsubscribe send an email to time-nuts-leave at lists.febo.com
_______________________________________________
time-nuts mailing list -- time-nuts at lists.febo.com
To unsubscribe send an email to time-nuts-leave at lists.febo.com
Â