GPS Time - Free Windows Tool

[Bicurico]

Hi,

I have purchased a cheap USB GPS receiver, mainly to get the accurate GPS time:

https://www.aliexpress.com/item/BEITIAN-USB-GLONASS-GPS-receiver-Dual-GNSS-receiver-module-antenna-FLASH-laptop-PC-1-5m-BN/32981069570.html

Because I did not find a tool that was easy to use and discrete on the screen, I did my own tool. Note that LadyHeather is a much better tool, but kind of too much for a simple GPS receiver...

Anyway, grab your free copy if you want. It is just one .Net executable.

Regards,
Vitor

[bob91343]

That looks interesting.  How does one use it?  I see no software listed.

What I want is not only accurate time but also frequency.  I have a rubidium standard but if this can do the same job more simply I might go for it.

[Bicurico]

You just execute the executable inside the ZIP archive I attached. Then you select the COM port of the GPS receiver. That's all.
Note that this tool is to be used with GPS receivers supporting NMEA at 4800 baud over RS232 or virtual COM port over USB.
The tool will ONLY read the time (and LAT/LONG) out of the NMEA protocol. Nothing more. It is NOT suitable for rubidium standard! For that, please use LadyHeather.

Regards,
Vitor

[bob91343]

Thanks Vitor!  While I don't like the answer, I appreciate the information.  I can still get time accurately enough from WWV but it's been a long time since its carrier frequency has been satisfactory for my use.  So my rubidium box is what I need.

My counter can now be relied upon for any of the frequency measuring tasks I have, with the rubidium standard as a reference.  I could use it as a master oscillator but for that purpose it uses too much precious bench space.  So it sits in the closet quietly until such time as I deem a recalibration may be in order.  That's better anyway, as its expected life is much longer.

[texaspyro]

Heather does have a nice full-screen zoomed clock/date display that can run off a GPS receiver or the system clock.   It can either zoom the digital clock or analog watch to full-screen.  The clock supports time zones and daylight saving time corrections.  There is also an audible tick clock mode that provides ticks on the second and/or minute... useful for setting other clocks.

If you have an environmental sensor attached, it can also display humidity, temperature, and air pressure. 

[bob91343]

Well yes apparently there are lots of functions available.  But my needs are simpler.  I only want two things: accurate time and accurate frequency, arguably the same thing in different forms.

Time is more complete, as it is basically an accurate frequency but referenced to a time frame.

I want time accurate to about one second, easily available with WWV.  I want frequency to perhaps one part in a billion.  One part per billion means uncertainty of one Hz at one GHz.  Close enough to set my radio dial.

I can measure frequency two ways that are accurate.  My HP counter will go up to perhaps 400 MHz and can resolve 0.1 Hz so that's better than 1 ppb.  My rubidium standard is fine for setting up the master 10 MHz oscillator to that precision.

Above that frequency I am limited by a cheap counter that can go to around 2.4 GHz but its reference isn't anywhere near that good.  To cross over, my signal generator has an OCXO that I can set closely, and after about 45 minutes warmup it is good enough for that precision.  What I do is measure its 10 MHz master oscillator with my counter (or compare it to the rubidium standard) while it generates 2.06 GHz (its maximum frequency).  So I can trust the 2.06 GHz to be within the same precision as the master oscillator.

While these things are of almost no practical value to me, I find it is satisfying to do such precision work at home with gear that I obtained cheaply.  It is, after all, a hobby, and if I get my jollies staring at a 10 digit frequency display or timing the revolution of a Lissajous pattern, at least I'm not bothering anyone.  And yes, I do come out of my cave and do other things, like weight lifting and playing jazz and working DX and cooking and yard work.

[Electro Fan]

Nice post and well said. 

My HP counter will go up to perhaps 400 MHz and can resolve 0.1 Hz so that's better than 1 ppb.

Which model HP counter are you using?

[bob91343]

HP 5328A I cobbled together from three nonworking units.  It has the 10811 OCXO.

[texaspyro]

You should try and add the time zone and DST functionality...  those make for a much more generally useful time display program.  Also, an alarm clock feature is quite useful... ahhh, the subtle wonders of feature creep...   before you know it, you will have a 100,000+ line program... now, what time is it on Mars in the Aztec calendar?   

[bob91343]

No thanks.  I am not interested in features.  Just time and frequency, thank you.

This stuff isn't to impress visitors.  It's just for my pleasure and, now and then, a bit of bragging to the very few who have any idea what this is about.  I did show my nearly stationary Lissajous pattern to one engineer who said wow but then he left and I am still playing.

Back in the day I would look in General Radio catalogs and see the 7 foot rack with the clock on top and think that I want something like that.  Well, I kind of do have that capability but I rely on an 'atomic' clock that receives, I think, WWV on some low frequency.  Its displayed time is within one second of the standard, just what I want.  And the rubidium oscillator gives me very good frequency calibration.

[bitseeker]

Because I did not find a tool that was easy to use and discrete on the screen, I did my own tool. Note that LadyHeather is a much better tool, but kind of too much for a simple GPS receiver...

Anyway, grab your free copy if you want. It is just one .Net executable.

Neat little tool, Vitor.

Bonus points for being able to set the OS clock. No need for Internet time anymore.

[Bicurico]

I must look into how to set the OS clock - no idea how to do that or if you need special permissions to do so.

By the way, I realized yesterday that I am commiting a calculation error when converting the GPS LAT/LONG positions...  Need to fix that ASAP.

As I said - this tool is just a simple tool to show the clock and generally see if the device is working. Inbterestingly I did not find any such tool for free, hence why I did this one.

Regards,
Vitor

[Bicurico]

Hi,

I fixed the bug in the LAT/LONG conversion from NMEA to Decimal format and, why not, added a new function that draws the variation of the geographic position:

The first position is centered in the image and then a line is drawn from previous point to each new position.

This shows the random error introduced for civilian use.

I could not resist uploading this picture, obtained by a 1 minute trial run: doesn't it look like a cool dude walking? 

It actually reminds me of Mortadelo y Filemon. Or Mort and Phil to English readers, Clever & Smart in German.

Perhaps just my imagination...

Attached new version.

Regards,
Vitor

[hwj-d]

Yes, Clever & Smart. 

[bitseeker]

I must look into how to set the OS clock - no idea how to do that or if you need special permissions to do so.

At least in the Control Panel, there is a shield on the button to change the time/date as well as the Internet time settings.

I fixed the bug in the LAT/LONG conversion from NMEA to Decimal format and, why not, added a new function that draws the variation of the geographic position:

The first position is centered in the image and then a line is drawn from previous point to each new position.

This shows the random error introduced for civilian use.

I could not resist uploading this picture, obtained by a 1 minute trial run: doesn't it look like a cool dude walking? 

It actually reminds me of Mortadelo y Filemon. Or Mort and Phil to English readers, Clever & Smart in German.

Perhaps just my imagination...

I see it, too. Pretty funny. The brain is very good at pattern recognition, whether or not it's technically correct.

[texaspyro]

I must look into how to set the OS clock - no idea how to do that or if you need special permissions to do so.

Under Win XP you do not need permission.  I think later Windoze versions require admnistrator rights.   Linux, etc requires super user status.

Lady Heather has a clock setting feature.  It can set the clock on demand or whenver the GPS and system time diverge by more than a set threshold.  Take a look at the Heather source code (function set_cpu_clk() in heather.cpp) to see how it does it.   There are some subtleties in setting the system clock accurately.

[SilverSolder]

This shows the random error introduced for civilian use.

"In May 2000, at the direction of President Bill Clinton, the U.S government discontinued its use of Selective Availability in order to make GPS more responsive to civil and commercial users worldwide."

https://www.gps.gov/systems/gps/modernization/sa/

[Bicurico]

Hi SilverSolder,

I know about that, yet the GPS position is only accurate +/- 10m.

When I leave my software running it will form a big cloud due to the randomness of the location received.

Of course I don't know WHY this randomness occurs and it may not habe to do with the Selective Availability.

Thank you for pointing this out, though.

Regards,
Vitor

[rsjsouza]

It actually reminds me of Mortadelo y Filemon. Or Mort and Phil to English readers, Clever & Smart in German.

In Brasil they were Mortadelo e Salaminho. Pretty cool comics that I read together with Asterix e Obelix.

(now, carry on with the normal programming)

[Conrad Hoffman]

There used to be a program called sawatch that might be interesting to look at. It's long gone, but there's still a page with a screen shot and some description- http://www.gpsinformation.net/main/sawatch.htm I used it for a while and when I tried to run it recently, my new machine said no way no how. That's when I discovered Lady Heather, which does a nice job even though I just use it with an old Magellen Explorist 210. My main interest is averaging to get a more accurate position. One question I've always had is how much waas helps with elevation. GPS elevation is pretty lousy. Barometers work better if you have a good reference to set them to. Up the road I have a handy HARN station- https://www.ngs.noaa.gov/datums/harn.shtml It's really nothing more than the usual brass benchmark set in concrete, but with a much better downloadable data sheet.

[Bicurico]

I actually know very little about GPS...

Here is a sample picture showing how the position changes over time. Of course this is a GPS receiver connected to my home PC and it is not moving at all.

The graph works like this: the first position is set as centre of the picture box. Then I just consider the relative change in position and draw it on the graph with line segments. The graph essentially shows the last 4 digits of the position.

Regards,
Vitor

[Bicurico]

Left it running over night - it is clear that the position hovers around, but could be optimized if one would take the average.

Never mind the red lines - these are wrong and due to false readings or, most probably, some bug in my software.

Regards,
Vitor

[bitseeker]

Your lab sure gets around.

[Bicurico]

Lol
But have I mentioned that my lab is now wife-approved?
After some home remodelling with new furniture she now loves the space as my desk is ingeniously hidden and no cables in sight.
I should become an interiour architect.

[hamster_nz]

Hi SilverSolder,

I know about that, yet the GPS position is only accurate +/- 10m.

When I leave my software running it will form a big cloud due to the randomness of the location received.

Of course I don't know WHY this randomness occurs and it may not habe to do with the Selective Availability.

Thank you for pointing this out, though.

Regards,
Vitor

The actual accuracy was designed to be +/-10m, 95% of the time. There are quite a few reasons for errors.

* Less than ideal reception (e.g. poor antenna placement)

* Without SBAS (Space Based Augmentation System), the current atmospheric, ionospheric and space weather conditions cause most of the 10m of 'standard 'error.

* The receiver picks which space vehicle signals to use in computing a fix, based on quality of signal (e.g. S/N ratio, height in the sky...). As it a selects to use different satellites the solution will change.

* The signal tracking loops are tuned for a moving receiver, allowing the frequency and phase to change as the receiver moves around and change direction, so they must have quite a bit of 'slop', making them susceptible to noise.

* Multipath / signal reflections can cause errors in the phase of the received signals.

* Uneven coverage in the sky - The orbit of the satellites never take them into the far north or south sky - you need a clear sky facing the equator for best reception.

* The satellites orbits and clock errors are predicted in advance, so are not completely accurate. This too adds to errors.

The solutions are pretty mundane:

* Use a good GPS module

* Use a good external antenna, with a good ground plane. Even a large coffee tin lid will make a good difference. Try it and see the improvement of S/N in the NMEA sentences.

* Make sure it has a clear view of the sky, especially towards the equator.

* Try to stay away from large building that might cause multi-path errors

* See if the module can be told that it is stationary, allowing it to use more appropriate parameters for tracking and filterering

* Use a RTK (Real Time Kinematic) receiver with an external service that publishes known errors for you (e.g. https://www.linz.govt.nz/data/geodetic-services/positionz/positionz-real-time-service)

I've just attached an image of a cheap module with poor placement on a windowsill over 58 hours. The green grid is 100m.