Using GPS with RISC OS

The receiver will first be looking for satellites.

Yes Rick, I should have added ‘This is also an early indication that serial data is being received by SatNAV.’ to further explain.
But related to /your/ findings, my GPS128 manual states that if the unit has been moved more than 500 miles since last switched on the acquiring will take longer. In this case one can enter the country so that acquiring is quicker. This suggests some sort of memory, perhaps I should open it up and check for a battery? (even if only for corrosion)

My GPS module has QSPI flash and a CR 1220 battery. This means it keeps its own RTC going as well as satellite position data for a few days. Another module has a 15uA drain to keep its memory going and no replcaeable battery – not sure whether its a capacitor that it uses.

I should open it up and check for a battery? (even if only for corrosion)

And yes, it has a coin sized battery soldered to the circuit board. It measures 3V and looks clean but will check it occasionly as it is getting old.

Rick said

The receiver will first be looking for satellites.

Nothing is sent on the serial until some satellites are found.

That is not true of all GPS modules. Some send NMEA sentences over the serial port with just time and date to start with, even though they can see no satellites.

Have made any improvements and found a unused tablet-holder in my garage…

Now I have a PiTab-Nav ;-)

Looks really cool. :-)

Was your battery running low? The screenshot photo, there’s the rainbow square at the upper right.

How long does it run on a charge?

Always on the hassle. ;-) A point which I work on.

I use two 2500mA/h powerbanks but if you look at the fist picture, I was using a 2,5A power supply (have 230V in my car for the professional use measuring instruments) and the square are also visible. Not sure. I have this “problem” only with the DSI display.

The powerbanks are working round about 4h with a activ GPS, than SatNav tells a HEX-error and I shutdown or plug in the powersupply.
With a USB surfstick online… round about 3h.

Charging takes 6h. Powermanagement is a problem I work on. At this time the LED of the powerbank (and also GPS) is visible behind the case. But is not a good solution. Not a really control. I have two other. The first with a “percent-display” without OS control should be possible easy but the second one with a I2C-control by the OS not for me. Hardware yes, software no.

Have “capture” my parts of “my way”…mp4 and mpeg … not a good quality but any problems are viewable.
All not optimized. Not sure if it is possible to disable not needed layers.

than SatNav tells a HEX-error

I think I fixed this in version 1.01? [The problem was an EVAL statement to check the checksum against the remainder of the NMEA sentence after the asterisk. EVAL(“&”+some text) was an unwise way to do this. Much better to check that the next two hex characters were what was eexpected.] Are you still getting it? The other thing I fixed in 1.01 is a serial bug that caused an indefinite loop with a buffer not empty but no character available. I now have a battery life of circa 36 hours with a 5200mAh battery and a 16×2 text display (only).

I will try 1.01 but I have no problems with 1.00. If I get the hex-error, the powerbank LED change the colour to “low”. So I think “accu down” was the reason. Can also be I am wrong and it was just a coincidence.

I am very impressed what already worked. Thank you very much.

Have it. I2C as I wrote. Look at the DSI thread.
https://www.riscosopen.org/forum/forums/5/topics/3591?page=4#posts-59762

I think I fixed this in version 1.01?

Have try “mobile” in my car. No problems. No HEX-error and after the last changes no “rainbow square”.

Version 1.20 of SatNav now released. The PiSerial drivers are no longer supported or included as there is now serial support on all platforms, in ROM, including Pi. An OLED display (128×64) and LCD display (16×2 text) are both supported if present on the IIC bus to display GPS information.

No changes for me. ;-)

Works fine with Pi-Top, PiTab and Pandora.

Thanks a lot.

I’ve read through the thread but I’m not sure which GPS units work with which boards?

The Adafruit unit says it does not work with a Pi 3, but Chris said higher up that might just mean it doesn’t work with Linux. Does the Adafruit unit work with RPi 3 and RPi 2 on RISC OS now, or only RPi 1? (Can’t remember all the variants, e.g. B+)

Chris mentioned another GPS device. Is that working now, and what boards can it operate with?

The Adafruit unit says it does not work with a Pi 3,

that is the Linux software had not been updated for Pi 3. SatNav is written in BASIC and so both Adafruit GPS boards work on Pi 3 under RISC OS. I have two other GPS modules and I think one of them does work (but I can’t be sure I remember correctly).

I use the “Adafruit Ultimate GPS HAT for Raspberry Pi” (Pi3 in the Pi-Top and Pandora) and the “Adafruit Ultimate GPS Breakout” (Pi A+ in the PiTab). Both working with RISC OS and SatNav without problems. Have also try other boards like Pi B, B+ and Zero, Beagle(xM) and MX6. Both GPS are also working.
In Linux the first one is not supported for the Pi 3 and the second one is officially not Pi supported.
Both need a external active antenna for better results.

Have try the latest SatNav and the latest RiscOSM (both download yesterday). The RiscOSM autostart and continous map redraw, if SatNav found the right position, no longer works.
Normally I start SatNav and if the satellite are tracked, RiscOSM starts automatic. Now I must start RiscOSM and than SatNav. The first tracked position is displayed but no refresh if I move.
If I use SatNav 1.20 all things fine.

Puzzled. SatNav 1.40rc10 and RiscOSM 1.43 seem to work together for me, exactly as before.

I only try with the PiTab. Can have the DSIScreen and the touch function via “mailbox event” a influence as a side effect with your screen support?
I have not a idea what I should do on a other way.
If I disable (power off) the GPS hardware, your “demo routine” works.

Is SatNav displaying updated positions in its window, when you move?

It looks like yes. For me as a “stupid user” no difference. The SatNav Window is closed if RiscOSM begins to display… If I reopen the “Sat points” are moving and the coordinates are changed.

There are some complications in 1.40 to account for different hardware. Another change is that the SatNav status window no longer closes itself after the first valid fix. Provided the icon bar icon is in place and shows (in your case) ‘No grid’ [meaning the OSGB grid system does not extend to where you are] but the status window shows the changing lat/long, it should be sending messages to RiscOSM.

However use in an area with no OSGB grid is something I haven’t tested yet… I’ll do a trial version that thinks there’s no grid available in GB and report back. It must be something simple…

“No grid”, I know, is right for Germany. Lat/long is changing, also the “Sat-points”. All the same (for me) like the older version but looks like sending/receiving the messages is the problem.
With SatNav 1.20 all things fine on my PiTab (Pi A+ with DSI). I use the Adafruit Ultimate GPS Breakout here.
I will try also on my Pi-Top. Pi3 and a Adafruit Ultimate GPS HAT.

I seem to have slightly cocked up the logic – I have tried SatNav 1.40rc10 with the grid function set to return ‘no grid’ whatever the lat & long. It does seem to have turned off the messages… Watch this space.

I have now corrected line 18540 – the geo: message and update messages should now work in locations where there is no OSGB grid value as well as the ‘first fix’ message (if RiscOSM is already running) which was unaffected. I had made a subtle ‘improvement’ that didn’t bother to update the position by messages if the grid reference was unchanged but forgot about places where the grid reference is always zero (out of bounds).

Version 1.40rc12 (4-May-2017) is now available from !Store and my web site.