Number of satellites to 0 after datatransfer

RE: Number of satellites to 0 after datatransfer

Didrik Rokhaug — Fri, 22 Dec 2023 15:24:37 GMT

[quote user="ralphvn"]When the nrf_modem_gnss_fix_retry_set() is set to 0 it will search infinite. What is the downside of this, has it something to do with the datatransfer or is it only out of power perspective hat you would put a max time instead of infinite on it? If I use a continues power supply and I want a location every hour is nrf_modem_gnss_fix_interval_set(3600) and nrf_modem_gnss_fix_retry_set(0) a good practice?[/quote]

Sorry, I misunderstood the question.

Yes, it is primarily for power saving reasons that you might want to stop the GPS after a certain time if it hasn't produced a fix yet.

The LTE stack has priority, so the GPS stack should not interfere with any data transfers.

If power consumption is of no concern, but GPS performance is, you can also use continuous mode, so that the GPS can keep tracking satellites between the fixes you are interested in. That way, when it becomes time to get a new fix, the GPS already has most, if not all, the information it needs, and can give you a fix much quicker.

When you use periodic mode, the GPS will turn off after it has produced a fix. The infinite search time is only relevant in the cases when it is not able to produce a fix. This means that when the GPS is turned back on, it will need more time to re-synchronize the clock, and it might need to download new data from the satellites.

P.S. Due to Christmas vacations, we will have reduced staffing next week, so responses might be delayed. We apologize for the inconvenience, and wish you a merry Christmas.

RE: Number of satellites to 0 after datatransfer

ralphvn — Fri, 22 Dec 2023 08:48:45 GMT

If it is downloading ephemerides or almanacs from the broadcast with an interval of 13/14 minutes that would be a good explanation, I will try to find out further.

I understand the quick search and the more intensive search now, thanks for that, but I do not think it is an answer to my question:

When the nrf_modem_gnss_fix_retry_set() is set to 0 it will search infinite. What is the downside of this, has it something to do with the datatransfer or is it only out of power perspective hat you would put a max time instead of infinite on it? If I use a continues power supply and I want a location every hour is nrf_modem_gnss_fix_interval_set(3600) and nrf_modem_gnss_fix_retry_set(0) a good practice?

RE: Number of satellites to 0 after datatransfer

Didrik Rokhaug — Thu, 21 Dec 2023 19:07:37 GMT

[quote user="ralphvn"]My last question is still open, if I put it on infinite search and the device is for example inside and will never get a fix, does this unlimited searching have any downsides except power consumption what is not really an issue for our use case?[/quote]

It might slow down getting a fix when the device moves back out. When the GPS is first started, it will first do a quick search, before it does a more thorough search. If the device is indoors while the quick search happens, it will move on to the thorough search, which will spend more time on each frequency before moving to the next. This means that if the search has "gone past" one or more satellites when the device moves outside, it will take longer before the GPS searches those frequencies again. If, on the other hand, the GPS is started outside, it might be able to find all the satellites it needs with just the quick search.

[quote user="ralphvn"]What are these unexpected GNSS interactions in between my nrf_modem_gnss_fix_interval_set(3600) setting? It is next to the configured hour cycle some other cycle that shows up every 13/14 minutes? It is not a big deal I think but I just don't get it, it also looks like it finds a quick faster then it does at my hourly fix interval?[/quote]

It might be that the GPS is downloading some data from the satellites, but gets interrupted by the LTE stack, so it has to try again.

See the note here for more information: https://developer.nordicsemi.com/nRF_Connect_SDK/doc/latest/nrfxlib/nrf_modem/doc/api.html#c.nrf_modem_gnss_fix_interval_set

Best regards,

Didrik

P.S. Because of Christmas vacations, responses might be delayed for the rest of the year.

RE: Number of satellites to 0 after datatransfer

ralphvn — Thu, 21 Dec 2023 10:19:54 GMT

devzone.nordicsemi.com/.../serial_2D00_terminal_2D00_2023_2D00_12_2D00_21.txt

I attached a logfile with debug output with some expected and unexpected behaviour. It logs every minute the modem temperature to AWS and every hour the location to the AWS shadow.

The start of the file shows the GNSS looking for a fix, this started at 17:00:00 and is expected behaviour, after 3:33 it has fix and send this to the shadow. After that it starts sending the temperature every minute, also as expected.

At 18:00:00 I expect a new event to find a fix but at 17:10:10 suddenly GNSS events are showing up. After that it prints out 6 times different fixes of the NRF_MODEM_GNSS_EVT_FIX event, why is this? Why is that? I do not see a NRF_MODEM_GNSS_EVT_PERIODIC_WAKEUP or a NRF_MODEM_GNSS_EVT_SLEEP_AFTER_FIX event. Like in the first situation, the SLEEP_AFTER_FIX is my event to make the device log to prevent it logging the 6 fixes to AWS.

Then it goes back to for expected behaviour of sending the temperature till this unexpected fix happens again at 17:23:41, 17:37:10, 17:50:10 and then the expected fix of 18:00:03, an unexpected one at 18:03:34 and one again at 18:18:10.

What are these unexpected GNSS interactions in between my nrf_modem_gnss_fix_interval_set(3600) setting? It is next to the configured hour cycle some other cycle that shows up every 13/14 minutes? It is not a big deal I think but I just don't get it, it also looks like it finds a quick faster then it does at my hourly fix interval?

RE: Number of satellites to 0 after datatransfer

ralphvn — Thu, 21 Dec 2023 08:54:33 GMT

I found out AWS_IOT_EVT_PINGRESP interrupts the GNSS, from +CSCON: 1 there is an NRF_MODEM_GNSS_EVT_BLOCKED event and the number of satellites goes to 0, after +CSCON: 0 and the connection is idle there is an NRF_MODEM_GNSS_EVT_UNBLOCKED event and the number of satellites goes kind of back to the old amount of satellites and it continues trying to get a fix.

My last question is still open, if I put it on infinite search and the device is for example inside and will never get a fix, does this unlimited searching have any downsides except power consumption what is not really an issue for our use case?