Configure fix-float time after losing corrections.
I’ve seen that after losing corrections through ntrip, the postcard RTK remains in fix mode for up to a minute or more on some occasions.
How can I configure the acceptable time to rely on fix mode after stopping receiving corrections?
Should I set it to 10 seconds, for example?
Do you understand what I’m saying?
Sometimes my translator doesn’t say what I need to say.
Regards
Angel
I just checked how long it takes to switch from fixed mode to single mode after losing internet and ntrip corrections.
It was exactly 2 minutes.
That’s a long time, and a fix mode almost 2 minutes after losing corrections isn’t a reliable option.
In professional point collection programs, that time is generally 10 seconds.
Yes, this is a setting for various modules. The ZED-F9P will loose RTK Fix after 30 seconds of no corrections, where the LG290P is 120 seconds. It can be set 1 to 600, but defaults to 120 (see command PQTMCFGRTK). I don’t think the ZED-F9P’s setting is configurable, UM980 is. It varies.
0x201100C4 CFG-NAVSPG-CONSTR_DGNSSTO
Default is 60, Range 0..255
Times out DGNSS derived from SBAS, RTCM3, SPARTN, SLAS, etc
Of course…this feature was introduced on lc29hea (PQTMCFGRTK) and improved into _v4 on 290p adding timeout.
Well, are they “Corrections” ? ie differences and rates-of-change, so a forward extrapolation of diminishing relevance.
Whilst this is true of older DGPS/DGNSS from RTCM2 beacons only using pseudo-range, I don’t think that’s the methodology at work here. The mechanics are notable different.
The RTCM3 MSM messages are raw observations from the base, and the rover is using these to affirm/contain carrier based measurement. If the rover maintains carrier lock, and doesn’t cycle slip, the containment is still sub-mm.
The receiver is going to degrade quicker out of Fixed and Float if it loses lock on more of the signals, and derates the accuracy estimate.
My confidence in these does diminish with time, but still we’re likely orders of magnitude better than the SBAS/DGPS
Thank you. I understand the LG290p firmware already allows you to configure the time.
Is this a setting available in Postcard version 2.1?
On the other hand, I’ll start running precision tests with known points in the 120-second time lapse and report back.
Regards
Angel
1 Like
@clive1 - Excellent! Thanks! Helps me find the docs faster.
@bamarcant - Yep, I just noticed PQTMCFGRTK in v1.0 of the protocol didn’t have the option. I see it is mentioned in v5 firmware, but not v4.
Above, v5 release notes (Quectel_LG290P(03)_Firmware_Release_Notes_V0105S.pdf).
@amlago - Sorry no, the max differential age of an RTK fix is not a setting in the RTK Everywhere firmware.
I think that might be instructive, if you can record/capture the self-reported accuracy vs physically observed one, that would give some insight into how effective/realistic the internal metrics are.
If you’re moving the ability to maintain carrier lock will be more challenging. The receivers should be able to score this, and knock down the fix-type more quickly than the time-out in situations that warrant it.
Trust but Verify..
@Sparky. Can this be added to a future version of the Postcard firmware?
@Clive1. As soon as I have precise data to share, I’ll post it here.
Preliminary accuracy test within 120 seconds after removing corrections from the Postcard.
The test was performed with the Postcard fixed at one point.
The corrections were then canceled via ntrip.
fix 812.328,76.896
22 sec 812.276,76.887 0.05
36 sec 812.268,76.894 0.06
53 sec 812.319,76.903 0.01
77 sec 812.291,76.944 0.06
118 sec 812.109,77.019 0.25
The differences were obtained by comparing the coordinates in fix mode.
A very preliminary conclusion that coincides with the time range of professional collectors. Do not exceed 10 seconds after losing corrections. The result at 53 seconds, where the differences were minimal, falls within the probabilities. I don’t think a measurement at 53 seconds without corrections can be trusted.
1 Like
This was a very good experiment and good info. I have yet to use the postcard to tie monuments into a survey because I don’t get the warm fuzzy feeling it being homemade.
I will say after adding more mask angle I am seeing satellites that I find more reasonable compared to professional gear I use.
Hi Aorpls.
It was a very quick comparison experiment to get preliminary data.
I’d like to set a more realistic time, something between 10 seconds and 30 seconds.
It would give me more security.
The professional equipment I own would never use a 120-second time, and I can still report in a fixed way.
I haven’t updated the postcard yet, and I’d like it to be easily configured via serial or Wi-Fi configuration.
I’ll have to request the update on GitHub.
Regards
Angel
@Sparky.
Will you consider adding this to the new firmware?
Should I request it on Github like we did before?