Thread: About 37xx converted to 34xx satellites reception accuracy

Originally Posted by Neil

Thanks Angelo. I would have done the tests using off-road mode to disable 'lock on road' in suburban twisting circuits, but would have neglected to use 'Pedestrian'. I can easily do the test today in the outer 'burbs with plenty of trees, but if you require big city buildings that test will have to wait a few days.

I'll also enable the trip log in map layers so it's visible during the test, but doubt i'll have much opportunity to compare the 4 trails in real time as i'll be alone in the car and will need my attention elsewhere mostly.

In regard to the log data, do you want bin as well as gpx files? I would have logged only gpx, but if you also want bin file let me know what Binary Log types you want enabled. This stuff is over my head, i know how to view the gpx coordinates etc in G7toWin but that's about if for me. I'll need some precise guidance from you blokes if you want any fancy data to play with later.

In the absence of anyone replying with further directions, i'll start a half-hour test after about 4 hours from the time of this post [it's about 21:50GMT 04 May now].

Neil, thank you for your feedback, in fact I just gave some suggestions for the records, but particularly I am absolutely convinced the accuracy on the screen has direct relation with the record tracklogs. I've done several comparative tests with different gps tracklog of Etrex 30, Oregon 450 and 3790 not modified, and especially observations of in real time when there was an abnormality, such as when the vehicle jump to another street, for example. In this case when the accuracy was poor, evidenced a direct relation between them. I did the tests in various situations and on different days. I do not published because I thought it was surpassed a question and subject only to an interested of a person.. Anyway your work may be elucidator and all, once again, we have to thank for his good pleasure.

Is important to note that in general there are not major differences in tracklogs, but appears some aberrations and inaccuracies not pleasant when driving, like delay in detecting conversions, among other problems, especially in places of poor satellite acquisition. Obviously the tracklog show these anomalies with lines that do not obey the real path that we do and the map design (not always accurate).

Finally I ask where you found the option for enable/disable the off road mode in 3490 sw? Similar option I just found in Navigation/Calculation Mode, but from what I understand these options (Faster Time, Shorter Distance, Less Fuel, Off Road) are valid only for calculation of routes. In normal use the default mode remains lock on road. In my view, any other use that want to do from real off road mode has to be in Pedestrian mode. If I am wrong, please correct me ..

Originally Posted by Kanopus

Use gpx from hidden tracklog. Format of bin files is unknown, we can't extract data from them.

What about nutrak program?
http://www.gmaptool.eu/en/content/nutrak

I think both .bin & .gpx can be recorded at same time n'est-ce pas? Although i have never tried that out, both can certainly be enabled simultaneously in 37x0 and 35/34x0. If the info from bin files will be useful please let what binary logs to enable if any. If no reply i'll just do .gpx only.

PS: Sorry Angelo, forgot your question:
In 34/35 - Settings>Navigation>Calculation Mode>Off Road.
In 37 - Settings>Navigation>Car>NuRoute>Calculation Mode>Off Road.

I guess you right though and selecting Pedestrian Mode overides anything set in Car Mode, but i'll enable Off Road anyway [can't hurt anyway can it?].

Ok guys [& gal], he's the results. I've simply attached the track logs & screenshots rather than post them for obvious reasons. Sorry, but i didn't think to do screenshots until almost 18 mins into the test. Total time is nearly 40 minutes to there's a lot to compare on the logs in 1 sec increments. I sat for about a min or so to give the units time to properly lock. Prior to the test all units had nv cleared and are using the same map.

As i said, i'm a bit of a noob using track logs and making deductions from them, but it seems to me that they prove the lack of accuracy in the 2 converted units when compared to the 3590 and unconverted 3790 (Nos. 1 & 4 respectively). That is, the 3590 & 3790 appear to agree more than disagree. No 2 is the 3490 newly-converted flashed with kunix's rgn and No 3 is the 3490 convert which has been flashed back and forth 37>34 numerous times. The vertical accuracy in the converts is crazy-inaccurate. Even the 3590 & 3790 don't agree totally there, but they i think are within acceptable limits for automotive units (although i wouldn't want to be doing a precision instrument landing approach using any car gps, lol!)

3590 & 3790 showed mostly 3 m accuracy as expected, occasionally with one displaying 4 m for a mere couple of seconds or so [but never both together]. I still don't believe those constant 3 m accuracy reports. Hotfix 'mirrors & cigar smoke imo. Both converts averaged around 8 m mostly i suppose, but there was disparity between them at times, with them alternating showing the more accurate reading.

Conditions were semi-overcast. Except for the last straght run along a main road, I kept to a well-treed suburban area with lots of curvy streets and cul de sacs, so direction was changing often. Very tall trees in places. Before the straight run at the end, I took the car into a shopping centre underground car park and was amused to see 3 m still reported with no sats visible.

[Only registered and activated users can see links. ]
[Only registered and activated users can see links. ]

All Tracklogs Combined in GDB Format.

[Only registered and activated users can see links. ]

@kunix
That isn't a conclusion that i could draw definitively but it on the surface it does seem like the flash hasn't worked or it should show better accuracy than the other 3490 if they are equivalent units. At best accuracy is the same, maybe, but .... ? The big question is the gps hw in the two 37>34 converts. 'Coz the other consideration is are they the same gps hw? Buggered if i know. It could be that the one i flashed with your rgn is like the ones that angelozip, rfb et al have which would otherwise constantly show 12-18 m accuracy. I didn't test it as a conventional 37>34 convert before i flashed your rgn [Bad me! Lack of foresight is my strong suite]. If you want, i'll reconvert it and then convert it to 34 normally to retest. Maybe then it will show up to 18m av.? Otherwise i just may have been comparing 'an apple with an orange'.

Ok, let me know. [Painlessly trashed = still under warranty]