About 37xx converted to 34xx satellites reception accuracy

**Neil** · 7th May 2013, 12:09 AM

Ok, i'm maybe gonna chuck '[Only registered and activated users can see links. ]'. My habit of making outrageous statements might be irritating to some ppl but i believe we need to be openly polarized about this 'problem', i.e. I think there are 3 main camps:

(1) those who think there is no real problem and therefore nothing needs to be done about it;
(2) at the opposite end are those who not only believe that the accuracy problem is real but who passionately seek a solution;
(3) in the middle are those who find this investigation interesting and are participating mostly out of interest and curiosity.

I started in Camp 1 as evidenced by my statement in post #6 linked above. I've now moved to Camp 3. I may be one of the few users here who have used GPS (i) on road in cars, heavy trucks & on bikes; (ii) outdoors both day and night [in military situations on foot navigating in difficult terrain]; and (iii) in aviation. I can state categorically right here and now that automotive use on road does not require the same degree of horizontal or vertical accuracy as the other two. Normal auto use particularly does not need highly accurate vertical readings, as evidenced by the fact that nobody except me has commented on the wildly varying altitude readings between the converted and unconverted units. Have a look at the readings. They generally don't agree. Maybe none are correct, i haven't checked but that's not my point. It's that the divergence is much wider than their horizontal 2 D divergence and i believe that is perhaps to do with hotfix probably not geared to work on altitude, only on coordinates. None of these unit rely on SBAS (WAAS) or LAAS or any other physical (real) means of positional error correcting. What they have is software help with that, e.g. lock on road, hotfix.

To some degree i think this is a 'storm in a teacup'. Nobody is going to die in a fiery plane crash because he used his 37>34 convert to make a rainy/foggy/night landing approach only to find out too late that he was 50' to the right/left of the centreline and/or 100' lower than he thought. Nor die because he accidentally stumbled onto an enemy firepit during a nighttime recce intending to set up an ambush or sniper post. Where such close accuracy is important there are specialized precision gps units designed for those jobs. I'm not being dramatic here, just trying to put this in context and perspective. So i believe this absolutely: Vertical (3D) precision is of little importance for auto devices and 15 metre/50 feet 2D accuracy is good enough. No doubt others hold a difference view and i'm not saying they are being precious about wanting 3 metre/10 feet 2D accuracy in an auto device. But if you do want it, and kunix cannot sort this out with a fw mod, then convert back to 3790.

Regardless, i'll do a side-by-side of virgin and convert in the city when i can. If kunix has come up with some new tweak i'll test that too. If not, i'll revert the 3490_kunix to regular 37>34 convert and include that just in case it's different gps hw to the others [as kunix said, maybe his mod didn't actually flash anything]. I only wish i knew someone locally with an original 3490 too. God know's dasivarsa won't send his over to me.

PS: Angelo, do you wanna buy my 3590? Make me an offer. I must say i'm a little surprised that the smaller-cased 3790 is more accurate than it.

**catymag** · 7th May 2013, 12:18 AM

@isg
Reverting to 37xx firmware, in my experience, is enough to get 37xx usual reception.
37xx -->34xx and other Nuvis comparison tests are welcome

Put them off road.

More interesting should be to know if someone has two 37xx to compare,may be one provided by a friend: one 37xx original and a 37xx/34xx.

**isg** · 7th May 2013, 05:48 AM

@ Neil
Yesterday I had a 500 km trip back home. The terrain differed from flat and open fields to mountain passages, many trees, tunnels,towns and a few cities. In driving conditions the best accuracy was 15 m. and this was best case scenario. Usual accuracy was 35 m + and that not depended from terrain at all. In open field, clear sky to hear Lost satellite reception on each 5 minutes is too much. So many times have left the road on gps screen that for me is a really bad feel to relay on gps in certain situations.
The bad thing in my case is that even go back to 3790 the accuracy is still questionable.
So count me in camp 2 and soon in camp 3 because I'll buy 3597 immediately after its appearance on local market. May be and because of above I'll take harder way and will format unit and will start from ground 0.

**Neil** · 7th May 2013, 06:32 AM

Originally Posted by isg

........... The bad thing in my case is that even go back to 3790 the accuracy is still questionable.
So count me in camp 2 and soon in camp 3 .....

The only reason i'm not in camp 2 is because i'm not having the accuracy problems you're having. My 1st convert functions well enough and it exhibits none of the behavour you and others are reporting. It seems to me that there may be gps receiver hw differences in the 37x0 productions which it's native fw allows for, but the 34x0 fw either barely copes with [i.e. mine and Caty's and some others] and other hw where it doesn't work well at all [i.e. yours and others].

Originally Posted by catymag

@isg
Reverting to 37xx firmware, in my experience, is enough to get 37xx usual reception.

I can confirm that info by Caty. After converting and reconverting many times, my 1st 3790 returns to reporting 3m precision. NV cleared both before and after conversion/reconversion each time. I haven't yet reconverted the 2nd 3790 [3490_kunix] to find out.

37xx -->34xx and other Nuvis comparison tests are welcome

Put them off road.

In fairness, although any comparison is welcome you will probably only show that a 37>34 lacks accuracy compared to any other unit. My old SP 2610 is better than 37>34 anecdotally. All my other units show better precision than 37>34 converts do i.e. 13x0/14x0/1690/2455 etc. I've also done informal tests with 2460 as original and converted to dezl 560 which reports mostly 3m both ways. Also 255/265 converted to 465 doesn't suffer any accuracy loss by casual observation, nor does 40/50 > 42/52. The 37>34 is different. The other 3 conversions will both flash and work equally well with original boot.bin or with the full fw flashed. The 37x0 won't flash using original boot.bin but will flash full 34x0 fw. Whatthe! Weird hey? You bet it is, and my layman's guess is that's simply because the 3790 and 3490 are not completely identical physical hardware. So flashing a 37 to 34 is like trying a kid's pony saddle out on a racehorse, you might get it to go on but it won't work property.

More interesting should be to know if someone has two 37xx to compare,may be one provided by a friend: one 37xx original and a 37xx/34xx.

Well, i still have one virgin 3790 [original & never converted] and 2 x 37>34. The only question is will i do the city test run with the 3490_kunix as 3490_kunix#2, as 3490 with unaltered 34x0 fw, or as 3790.
Suggestions are welcome, but if kunix comes up with another fw mod that's what i'll be using regardless [if it doesn't brick it of course]

.

**catymag** · 7th May 2013, 11:10 AM

Originally Posted by Neil

Well, i still have one virgin 3790 [original & never converted] and 2 x 37>34. .

Neil ,I lost count of your owned Nuvis ,lol you are our hero

Kanopus · 7th May 2013, 03:25 PM

Neil, if you are still willing, take as many 3790 as you have and record some tracks, so we could see what is typical deviation. Additionally you could swap nuvis in mounts. All this could help to establish some measurement reference.

**kunix** · 7th May 2013, 05:23 PM

You guys are so verbose

I've pumped through my brain so many english words today

Not sure I remember all the important things you've said.

Originally Posted by Neil

Suggestions are welcome, but if kunix comes up with another fw mod that's what i'll be using regardless [if it doesn't brick it of course]

.

I'm looking at this topic and your accuracy tests. I'm afraid the next mod would require digging into how fw_all.bin processes GPS chip output, as this seems to be the only place which could cause problems. So this won't happen too soon for sure.
I have a thought that maybe 34xx firmware processes signals from MT3328 chip in a different way (maybe somehow retarded way), as this chip is only present in older 34xx devices. It would be very interesting to find such 34xx device (with MT3328 inside) and run 37xx and 34xx firmwares on it and compare the accuracy with other 37xx and 34xx devices.
Also I don't like that using GPS firmware from 37xx with 34xx fw_all.bin makes the accuracy even worse. It could indicate that GPS firmware and fw_all.bin are tightly coupled somewhow and everything is even more complicated.
Also it would be nice to see the output of different levels of GPS chip signals processing. I don't really believe in track log, as it was already processed with different filters and probably adjusted to match the nearest road.

Originally Posted by Neil

The 37x0 won't flash using original boot.bin but will flash full 34x0 fw. Whatthe! Weird hey? You bet it is, and my layman's guess is that's simply because the 3790 and 3490 are not completely identical physical hardware. So flashing a 37 to 34 is like trying a kid's pony saddle out on a racehorse, you might get it to go on but it won't work property.

Yes, they seem to be quite different in some places. But 37xx bootloader doesn't work with 34xx fw_all.bin just because 34xx fw_all.bin is too big, no mystery here, and gfwpack solves this problem by making fw_all.bin smaller, though it's quite useless for us.

**angelozip** · 9th May 2013, 10:19 PM

Well, I am sorry once again contradict our competent collaborators friends. My conclusions (partial and subject to more tests, as always) were based on data collected in gps that I tested. As Neil has concentrated its tests in Kunix 's fw # 2, I decided to focus on #1. In despite of the statements of Neil and Kunix that the boot.ini of 3790 + 3490 fw_all is inefficient in accuracy, was not quite what I noticed in my tests. In fact the precision shown on the screen of this fw is similar to the 3490 full_fw, but the only Kunix_fw that proven unequivocally increased inaccuracy was the # 2 tested by Neil, as we see in the tracks that he published. In my case I compared the kunix # 1 fw (post 310) with a nuvi 3490 real and once again I was disappointed with the results, but this time not with the fw Kunix’s, but with the performance of my 3490, lol .. Turns out he had practically the same performance on 3790 mod with fw mentioned, same margin of error. Importantly, the tests were conducted in urban area with many buildings and trees, ie, areas of poor signal and also to emphasize that the GPSs was not set in the windscreen, but in horizontal position on the panel, really a condition not very suitable.

The basis of analysis were 2 tracklogs (per gps), recorded simultaneously in each gps (4 tracklogs total): off-road (pedestrian mode, 1 in 1 sec) and Lock on Road (trip log enable). Thus in addition to the “real” track record in the "Logs/Gpx" folder, we also have the track lock-on-road on “Current.gpx” recorded in the folder /GPX, that can be downloaded by Basecamp (the other track has to be imported from the Logs folder). Having two types of tracklogs from a single gps and recorded simultaneously can be very interesting to analyze. The lock on road catches the track, based on the map in use, and the off road track is independent of it. When we open the tracks in Google Earth with the track Lock on Road, we first give a margin of accuracy of the map, because if the tracks is locked out of the roads, certainly the map is misaligned (higher probability of error than photos satellites). But what matters is the right relation between accuracy "real" and apparent (shown on screen). We can observe behavior of these crazy tracks when the satellite signal is weak or is lost and its relationship with the "accuracy" shown on the screen usually with high values of inaccuracy. Garmin's software tries to position the vehicle in a road and sometime jump in other roads or with deviations amazing. Are situations that when we follow the routed gps navigation may involve errors such as wrong indication of streets to below, delays in recalculating when we left the route and other things not so pleasant symptoms that I and other users have experienced in practice with the firmware 3490 full in 3790. Fortunately the lock on road usually corrects these distortions, which bother more for those who make frequent use of the routing gps. Even with this drawback I prefer to use the firmware of 3490, to be clear, because it has advantages, for me the accuracy to drag the screen is the biggest one (very bad in 3790).

But it is from of the off road tracklog that we can extract information more effectively and realize the relation between accuracy "real" and apparent. When analyzing such tracks in Google Earth, the deviations of this tracks from the midline of the track “lock on road”, if it is centered in the middle of the road (therefore, map and satellite photo are right), it is possible to check the accuracy of the deviation in meters (measure tool of G.Earth). Most likely the average values of these measurements must match, or come close, the values in meters of "accuracy" shown on the screen of the GPS. Would have to make such observations at the time of testing, which is difficult for just one person that's driving.
Besides the comparative test between 2 gps with based on tracklogs I also did a practice test in routing with Kunix’s fw #1 in 3790. I drove for over 2 hours straight in a metropolitan area that I did not know following the gps routing and, surprise, there was not a mistake! Honestly I elect this firmware as the best I”ve ever experienced for 3790. I suggest that who have 3790 to try this firmware and repeat this tests, this version has no risk (not even have to install the "logo" of the Garmin 3490, it became something different, but I left anyway). For me 100% satisfaction, despite its "precision" of screen be below of the 3490 real.

The previous tests I did with 3790 with full 3490 fw were practical, day to day use, so would need to repeat the tests accurately analyzing tracklogs (now with real mode off Road in pedestrian). The problem is that lent this gps to a friend wich was travelling and only have back in 2 weeks.But the good thing is that I bought another 3790 and so in 15 days to have two 3790 and one 3490 to play, at least for awhile. At last, if there is a passion here, everything indicates that it is by knowledge and reason, a pleasure without nasty side effects, except, sometimes, when we went wrong our guesses

**3790 (V6 8GB 035 DA TC HP GM) mod Kunix #1 fw
3490 real (V7 2GB 036 DS TC GTL) sw 7.90

Legend tracks:
- 3490 off road: blue light
- 3490 lock on road: blue

- 3790 off road: red
- 3790 lock on road: red light

(click to increase)
[Only registered and activated users can see links. ] Path out of the official road, place of low reception satellites signal. Garmin software trying to lock on road with great distortion.

[Only registered and activated users can see links. ]Another example of track "lock on road" crazy due to loss or weak satellite signal which can result in errors while navigating with GPS routing. Unfortunately in this case was failure of 3490 real (blue track)

[Only registered and activated users can see links. ] Attempt to have approximate accuracy in meters from the track in relation to the terrain, taking the middle line of the road or track "lock on road"

[Only registered and activated users can see links. ]

[Only registered and activated users can see links. ]

**catymag** · 9th May 2013, 11:36 PM

I've put your tracks in GPSMapedit like I did with Neil's tracks

3490_tracklog_lock_on_road_5_5_13 Dark green
3490_tracklog_pedestrian_5_5_13 Light green
3790_lock_on_road_05_05_2013 Dark red
3790_pedestrian_ 05_05_2013 Red
some examples:

88zviWk

**Neil** · 10th May 2013, 01:47 AM

Originally Posted by angelozip

Well, I am sorry once again contradict our competent collaborators friends. My conclusions (partial and subject to more tests, as always) were based on data collected in gps that I tested. As Neil has concentrated its tests in Kunix 's fw # 2, I decided to focus on #1. In despite of the statements of Neil and Kunix that the boot.ini of 3790 + 3490 fw_all is inefficient in accuracy, was not quite what I noticed in my tests. In fact the precision shown on the screen of this fw is similar to the 3490 full_fw, but the only Kunix_fw that proven unequivocally increased inaccuracy was the # 2 tested by Neil, as we see in the tracks that he published. In my case I compared the kunix # 1 fw (post 310) with a nuvi 3490 real and once again I was disappointed with the results, but this time not with the fw Kunix’s, but with the performance of my 3490, lol .. Turns out he had practically the same performance on 3790 mod with fw mentioned, same margin of error. ...........

Well, actually what I said was that from my brief side-by-side of Kunix's 37 boot + 34 fw_all tested [kunix #1] against my conventionally modded 37>34 is that they appeared to have the same accuracy, not 'the boot.ini of 3790 + 3490 fw_all is inefficient in accuracy'. From my [Only registered and activated users can see links. ]:

Sat accuracy seems the same as my other convert, varied 4 to 10 metres today in mostly overcast suburban running (no tall buildings) in half-hr test. This unit and my other 2 showed rock solid 3 m before as 3790s which i find to be unrealistic in any case [just artificially enhanced with hotfix?]. From full off, about 4 mins until 'ready to navigate'. From standby 60-70 m quickly backing off to 8/9 m between 30 sec to a minute. I'd suspect hotfix is not working on these converts because while navigating there seems to be no problem, at least with both my 37>34.
Sat accuracy seems the same as my other convert, varied 4 to 10 metres today in mostly overcast suburban running (no tall buildings) in half-hr test. This unit and my other 2 showed rock solid 3 m before as 3790s which i find to be unrealistic in any case [just artificially enhanced with hotfix?]. From full off, about 4 mins until 'ready to navigate'. From standby 60-70 m quickly backing off to 8/9 m between 30 sec to a minute. I'd suspect hotfix is not working on these converts because while navigating there seems to be no problem, at least with both my 37>34........

Keep in mind that my conventionally modded 37>34 [full 34 fw] and Caty's both have far better reported accuracy than yours and isg's. I still regard my conventionally modded 37>34 to have acceptable accuracy. I also absolutely conclude that yours doesn't. I also continue to ponder the effect of hotfix not working on any of the converts [if that is the case]. I wish i had a real 34 to compare, but the 3590 was the closest i could think of and it didn't appear to compare well to 3790 original. The other thing is that I have never properly tested my 2nd 37 as a conventional convert, only as both kunix's mods, so i don't know if it's accuracy as that will be 'good' or 'bad'. I do want to keep the virgin 37 totally original as a standard, so that only leaves me with 2 other units to compare as various mods.

The ultimate definitive test imo would be simultaneous track logs of (1) 3790, (2) 3490, (3) 37>34 conventional convert [with reported good accuracy like mine & Caty's], (4) 37>34 conventional convert [with reported bad accuracy like yours & isg's], (5) 37>34 Kunix #1;(6) 37>34 Kunix #2, and (7) [if it happens] 37>34 Kunix #3. But who's gonna manage that, unfortunately.

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

Thread Tools

Display

Advertissements

Bookmarks

Bookmarks

Posting Permissions