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

Originally Posted by lolypop000

This topic is not forgotten, we still search for solution

Me too, but I am part of the manual workers,
Well, I keep checking the apparent accuracy of my 3790, became routine and fun, lol .. The difference is that now I always have other gps to compare, could be an 3790 original, but is a Etrex 30 or Oregon 450, also good references. For example, right now I just do static measurements for about 15 minutes and accuracy values ​​between 3790 and Oregon 450 were equal, yes! (less the height, disparate). Comparisons with the Etrex 30 made ​​at other times were more rigorous and in moving. In these situations the Etrex 30 almost always fared better, especially in motion, but the 3790 values were not as bad as the comparison I made weeks ago with another 3790 original (see video in previous post). Everything indicates that the effect of the real Cold Start + delete GPS folder was effective, although it has not yet matched, apparently, the accuracy of the 3790 original (lack do these tests with it). To better clarify recorded several tracks of both and at the same time, I believe that good indication discrepanciais between both. Opportunely, as soon as properly analyzed, will present the results.

Everyone knows, I believe, that the values present in gps are results of calculations of complex algorithms to determine the position with some degree of accuracy, from the raw data received of the satellites. Corrections satellites data vain from ionospheric distortions, orbits, reflexes, timing .. and also distortion of the oscillator, according to the different working temperatures of the device (internal and environmental). In other words, if we have the same hardware (3790) and the precision values ​​presented are quite different between two different firmwares, it is more logical to assume that the basis of calculation (tables, constants) or the algorithm of firmware of them are different. At the start of the topic has been discussed this issue from an important text of Engineering of Garmin (indicated by Catymag) which dealt specifically of a dynamic table with correction factors for the oscillator (which enables synchronization with the frequency of the satellites) in according to different internal temperatures of the gps. This table with the correction factor or factors, would be pre-installed at the factory. Question: would also be installed with the firmware of 3490 when flashing the 3790? Or it would be different calculation algorithms in both firmwares? Answer not very easy, but the most important is to know where and how to identify such region table, factor (constant) or algorithm and who knows thus be able to make some any "transplant".

Really looks like a great challenge, but I believe that I have a clue that may indicate at least the region: the mt.nvm file that is in the folder .System/GPS. The first hint is the proper termination of the file .nvm, it seems Non Vol Memory, but that's just supposition. Significant is that it is perceived that this file is created before of tunning of satellites, because its creation time is always 1 or 2 minutes shorter than the Arc.bin and Bee.bin files, that seem to be created at the time wherein the satellites are tuned. In other words, the mt.nvm is created from reading some region of the Non Vol Memory, and everything indicates that the region is not 41. So how to find out of which region he comes? Examining content's file so that the GPS is turned on, and seeking similar patterns in firmware gps, a way of knowing that the memory region they come. The data in this file, stored in non vol memory, plus current data of ARC.bin and BEE.bin files, should be the basis for the calculation of precision made ​​by the algorithm of positioning (accuracy), probably.