I have a question that i hope some could help.
I acquired a MDEK1001 and developed my own firmware to have more control on the system.
I programmed two devices to communicate in conditions 1 → (128 preamble, 6.81Mpbs, channel 5, 64MHz) e 2 → (1024 preamble, 110kbps, channel 5, 64MHz).
I saw here in some posts that MDEK1001 comes with the hardware adapted for channel 5 comunications and it is calibrated from factory (opt values stored). On loading firmware, i used thoose values for the configuration (antenna delay tx/rx, channel power and crystal trim value). I only did this values and correct one or another after fallowing the User Manual from Decawave repository. On the trim of the crystal i managed to make an adjustment on each good communication, so everytime i have a success range i validate the trim os the cyrstal.
I did a LOS test (outside of building - parking lot), on different distances and i want to show some results. The first plot is about mean error between distances and the second is about the standard deviation error on distances. Note that the ss-twr and ds-twr means single-sided and double-sided communication respectivally. Trim is about the crystal trim.
Did anyone get something similar like this or better? I mean, what could be possible more suitable to do to improve or if it necessary to do it? Since the error is positive it will get “better results” on NLOS situation and that will be the typical usage of my implementation. I know a lot of things could influenciate the results like ground signal reflection and even the difference from the devices hardware (in this particular LOS case)… but considering this results it is possible to say that they are “correct”?
Thanks in advance.