edgetx
edgetx copied to clipboard
EdgeTX
Calibration screen improvements
Is there an existing issue for this feature request?
- [X] I have searched the existing issues
Is your feature request related to a problem?
The current calibration process is quite annoying - it doesn't do some things that it could quite easily, and doesn't always tell you everything you need to know. While I would also like to attack the general onboarding experience as part of this (partly due to the radio setup corruption PR), I think that needs to be done separately.
Describe the solution you'd like
Generally, it should be
- touch and hard key friendly
- tell you to calibrate the 6POS (until/unless that requirement is removed)
- to allow the 6POS requirement to be removed - data is needed for TX16S, T16, T18, X10 and X12 on the unfiltered raw analog values Spreadsheet with the data collected so far, and video on how to get the values needed.
When doing the calibration of analog inputs, it should, in the case of a first run calibration:
- tell you to move the gimbals in a specific order, so that the stick mode can be auto detected.
- prompt for the default channel order
When doing the calibration of analog inputs, it should, in the case of a re-calibration:
- skip any start screen - i.e. jump right to centering
- doing next/ENT on the final screen should not start the calibration process off again
tell you to calibrate the 6POS
that requirement really needs to go away! @rotorman we need data 🤗
that requirement really needs to go away! @rotorman we need data 🤗
Thanks to raw unfiltered analog output display in EdgeTX, we could easily use the community to gather some sample raw ADC values from radios that have analog 6-pos (TX16S, T16, T18) to make some stats to decide where the borders for detecting the correct position need to be. I am myself not near a radio with 6-pos for next 2 weeks, thus cannot help out quickly with some values at the moment.
On my three, (raw unfiltered analogs) - Inp 6 TX16S Mk1 0,284,594,914,1223,1546 TX16S Mk1 0,283,601,920,1238,1562 TX16S Mk2 0,257,547,834,1100,1390
on mine: TX16S MK2 40,281,570,860,1150,1432 TX16S MK1 0,283,599,907,1223,1528 TX16S MK1 0,283,594,907.1219.1531
RM TX16S MK2: 0, 263, 550, 847, 1116, 1402 Eachine TX16S: 0, 248, 539, 836, 1106, 1397 X12S 6 pos knob: 502, 814, 1018, 1357, 2040, 4091
RM TX16S MK1, 0, 293, 610, 928, 1245, 1561 RM TX16S MK1, 0, 281, 586, 899, 1205, 1519 RM TX16S MK2, 0, 264, 553, 844, 1112, 1403 RM TX16S MK2, 0, 253, 544, 836, 1103, 1393
These are the results so far : https://docs.google.com/spreadsheets/d/1p3J7AUZ4oDhM3nNM5vfk0bHZHL13SXwb59qaR2ldnQg/edit?usp=sharing
From the data so far (we are still missing T16, T18, X10. X12S is also represented by a single specimen), we see that a uniform classification borders across radios is not possible, but on TX16S (16 radios were measured) at least skipping the end-user calibration is possible - there are no overlapping regions:
On TX16S: Pos 1 - a value equal or lower to 140 Pos 2 - a value between 141 and 429 Pos 3 - a value between 430 and 729 Pos 4 - a value between 730 and 1029 Pos 5 - a value between 1030 and 1327 Pos 6 - a value equal or higher than 1328
TX16S 1: -100 0 2: -60 294 3: -20 612 4: 19 932 5: 59 1251 6: 99 1573
edit: Updated per followup comment with corrected values
On TX16S: Pos 1 - a value equal or lower to 140 Pos 2 - a value between 141 and 429 Pos 3 - a value between 430 and 729 Pos 4 - a value between 730 and 1029 Pos 5 - a value between 1030 and 1327 Pos 6 - a value equal or higher than 1328
And with @ulfhedlund s values, that won't work anymore :(
@ulfhedlund Is that from the third tab - the "unfiltered raw analogs" - as that looks more like the values from the first tab - "calibrated analogs"
You are of course absolutely right, sorry. The correct values should be, from 1 to 6: 0, 294, 612, 932, 1251, 1573
Phew! you had us worried there! Thanks for the update :) That brings things back into alignment, so Risto's value brackets should are still ok so far.
I tested four of my TX16Ss as well, here the combined results with the detection borders of all of our data:
Looks very good for hard-coded classification borders - all 6 positions over all samples can be separated correctly. The updated classification borders for TX16S (Gen.1, mkII and Eachine):
Position 1 - 138 and below Position 2 - from 139 to 427 Position 3 - from 428 to 719 Position 4 - from 720 to 1019 Position 5 - from 1020 to 1326 Position 6 - 1327 and above
Thanks! With these two TX16S additions, the classification borders remain exactly as 3 posts above.
Guys, I do not have the transmitter at home right now... anyway I have Radiomaster TX16S Mk II. I will try to send the values in 1-2 weeks when I get to the transmitter.
Thanks!
We really need more data from Jumper T16, Jumper T18, FrSky X10 and FrSky X12S users (currently only a single X12S data record from rburrow87 above in addition to numerous TX16S samples).
EdgeTX users just love quality and open systems :) Thats why we have so many TX16S samples :)
For X12s is the hardwarescreen not useble yet. Only emti values. The process of calibratin are not working...
@rotorman we have a live session planned for next Sunday with @1Sven