* fix: SE and SF could not be set 3POS
*Somebody* (me) forgot ternary operator only compares two value in #1303
* fix: TX12 default contrast is too high
There is no 'perfect' value due to manufacturing variance, but 25 is simply too high,
and 15 a bit too low.
* ADC Filter - On/off state configurable at the model level
* Updated YAML configuration to match change from ADC Filter feature commit
* Companion changes to support ADC jitter filter to be a model configuration item.
* Renamed jitterFilter field to noJitterFilter.
This better describes the semantic, which is inverted.
* Updated after review
* Re-run generate-yaml
* Changed ADC Jitter filter control on Model to be tri state.
Off, On or use the radio setting
* Adde NOBACKUP() to the added field
* Re-generate yaml_datastructs
* Making the default state to use radio setting/Global value
* Updated after review:
Changing to more consistent naming across firmware and companion.
* Reverted files that was accidently added to the commit.
Co-authored-by: Peter Feerick <peter.feerick@gmail.com>
* [colorlcd] Ensure active internalModule is reset if different to radio level setting
* [B&W] Ensure active internalModule is reset if different to radio level setting
* storageDirty only when model setting cleared
* [b&w] Remove reusablebuffer, code re-order
* [colorlcd] More direct moduleType comparision
Consistent with B&W test
* Fixed support for Radiomaster TX12 additional switches (SI and SJ)
Two connectors are present on TX12 main board (near top of internal TX module) and are mapped as SI and SJ, but were not read out by the firmware.
With this update they might be used for some further hacking of the radio (ex. adding additional 2-pos switches, adding detector of external module insertion etc.).
* Radiomaster TX12 - preventing selection of Si and SJ as 3POS
* Radiomaster TX12 - corrected SI and SJ switch names in Keys Debug screen
Switches were displayed in test screen as SG and SH, due to fact that firmware did not take into account no presence of SG and SH in TX12.
The update provides correct and consistent naming of the additional switches across whole radio.
* Alignment tweak + fix rot_enc info overlap
Co-authored-by: Peter Feerick <peter.feerick@gmail.com>
* Relax assumptions about when Lua is called in foreground and background
* Start Lua cycle timer both in foreground and background
Do not assume that background/foreground alternates
* Lua is called at every Standalone GUI cycle instead of 5 Hz
Since Lua preempts every 50 ms or so, the GUI should remain responsive.
* In standalone script, only call Lua once per GUI cycle, to avoid messing up buffer handling.
* Revamped error reporting for Standalone and Telemetry Lua scripts
* Font size for Lua error is reduced from XL to L on color if LCD_W < LCD_H
* In drawTextLines, also possible linebreaks for opening brackets.
* Smaller font for error message on tall color screen
* Make titlebar height match font size on narrow LCD
has a slow inverter circuit by changing the UART sampling behaviour.
* It was found that the X9D+ and X7 (maybe others) have a slow inveter
that messes up reception of fast (400,000baud) UART signals. On the
scope the rise time of the signal is very long.
* See RM0033 24.3.3 Receiver / Selecting the proper oversampling method.
Page 631.
Also note that reducing the pull-up resistor from 10k to 1k at R82
(X9D+) / R406 (X7) will also help with the rise times. However on the
X7 changing the resistor is not enough, OneBit is needed too.
