Flysky gimbal support for Boxer

[richardclli]

Fixes #3066

Summary of changes:

[richardclli]

Seems not working yet, discovered that the SD card DMA may affects the UART4 DMA, disabling the SD card DMA still no luck. Try to use the UART3 instead but seems not working as well.
@rotorman Can you spot any errors in my changes?

[rotorman]

The RCC_APB1PeriphClockCmd for UART4 needs to be enabled as well.

[raphaelcoeffic]

@richardclli you should not try to use TX DMA on UART4 ;-)

[raphaelcoeffic]

It should be much easier to setup once #3050 and #2631 are merged.

[rotorman]

Tested with Boxer:

1. analog gimbals still work as expected.
2. Flysky digital hall gimbals do not work yet, need to investigate.

[rotorman]

PA0 (UART4 TX) and PA1 (UART4 RX) on Boxer are on the right stick header (not on the left stick header as with T16/TX16S/T18 in https://github.com/EdgeTX/edgetx/wiki/Flysky-Hall-Sticks-Mod ).

Also the RX and TX pins are swapped between vertical and horizontal stick, thus in the image in https://github.com/EdgeTX/edgetx/wiki/Flysky-Hall-Sticks-Mod (here for T16/TX16S/T18):

For Boxer, need to swap the RX and TX as below:

I am getting data in on Boxer with Flysky gimbals. Need to check/fix their scaling though, but we are almost there.

[richardclli]

@rotorman Just spot that when you disabled the DMA_Stream for TX, the driver just skip the command sending, will this cause problem because the gimbal is not properly reset by the reset command?

[rotorman]

The reset command is not required, the Flysky digital hall gimbal will still stream out it's data. Thus IMO, TX will only be required for possible future gimbal upgrade, otherwise for normal operation only RX on the radio side is needed.

[richardclli]

I let somebody test the boxer with Flysky gimbal. It is quite strange that the raw values displayed do not have -ve values and the range only has around 2500.

[rotorman]

The raw values from Flysky digital hall gimbals are available to Boxer radio target OK, as can be seen in the raw analogs diagnose page (the range is from ca. -2500 to +2500, similarly to color targets). Thus the issue likely stems on how black-and-white targets process the values above 11-bit range. I have not yet had time to look into it, but will try check this in the near future.

[rotorman]

Installation of Flysky digital hall gimbals into RadioMaster Boxer, using the Hawk-RC Adapter ring set:

Step 1: mounting the bottom rings using stock RadioMaster analog gimbal screws:

Step 2: mounting top rings using 2x2 M2x6 screws:

Step 3: mounting Flysky digital hall gimbals HZCZ03-ADJ (here HCZC03 from an EL18) with 2x2 M2x6 and 2x2 M2x14 screws to the adapter rings and attaching the gimbal connection cable:

Do note that the RX and TX lines of the gimbal connection cable need to be swapped for Boxer, when compared with T16/TX16S/T18 cable, that comes with Hawk-RC adapter rings.

Here a close-up of the correct cable for Boxer - note especially the white (Gimbal RX) and yellow (Gimbal TX) data lines:

Further, please do note that on Boxer, the UART4 serial connection for digital gimbals is on the right gimbal header (when radio is looked from the back, then left side, as in the above picture), whereas on T16/TX16S/T18 the digital gimbals need to be connected to the left gimbal header.

End-result:

Video showing calibration and proportional channel diagnose pages output:
https://www.youtube.com/watch?v=ifFAXC0KMwM

[pfeerick]

Just a general though, as it seems counter-intutive to have non-radio specific code conditionally marked as such. Tie it to the feature, not the radio, and enable it at the target level perhaps?

[pfeerick]

Thanks for that @rotorman, looks a lot cleaner now :)