Radiolink AT9 as a platform for Deviation?

Everything is possible, the question is what efforts it requires. Someone in this forum implemented a build with speech support, apparently there are text-to-speech modules on the market which can be used with minimal effort. I did not look into this though.

Ok, Thank you

If you're curious, I found this thread: www.deviationtx.com/forum/builds/5089-voice-alerts-test-build

OK, I have full color Deviation working on my AT9, including radio. I did not test it extensively, just bound and flew couple of helis, but it's functional. Thanks PhracturedBlue for your grandiose effort on providing UI for Radiolink's totally different controls, it feels natural.

Hey and thank you Victzh for the link. I will try this one's the deviation completed.
Btw I don't know what are the steps left to comptel It? The UI is completed now. We can use it, is the radio working? I can help you for the test if you want .

Sure I will, but the problem is - I merged common-proto in my at9 - I needed CS_HI/CS_LO code concentrated in one place, not spread over modules. I'll put my modifications into victzh/deviation/at9 and can make a pull request against main deviation/at9. I made a separate target, at9-mod, but I can try to merge my modifications into regular at9 - I have conditional HAS_4IN1_DL_SUPPORT anyways. DL stands for Discrete Logic - my imagination for names apparently failed me here

@anasm2010 what do you mean here?

AT9 Deviation port? It is far from completion, it basically works but is in a very fluid state. Radio works, but only with my board and 4-in-1 module from Banggood/TaoBao mounted on it. Full color works with the same board because it brings a substantial flash upgrade to AT9.

Sound support is another matter, I have not looked into it myself, just found a discussion on the forum.

Yes, I'm talking about the deviation port!
Thank you for your answer.
It's better to add the Board, as you did, to enjoy the more flash and make a better use of the color screen of the AT9? Or is It difficult to add (solder)?
Do you think that you will add the support for the AT9 stock protocol, or the deviation will just use the external 4in1 RF module?

To add the board is now the only way to make radio work, so you have option to add memory (around $1) to the board or not, but realistically not the option to add the board or not.

Stock AT9 protocol is not reversed, and may be hard to do so. They use 802.15.4 RF chip which has its own MCU on board. So the actual protocol is very hard to figure out. I have no plans doing this - many more interesting and more high impact projects are there.

I did play a bit with AT9.
The internal module protocol seems to be very simple - 41 bytes are exchanged, with channel values written and telemetry read out. It seems to be completely stateless - no binding, no TxId. It should be quite easy to support this module ...

But almost no point - we're looking for a replacement of Devo line of transmitters, apparently Walkera is not interested in developing it further and moving into FPV quads. Supporting this module does not buy much - you don't have any other source for it so you can't install it into any other TX. It uses a chip with MCU, so the actual protocol is implemented inside the radio module. So this internal module can not be used for anything else beyond its native protocol. To add fifth module to our collection for the privilege of covering one more rarely used protocol seems to be excessive.

victzh: PhracturedBlue accurately described my idea - I bought AT9 as second transmitter mainly for simulator and tariner use (with Taranis as primary TX). So modding it not that important for me now, but Deviation firmware is nice feature for me. And supporting existing module is another bonus ...

Hey,
Personally, I use the AT9 as my primary TX, and I have a multiple radiolink receiver that I get. That will be so cool If I can get the deviation working on It + The choice to use also the stock protocol of the AT9.
By the way, Is it possible to install the Deviation for testing on the AT9 without adding any hardware? and is It reversible to get back to the stock firmware?
Thank you

module TX format:
<0xaaaa><0x2901>[<CH1><CH2><CH3><CH4><CH5><CH6><CH7><CH8><CH9><CH10><FS1><FS2><FS3><FS4><FS5><FS6><FS7><FS8>]<CSUM>

0x2901 seems to be fixed and is not part of checksum. I will try different FW version to see if it does change.

<CHx> 16bit big endian, encoded as value 0 - 4000, center is 2000.
<FSx> 16bit big endian, failsafe value, either <CHx> for NORmal mode, or failsafe value for F/S
CH9 and CH10 are either 0 or 4000 and don't have failsafe. R9D RX does not output these channels on SBUS.

<CSUM> is 8-bit sum of <CH1>-<FS8> modulo 256

By default THR and ELE channel is reversed (at least with respect to what I am used to) - min value / pulse length is with sticks up. This is consistent with value sent over SBUS, I don't see any reason why ...

Module RF is enabled (RF-MODE: ON) by high level on REST pin. Module will communicate even when REST low.

Telemetry is quite similar, there is fixed position for RSSI and Rx voltage, don't have other sensors. There is no checksum.
There is bug in telemetry implementation - module will return correct value only if it is read immediately after preceding byte. When there is small delay between bytes, next byte is read instead. This happens quite frequently with AT9 firmware. New value is probably placed into TX buffer from wrong interrupt, creating race condition.

BTW: at9-fs.dfu is missing / is not included in nightly builds ... Flashing works fine when this file is build using docker. And flashing back to Radiolink firmware works too.