TARANIX X9D: New hardware platform for Deviation?

Hi guys,

Came across this

www.frsky-rc.com/download/view.php?sort=...0For%20Taranis%20X9D

From having a quick look at the source it is based on the STM32F2 processor using the coocox operating system and its open source development environment. It might be a suitable HW platform if we get issues with the Devention HW.

You'd have to port it to the new mcu (maybe trivial, maybe not), and redo all the gui.

Taranis display is 212*64, but could be very interesting if a skilled programer would find some interest in doing it.

And as you say, if ever devo line gets deprecated, it's an interesting candidate, price is good to.

If I could fly devo, dsmx and frsky I'd a very happy camper indeed.

blackmoon wrote: You'd have to port it to the new mcu (maybe trivial, maybe not), and redo all the gui.

Taranis display is 212*64, but could be very interesting if a skilled programer would find some interest in doing it.

The MCU/CPU is the same family, runs the same code, so porting will be mainly the HW layer/drivers. The work done for the larger Devo 12 Display should carry over to the new display as well.

But as you say, that is a job for another day.

@PhracturedBlue - can it be because in one of particular PA CC2500 modules either TXEN or RXEN should be active at one time, but not both? Some PA/LNAs interpret it as forbidden state and turn off the both amplifiers. We need an extra pin (or some kind of signalling) for this.

I understand. I don't have any experience with CC2500 and don't have any models/RX's for it to test myself, and my recent debugging of nRF24L01 showed that there are intricacies of the chip functioning which are not well documented, so it takes time to feel it. So if you have time for this, I'd concentrate on the universal board design with the ability to drive these extra pins accordingly.

victzh wrote: I understand. I don't have any experience with CC2500 and don't have any models/RX's for it to test myself, and my recent debugging of nRF24L01 showed that there are intricacies of the chip functioning which are not well documented, so it takes time to feel it. So if you have time for this, I'd concentrate on the universal board design with the ability to drive these extra pins accordingly.

Victzh

Sorry to disturb your ultra-specialist talks but I got a question concerning the V2x2 protocol. In pratice, on the RX board of such WLtoys quad, you can find a BK2423 chip.


I read carefully the techdata, and I am wondering if they activate the "high sensitivity mode" (RX_SEN) in MSB of byte 4 of register 1.
Is this value hardly coded ? or it can be changed by the protocol initialisation ?. Sorry maybe my question is stupid...

@SeByDocKy - can't tell easily, I need to remember how I reversed V202 - from RX side or TX. I have their TX equipped with SPI pinout for logic analyzer, but I forgot about the RX. I need to check in my logs, I must have them collected on the RX side and they have the initialization code for Beken 2423.

victzh wrote: @SeByDocKy - can't tell easily, I need to remember how I reversed V202 - from RX side or TX. I have their TX equipped with SPI pinout for logic analyzer, but I forgot about the RX. I need to check in my logs, I must have them collected on the RX side and they have the initialization code for Beken 2423.

Should be great to know this point coz if they didn't activate this option, we could expect a range control extension for almost free ...

@PhracturedBlue - that's what I plan to concentrate on. I had a mock design for the boards and placement inside Devo7E (it's more challenging - less space) when I thought about making it a bit more versatile. But maybe even simple initial design would be good enough to try.

Victzh

This document is very interesting for the V2x2 protocol I guess :

www.inhaos.com/uploadfile/otherpic/AN000...bps%20Air%20Rate.pdf

Yea, that's interesting document - more detailed than their "datasheet" which mainly refers to nRF24L01.

What is your final purpose studying all this?

victzh wrote: Yea, that's interesting document - more detailed than their "datasheet" which mainly refers to nRF24L01.

What is your final purpose studying all this?

To know if for all V2x2 quad, they activated the extra sensibility of the RX. I found the actual range of these quad not fantastic (approx 100m, up to 300m after tuning) but for fpv, should be interesting to extend range if possible.

More, the speed of communication is hard-coded (250Kb/1Mbits/2Mbits) ? or it's evoluting according to RSSI ?. If actually it's hard-coded to 2Mbits, do we really need such speed to fly the quad ? 250Kb won't be enough coz there are 10dB of difference in sensibility between 2Mbits and 250Kbits. I am not specialist in electronic, maybe my questions are stupids

Unfortunatly, I don't have any AMtel to read SPIout of the RX board

This is exactly what happens upon the power up on V202:
0 W_REGISTER(CONFIG) 0C
1 W_REGISTER(EN_AA) 00
2 W_REGISTER(EN_RXADDR) 3F
3 W_REGISTER(SETUP_AW) 03
4 W_REGISTER(SETUP_RETR) FF
5 W_REGISTER(RF_CH) 08
6 W_REGISTER(RF_SETUP) 05
7 W_REGISTER(STATUS) 70
8 W_REGISTER(OBSERVE_TX) 00
9 W_REGISTER(CD) 00
10 W_REGISTER(RX_ADDR_P2) C3
11 W_REGISTER(RX_ADDR_P3) C4
12 W_REGISTER(RX_ADDR_P4) C5
13 W_REGISTER(RX_ADDR_P5) C6
14 W_REGISTER(RX_PW_P0) 10
15 W_REGISTER(RX_PW_P1) 10
16 W_REGISTER(RX_PW_P2) 10
17 W_REGISTER(RX_PW_P3) 10
18 W_REGISTER(RX_PW_P4) 10
19 W_REGISTER(RX_PW_P5) 10
20 W_REGISTER(FIFO_STATUS) 00
21 W_REGISTER(RX_ADDR_P0) 66 88 68 68 68
22 W_REGISTER(RX_ADDR_P1) 88 66 86 86 86
23 W_REGISTER(TX_ADDR) 66 88 68 68 68
24 R_REGISTER(STATUS) 0E
25 ACTIVATE(53) bank switch to 1
26 W_REGISTER(00) 40 4B 01 E2
27 W_REGISTER(01) C0 4B 00 00
28 W_REGISTER(02) D0 FC 8C 02
29 W_REGISTER(03) F9 00 39 21
30 W_REGISTER(04) C1 96 9A 1B
31 W_REGISTER(05) 24 06 7F A6
32 W_REGISTER(06) 00 00 00 00
33 W_REGISTER(07) 00 00 00 00
34 W_REGISTER(08) 00 00 00 00
35 W_REGISTER(09) 00 00 00 00
36 W_REGISTER(0A) 00 00 00 00
37 W_REGISTER(0B) 00 00 00 00
38 W_REGISTER(0C) 00 12 73 00
39 W_REGISTER(0D) 46 B4 80 00
40 W_REGISTER(0E) 41 10 04 82 20 08 08 F2 7D EF FF
41 W_REGISTER(04) C7 96 9A 1B
42 W_REGISTER(04) C1 96 9A 1B
43 R_REGISTER(07) 8E
44 ACTIVATE(53) bank switch to 0

You can check whether it sets the high sensitivity bit or not.
Speed is hard coded, moreover, nRF24L01 and, consequently Beken 2421/2423 do not have proper RSSI, only carrier detect, so such smartness in protocol is not for them

I'm afraid we're limited with what is there, and their (V2x2 series) avionics algorithms are better than all their cheap competitors. You can try to tune the antenna to be exactly 31mm starting with soldering place (taking into account board thickness) - someone said it matters. But the range is limited.

victzh wrote: This is exactly what happens upon the power up on V202:
0 W_REGISTER(CONFIG) 0C
1 W_REGISTER(EN_AA) 00
2 W_REGISTER(EN_RXADDR) 3F
3 W_REGISTER(SETUP_AW) 03
4 W_REGISTER(SETUP_RETR) FF
5 W_REGISTER(RF_CH) 08
6 W_REGISTER(RF_SETUP) 05
7 W_REGISTER(STATUS) 70
8 W_REGISTER(OBSERVE_TX) 00
9 W_REGISTER(CD) 00
10 W_REGISTER(RX_ADDR_P2) C3
11 W_REGISTER(RX_ADDR_P3) C4
12 W_REGISTER(RX_ADDR_P4) C5
13 W_REGISTER(RX_ADDR_P5) C6
14 W_REGISTER(RX_PW_P0) 10
15 W_REGISTER(RX_PW_P1) 10
16 W_REGISTER(RX_PW_P2) 10
17 W_REGISTER(RX_PW_P3) 10
18 W_REGISTER(RX_PW_P4) 10
19 W_REGISTER(RX_PW_P5) 10
20 W_REGISTER(FIFO_STATUS) 00
21 W_REGISTER(RX_ADDR_P0) 66 88 68 68 68
22 W_REGISTER(RX_ADDR_P1) 88 66 86 86 86
23 W_REGISTER(TX_ADDR) 66 88 68 68 68
24 R_REGISTER(STATUS) 0E
25 ACTIVATE(53) bank switch to 1
26 W_REGISTER(00) 40 4B 01 E2
27 W_REGISTER(01) C0 4B 00 00
28 W_REGISTER(02) D0 FC 8C 02
29 W_REGISTER(03) F9 00 39 21
30 W_REGISTER(04) C1 96 9A 1B
31 W_REGISTER(05) 24 06 7F A6
32 W_REGISTER(06) 00 00 00 00
33 W_REGISTER(07) 00 00 00 00
34 W_REGISTER(08) 00 00 00 00
35 W_REGISTER(09) 00 00 00 00
36 W_REGISTER(0A) 00 00 00 00
37 W_REGISTER(0B) 00 00 00 00
38 W_REGISTER(0C) 00 12 73 00
39 W_REGISTER(0D) 46 B4 80 00
40 W_REGISTER(0E) 41 10 04 82 20 08 08 F2 7D EF FF
41 W_REGISTER(04) C7 96 9A 1B
42 W_REGISTER(04) C1 96 9A 1B
43 R_REGISTER(07) 8E
44 ACTIVATE(53) bank switch to 0

You can check whether it sets the high sensitivity bit or not.
Speed is hard coded, moreover, nRF24L01 and, consequently Beken 2421/2423 do not have proper RSSI, only carrier detect, so such smartness in protocol is not for them

I'm afraid we're limited with what is there, and their (V2x2 series) avionics algorithms are better than all their cheap competitors. You can try to tune the antenna to be exactly 31mm starting with soldering place (taking into account board thickness) - someone said it matters. But the range is limited.

Thanks

I will try to understand and to match with the documentation.


According to your measurement 30 W_REGISTER(04) C1 96 9A 1B

the activation is performed on the third hex number, here 9A

and according to the pdf I linked p9; you have an example
when it's BE High sensibility is on when it's 9E it's normal ..
so I extrapolate since we have 9A and not BA it's in normal sensibility


EDIT : Ok RX_SEN is byte 21 and TX_PWR is byte 20. We have 9 for byte 23:20 so 1001 means that RX_SEN is set to 0 and TX_PWR is set to 1.

I don't understand why they activated TX_PWR for the RX board. Maybe in the binding process, there are some informations exchange ?