JD 395 cx-10

Sorry to disappoint, but it uses 3 different protocols and 2 of them use a chip not totally compatible with nRF24L01+. If you're willing to add another radio chip to your Devo - it's available yesterday (actually, for quite some time). The red board works with already existing Deviation protocol and nRF chip.

You guys never disappoint.

I didn't realized the green board used a different chip. That's lame. Especially since the red board was fine already. I thought only the blue board used a different chip. Thanks for the info and getting me up to speed.

Actually, by chance I have red and green board. I like how green flies better, despite flying red with Devo. Also, sometimes red flips for no reason, may be an error in protocol implementation and some stick values leak into flags.

Flag data is completely separate from stick channels in the YD717 protocols. If your red board CX-10 is flipping when channel 5 is less than zero, then we need some SPI captures from the red board CX10 to see what the protocol difference is.

Does the red board CX10 stock tx have a flip control? The original yd717 does not. The protocol data showed flag bits being set when in high rates, and it turned out the flag byte controlled the flips separate from the rate control. Maybe the CX10 firmware is not exactly the same.

I checked my config file - channel 5 is mapped to AUX4, which - you guess - happened to be positive. It leads to FLAG_FLIP to be constantly on and heli reacting to it, let's say, less predictably

And I checked behavior turning AUX4 back and forth - it is consistent.

Sorry for the confusion.

victzh wrote: I checked my config file - channel 5 is mapped to AUX4, which - you guess - happened to be positive. It leads to FLAG_FLIP to be constantly on and heli reacting to it, let's say, less predictably

And I checked behavior turning AUX4 back and forth - it is consistent.

Sorry for the confusion.

No worries. Thanks for testing. Wouldn't surprise me if some differences appear someday among the quads using the protocol.

I had another setback with XN297 - this time it does not work in my test setup, Arduino. I suspect it gets too much voltage, my Arduino works on USB power and does not have down regulator to 3.3v. Or I am doing something radically wrong - it just does not appear to be alive for me - no response, or one response only - then silence. I resoldered it back to check whether I ruined the module with 5v signals (I put 3.3v power, but Arduino uses 4.8v or so on signal lines) - no, the original TX works fine.

Not DeviationTx related, but I want to thank you closedsink for his Arduino code and Hexfet for his protocol Analysis.
Cheers guys!

I have an intermediate result - maiden flight of CX-10 green board with my TX code using nRF24L01+/Beken.

I will finish up the XN297 emulation layer and make Deviation module in coming days. There can be problems with bidirectional protocols - it's easier to generate acceptable packet than to decode with Nordic/Beken.

Great Job !!!!
So you manage to scramble the packet by software just like XN297 does...
My green board CX10 is waiting for your protocol !!!
Thanks for your research !!!

victzh wrote: I have an intermediate result - maiden flight of CX-10 green board with my TX code using nRF24L01+/Beken.

I will finish up the XN297 emulation layer and make Deviation module in coming days. There can be problems with bidirectional protocols - it's easier to generate acceptable packet than to decode with Nordic/Beken.

Wow, that sound great!
So probably finally we could replicate XN297 protocols using our beloved nRf24L01 in our Devo TXs?

Couple of kinks, and no blue board yet (I don't have blue board and as you know it is impossible to tell what they gonna send you). As soon as my Devo10 works reasonably well I'll put it in Builds subforum.

I wish too I actually plan this, but I'm a lousy blogger. We'll see.

Yes, the reading is totally untested, even conceptually. I think that it's possible, but not sure.

goebish wrote: Awesome !
I wish you make a blog or forum post with some in-depth explanations of this "hack" if you have a chance
The blue board is bi-directional (at least for bind stage), do you think there's a chance of making it work ?

What I can say He did amazing and tricky stuffs

victzh: Would it help if I mail you a blue board?

Generally, I prefer to have a model I'm deviating (though I had a good experience of making a model just by traces - SeByDocKy was very helpful in providing them for me). In this case I need to make sure first that nRF24 is able to receive ANY packets from XN297 at all. Theoretically it's possible, but not verified yet.

I may ask you about the board later, now I still have problems to solve with hardware I already have. But thanks for the offer.

Check out Builds subforum, I uploaded Devo 10, Devo 7E, and my Devo 10 model file for CX-10 green board.

The build is based on hexfet's code with my addition of XN297 emulation layer.

Please report success/failure.