HontaiTec Quadcopters (HT F801, HT F803,...)

Thanks to the great work from Victzh emulating XN297 transceiver using our beloved nRF24L01, we have the chance to add a lot of new protocols using XN297.
I captured SPI traces from my HTF801 quadcopter using XN297. This quadcopter uses the same protocol as HT F803, HTF802 (not completly sure ) and his variants with camera or Wifi FPV.
I captured without bind with the quadcopter, but the captures are very strange, so not very sure I don't have a problem with my Saleae Analyzer. Anyone could take a look to confirm my captures are ok? Could be a problem related with a bi-directional protocol because I captured without bind with the Quadcopter?
If anyone could confirm my SPI traces are ok, I will capture more samples in a few hours with the Quadcopter binded.

www.dropbox.com/s/wuvjaiqnracp72f/HTF801...hout%20Bind.zip?dl=0

goebish wrote: This capture is correct.
Looks like the tx is waiting for a response from the aircaft

Ok, perfect! It's the first time I capture a bidirectional protocol.
I will capture some samples asap

Thanks Goebish!

Ok, I captured some samples.
1 capture with only bind process
After bind with the TX, I disconnected battery at the quadcopter (for security)
2 more captures, one of them full stick movements. The other one pressing all function buttons (Rates, Flip, Headless mode, RTH, video and photo).

dl.dropboxusercontent.com/u/14941708/HT%20F801.zip

New SPI samples added.
This time I captured an "Special" Bind sequence. This Quadcopter enter in calibration Gyro/Acc with a special bind process. You need to start the TX pressing the left shoulder button, and finish the bind process as usual (Throttle Up, Throttle Down). Starting the TX with this method, the TX send an special command to the quadcopter and resets his Gyro/Acc.
Also I added independent captures for every channel. First in low rate, then in high rate.

www.dropbox.com/s/xvrscbf2wv8sg5i/HT%20F801%20channels.zip?dl=0

Would you do the flip, headless etc again please? I wound up one short matching the changing bits with the function list. You've got the channels set up correctly so you can just export the csv (under the SPI settings button) instead of saving the logicdata file.

This protocol uses a 5-byte address and doesn't receive from the aircraft so not sure how easy it would be to add to CX10.

At least twoone pieces still unknown. Haven't looked at tThe checksum is X.25 according to reveng. And what appears to be the tx info in the bind packet is not the actual tx address used to send bind packets.Any chance of getting a bind capture from another tx?

Need captures of the trim adjustments

Binding happens on rf channel 0. Data phase alternates between 5, 19, 28 (tx dependent?)
Bind tx address is D2 B5 99 B3 4A (LSB first)
Bind packets are 10 bytes, data packets 12
Bind packets are sent for about one second
No retransmits, no AA.
Packet interval is about 15ms in both phases.
Low rate is implemented in the tx by limiting the channel range to 10-2F.

Data phase packet bytes
0: 0B
1: 00
2: 00
3. Throttle, bits 7-1: 0-7F, bit0-picture
4. Aileron, 0-3F (0 right), bit7-RTH, bit6-headless
5. Elevator, 0-3F (0 up), bit7-calibrate, bit6-flip
6. Rudder, 0-3F (0 left), bit7-video
7. Aileron Trim, -20 to 20, +left
8. Rudder Trim, +left
9. Elevator Trim, +forward
10-11. CRC-CCITT (X.25) MSB first

Thanks Hexfet!

I added new independent captures for functions:
Flips. I pressed flip button and then push right stick for every side to simulate flips at every side. So four different flips simulated.
RTH. I pressed RTH function and then pushed forward right stick to stop RTH. 2 times the same manoeuvre.
Headless. I activated and then deactivate headless mode 2 times.
Video. I started and then stopped video. 2 times.
Picture. 2 pictures taken.

dl.dropboxusercontent.com/u/14941708/Function%20buttons.zip

Trim adjustment:
I captured independent trim adjustments. Starting from center, then full to one side, full to the other side and finishing at the center.
I didn't capture Throttle trim because I guessed is not needed.
dl.dropboxusercontent.com/u/14941708/F801%20Trims.zip

About the chance to capture another TX bind, difficult to say... I will ask to RCgroups guys in the hope somebody could take another capture from his TX.

Thanks !!!!

Thanks, the individual button captures is much easier. Next time you can just upload the exported csv

Post is updated with the latest info. The picture button didn't have any effect - is it working on your aircraft?

For the yaw/pitch/roll min-max captures is "min" left/down and "max" right/up?

Sorry about don't share the CSV file I misunderstanding you
My bad english is a problem sometimes
I always guessed is better to post the logicdata file than CSV file.
I have the no camera version, so really can't test camera functions, but the TX beeps pressing his buttons. I guess function is working but really don't know...
The Yaw/pitch/roll trim captures order are written under every name's file.
Yaw. Starting at center. Full right, then full left and finishing at center.
Pitch. Starting at center. Full forward, then full backward and finishing at center.
Roll. Starting at center. Full right, then full left and finish at center.

I checked the manual about how to take pictures, and you are right!
Don't have picture function. The TX beeps pressing the left stick (pressing the right stick is start/stop video), and I guessed this function is for picture, but no, no function at all!
Sorry for the confusion...

Thanks!!!!

No worries. If the SPI is decoding properly the export csv is all that's needed. Saves the steps of loading/exporting later.

I've updated the into , but likely won't have time to work on an implementation until next week (if no one else does it before then )

Thanks guys!

I posted at RcGroups looking for help from any HonTai owner with a Logic Analyzer.
I will ask to a coworker who bought one F801 some time ago to gift them. Maybe he could capture SPI traces yet...

goebish wrote: Would still be nice to have captures from a few other stock transmitters to try to understand tx id / address / frequencies relationship.

I just started working on capturing the protocols for my HT F803C. I'm new to this so I will probably need some help. I should get in a proper analyzer next week.

I've already cracked opened the TX and found the SPI debug pads.

I've uploaded a Test Build for the devo10. Look for the HonTai protocol.

Code changes are in the hontai branch in my repo. Based on the CG023 code - nice scale_channel function

Compiling for 7e.....

Thanks!!!!

I will report in a few minutes, but it's too late here to properly test outside.

Just pushed an update. Had forgotten to change tx address for data phase.

Ooopsss...
The Hontai module is not building. I think you forgot to add the module at Makefile.inc
Is posssible?

Yes, I've pushed an update. I can't build 7e at the moment. Need to look in to using the 4.8 compiler.

No prob my friend I learn to compile builds yesterday
Compiling and flashing last update...