H377 protocol

More
02 Apr 2014 03:06 - 02 Apr 2014 03:50 #22030 by ansheng
H377 protocol was created by ansheng
Dear All,
I have implemented the H377 protocol.
I have tested it on DEVO7E.
Could anyone help me to test it more and fine-tune the model6.ini.
Any question, please contact to me.

Thanks.
PS.
1. This protocol requires the addition of a ‘NRF24L01’ hardware
module to function.
2. DEVO10 please use devo7e`s protocol directory

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com
Attachments:
Last edit: 02 Apr 2014 03:50 by ansheng.

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 03:50 #22033 by ansheng
Replied by ansheng on topic H377 protocol
Dear All,
Because DEVO10 has more memory, need not the protocol directory.
Thanks...

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 04:45 #22034 by PhracturedBlue
Replied by PhracturedBlue on topic H377 protocol
Please provide the source code for your release. I have no idea what version of Deviation your changes are based upon here. Being able to look at the code is critical to being able to provide feedback.

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 04:59 - 02 Apr 2014 05:01 #22036 by ansheng
Replied by ansheng on topic H377 protocol
Hi PhracturedBlue,
I modified the code base on 4.0.1
Hopping frequency is fixed.
Because i don`t know the algorithm of h377 like the function "calc_fh_channels(u32 seed)"
Hopping frequency is according to the SPI log from H377 transmitter.
Do you have any advance about the algorithm of h377??
Thanks...

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com
Attachments:
Last edit: 02 Apr 2014 05:01 by ansheng.

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 05:15 #22037 by PhracturedBlue
Replied by PhracturedBlue on topic H377 protocol
Thank you for posting the code.
From what you've written, we can use any algorithm we like to choose the hopping sequence.

There are 2 likely paths:

1) Take a look at the cyrf6936 'CYRF_FindBestChannels' function. This uses the RSSI capability to determine the best available channels. I haven't worked much with the nrf24L01, so I'm not sure this is a viable option

2) Use the txid along with the LFSR togenerate a unique set of channels for each txid. You can look at the v202 code for an example (though I recommend looking at the latest code here, as we've unified the lfsr code recently):
bitbucket.org/PhracturedBlue/deviation/s...rf24l01.c?at=default

It looks like you've already got #2 implemented (though disabled). Is there a reason you disabled it? There is no reason to copy the hopping sequence from a real Tx if it doesn't actually matter.

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 05:46 #22038 by dc59
Replied by dc59 on topic H377 protocol
Hi ansheng,
Thanks for great job on H377 protocol.
Is it possible to provide Devo8S version F/W for test?

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 06:02 #22039 by ansheng
Replied by ansheng on topic H377 protocol
Thanks for your reply..
I think there are 3 important Variable:
binding_adr_rf : it is fixed
rf_adr_buf : according to lfsr
hopping_frequency : Randomization (in function calc_fh_channels(u32 seed)) , according to lfsr

If I use the those algorithm to calculate the rf_adr_buf & hopping_frequency,
DEVO can`t bind h377.

Fixed rf_adr_buf & hopping_frequency is the only way to success.
And according the SPI log from h377 transmitter, rf_adr_buf & hopping_frequency is fixed every time.
I don`t use any code about lfsr. Thit is the reason I marked the function update_lfsr.
I am afraid that those rf_adr_buf & hopping_frequency just bind my h377.
It can`t work on other h377.

Do you understand what i mean??
My English is not very good, sorry about that...

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 06:08 #22040 by ansheng
Replied by ansheng on topic H377 protocol
I don`t see any compile condition about devo8s.
I upload the F/W of devo8.
Thanks.

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com
Attachments:

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 07:00 #22041 by dc59
Replied by dc59 on topic H377 protocol
Thanks ansheng.
I will report result after testing on Devo8s.

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 07:25 #22043 by ansheng
Replied by ansheng on topic H377 protocol
Thanks

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 13:54 #22052 by PhracturedBlue
Replied by PhracturedBlue on topic H377 protocol
I looked at the code a bit more, and I don't understand all of it.
You have a 2-d array: bind_buf_arry but the 1st index is always '0'. This appears that you never send the hopping-table to the Rx, only rf_adr_buf.

It also appears that the frequency hopping occurs during binding. If so, I'm not sure how you could ever change the hopping table. I would expect either to use a different hopping table during binding vs flying or no table for binding if it is changeable.

Are you sure binding_adr_rf is actually fixed? could it be that it is related to rf_adr_buf?

With only 1 Tx available, it can be difficult to figure out this kind of stuff. I would normally try wiring SPI up to the Rx, and watching how it does frequency hopping after changing the bind_buf_arry values (if it does) to try to detect a pattern. Or find someone else with a 377 that you can snoop the Tx.

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 15:14 #22054 by ansheng
Replied by ansheng on topic H377 protocol
update model file...

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com
Attachments:

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 15:54 - 02 Apr 2014 15:55 #22055 by ansheng
Replied by ansheng on topic H377 protocol
Dear PB,
In bind_buf_arry, I just use the bind_buf_arry[0].
bind_buf_arry is used to send information about hopping table.
According to the log of SPI, transmitter just send 10 bytes to H377.
So I just use the bind_buf_arry[0].
Hopping-table has 20 bytes.
It has regular. That is why the information about hopping table just need 10 bytes.

Just as you say, rf_adr_buf and hopping_frequency are not fixed.
rf_adr_buf : according to lfsr
hopping_frequency : Randomization, according to lfsr
bind_buf_arry: according rf_adr_buf and hopping_frequency

I have only one h377 transmitter .
I just get a set munbers of rf_adr_buf , hopping_frequency and bind_buf_arry from SPI log.
If i have more transmitter, I could get more relationship between them.

I send the fixed number from transmitter to h377.
This is why that maybe I just could control only one H377 at the same time.

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com
Last edit: 02 Apr 2014 15:55 by ansheng.

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 16:06 #22056 by PhracturedBlue
Replied by PhracturedBlue on topic H377 protocol
with only one transmitter, how do you know how the hopping-table is built from the rf_adr_buf?

I still don't see how the Rx can find the Tx if you use the hopping table during binding. Looking at the SPI trace from the Rx would likely help.

You may need to wait until someone can grab SPI logs from a 2nd Tx.

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 19:20 #22058 by victzh
Replied by victzh on topic H377 protocol
RX seems to be using nRF24LE1, so tracing it can be problematic.

www.rcgroups.com/forums/showpost.php?p=26510850&postcount=9

Second TX is the best chance, but in general it's discouraging.

Please Log in or Create an account to join the conversation.

More
02 Apr 2014 19:59 #22060 by victzh
Replied by victzh on topic H377 protocol
Great effort on reverse engineering it, if you allow me some notes on the code and further direction.

I don't understand why do you initialize 4 copies of binding buffer. In the original HiSky code it makes sense - they are all different, so the frequency hopping is passed in several packets. In your case you pass the same binding buffer every time.

Frequency hopping sequence is pretty simple - it consists of channel monotonically raising with step 3. So there is probably only one dependent parameter - the beginning of the sequence.

Also there is unclear contents of the binding packet - the first 5 bytes is the address (which better be clearly reflected in the code), and then 5 bytes more. I would try modifying these bytes first to figure are they important or just random filler.

May be the perspective is not so glum, but the protocol is less interesting and interference proof than it could be with minimal efforts of its creators.

As a benefit one should note 10-bit precision of all 8 channels.

Please Log in or Create an account to join the conversation.

More
03 Apr 2014 00:08 #22065 by dc59
Replied by dc59 on topic H377 protocol
Good News!
I tried this protocol on DEVO8S,it worked well!
Thanks for great job ansheng !
Attachments:

Please Log in or Create an account to join the conversation.

More
03 Apr 2014 00:48 #22066 by ansheng
Replied by ansheng on topic H377 protocol
dear victzh,
I just use one copy of bind buffer
I modified the code from hisky. It original initialize 4copies.
"the beginning of the sequence" is 0x1c at bind_buf_arry[0][5] and [0][7].
just as you say, bind_buf_arry[0][0]~[0][4] is the address
But I don't know bind_buf_arry[0][6],[0][8], [0][9] how to calcular.
If i can get 2nd Tx, may i could know how to get them.

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com

Please Log in or Create an account to join the conversation.

More
03 Apr 2014 02:05 #22067 by ansheng
Replied by ansheng on topic H377 protocol
Dear all,
Does anyone know who lmplement the hisky protocol in deviation?
Thanks!

Email:ansheng.chang@gmail.com
MSN:samchang28@hotail.com

Please Log in or Create an account to join the conversation.

More
03 Apr 2014 02:30 #22068 by PhracturedBlue
Replied by PhracturedBlue on topic H377 protocol
victzh (who posted just above) implemented it.

Please Log in or Create an account to join the conversation.

Time to create page: 0.119 seconds
Powered by Kunena Forum