Deviating a JoySway

More
03 May 2015 16:18 #32009 by mwm
Replied by mwm on topic Deviating a JoySway Tx.
Ok, I got everything hooked up, and tried a capture. For the record, I'm using the Pulseview tool with an fx2law device.

Got nothing. One channel at 1 the entire time, the others all at 0.

However, it seemed to stop the capture early whenever it got a signal: either when I powered on, or when I turned on the Rx. Reliably. This makes me think that I've got a connection issue. Possibly what I have as "Ground" isn't? Or one of the two grounds isn't - I connected up both of the things I think are the proper corners.

If that seems likely, could you repost the diagram, with indications as to the board orientation? I.e. - assuming I'm looking at the side that's away from the main PCB, which column of pins is the one close to an edge?

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.

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

More
05 May 2015 01:49 #32087 by victzh
Replied by victzh on topic Deviating a JoySway Tx.
I thing in PhracturedBlue's diagram the position of GND is incorrect. It's not top left, but top RIGHT pin as on your photo. As I had a chance to experience, missing GND can lead to very strange Logic Analyzer results. I figured that something is wrong only when used oscilloscope to verify these crazy results.

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

More
05 May 2015 13:34 #32102 by Deal57
Replied by Deal57 on topic Deviating a JoySway Tx.
I think Victzh is right. Here's how I look at the pins from the side with the A7105 chip (e.g. away from the main board).



Deviation Devo7e 3way switch mod, A7105, NRF24L01
Devo6s 2x2 switch mod, trim mod, haptic, multimodule, A7105, NRF24L01, CC2500
Devo12e 4-in-1 with voice mod -- it speaks!!
Attachments:

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

More
07 May 2015 22:48 #32231 by mwm
Replied by mwm on topic Deviating a JoySway Tx.
Ok, moving the ground to the indicated pin got me some signal. I managed to get captures of all the things PB asked for. They are in the attached zip file, with names indicating what they are. Names indicate the standard mode 2 sticks, not what the Orion uses them for (you really want a "sheets" channel?).

16MHz capture. Probes 0-5 on the five inside pins, so that 1, 2 and 3 should be SDIO, SCK and SCS. Pins 0 & 4 don't seem to have a signal, except for going up on poweron, but I didn't do a full search. If you need a capture with more pins on it, let me know.

One odd thing: the Rx has a "bind" button. The Tx doesn't. Just hit bind on the Tx, the led flashes, and it reconnects.

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.

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

More
08 May 2015 13:14 #32245 by PhracturedBlue
Replied by PhracturedBlue on topic Deviating a JoySway Tx.
your zip is too big and isn't included. Either make it smaller or put it somewhere else.

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

More
08 May 2015 13:32 #32246 by mwm
Replied by mwm on topic Deviating a JoySway Tx.
Didn't seem like I could reduce it by much more than 50%. So I put it over here:
drive.google.com/open?id=0B72nhihrc01MUU...VtODM2U1E&authuser=0

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.

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

More
08 May 2015 18:19 #32261 by PhracturedBlue
Replied by PhracturedBlue on topic Deviating a JoySway Tx.
data doesn't make much sense. there is nothing that looks like csn in your data stream

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

More
08 May 2015 20:14 #32264 by mwm
Replied by mwm on topic Deviating a JoySway Tx.
Ok, I only had probes on 5 of the 10 pins , so it's not a complete shock I missed one.

Just to check, CSN should be obvious in the power on capture, right? If not, then is there a single capture I can use?

I had probes on channels 0-4. My cursory indication of the data is that channels 0 & 4 had no useful data on them. Is this correct?

Thanks!

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.

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

More
08 May 2015 21:01 #32267 by victzh
Replied by victzh on topic Deviating a JoySway Tx.
CSN is what marked as SCS on the above diagram. If there are no changes on it, something is wrong. May be insufficient contact.

How did you solve the connection without soldering problem?

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

More
09 May 2015 00:11 #32270 by PhracturedBlue
Replied by PhracturedBlue on topic Deviating a JoySway Tx.
the csn pin should go low and stay low while sck (channel 1 I think) toggles 8 times , and then it will go high again. It is difficult to identify the packets without having a proper csn signal.

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

More
09 May 2015 02:19 #32276 by mwm
Replied by mwm on topic Deviating a JoySway Tx.
My apologies, but I forgot which way I hooked up the pins. Only five channels are connected, in order from 0 to 4 on the row of pins that didn't include a ground in Deal57's diagram. Ground is hooked up where he indicated, and I only saw signals on channels 1, 2 and 3 - but did see signals on all of them. The signal from the SCS pin should be on channel 1 or 3.

If you haven't checked it, the powerup csv file is probably the most reliable, as it was done first.

I unplugged the module, wire-wrapped all 8 pins, and then clipped onto those. I'm going to work on the idea the the clips may have had intermittent contact, and try a couple of workarounds for that. I'll also double-check that the wire wraps lines are on the pins I think they are on. I'll try some other pin hookups until I find the pattern that PB mentioned. I do have one question about that, though: is the "it" that goes high again after sck(?) toggles 8 times sck or csn?

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.

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

More
09 May 2015 02:35 #32278 by PhracturedBlue
Replied by PhracturedBlue on topic Deviating a JoySway Tx.
what you should expect to see:
MOSI will look somewhat random. If you are lucky it will only transition while CSN is low, but that depends on if the bus is shared. It should always transition while SCK is low though, and should only transition when SCK is also transitioning.
MISO will look similar to MOSI, but will change less frequenctly
SCK will be very regular. You will see it toggle in bursts of either 4, 8 or some multiple of 8 pulses at a time with long spans where it doesn't transition.
CSN will go low before SCK starts to transition and stay low through a set of SCK pulses and then CSN will go high again. If the SPI bus is shared, but may also see SCK toggle while CSN is high.

We specifically are only interested in reading the values of MOSI and MISO each time the SCK transitions while CSN is low. All other data will be ignored, as it isn't relevant to the trasnceiver.

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

More
25 May 2015 16:50 #32969 by mwm
Replied by mwm on topic Deviating a JoySway Tx.
Ok, I think I see what you're looking for, but I'm not sure which - if any - of these are CSN. Looks like the third channel is SCK, with MOSI and MISO on either side of it.

This is the Tx power on, and the second one is the Rx power on and connect. Let me know if these look right, and I'll get the other needful things.

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.
Attachments:

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

More
29 May 2015 16:48 #33188 by mwm
Replied by mwm on topic Bumping Deviating a JoySway Tx.
Any chance I can get someone to take a look at those new captures?

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.

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

More
29 May 2015 21:00 #33210 by victzh
Replied by victzh on topic Bumping Deviating a JoySway Tx.
Looks more normal - three signals change and with seemingly correct frequencies - 2nd signal is data, 3rd is clock, and 4th is chip select (CSN). And CSN is asserted for correct number of clock transitions - either 4 (strobe) or divisible by 8 (command). So the electrical connection seems to be OK.

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

More
31 May 2015 13:50 #33281 by mwm
Replied by mwm on topic Deviating a JoySway Tx.
Well, the new captures don't upload. So I've put them on drive like the first set. This tarball includes the ones in the uploaded tarball - power on and connect - if you haven't previously downloaded them.

Note that BINDING IS INITIATED AT THE RX, not the TX.

On names: Joysway is a product line from the company Ripmax. The receivers and transmitters are shared across other lines. The protocol is FHSS. Yes, it's Futaba's FHSS, but it's from a third party, and binding has been reported as being hit or miss.

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.

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

More
31 May 2015 18:02 #33289 by victzh
Replied by victzh on topic Deviating a JoySway Tx.
How can it be Futaba's FHSS (FHSS by itself is a well known acronym - Frequency Hopping Spread Spectrum) if it uses different chip - A7105? Futaba reportedly uses CC2500. Do you have any links reporting successful binding of Ripmax toys to Futaba TX or vice versa?

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

More
31 May 2015 19:36 - 31 May 2015 20:01 #33291 by mwm
Replied by mwm on topic Deviating a JoySway Tx.
When checking to see if Joysway and Futaba's FHSS were the same protocol, I found a post on the FliteTest forum indicating that binding Joysway FHSS with Futaba FHSS is "hit or miss": forum.flitetest.com/showthread.php?8351-...viewfull=1#post93206 . Since that's hit or miss, I wouldn't put much hope in our reverse engineering their version of the protocol working with actual Futaba gear. Since Futuba has moved on from FHSS to FASST, it's probably not all that interesting anyway.

As for the chips, those are just tools for translating for digital signals to RF signals. Whether or not they can talk a given protocol will depend on the chip features. All the stuff I see here reverse engineers the digital signals, which means you have to have the same chip. Given the right equipment (which I don't have), I'm sure someone with the appropriate skills (again, which I don't have) could reverse engineer things at the RF level, and then tell you whether or not the A7105 could do what Futaba is making the CC2500 do.

Please don't let this sideline slow down the work on the protocol. I now have a couple of models I'd like to sail with this.

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.
Last edit: 31 May 2015 20:01 by mwm.

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

More
01 Jun 2015 03:57 - 01 Jun 2015 03:57 #33303 by PhracturedBlue
Replied by PhracturedBlue on topic Deviating a JoySway Tx.
I haven't actually looked at the data much yet, but I did fix the parser scripts to be able to handle your data.
You can run:
utils/spi/rawcsv_to_spi.pl -mosi 2 -miso 2 -clk 3 -enable 4 -sample 0 tx-poweron.csv > tx-poweron.spi
utils/spi/format_spi.pl -tx a7105 bind.spi
to convert your logs into something useful.

You can easily see the initialization code in tx-poweron.csv using this
You can also see the channel pattern:
0.121050   = STROBE_STANDBY        a0 => a0
0.121078   > PLL_I                 0f 30 e0 => 30
0.126057   > FIFO_DATA             05 ff f8 2d ca a0 00 5b 86 64 64 c6 66 64 64 aa d5 d0 => ff
0.127101   = STROBE_STANDBY        a0 => a0
0.127131   > PLL_I                 0f 30 e0 => 30
0.132108   > FIFO_DATA             05 ff f8 2d ca a0 00 5b 86 64 64 c6 66 64 64 aa d5 d0 => ff
0.133153   = STROBE_STANDBY        a0 => a0
0.133181   > PLL_I                 0f 30 e0 => 30
0.138158   > FIFO_DATA             05 ff f8 2d ca a0 00 5b 86 64 64 c6 66 64 64 aa d5 d0 => ff
0.139203   = STROBE_STANDBY        a0 => a0
0.139234   > PLL_I                 0f 31 e0 => 31
0.144209   > FIFO_DATA             05 ff f8 2d ca a0 00 5b 86 64 64 c6 66 64 64 aa d5 d0 => ff
0.145253   = STROBE_STANDBY        a0 => a0
0.145281   > PLL_I                 0f 30 e0 => 30
0.150260   > FIFO_DATA             05 ff f8 2d ca a0 00 5b 86 64 64 c6 66 64 64 aa d5 d0 => ff
0.151305   = STROBE_STANDBY        a0 => a0
0.151335   > PLL_I                 0f 32 e0 => 32
0.156311   > FIFO_DATA             05 ff f8 2d ca a0 00 5b 86 64 64 c6 66 64 64 aa d5 d0 => ff
0.157356   = STROBE_STANDBY        a0 => a0
so we see data being alternately transmitted on channel 30h and some channel between 30h and 44h. It appears data is sent every ~6ms
Data packet size is 17bytes.
I looked at aileron.csv, and see that byte 7 is the aileron data with values from 1eh to aah (mid-point is 5bh). Byte 16 seems to be a checksum. Very occasionally, I see a 'ddh' in the 1st byte (otherwise it is 'ffh') I wonder if this is used by the Rx for binding.

That is all for now.
Last edit: 01 Jun 2015 03:57 by PhracturedBlue.

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

More
02 Jun 2015 16:37 #33365 by mwm
Replied by mwm on topic Update on FHSS
I've done some more research on the protocol. .

Ripmax seems to be a lot like WLToys. Except they also sell some FTR aircraft under the Joysway label. These are what have "hit or miss" binding with Futaba Tx's. They also use component-based electronics, so you can get standalone Rx's as well as Tx's, at least for some of their craft. However, you are expected to either buy a Tx/Rx Pair, or a Tx designed for your craft. Beyond that, compatibility isn't guaranteed.

So victzh is right - this is just Frequency Hopping Spread Spectrum, and not Futaba's protocol.

I have verified that this Tx works with the Rx in their two smallest yachts, the Orion and the Caribbean. The Rx used in the Orion is also used in their most popular yacht, the Dragon Force, as well as the recent Explorer yacht and the Force 2 and Binary catamarans. It is not used in the older Focus, which is their largest yacht, and I don't believe they are compatible. However, the Focus appears to be out of production.

Do not ask me questions via PM. Ask in the forums, where I'll answer if I can.

My remotely piloted vehicle ("drone") is a yacht.

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

Time to create page: 0.121 seconds
Powered by Kunena Forum