pascallanger
Evaluate Radiomaster RC3S protocol
Per our conversation on rcgroups. Here are the images from the radio to evaluate if it may even be possible to implement the earlier Radiolink ground protocol: https://www.radiolink.com/newsinfo/664430.html
The RF chip is under the glue so we don't know what it is... The SPI bus is well exposed and can be connected to an analyzer. Do you want to purchase and connect an analyzer (around $10) to your radio or proceed differently?
Let me know what I need and to do (I might even have one here)
Happy to give it a try!
On Mon, Feb 5, 2024 at 11:44 AM pascallanger @.***> wrote:
The RF chip is under the glue so we don't know what it is... The SPI bus is well exposed and can be connected to an analyzer. Do you want to purchase and connect an analyzer (around $10) to your radio or proceed differently?
— Reply to this email directly, view it on GitHub https://github.com/pascallanger/DIY-Multiprotocol-TX-Module/issues/938#issuecomment-1927579046, or unsubscribe https://github.com/notifications/unsubscribe-auth/AD36JTY47JQW4LMCYZJR6E3YSELBTAVCNFSM6AAAAABC2RRGASVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRXGU3TSMBUGY . You are receiving this because you authored the thread.Message ID: @.*** com>
A small 8 ch Saleae analyzer (https://www.ebay.fr/itm/262812296176) connected to MOSI, CLK, MISO, CSN and GND.
That's funny, I happen to have one of those right in front of me, though I have never used it...
I can solder that up over the next few days. (I found the software)
I assume, record from power on to bind?
On Mon, Feb 5, 2024 at 12:11 PM pascallanger @.***> wrote:
A small 8 ch Saleae analyzer (https://www.ebay.fr/itm/262812296176) connected to MOSI, CLK, MISO, CSN and GND.
— Reply to this email directly, view it on GitHub https://github.com/pascallanger/DIY-Multiprotocol-TX-Module/issues/938#issuecomment-1927679106, or unsubscribe https://github.com/notifications/unsubscribe-auth/AD36JTYKVEKQPIZ4KAHB7G3YSEOFRAVCNFSM6AAAAABC2RRGASVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRXGY3TSMJQGY . You are receiving this because you authored the thread.Message ID: @.*** com>
Please use the logic software v1.2.14 I would need different capture files for:
- power on
- power on + bind
- throttle mid->low->mid->high-mid
- steering mid->low->mid->high-mid
- same for any other channels
CSN = Chan2 MISO=Chan3 CLK=Chan1 MOSI=Chan0
Can you see if these are what you need, or if I did it correctly? (I expected to see more signal) Model bound and is operable...
Thx
Sorry but the sample rate is too low, the files are unusable as they are... You need to increase it quite significantly (*4 or *8). GND is well connected between the GND of the radio and the GND of the Saleae, right?
Yes ground is connected. Will up sample rate and try again. First time with this so learning how to use it correctly. Thx!
On Tue, Feb 6, 2024, 2:49 AM pascallanger @.***> wrote:
Sorry but the sample rate is too low, the files are unusable as they are... You need to increase it quite significantly (*4 or *8). GND is well connected between the GND of the radio and the GND of the Saleae, right?
— Reply to this email directly, view it on GitHub https://github.com/pascallanger/DIY-Multiprotocol-TX-Module/issues/938#issuecomment-1929038991, or unsubscribe https://github.com/notifications/unsubscribe-auth/AD36JT4LSS24RHZFVFGSYRTYSHVANAVCNFSM6AAAAABC2RRGASVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRZGAZTQOJZGE . You are receiving this because you authored the thread.Message ID: @.*** com>
Here are the files a 8MS/s, let me know how these work. I did not initiate a bind to new RX yet, want to make sure this works so far. 8MS per s.zip
Thx!
Getting there, but still not fast enough... You should see 8 pulses of clock per byte, there are only 5-6 right now. If you add a SPI analyzer (on the right side) and configure the inputs (MOSI, MISO...) with everything else by default, it should be able to decode which is not the case right now. I don't think you need to capture the bind of a RX. From what I can see the radio sends bind packets for 2.5s then switch to normal mode. So just a long power on >5s would be good.
Ok, tried again at 24MHz SPI analyzer finds stuff, but the clock has me confused - I'd expect to see a consistent pulse Power On.zip
It works this time. I need to find out which component they are using now.
It looks like they are using a CC2500. The startup or bind phase looks strange from a first look so I'm wondering if it could be another CC2xxx. I need to spend more time on it.
Cool, please let me know if you need any further data, happy to run more scenarios for you.
Note: I did not put the RX in bind mode - it connected to an already paired RX
Now that I've removed the excessive polling that they are doing, I can see what 's happening. At power up, they scan the RF channels surely to find 2 unused and then start to send data on them every 14.4ms.
For "Power and Bind", what are you actually doing on the TX and RX side? I'm asking since I don't see a difference between bind and no bind... Is there a bind button on the TX? My guess is that there is none and that if you put the RX in bind mode, it just binds to whatever TX is powered at that time. Do you have multiple TXs? If yes a dump of the power on from at least one more TX (more is better) would be great.
Power and bind is just turning on the TX and turning on the RX (it is already bound)
I was waiting to see if this worked and then going to do an actual bind - which I have to refresh my memory on how to do it with this setup. I know there is a bind button on the RX. Will look into that in a bit
The TX and RX don't "talk" at least in normal mode, TX->RX only. Please do a dump of the TX in bind mode. I doubt that the RX being off/on/bind matters but may be I'm wrong...
I didn't see a bind mode in the TX: per the instructions:
In fact - I switched to a different model powered up and it connected to the RX without issue
I'm sure you are right - RX on doesn't matter - I wasn't thinking
What do you mean by "different model"? Can you select a model number on the TX? If yes I would need a dump of each model numbre.
Yes, I selected a different model on the TX and it had no effect
It connected to the RX automatically - did not have to put the RX in bind mode
Ok different model does not mean that there is "model match", ie the model and the RX have to match
Hummm, the model which was loaded was not default right? By this I mean there are stuff like end points adjustments, trims and may be more?
Also in the "Gyro" dump I see 2 things moving, like 2 channels. Do you have a mixer or were you moving 2 knobs in the same dump?
In the above upload 'Scans 16MHz.zip' those were with the programmed model. EPA and likely some trim may have been not standard.
There is no 'mid' notch on the gyro channel, it was best guess - that is 100/100 EPA no trim Aux channel 4 was just an on/off button
Would you like me to re-do any - I can do another gyro scan - I may have bumped the steering, it is close to the knob.
It's not steering moving at the same time as gyro as I can differentiate the channels. It's something else that seems to follow the gyro value in some way. Could it be an unconnected 5th channel which follows the Gyro just because it's closed to it?
That is weird, this is only a 4ch TX.
Mixing is off - I watch output from ch1-4 when adjusting gyro(3) and none move.
This extra output does not happen on other channels? I can re-do the gyro scan if you would like
Let's test. Reset a model to its default values and redo a gyro dump. Please move it slowly from one end to the other (you have been a little quick on the dumps).