Scavenge RF module from Toy grade TX?

I've tried options 1 and 2. Both allow the 7E to boot without error and recognize both the internal module and the Beken as an nRF24L01.

So the CE connection to VDD was causing the CCYRF6936 recognition problem and that error is now fixed. However, neither wiring option allows me to bind to the H8 or E010.

Now what? Is it likely to be a model definition problem or user error?

Good to know you have overcome the detection issue.

I looked at BK2423 datasheet and noticed it got 2 banks of registers whilst nRF24L01+ only got one. Those additional registers have to be initialized for proper operation. I am unsure if existing Deviation can manage BK2423. Chances are that you need to modify the code. Better let other gurus or developers comment on this.

Aha. I thought I'd read that the code handled the Beken initialization, but perhaps that's only for the Beken 2421. I've got an nnRF24L01 module on order, but it's being shipped slow boat and they said up to 40 days.

If I check out the code base, do you think the comments will tell me if the Beken 2423 is detected and initialized? Or will I need to closely read the code and compare to the Beken spec sheet?

I can ship you a nRF24L01+ module (with PCB antenna) if you PM me your address. It carries the genuine chip so that you won't get into any potential troubles other hobbyists encountered before.

Regarding the code, I cannot tell you more since I just started writing codes a few weeks ago. You have to check at the repository yourself, sorry.

All the nrf protocols include a check for the second register bank and initialize it if found. I don't think all the protocols use the same values though.

The initialization values in the 2421 and 2423 datasheets are almost identical. Don't remember anyone scavenging a 2423 before though.

@ajtank - I PM'd you my address. Thanks for the offer. A real Nrf24l01would be great, even with just the pcb antenna. I can only fly indoors anyway, so don't need much range.

@hexfet - The MCU on the scavenged TX has a pin connected to CE on the BK2423. The BK2423 datasheet says that setting CE high puts it into TX mode and low into RX mode. Clearly the scavenged TX MCU must be using both TX and RX modes g why else would they waste an MCU pin. Is it even possible for the 7E to use the BK 2423 with it permanently in only TX mode (due to connecting CE to VDD)? IOW, does the 7E need to receive data?

BTW, I have a spare Eachine H8mini TX with a small 8pin module that is easily removable. I haven't yet figured out what chip is in that module (it's one of those chip on pcb devices covered in black sealant.) Anyone know if it could be scavenged and used?

CE on the 2423 works the same as on other nRF clones. It must be high to enable both TX and RX modes as seen in the state machine diagrams. Driving CE low is a fast way to force a switch to the standby state and is used by some stock transmitters. With Deviation it's always tied high. The device can still be commanded to either TX or RX mode but it's slower because you have to force a power-down in between.

Thank you. I didn't read the datasheet closely enough to realize that. Does the fact that I'm not getting a missing module error mean that the BK2423 is communicating with the 7E?

Ajtank has kindly offered to send me an nRF24L01 module, but I'd still like to try to get the BK working. Are there any tests I can do that might help me find where the problem is, other than just trying to bind to the H8 or E010? I've never used the 7E to bind to anything yet, so I'm not completely sure what I should be seeing even though I think I'm using correct models from the forum.

I've got a 10MHz analog oscilloscope if that might help.

There's a good chance the Beken communication is working correctly, but the code to check for module presence may not be infallible. Checking the MISO line with the o-scope should indicate if the 2423 SPI interface is responding. Probably need a logic analyzer to really decode the bytes.

Make sure the models have the correct protocol and protocol options.

I'm using models written by others and posted in the forum, but I've checked that the H8mini uses Bayang and the E010 uses MJQx with E010 format.

I haven't figured out what the "PPM in" option does, but it's set to None.

@ajtank - Don't send me your Nrf24l01! I just received the one I ordered. They went to great effort to tell me it would likely take 40 days and it arrived in 6!

Thanks for the offer of help anyway. It's people like you that make this hobby fun.

Success! The new Nrf24l01+ bound on the first attempt to all models. Two Eachine H8s and an E010. So at least I can use the 7E for something other than a paperweight, and I know I was doing things correctly.

Perhaps the BK2423 requires some initialization that differs from the 2421.

Congratulations! Fun only comes after frustration, no gain without pain
Your posts also let me learn something new.

I just finished my first flight with the Deviationed 7E. That was fun! The control was so much smoother and precise as compared to the tiny stock TX each quad comes with. I loved the H8 model I flew. The E010 needed to be spiced up... and that's why I bought the 7E... to give me more control over rates and speeds.

I'm reading up on model creation and editing now. Can you tell me - do people build models on the TX? Or do they write them in a text editor directly into a model.ini file. Or is there a separate model creation program that runs under Windows or *nix?

Eventually I may try again to get the BK2423 to work by reading the datasheet and modifying the DeviationTX code, but as a newbie, that's probably a ways down the road.

Thanks again for the help.

You can do it either way. The download section offers emulator builds such that you can try on the PC. You can also create directly on the TX and test with the channel monitor option.

Thanks for the help.

It happened that I got an idle transmitter with a BK2423 and I installed it on my Devo 10.

Tarantual X6: V2x2 protocol, sometimes binds but sometimes not, chance is less than 50%
Eachine E010: MJXQ E010 protocol, unable to bind
Cheerson CX10WD: Q303 CX10WD protocol, sometimes binds but sometimes not, chance is less than 50%

All of the above protocols write magical values to bank1 register if BK2421 is detected. I am unsure how well BK2421 works but at least it seems unreliable to use BK2423 as an alternative.

These register are not well described, programmatically you can't tell BK2421 from BK2423, so all this Beken business is pushing your luck. Better stick to nRF24 if possible.

According to information here , BK2423 does not require initialization of bank1 except for special needs. I set up 2 modules on Arduino, the Rx being an nRF24L01 whilst the Tx an BK2423. The Tx can successfully send packets to Rx even without initialization of bank1 registers.

Anyway as hobbyists do not enjoy cost advantage for small quantity, it is still better to always use nRF24L01 for best compatibility.

I was never able to get the BK2423 to bind to either the Eachine H8 or the Eachine E010. I received an Ebay nRF24L01 ($0.72) and it worked immediately.