I am looking for an alternative radio that can communicate 2 ways (with ack packet) at a range of 200-400 feet min, that is NOT in the NRF24L01 family. I need transmission of 50 packets per sec reliably with less than 10 bytes. Seems simple right? Cost is not really a concern.
I have spent possibly over 100 hours messing with the NRF24L01 family of radios over the last few months. I have bought several of every single type of board to experiment. Followed all of the tutorials on here. Done self shielding with tin foil. Bought the expensive pre-shielded boards. Done all types of caps. Tried powering with external 2xAA. Tried external USB power to main board. Computer USB power to main board. 12v power to main board. Tried voltage down converters. Tried breakout board. Tried no breakout board. I am literally banging my head against the wall on this problem. I can get them to work but after about 2 min they would disconnect for 5 seconds then reconnect or disconnect endlessly. This is not acceptable to disconnect unless > 50 feet apart. I've tried what seems like all the settings. High power, low power, min. High channels. High/LOW BPS, Large packets. Small packets. Struct data. Char Data. Tried limiting the connection speed to not go more than 50, 100, 150, 200 packets per second. Tried letting the Packets per sec go up to 400 per sec. Nothing works reliably for my use case.
I've concluded that these radios are not reliable enough for my use case.
I have several voltage regulators from digikey on the way. This is my last ditch effort to get these half a$$ boards to work.
However, I would love to just throw money at the problem and get a more reliable solution. My search has come back with nothing.
It is difficult to make any general statement about these NRF24L01 boards because there are a number of types and clones/fakes in circulation. It does appear, however, that you have up substantial effort into problem solving in most of the areas known to cause trouble.
My own experience is that with the TMRH20 library and adequate power with good connections (not via a breadboard) and the simple test sketch from @robin2 these can be made to work reliably. Sometimes, an additional say 10uF capacitor soldered directly on the board helps. Some versions don't work down to 250KBPS so it is best, if there is any doubt, to omit the data rate parameter to fall back to the library default. In the case of some clones, they talk to each other but not well with other devices so it is best to pair them. Usually the modules without external antenna and the "LNA + PA" suffix are easier to get started with. Some clones have difficulty with a payload in the ACK message. Some clones can have a high quiescent power consumption.
Define 'reliable'.
Radio links, especially those using the shared ISM bands are fundamentally unreliable so you need to define what percentage packet loss and what latency in milliseconds is acceptable in your particular application.
Given that you now have have multiple transceivers, one strategy to reduce packet loss is to setup Tx/Rx transceiver pairs on different channels spread evenly throughout the 2400 MHz to 2500 MHz band, then transmit the same data through each pair and collate the results. It's unlikely, that the whole spectrum will suffer interference at the same time, unless you are in the path of an airport radar or something like that.
Below is all of the modules I purchased. How would I know if they were clones?
Here is the one I ultimately settled on:
What I've tried below. The forum won't let me put in all the links. Just paste the code into Amazon search bar to see what I'm using.
B06WLH4ZG6
B01ICU18XC
B01IK9GCPE
B07QC1SXJ8
B01ICU18XC
B082VLQK1M
B00O9O868G
B06WLH4ZG6
B07KD4GR18
I did a quick test of the basic NRF24 a few days ago and managed to transfer 1024 packets (@32 bytes each) for around 0.75 second (for a throughput of ~350 Kbps). Granted I wrote my own C-drivers for these that is efficient and lean.
If you want, I can write you a test application that you can load and see you are still seeing the issue you are encountering...
Just post the platform you are using here. I might be able to modify some of my code already if you are using Atmega328P based MCU. I also have one for Attiny series and Xmega... Either way, you only really need a couple pins (SPI, IRQ, and CE) from MCU to NRF
Hey guys, after WAYYY too long messing with this, I think I've been able to isolate the problem on why the RF24 is not staying connected consistently. I am controlling a motor via ESC via PPM signal. This ESC has 5v,grd,servo ports for PPM. I don't need 5v so I'm connected to GRD and SERVO only. (I've tried using 5v as well but no diff) With everything hooked up except the GRD going to the ESC, as soon as I hooked up the GRD to the ESC, I start loosing RF24 connection. Of course I've tried to power RF24 via AA batteries serparately. However when doing this we still have to connect the ground back to the Arduino. This then exposes the RF24 to the GRD noise? Even if I don't connect the RF24 grd to the Arduino when using AA batteries to power RF24, I still have the problem. I believe the noise could still be coming from the miso,mosi,ce etc.
I've even tried introducing a Nano board with the sole purpose to controlling the PPM signal. So, the RF24 is on the mega, which sends the throttle value to the Nano via serial, and then the nano would have 1 purpose. Send the PPM signal to the ESC. However we still have to expose the ground to the nano. Still not solved.
How can I isolate this PPM signal? I've tried a ferrite. No luck. I'm considering an Opto-isolator but I haven't figured out how to hook it up yet. Any other ideas?