help-with-radio

[calm thistle]

Hello can anyone tell me if it is possible to connect the rfm69 radio to a raspberry pi pico ?

I cannot find to which pin I must connect the interrupt pin G0 from the RFM69

https://rpi-rfm69.readthedocs.io/en/latest/hookup.html#pinout-guide

[granite spear]

I'm not familiar with this off the top of my head, but my assumption is that you could use almost any GPIO, and you would just specify which one you chose in the initialization parameters for the RFM69 library.

[normal drift]

There should be no problem using the RFM69 with a pico. As noted, G0 can be any available GPIO pin. If you are using CircuitPython, then G0 is not needed. It is needed for Arduino libraries.

For the Pico you will need to select SPI pins see https://learn.adafruit.com/getting-started-with-raspberry-pi-pico-circuitpython/pinouts fro an example and a pinout diagram.

[crude gust]

Hi

I'm using lora sx1278 E32 433T30D which is 1watt model

and at datasheet it says it can transfer 58 bytes data

I'd like to know about, in a second how many packet I can send

how can I calculate this

[granite spear]

LoRa has a lot of different speed modes depending on how much range you need. As I recall it maxes out at about 40kbps throughput at the highest settings.

[crude gust]

I need to transfer 5 packet per second, And packet will include 3 float data

so, its possible with lora right?

[granite spear]

Yep, that should be easily possible.

[crude gust]

thank you, really

[crude gust]

hi

I'm using lora

Im using lora, and trying to transfer mpu6050 gyro datas
I can get gyro data, and make packet for transmit
but my code just sends the first value of gyro
the other packets are constant firsst values
doesnt change

#include <Servo.h>
#include <Wire.h>
#include <MPU6050_light.h>
#include "SparkFunBME280.h"
#include <TinyGPS++.h>
int offset =20;
int a;


struct Signal {

  byte GyroX[4];
  byte GyroY[4];
  byte GyroZ[4];
  byte Nem[4]; 
} data;

static const uint32_t GPSBaud = 9600;
TinyGPSPlus gps;
BME280 mySensor;
MPU6050 mpu(Wire);
long timer = 0;
Servo sg90;


void setup() {
Serial.begin(9600);
Serial2.begin(9600);
Serial3.begin(9600);

sg90.attach(8);

sg90.write(180);
  Wire.begin();
    byte status = mpu.begin();
  if (mySensor.beginI2C() == false) //Begin communication over I2C
  {
    Serial.println("The sensor did not respond. Please check wiring.");
  }
}


void loop() {
   mpu.update();

  while (Serial2.available() > 0) {
       if (gps.encode(Serial2.read()))
       displayInfo(); 
    }
}

void send(Signal *table)
{
  
  Serial3.write((uint8_t *) table, sizeof(Signal));
  
}

void displayInfo()
{
  if (gps.location.isValid())
  {  

   *(float*)(data.GyroX)=(mpu.getGyroX());
    *(float*)(data.GyroY)=(mpu.getGyroY());
    *(float*)(data.GyroZ)=(mpu.getGyroZ());
    *(float*)(data.Nem)=(mySensor.readFloatHumidity());        

 Serial3.write((byte)0x00);
      Serial3.write(0x3f);
      Serial3.write(0x17);
      send(&data);    


 delay(1000);
   

  
}
 
  Serial.println();  }
}

this is my code

[primal warren]

You might need to say struct Signal instead of Signal. You could declare it with float values instead of doing all the casting.

[granite spear]

It looks like your code might be getting trapped in the GPS character-receive loop, and never getting back to the main loop() to run the IMU update routine. It might just send the same LoRa packet over and over again once the GPS location becomes valid, if new GPS characters come in faster than the time it takes to send a packet.

[crude gust]

thanks, I fixed my code and now working

[ebon needle]

I am trying to establish a P2P connection between LoRa e5 mini but the serial monitor is showing error as "LoRa sender failed" and same error for receiver part too.
Code for sender 👇🏻

#include <LoRa.h> 
  
 int counter = 0; 
  
 void setup() { 
   Serial.begin(9600); 
   while (!Serial); 
  
   Serial.println("LoRa Sender"); 
  
   if (!LoRa.begin(868E6)) { 
     Serial.println("Starting LoRa failed!"); 
     while (1); 
   } 
 } 
  
 void loop() { 
   Serial.print("Sending packet: "); 
   Serial.println(counter); 
  
   // send packet 
   LoRa.beginPacket(); 
   LoRa.print("hello "); 
   LoRa.print(counter); 
   LoRa.endPacket(); 
  
   counter++; 
  
   delay(5000); 
 }

Code for receiver👇🏻

#include <SPI.h> 
 #include <LoRa.h> 
  
 void setup() { 
   Serial.begin(9600); 
   while (!Serial); 
  
   Serial.println("LoRa Receiver"); 
  
   if (!LoRa.begin(915E6)) { 
     Serial.println("Starting LoRa failed!"); 
     while (1); 
   } 
 } 
  
 void loop() { 
   // try to parse packet 
   int packetSize = LoRa.parsePacket(); 
   if (packetSize) { 
     // received a packet 
     Serial.print("Received packet '"); 
  
     // read packet 
     while (LoRa.available()) { 
       Serial.print((char)LoRa.read()); 
     } 
  
     // print RSSI of packet 
     Serial.print("' with RSSI "); 
     Serial.println(LoRa.packetRssi()); 
   } 
 }

https://github.com/sandeepmistry/arduino-LoRa

I am referring this for the circuit connection and codes.

Above github page suggests us to use sx127x lora boards and I am wondering 🤔 why ony sx127x, why not lora e5 mini. (Developed by seeed studio)

[normal drift]

Unless I am mistaken, the E5mini does not use SPI to communicate with the microcontroller. It uses a serial UART with AT commands. the arduino-LoRa library is expecting an SPI device.

Looking at the Wiki https://wiki.seeedstudio.com/LoRa_E5_mini/ there "may" be a way to use it via SPI, but I have never tried it.....

[celest fable]

Good Morning! I have a feather_m0_rfm9x feather board, and I noticed the circuitPython modules omits the 'onewireio' module. Is there a way to get one wire working on this feather?

[half plover]

In the Adafruit_circuitpython_bundle should be the library you’re needing

https://circuitpython.org/libraries

[celest fable]

I have copied the adafruit_onewire over to the lib folder, but when I import the library, I get the following error

Auto-reload is on. Simply save files over USB to run them or enter REPL to disable.

Press any key to enter the REPL. Use CTRL-D to reload.

Adafruit CircuitPython 7.3.2 on 2022-07-20; Adafruit Feather M0 RFM9x with samd21g18

import board

from adafruit_onewire.bus import OneWireBus
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "adafruit_onewire/bus.py", line 17, in <module>
ImportError: no module named 'onewireio'

I noticed the 'onewireio' module is not listed for this board, and when I import the library in the lib folder, I get this error. Everything is as current as I believe is available to download.

[normal drift]

onewireio is only include in the "FULL_BUILD" and the feather_rfm9x cannot handle the FULL BUILD. You can try a custom build to attempt to add it, but it's tough to cram more into the m0 builds.

Wait - I may be incorrect on that -- looking...

[crisp rampart]

Hi everyone! I am new to programming for BLE/Bluetooth. I'm making an RC car with a Feather ESP32-S3 (with 2MB PSRAM) and I have tried using Adafruit IO and it works very well but I'm only allowed a certain number of requests a minute, so I thought BLE would be perfect. I have the Bluefruit Connect app on my computer (and phone) but when I run the code from this learn guide: https://learn.adafruit.com/circuitpython-nrf52840/bluefruit-le-connect-basics the feather raises a NotImplementedError when I initialize a UARTService.

I think it only works with an NRF52840.

But then I found this learn guide (for an ItsyBitsy, I know): https://learn.adafruit.com/adafruit-airlift-bitsy-add-on-esp32-wifi-co-processor/circuitpython-ble and I couldn't find the adafruit_airlift library so I used the esp32.mpy library instead. It worked, but it's using the wrong pins, it's using a board.ESP_RESET pin as the reset button but it's not working. When I edit the .mpy file it raises some memory error. Do you know what is happening?

[slender current]

ESP32 as a coprocessor is different than the S3. BLE on the S3 is incomplete unfortunately.

[celest fable]

Is it correct that if oneWire is not in the FULL_BUILD, then adding the module (adafruit_onewire) to the LIB folder won't actually add oneWire?
thanks

[normal drift]

Yes, but I am still trying to confirm (or deny) that onewire is not in the rfm9x build.

[normal drift]

What version of CircuitPython are you using?

[crisp rampart]

You might be right. However, after replacing the UARTService() with board.UART(), I at least got it connected to my phone but it can't detect anything about the feather except for that it is called CIRCUITPY.

[celest fable]

7.3.2

[normal drift]

It looks to me like it "should" be there -- still checking.

[normal drift]

ah -- no -- it is not -- see https://docs.circuitpython.org/en/latest/shared-bindings/support_matrix.html

adafruit_bus_device, analogio, board, busio, digitalio, math, microcontroller, neopixel_write, nvm, os, pwmio, rainbowio, random, storage, struct, supervisor, time, usb_cdc

Frozen Modules: adafruit_rfm9x```

[celest fable]

I have seen this and can confirm that the onewireio is not in the firmware file feather_m0_rfm9x. I was thinking, okay, we just add it to the lib folder, but alas I see there is more to the module's nature than simply dropping into the lib folder.

Thanks, this clarified that onewire is not available on the rfm9x firmware for circuitPython, and that's okay. I have a M4 express standing by, but will have to use a separate lora module, and that's okay too. I was trying to use the ds18x20 temperature library to read the temp and then push the value upstream with the radio, but this is not the board for that task.
Thanks for the help!

[normal drift]

I think the adafruit_onewire library still depends on the underlying onewireio moculde

yes https://github.com/adafruit/Adafruit_CircuitPython_OneWire/blob/main/adafruit_onewire/bus.py#L17

The feather_m0_rfm9x is a challenge with CircuitPython -- the M4 + external rfm9x will be a much better experience!

[crisp rampart]

Looks like I'm a little stuck. Is there a timeline for when BLE for the S3 will be complete?

[slender current]

no, unfortunately it requires changes to the underlying nimble library

https://github.com/adafruit/circuitpython/issues/5926 will be updated when things are added

[crisp rampart]

Thank you!

My idea is to take the code from the ItsyBitsy ESP32 (the adafruit_airlift.esp32 library) and use it on my feather, changing the pin names so it's compatible. Do you know the pin mappings/names for "reset" "gpio0" "busy" and "chip_select" - for example on the Airlift Shield it's D5, D6, D7, and D10 - and do you think this approach might work?

[woeful vector]

hi all,
any good GSM module to hook up to a microcontroler

[primal warren]

There are a few choices, the Simcom SIM7600 series (different variants for different countries), some Mikroe/Quectel modules (don't know if they offer 4G), and the U-Blox modules.

[buoyant elbow]

Any pointers to connect 2 cc1101 to a feather m4? I'm new to Microcontrollers and spi

[narrow pike]

Hi folks! I'm having trouble with my various rfm96 radios, I'm trying to use the addressed "Reliable Datagram" mode. I have two feathers running Arduino with the Radiohead module (addressed as #1 and #2) and a kb2040 with circuit python (set to #4)

The kb2040 (#4) is receiving a tx from #1, but #1 says it doesn't get an ACK. When the two feathers send to each other, the ack works, but that code does manually send the ack. The circuit python code appears that it sends the ack for you (with_ack=True).

I've tried debugging by playing with the timeouts/wait times, and I can't get #1 to see an ack from #4. #4 only needs to get messages, aside from the ack.

Any tips to debug further?

More info: The line:
rf69_manager.sendtoWait((uint8_t *)radiopacket, strlen(radiopacket), DEST_ADDRESS)
returns false, even though the other radio got the message

[normal drift]

do you have rfm96 or rfm69? you mention both.

It can be hard to get a Circuitpython to see an ack from an Arduino board. They just come too fast.

The current RadioHead library does not have a delay for that ack. There was talk of allowing it, but I don't think it has been added to the released version yet.

[narrow pike]

Typo, I keep making that. They're all rfm69

It's the other way, the CP board is supposed to send ack to ardunio

[normal drift]

It sounds like your issue is the opposite, so it should work. Try setting the ack_delay to 0

[narrow pike]

Will do - is there a way to increase how long rf69_manager.sendtoWait waits before it says it failed?

Okay with the CP board set to ack_delay of 0, the feather is seeing some of them - but some still are not considered ack'd

[normal drift]

glad it is improving 😉

I can't access the Radiohead documentation now -- website is not responding.

[narrow pike]

no worries, ty for the help

[normal drift]

now its back

it's very slow -- I thought there was a setting for the timeout on the Radiohead side.

[narrow pike]

it seems to be around 1000 ms

but I can't figure out where to set that

[normal drift]

I meant the website is slow...

[narrow pike]

hahaha

😄

[normal drift]

I have to go offline now. I can try some tests maybe tomorrow to see if I can reproduce the problem. Can you post the code you are running?

[narrow pike]

Sure, I'll get it into a minimal repro and share

thanks again for the help, this is at least more stable than it was 😄

[normal drift]

Great -- glad to help.

[narrow pike]

Oh I found setTimeout method

That looks like it's what I needed, I set it to 2000 and they're all passing! I will play with the times, nothing I'm doing is super time-sensitive so this works great

[normal drift]

Great! Good luck!

[fervent plank]

oh dang a channel for my next project, a diy car audio system. i wanna exit the world of car stereos with useless features and make my own thats basically all inside the dash except an aux in port

so basically bluetooth receiver and 3.5mm in to rca to amp to speakers

im having trouble sourcing what seem to be good components tho

any recs audio boards with bluetooth ? if the bluetooth thing doesnt have any direct input line how should i handle switching? any other oversights?

[vale star]

I wanna do something for fun....make a little AM Radio receiver that only picks up Jupiter's Radio waves - is there a circuit I can build that I can adjust to this Freq?

It's kinda weird thing I wanna do but wish to do as a weird project

[granite spear]

I don't think Jupiter has a particular narrowband frequency range. You'd more want a satellite dish that you can point specifically at its position in the sky. Though no guarantee that would work either, as I'm not sure of the signal strengths involved or the other noise sources in those bands.

[vale star]

trust me it's all the way down of the range it has so much radio waves it's freaky

[primal warren]

Many radio circuits us an LC tank circuit for tuning and by changing the values, you change the frequency

[floral onyx]

I am curious if I soldered the spring antenna onto my RFM95W, LoRa bonnet correctly. I couldn't find a close up of what was meant by the ant pad. After a few hours I have only picked up two packets, I was expecting to scrape more. Did I put this in the right location? Also, I accidentally got solder in the little hole behind the SMA mount, is that a problem?

[primal warren]

I think that's probably right, but I'm not sure if there's a ground pad nearby that could be shorted. Normally an antenna or the connector is populated, not both, but that's probably okay. That little spring antenna is not a very efficient antenna.

[normal drift]

I think the antenna should go through the hole next to the ufl connector. You have it connected to the ground pad, I believe.

The top picture is just a wire antenna the lower 2 are the spring version. These work well for me.

[normal drift]

the schematic shows the connection point as JP6

[normal drift]

and as 1ANT on the layout

[floral onyx]

This is exactly what I was looking for. Thanks! Ill try and clean it up and resolder to just the hole.

[floral onyx]

I was hoping to multipurpose my RFM95W bonnet and use it to listen for my utility meters that broadcast in the 900mhz range. I am finding a lot of SDR related libraries that should be able to decode the messages, like https://github.com/merbanan/rtl_433, but these seem to be SDR specific and expect USB dongles. Is the RFM95W bonnet considered an SDR? Does anyone know if I can get a SDR utilities to work with it? Or am I going to have to dig in further and write my own decoding?

[granite spear]

No, it's not really an SDR, as it's receiving with specific frequencies and modulation schemes. So you'd need to match the exact channel, bit rate, etc., assuming the utility radios are compatible at all.

[floral onyx]

So does that mean with a true SDR I wont need to know the exact frequencies and broadcast details? Does an SDR somehow receive from a range instead of a specific configuration?

[granite spear]

Yeah, a software-defined radio basically digitizes the raw radio signals from a chunk of frequencies, and then can run analysis on all the data to decode whatever protocols it knows about. Whereas a "normal" radio receiver has specific circuitry for the signals that it uses.

[floral onyx]

Thanks. I was struggling to understand the difference between my bonnet that I can set via code and these SDRs. Appreciate it.

[short nimbus]

Hey my question is a bit raw but there is something stopping me from doing radio related stuff

my apartment is really small and I can't have much of an atenna, also it's concrete etc. Have miniaturization helped antenna gain a lot or you still need a tower antenna in your backyard ? If the 1/2 and 1/8 formula still apply the max I can hope for is for wavelenghts between 60cm and 240cm right ?

[remote pond]

the concrete might be a bigger barrier than lack of space

[short nimbus]

Could put an antenna up to 2m on my balcony only 5 months a year outside. Would have to be secured for wind/rail etc and even then the outside layer of my building is worked metal

[remote pond]

what sort of radio work are you thinking of doing? receiving? transmitting? amateur radio? or other services/unlicensed bands?

[short nimbus]

building my own AM/FM radio hopefully, checking for wifi/IR interference (this one would need to be portable), I don't have an ham license and I'm within LoS of the police station so I'll have to double-check the legality of anything I do

possibly trying radar indoor too

It's just that looking at AARL book not having a big antenna seems as much of a problem as a car that doesn't have wheels or they are too small

[remote pond]

the 1/2 wavelength thing is only a guideline for certain kinds of antennas (half-wave electric dipole). there are other kinds of antennas (magnetic loops, for example) that take up far less space for the same wavelength, but there are tradeoffs

[short nimbus]

magnetic loops is the one where you buy a huge winding of naked copper cable and try to cleanly wrap all of it around say a pole or a nail right ?

[remote pond]

there are also ones that are a single circular loop of copper pipe

but you can tune any random wire to receive/transmit on any frequency (within reasonable limits), if you have the right amount of capacitance and inductance. it might not be very efficient, or you might need horribly high voltages to transmit on it…

[short nimbus]

for the time being it would just be receiving

would like to 'listen' to space as well but apparently that's not the right word, it's more light being "converted" to radio waves and it assume it's going to very hard in the typical city (atmospheric scattering, lights around etc)

my most pressing concern will probably be to pick a frequency on a transmitter control for r/c and detect beforehand if somebody is going to interfere with me if there are several remote control going on

[remote pond]

where did you hear about light being "converted" to radio waves? there are radio telescopes, and they don't convert light (if by "light" you mean the visible spectrum)

[short nimbus]

Saw a video critic of contact by an astrophysicist and she kept getting annoyed seeing ellie listen to space with headphones

And she kept trying to explain we can't listen to them like this since it's light being converted to radiowaves

But I'm not an astrophysicist so I have no idea what that even mean

I can probably find the video but I've been very confused since then

Had one IRL tell me about the same in person and not even being to listen live because you have to rent time on it and it's processed by computers first so live isn't even possible

[remote pond]

can you "listen" to radio waves from space with headphones/audio? yes. is it useful for doing astronomy to do so? mostly not, and that's what they might have been complaining about

also, i gather that some radio astronomers have prejudices against SETI folks, for "wasting" valuable radio telescope time looking for signals from extraterrestrial intelligences instead of doing "real" astronomy

[short nimbus]

she wass a SETI species/kind of astrophysicist

wikipedia used to say that SETI was pseudo-science and was going to tell her in the live but couldn't find a source

[remote pond]

but even SETI does a lot of computational stuff instead of listening to the raw signal, so that part is true. i guess an eccentric SETI researcher could feel like additionally listening to the raw signal "just in case", but doing that instead of letting the computers work seems foolish

[short nimbus]

so if there is a lot of r/c in my area and they are interfering with my transmitter/receiver IR remote controls it would be illegal to jam them right ? I don't mean in competitions where it's against the rules but every day use 🤣

I rofl because I typed electronics RF on my public state library digital books collection and the first 150 books are about electronic warfare 🤣 Not sure if it's the right public library...

[remote pond]

um, radio and IR are different parts of the spectrum; how are you experiencing this interference?

[short nimbus]

When I go to R/C convention there are often transmitters interfering with each other, that would be my 1st project if it's easier than getting am/fm

Like something a transmitter will control someone else r/c etc so they usually will ask everyone to turn off their transmitter etc but I have an unsupervised track nearby with no rules

[primal warren]

The 2.4GHz spread spectrum ones are very good at playing nice with each other.

[blazing maple]

I should work on my gps tracker some more.
We have a couple JLGs at my work and they are always hiding somewhere...

I assume xbee were like wifi in strength? It's been a long time since I used these. The lora is way betterer yes?

[granite spear]

XBee would probably be intermediate between the two. WiFi is fast and short-range, XBee is medium and medium, and LoRa is slow and long-range.

[short nimbus]

canadian seller selling atomic clock receiver board for 50$ but the signal is 60KHz and from NIST, as if that will reach canada...

If I don't catch AM stations from the US don't see why this would work. Right ?

[granite spear]

The atomic clock signals are much longer-range than AM radio due to the lower frequency. There's only the one broadcast station in Colorado, but it can be received all across the U.S., so I presume Canada is in range too.

[short nimbus]

it can be but I really doubt it and they should says so

There are probably peoples doing ham also repeating the signal in their antennas set I guess

[granite spear]

I suspect that's probably illegal if so. It's likely a reserved frequency.

[short nimbus]

Ah I thought that certain peoples/organizations like universities/antennas parks owners might be authorized to repeat this signal to provide the signal as a public signal which would probably require a large/powerful/expensive transmitter

[remote pond]

i think repeating the signal defeats the purpose of having a continuous carrier wave broadcast of exact frequency broadcast from a single known location

[short nimbus]

Probably and apparently we have a similar canadian service on a widely different frequency

so shrug

[primal warren]

Most of the US: there's another station (WWVB) in Hawaii.

[quick ermine]

Late reply, but the FCC refers to jamming as intentionally causing harmful interference. It is always illegal, regardless of where you do it. Some folks think it’s only bad on licensed frequencies, but that is wrong. Even jamming on the ISM band is illegal.

[restive fjord]

hi, why there is no cc1101 hw from adafruit?

[restive fjord]

The very definition of WiFi. ;) From a timekeeping perspective. In practice, the deployment of cellular is so widespread now that it is effectively a good source of standard time.

CHU in Canada broadcasts on three frequencies.

https://en.wikipedia.org/wiki/CHU_(radio_station)
"CHU transmits 3 kW signals on 3.33 and 14.67 MHz, and a 5 kW signal on 7.85 MHz."

[primal warren]

The CC1101 is a fairly old chip, the CC1200 is a more current transceiver, and the CC1310 and CC1312 are the related MCUs with RF capability

[restive fjord]

Simple google search suggests much of FCC enforcement is concentrated on inintended radiators, including roadside LED/LCD displays (digital signage). We have traffic signals in town that render HF communications pretty much ineffective, while stopped for the light.

[edgy breach]

I run my FONA without a Lipoly battery 😈

[bronze lichen]

Legends say that will open a black hole so be careful

[lethal helm]

Whats up.

I got 2 rfm9xs. I wired them exactly as the tutorials (3v, ground, ce1, gpio 25 for reset, miso, mosi, and gpio 5 for g0). I have tried the basic test code and it used to work for both modules, now it does not. Sometimes I get the "failed to find rfm9x with expected version" and sometimes I get "failed to configure radio for LORA mode". I have tried everything and nothing works.. SPI is enabled The only other device I got connected to the pi is a sense hat that I got working with only 8 gpio pins but I doubt having that connected is the issue.

anyone know what else I'm missing. Im also new to this stuff. Thanks

and sometimes (like once every 50 times i try to run the script) it does work (and when I mean work i mean initalize)

[granite spear]

That sounds like it might be a flaky connection in the wiring, so I'd double check your solder joints or breadboard connections to see if something might have worked itself a little loose since the last time it worked consistently.

[waxen idol]

Hello, Im working with this radiofruit feather m0: https://www.adafruit.com/product/3177 and am trying to have it send radio packets, which I can then receive on my computer using wireshark. Does anyone have experience doing this and if so, how I could configure wireshark to do this?

[granite spear]

I don't quite understand. Wireshark is a piece of software for analyzing packets on a network interface or USB... it won't just receive 433MHz radio broadcasts.

[waxen idol]

I thought it could monitor radio traffic no?

@granite spear

[granite spear]

I believe it can look at WiFi and probably Bluetooth traffic if the PC is equipped with those radios, but it can't just pull in random radio waves without any sort of antenna or receiver circuitry, since it's just software.

[little forge]

Is it possible to have a simple website on wifi boards so as to control it within the same wifi network preferably running Circuit python

[granite spear]

I don't know what the status of Circuit Python support is for web servers, but the ESP-IDF framework does have library support for that sort of thing.

[little forge]

is there a way to send data over wifi to a wifi enabled board from a phone or tablet?

[primal warren]

Yes, WiFi is just a transport layer for TCP/IP, so you can open sockets and send data. One reason HTTP is popular is that it's widely available on phones and tablets, but there are apps available that support other protocols, or if you're into programming, you could roll your own.

[dusty onyx]

anyone up for helping sort out transmitting a data set from one nano to another using nrf24l01s? 😛

[dusty onyx]

I need to simulate lightning flashes created by one arduino and actually flashed by another. For example, turning an LED panel on for 410ms, off for 250, on for 337, off for 678, on for 635 then off. (LED string controlled via mosfet - that part works fine)

[primal warren]

I'm actually working on a similar project, although the nRF chips are not my primary option

[wraith panther]

I forget this channel exists sometimes 🙂
https://www.aliexpress.us/item/3256802523800409.html?spm=a2g0o.order_detail.0.0.22487d56P77zj2&gatewayAdapt=glo2usa&_randl_shipto=US

I bought these! They look "better" in some regards than the cheaper ASK modules. What I like is the receiver has a power enable pin which will come in handy for low power stuff. Just waiting for them to come in and I'll post a review for others

They also seem to be pretty well documented too

I wish it did have a "Wake on RF" pin though. I still have to figure out how I will put the receiver side to sleep and not miss a packet.

Maybe power up once per second and check for data?

And then there's these I bought too:
https://lcsc.com/product-detail/RF-Modules_Feng-Niao-RF-C2857839_C2857839.html
They also look...interesting.

I may have bought too many so Im gonna have to give some away lol

[primal warren]

I'll be interested in that review. I did some experimenting with some fairly minimal modules, they basically worked, but the RF flux was high enough to interfere with my other circuitry, so I ended up switching from 433MHz ASK to 2.4GHz DSS.

[wraith panther]

I dont have RF equipment on hand so I cant really say if it will interfere with other stuff-I dont have a particular way of doing so, but if you have a suggestion on how I can test it, Im open to hear about it. I have a scope that only goes up to 50Mhz though.

[primal warren]

You can make a simple RF sensor out of a loop of wire, diode, and capacitor to detect an RF field: it won't be calibrated, and it will be orientation sensitive, but might be useful for relative sorts of measurements.

[wraith panther]

Hmm

Couldnt I also use a receiver and just see how far away I get noise? Ive noticed on some of the modules I tend to get a lot of nonsense characters but once it picks up on the transmitting side, it tends to go away

I think I get what your getting at-a crystal RF receiver.

[primal warren]

That too, I suppose.

[little forge]

will BLE circuit python code written for board using the nrf 52840 work the same on an esp32-S3

[little forge]

can this be done using circuit python at this time while I've been a hobbiest programmer for 30 odd years I don't think I have the knowledge or experience to write libraries or protocols but if an app exists for Android devices and there is a Library that already exists for Circuit Python that would be great

[worn bridge]

There are many different ways to do this, depending on what kind of data you want to send. MQTT is popular with IoT and small data transactions, while HTTP is suited to sending more complex pages and documents for viewing convenience.

https://learn.adafruit.com/mqtt-in-circuitpython is a good starting point for the circuitpython side, and #help-with-circuitpython can help you with more circuitpython-specific questions.

As for the Android side, you can either find a broker app directly on the tablet, or set up a separate MQTT broker with the tablet as a client.

Google can help more with that, as I’m not familiar with any particular app selections.

If MQTT doesn’t seem like a good fit for your project, circuitpython also has an HTTP library somewhere, but I’m less familiar with it and don’t really know of any really good guides that help explain how to use it. https://learn.adafruit.com/wordle-personal-esp32-s2-web-server/create-the-web-server does, however, demonstrate that circuitpython capable Wi-Fi boards are capable of hosting small web servers, but I think this project uses Arduino…

[timid onyx]

anyone know of a good board for cellular in the US? Adafruit still hasn't replaced/updated the fona and it maketh me sad 😦

[primal warren]

There are some SIM7600 boards and Mikroe offers some 4G boards based on ublox modules

[wraith panther]

Got my RF modules in 😄

these are all "dumb" ASK/OOF modules but the circuits look a bit more complex 🤔

[wraith panther]

I have to figure out how to test them and come up with a method. Some of them use non standard pin spacing , ie not 2.54mm

8 and 9 (on the bottom) seem to be the smallest of the bunch. What interests me is that little DFN IC. Thats pretty complex for a simple receiver..

[primal warren]

Looks like one uses a SAW filter

[wraith panther]

yea, the middle one. BUT theres more circuitry on back

[wraith panther]

So far, on the receiving side this modules are pretty good! Im getting very little noise and no random characters.

One module so far only consumes 800uA and is probably the best one so far. 😁

[primal warren]

Ah, good findings! Thanks for getting all of these, testing them out, and sharing the results with us.

[wraith panther]

I could send you some if you trust me with your address. Its not like I need 5 each of them. I kind of had to order that way on purpose and I have more coming from Aliexpress. They are kind of mostly the same but theres a chinese seller who makes them. Their whole website is in chinese and doesnt translate well, but they seem to be where these modules come from.

Not much info on them on their website either

The only test I havent been able to do are the ones with non standard pinouts (I guess I could wire them) and long range tests. The only range test Ive done is say from my PC to out in the living room, which is relatively short (30ft?) and is LOS.

[edgy breach]

Can the FONA 3G make calls and send data at the same time?

If not, can someone link me a module that can? (CLASS A GPRS)

[wraith panther]

ASK/RF Module Testing thread

[primal warren]

While I'd trust you with my address, I already have plenty of these modules (bought from various surplus electronic vendors over the years). I recently tried to use them to implement a remote control robot, but found out the cheap ones radiate enough energy to interfere with surrounding circuitry, so I shifted to 2.4GHz spread spectrum instead.

[wraith panther]

Hmmm. How could I test that? Ive been thinking of using the Antenna that was sent with these modules, along with a germanium diode and small cap to form some rudimentary signal strength meter

They are interesting modules, Much better than the really cheap ones. But I had to dig around to find these.

[primal warren]

That seems like the right sort of idea. Maybe a few inches of wire, a Schottky diode and small capacitor. I did find, unsurprisingly, that it was very sensitive to orientation.

I had stuffed one of the UHF transmitters in a solderless breadboard along with a Teensy to generate the PPM waveform I was using, and it worked briefly, then died. I was annoyed to find out I had fried my Teensy.

[wraith panther]

Did you use a Schottky diode because you didnt have any germanium ones on hand?

[primal warren]

I used it for the very low voltage drop and fast recovery, also the small size (the germanium diodes I have around are larger 1N34A style through hole ones)

[wraith panther]

I think I have those too unless I got rid of them

but I too have a fair amount of Schottky diodes

[wraith panther]

https://vk3ye.com/gateway/noapr97.htm

is what Im building

https://vk3ye.com/gateway/fsm.gif

Not knowing much about RF, but whats the 50nF cap for? Ive seen values elsewhere in the pf range

OK So thats really cool. I can see the actual signal being sent out on my scope 😄

[primal warren]

The 50nF capacitor just acts as a DC block for a voltage shifter/doubler. Its value is not at all critical.

[wraith panther]

So testing is finished. I did not do long range testing as Im pretty confident theese would be OK. Updated my thread.

[wraith panther]

Huh, the bot stops me from uploading the code I used.

[primal warren]

Code tends to have character patterns that look a lot like problems, I think the usual advice is to use a service like pastebin

[wraith panther]

Its simple enough lol. I mean, its just using the equivalent command to send out ASCII characters via serial.

[grizzled vapor]

You probably want a bigger capacitor with low inductance is all you need so it can pass high frequencies. I don't know rf either, so I'm just making a very educated guess.

[primal jolt]

im having a very hard time connecting a nano every to an RFM69HCW (radio) using the RadioHead library, particularly it doesnt seem to recognize that theres even a device on the SPI interface

[granite spear]

The first thing I would double-check would be the pin assignments you're using.

[primal jolt]

what are the spi pins on a nano every? i couldnt find much info on that

[granite spear]

Looks like D11 (MOSI), D12 (MISO), and D13 (SCK).

[primal jolt]

which pins should those connect to on the radio chip?

[granite spear]

Aren't the radio pins labeled as MOSI, etc.?

[primal jolt]

yes, but im not great at this

[primal jolt]

nvm, im just dumb

[wraith panther]

Lol so I have a problem:

👀 I bought more modules than I need-because I didnt know if this seller on Aliexpress would get them to me, so now I have other modules to test. Guess its not a bad problem to have but, its not like Im going to use a bunch of wireless devices around the house lol

[dark tiger]

I see no problem, I see a hobby

[timid onyx]

anyone know of a decent z-wave 700 breakout?

thread is around the corner but still using 2.4GHz so i'd prefer to use z-wave

[haughty patrol]

Anyone know where to get, or build a 2m transceiver? I'm a licenesed ham, and want to build my own aprs weather station.

[primal warren]

Getting them isn't too tough, hamfests, surplus places, and ham radio stores like Ham Radio Outlet carry them. There are also the popular inexpensive radios like the ones made by Baofeng. There are also a variety of designs and kits by various radio clubs that can be instructive and satisfying to build.

[haughty patrol]

I guess I need to find those kits, cuz I didn't want a baofeng in a box hooked to a raspberry pi, it just feels really silly, I don't understand why all the ham stuff I find is like prebuilt radios.

Not when all other forms of radios out there I can buy chips and have.

[primal warren]

There are also some nice RX/TX modules by Dorji, NiceRF, Sunrise, etc, and some Arduino type shields (HamShield, ArduTrx, etc.) built around them

Some kit pointers here, but most of the kits I see are for HF, as VHF and UHF require special parts and techniques, so the designs are complicated and subtle and require somewhat exotic parts. No problem for manufacturers cranking out thousands of units, but hard to justify for building a kit you might sell a few dozen of.

https://www.hamradiosecrets.com/ham-radio-kits.html

[haughty patrol]

Ahhh that's some information I didn't have that makes some things make more sense.

[primal warren]

If you don't mind using the prebuilt modules as the heart of a transceiver, some useful part numbers to look for are the Dorji DRA818V, NiceRF SA818-V, and Sunrise SR_FRS_1WV

[haughty patrol]

I'm not really involved in the local club cuz I haven't bothered, and I can't ever seem to get a word in edgewize with the old men on the repeaters so I'm kinda just out here in ham world googling with no one to talk to.

[primal warren]

That's frustrating. There are several hams on this Discord, and we're happy to answer questions and help out how we can

[haughty patrol]

OMG those are some of the things I've been googling for for months with no luck.

Thank you!

[primal warren]

That's the result of me chasing a bunch of things down for a while, so I'm happy to share the results of my efforts

Most of the modules are based on the same chip (Auctus AT1846S or a clone), along with an RF amplifier of some sort, some switching logic, and antenna matching circuitry

Oh, I found a breakout board that could be handy http://hamshop.cz/rf-modules-c36/pcb-for-dorji-dra818v-or-dra818u-modules-i271/

Not much to it, but it looks really handy

[haughty patrol]

looks handy, now just to figure out how to order one of the chips LOL

[primal warren]

It looks like a couple of them are available on Tindie for good prices

[wraith panther]

Aliexpress has them too if you dont mind the month wait

[primal warren]

https://www.tindie.com/products/dorji_com/ham-amateur-radio-module-dra818v/ and https://www.tindie.com/products/sunrise/sr_1wv-ham-radio-amateur-radio-2-way-audio/

[wraith panther]

🤔 they have other RF modules lol

[primal warren]

Yeah, there are a variety to choose from, I figured I'd just give a couple of examples I'd happened to see

[molten torrent]

I'm using a bunch of nRF24L01s for my class to control simple electronics on a robot. We have the radios connected to Arduino running the RF24 library by TMRh20. (See here for the circuit and some example programs: https://github.com/michaelshiloh/resourcesForClasses/tree/master/src/arduinoSketches/nRF24L01). We have perhaps 7 pairs of these working at a time. Each pair has a different address so they should not be interfering with each other's data, although obviously they are competing for bandwidth.

We're having lots of problems with the radio link randomly dropping out completely, or receiving incorrect data.

Do you have any suggestions as to what might be causing this, and more importantly, what I might do to overcome these problems?

Program for example https://github.com/michaelshiloh/resourcesForClasses/blob/master/src/arduinoSketches/nRF24L01/minimalnRF/minimalnRF.ino

[little forge]

I know there is a way to set the ble Device name to something clear so I know what device to select when trying to use the ble connect app to control my playground bluetooth but I can't find this info anywhere maybe I'm just not seeing it because i'm to tired to be doing this at the moment currently the code for my project is unchanged from the example code from https://learn.adafruit.com/bluetooth-turtle-bot-with-circuitpython-and-crickit/code-with-circuitpython. I think name I want will be play-ble I have tried a few different things like ble.name = "play-BLE" no luck Please san someone help me

[primal warren]

I think those chips are using a spread spectrum modulation that hops around between different frequencies in a known pattern, both to dodge interference from other devices and reduce interference they produce. The problem you're seeing could be that they're all using the same modulation pattern so instead of dodging each other, they're all stepping on each other. What I don't know is how (or if) you can change the modulation pattern using that particular library.

[primal warren]

Bluetooth is complicated, so I'm mostly guessing here. I think setting ble.name just sets what is returned after the connection is made. In order to set the name that's selected before you connect, I think you need to create and configure an Advertisement object (which you can import from adafruit_ble.advertising). Then set the short_name property of that object to the one you want, and hand that advertisement object to ble.start_advertising()

[molten torrent]

Thanks for the insight. I'll study the documentation and see if there's a way to manipulate that.

[stoic glen]

I m using a bunch of nRF24L01s for my

[misty sleet]

I'm trying to play audio from a microcontroller over the radio using Adafruit's Si4713 breakout board. The transmitter works when I play audio from my computer, and the microcontroller plays audio when I hook it up to a speaker, but for whatever reason it doesn't work when I connect the microcontroller's audio output to the transmitter's input. Any idea what I'm doing wrong?

[primal warren]

Three possibilities occur to me. The first one is that the transmitter is coupling RF into the data leads and scrambling the signal (I had the same issue when trying to build a UHF remote control once: the data wiring acted as antennæ and picked up the transmission). The second is a ground problem, where the units don't share a common zero-volt reference for some reason (floating power supply, ground loop, etc.). The third is a power supply problem: it could be the power draw of the transmitter is dragging down the power supply to the microcontroller.

[misty sleet]

The microcontroller and the transmitter have a common ground, and the microcontroller is doing another task at the same time, which seems to be functioning, so I don't think it's a power or grounding issue.

I have noticed that my radio makes a slightly different sound depending on which way I wiggle the wires, so it could be the first issue you mentioned. How would I go about fixing this?

A few other things I should mention: When the circuit is off and my radio is tuned to the station, I hear static, whereas when the circuit is on, I hear nothing, so the transmitter must be doing something. In some cases I can hear the audio from the radio very faintly but with a lot of static and a screeching sound.

[primal warren]

You can try using shielded or twisted pair cable, or simply adjusting the length or arrangement of the wires

[restive fjord]

Hello everyone, I am looking for example code (circuitpython or micropython) the subject of receiving BLE beacons with an nRF24L01 module. who can help?

[sterile tundra]

Just to make sure I’m understanding this right, is the communication Beacon -> nRF24L01?

[restive fjord]

https://www.youtube.com/watch?v=BejQ1GyEmWU - this is what i want to do but using circuitpython on esp32s2 with no bluetooth build in

[sterile tundra]

What libraries are you using for the nRF24?

[sterile tundra]

I found what seems to be the go-to circuit Python library for the nRF24. When I look at the BLE API for the library though, it seems to indicate the nRF24 has some limitations for BLE support. In addition, the library only supports broadcasting Beacon packets. There’s no support in the library for receiving Beacon packets. It might make sense to look into a different module for Bluetooth support. Here’s the page that explains the Bluetooth library and it’s limitations. https://circuitpython-nrf24l01.readthedocs.io/en/latest/core_api/ble_api.html

[gilded isle]

hello community , i want to make a lora based network . i want to ask if someone know how destination and node function work in adafruit_rfm9x library , because i need to set source and destination addresses for my communication

in my scenario a drone talks to a base

and this base talk to a server via lora

all communication is based to lora

[granite spear]

The base level LoRa protocols doesn't include any addressing scheme (every node on the channel receives every packet), so you'd either want to implement that yourself as part of the data packet, or use a higher-level protocol like LoRaWAN.

[restive fjord]

https://tapr.org/ was a good resource for ax.25

[normal drift]

If you are using the CircuitPython adafruit_rfm9x library, it supports the same header used for the Arduino RadioHead library. You can specify addresses with this. There are examples here: https://github.com/adafruit/Adafruit_CircuitPython_RFM9x/tree/main/examples take a look at the rfm9x_node1.py and rfm9x_node2.py examples.

[deft lion]

Is there BLE support in CircuitPython for Adafruit ESP32 Feather V2? Or, is it coming in the future?

[slender current]

@deft lion There is a PR to enable it for some ESP32s. Note that it's still partial support like the S3 has.

[grim dove]

Does the https://www.adafruit.com/product/3072 do encryption and auto-retransmit or just the non-Lora RFM69's

[normal drift]

They are implemented in software. Are you using Arduino or Circuitpython? I think the Arduino RadioHead library does support encryption for the RFM9x. CircuitPython library does not. Both will support auto-retransmit -- it is referred to as "reliable datagram" in the RadioHead Library and the same protocol is implemented in the CircuitPython library. I hope that helps.

FYI -- here is an example of using encryption with the Arduino RadioHead library https://www.airspayce.com/mikem/arduino/RadioHead/rf95_encrypted_client_8pde-example.html

I have not tried it myself...

[sweet jungle]

Hello Folks. I have several Adafruit Feather 32u4 RFM95 LoRa "radiofruits" that I stack on Feather M4s for a variety of applications (weather station, IR animal camera, telegraph, etc.). Normally when I am testing these modules I can get -20 dB or better received signal power when I set the Tx within a few feet of the Rx (normally using spring antennas). I always run at max power setting (tx_power = 23). Lately, though, I can do no better than -50 to -60 dB with the same setup/configuration. One possible difference is that I am running with CircuitPython 7.3.3 on the M4s. Previously I think that I was running with CircuitPython 6.x. I have done some tests where I have tweaked the tx_power) down to 13 (default), but it does not change the RSSI. Any suggestions as to what might be causing my low RSSI? Thanks!

[main girder]

is this the better place to ask about FONA stuff rather than #help-with-circuitpython

because I'm decently sure it's a software issue

see here #1050380383188951040

[round acorn]

Hello! I'm using ESP32 RFM95W module to send data over LoraWAN. I'm using the latest version mcci-catena/arduino-lmic library. Although it works well, I'm unable to effectively set ADR or change SF. There are several issue of this type on github, but they are very old and apparently are already fixed. So, I don't know anyone here has faced similar issue and could help me please

[round acorn]

It's always working at SF7BW125

[round acorn]

Okey, so if anyone if interested on this, I found that the first uplink SF can not be changed. My ESP32 was deep sleeping after the first uplink, and as it reset itself after the deep sleep every uplink was the first one...

[woeful vector]

Hi all
Anyone worked with SIm800L for HTTP post ?

[primal warren]

I've used the SIM808 for that, but you're probably better off asking your actual question than whether someone has "worked with" it.

[sweet jungle]

Weak RFM95 LoRa Signal Power

[fringe isle]

Hi I have a question. I have a lora node (925Mhz) that placed under concrete box with metal plat on top. I know the best option for optimal RF transmission would be to place the antena outside the box. But if I made cut holes for vent would the signal penetrate, so I can mount the antena next to the metal plate? If yes how big the hole aperture/diameter so the 925Mhz can pass?

[restive fjord]

I'd be very suspicious that a metal plate improves signal transmission (or reception).

Normally I'd just say don't do it.

I usually reframe questions as if the device had 'wants' and that I had no wants. That's the right view, I think. (I know it is there but it does not 'know' that I am there ;)

[primal warren]

10cm would be large enough that 925MHz wouldn't be greatly attenuated

[fringe isle]

Thank you for the answers

[fringe isle]

Could you tell me how you come up with that number?

[restive fjord]

468 / f = y
https://www.radioqth.net/dipole

A 1/2 wave dipole (antenna) for 27 MHz is about 17 feet long.

 $ dc -e '5 k 468 27 / p q'
17.33333 # feet

 $ dc -e '5 k 468 925 / p q'
.50594 # feet

[primal warren]

The problem is equivalent to a "waveguide beyond cutoff", so I just used this calculator https://www.everythingrf.com/rf-calculators/circular-waveguide-calculator

[stuck portal]

Can a LoRa Radio transmit and receive audio?

[granite spear]

Not natively. If you compress it enough, you could send audio data over a LoRa link, but it's something that you'd have to implement as a custom protocol.

[umbral oxide]

packets are only 252 bytes

[stuck portal]

Ok. I'll make something at some point. Thanks.

[edgy swift]

I have run into a problem that is perfect for the NRF52840. I have several of the feather express boards, but I am struggling with the basics in Mu/CircuitPython I just need to transmit audio from several sources to a Pi and record the data via a python program. The demo programs (most up-to-date) to BlueFruit connect don't actually connect. Can anyone point me to an example, or should I drop back to Arduino IDE for more success?

[young cove]

Have you tried the CircuitPython 8.0.0-betas? A few bugs were fixed, though the basic demos should work with 7.3.3. Plenty of people have success with this.

Transmitting a lot of audio over BLE could be an issue. Are these short chunks?

CircuitPython BLE on the Pi requires using the bleak library under the covers. There have been problems in the past with the Bluetooth firmware on the Pi, though that was while ago.

Further questions are probably better asked in #help-with-circuitpython to get more eyes on the problems

[stuck portal]

Using Arduino IDE, can LoRa transmit and/or receive audio?

[primal warren]

As discussed above, LoRa probably doesn't have the bandwidth to do so effectively.

[wary arrow]

Is there a chip anyone could recommend for combined ADC + FFT around 10MHz?

[half plover]

How deep and precise of a sample do you want

Do you want like 8 bit ADC? 10 bit? 12? 16? You’ll want to use a SPI interface for speed. Though usually high speed Analog front ends feed into an FPGA for really high speed FFT.

If you took something like this: https://www.digikey.com/en/products/detail/analog-devices-inc/AD9283BRSZ-50/621671 and used PIO on an RP2040 in Arduino, you could probably get really close to 10MHz

The ADC uses a parallel interface which would either be piped to an fpga or pio input that could sample all the input pins synchronously

[primal warren]

How much bandwidth do you need? Is it like 0 to 10MHz, or maybe 9.5MHz - 10MHz? Many units use a heterodyne technique where the incoming signal is mixed with another frequency to produce a lower frequency result that's easier to process.

[wary arrow]

Thank you, I didn’t know about this. Bandwidth is 5-15MHz ish

[wary arrow]

Thanks!!

[untold needle]

Need to make a radio receiver for frequencies from 10-40MHz and need the data to store somewhere on the circuit to be analysed later. I have been advised to look at heterodyne receivers. I was wondering how to make this in MHz rather than what most people's projects frequencies are, which is usually in KHz. I was wondering if this would make a difference that I would need to account for.

[granite spear]

One obvious difference is the amount of data you'll need to digitize and store. That'll need a high-speed ADC and a lot of RAM, depending on how long you need to record the signal.

[untold needle]

right, but circuitry-wise, what components would have to be different as well and is there anywhere I can simulate this. I have an oscillator for those frequencies and an amplifier, do I need a mixer or anything else. Or can I just follow what most people have done for their AM/FM radio receivers.

[granite spear]

Sorry, I don't have that kind of experience to advise you on. Digital guy... 😅

[untold needle]

no problem, thanks though

[primal warren]

The differences between kilohertz and megahertz are mostly component values and construction techniques, the basic principles are largely the same.

You will need a mixer if you want to employ a heterodyne technique. There are other approaches, of course, but I'd recommend the heterodyne one as, while it sounds complicated, makes the circuitry simpler to build and more modular.

[untold needle]

right, how do I calculate these different component values

[untold needle]

Yes I think I will go with it

Will I need a high speed ADC chip as EdKeyes recommended or is there any other way to do it(I'm also thinking about the weight here since my restriction is 300g)

[primal warren]

Most things scale linearly with frequency, so as the frequency goes up, you'll see smaller values for capacitance and inductance. You can play with values and frequencies with this simple calculator to get a feel for how it works. https://www.allaboutcircuits.com/tools/tank-circuit-resonance-calculator/

[untold needle]

right right right

thanks

[granite spear]

BTW, the high-speed ADC would depend on the bandwidth of the signal you're recording. I was originally assuming you wanted to capture a good chunk of the 10-40MHz spectrum, but if you're just tuning into a narrow channel within that range to listen for some specific lower-speed data, you can digitize at baseband instead, and the ADC could be more modest.

[restive fjord]

The old problem with building a superhet from scratch was oscillator stability.
For that reason it was not considered a beginner's project at all.

621.384 is the call-number that's useful (in the lending library)

and gooogle said
https://w7dk.org/index.php?option=com_content&view=article&id=210&catid=67&Itemid=120

[untold needle]

I'm definitely a beginner

however

I do have a book on superhets

[untold needle]

Well mainly frequencies from 20-30MHz

[restive fjord]

I would wild-guess that you can get a stable oscillator today and interface it successfully in a modular build, but I've never done it. ;)
For some reason or another the theory behind receivers is something I iterate through quite often whenever I'm working with the technology (using a transciever to converse on the air, or just listening passively).

[untold needle]

I'm a very novice beginner

I have no prior experience with radio receivers

[restive fjord]

You can hold a second inexpensive shortwave radio very close to the one you're listening to and decode SSB on an AM-only radio for example. ;)
Or CW.

[untold needle]

hmmmm, I guess I'm definitely going to have to go down the rabbit hole of churning through material

Building my own radio receiver that is not AM or FM is going to be tought

tought

tough

[restive fjord]

I'd ordinarily recommend an inexpensive vacuum-tube based receiver but getting one in working condition isn't easy, anymore. ;)

[untold needle]

Right

[restive fjord]

AM is almost 'trivial' but FM is not. Today with all 'the chips' yea it is likely to be easier.

[untold needle]

well 20-40MHz probably isn't trivial

especially seeing as I have to create a receiver that can store the data under 300g

[restive fjord]

I've heard of this project type before (wx balloon) (radiosonde) but never once has someone (much) later came back with an 'I DID IT' report (pity if they had and I'm just not remembering it. ;)

My brief look into it said 'get a radiosonde if you want a radiosonde' as they've already solved the problems.

[untold needle]

this is very similar to my project but mine needs to operate at 20-40MHz

[restive fjord]

The nice thing about the upper end of that is the length of the radiating element of the antenna .. gets shorter. CB (27 MHz) is about 17 feet for a 1/2 wave dipole (L = 468/f?)

[untold needle]

I dont even know to start with the heterodyne

I have an oscillator

and capacitors

but I dont know what values I need for my components and I dont know how to connect them

[restive fjord]

signal comes into the antenna. it goes into one leg of the mixer. a local oscillator is also injected into the mixer.

[untold needle]

right

but how do I store this data

[restive fjord]

it's not data

[untold needle]

What is it then? after its in the mixer

[restive fjord]

you get the four products of the mixer (sum difference freq a freq b) out of the mixer.

[untold needle]

because ideally I would like to graph the frequencies Im receiving

[untold needle]

right

[restive fjord]

It's still a signal you've just moved the carrier frequency

[untold needle]

right

but can the mixer store different frequencies or just the first one that hits it

[restive fjord]

Signals consist of modulation and carrier (For basic signals)

[untold needle]

yes

[restive fjord]

The modulation has the information (intelligence) except for CW transmissions (morse code) where the modulation is simply the presence of absence of the carrier.

[untold needle]

right

ok

[granite spear]

Maybe back up a second. Do you know what transmitter is being used to send whatever signals you're trying to pick up?

[untold needle]

the project is to emit radio waves to bounce back off the ionosphere then receive them in my weather balloon

[untold needle]

I do

[granite spear]

So what does the transmitted signal look like? It is just a carrier, is it Morse code, is it video?

[untold needle]

a carrier

but for it to reflect off the ionosphere

it has to be between 1 and 40MHz

We are only planning to do 3 or 4 specific frequencies

like 21, 28, 32 and 36MH

[granite spear]

Gotcha. This starts to be easier, then, since your receiver just needs to detect the strength of the signal at a particular frequency, rather than try to decode any actual data, etc.

[untold needle]

yes

exactly that

[remote pond]

so something like ionospheric radar?

[untold needle]

I guess so yes

its quite simple

send a carrier

it reflects

that hits my weather balloon

and I want it to basically just store the frequency received and the strength

[untold needle]

That sums it up basically perfectly yes

[remote pond]

you might want to modulate the transmitted signal somewhat, to reduce noise. do you need to get the time of flight?

[untold needle]

time of flight?

[remote pond]

like do you want to figure out how far away the reflection is?

[untold needle]

no

just strength and frequency

[granite spear]

It'll likely be easiest if you somewhat hard-code the frequencies you're actually using rather than getting a general "show all the frequencies" spectrogram. So you can feed the signal into specific filters, or switch your oscillator frequency to tune into the right channel.

[untold needle]

thats the plan

do i need fourier

[granite spear]

No, you shouldn't need Fourier, just a filter circuit or few. I'm not an expert, so I'll defer to other people here who are, but my mental model would be like antenna -> mixer -> filter -> amplitude detection -> low-speed ADC -> store the data.

[untold needle]

the hard code frequencies will be 27MHz, 27.5 and 28MHz

since we have an oscillator for 27-28MHz

These our our components:

ADE-1 RF Mixer
TL051 IF amplifier
MAV11 amplifier
AD8307 log detector

so I have 2 amplifiers, a mixer and a log detector @granite spear

and an oscillator for 27-28MHz

[granite spear]

Great, the AD8307 is the "amplitude detection" part I was fuzzy about. That'll give you a good signal strength measurement.

[untold needle]

write but couldn't we store the signal as analogue rather than using an ADC, what are the benefits of an ADCA

ADC

[granite spear]

"Store" usually implies saving it on an SD Card or something digital. To store it as analog you'd need like a tape recorder or something like that.

[untold needle]

right yes yes yes

[untold needle]

how do I save it on an SD card without programming it

does it just automatically go from the AD8307 to the SD

but if we alter the frequency we transmit on the ground, we might not be able to change the oscillator's frequency in-flight, we have 3 hard coded frequencies to transmit, 27, 27.5 and 28, so will the oscillator still work

[remote pond]

probably not. you need to read the analog output from the AD8307 with the ADC of an MCU or something, and write that out to the SD card

[untold needle]

Right, understood

[granite spear]

"Oh my sweet summer child..." Nothing in electronics is automatic. You have to force the circuit to perform your will through design and programming and blood and sweat and tears... 😉

[untold needle]

Hahahaha

but apart from saving it to SD

[remote pond]

except maybe for the kinds of circuits with hidden feedback that makes them look like magic

[untold needle]

the rest seems pretty simple, or am I missing something

antenna -> mixer -> filter -> amplitude detection -> low-speed ADC -> store the data.

this doesn't seem so hard

will this definitely work? It seems way too good to be true haha.

[remote pond]

so your (variable-frequency) transmitter is on the ground, and your receiver is on the balloon?

[untold needle]

yes

we are only planning to transmit 3 frequencies

27, 27.5 and 28MHz

[granite spear]

There's definitely a lot of details to get right in the process, but as radio projects go, it's not too bad.

[untold needle]

yeah......

how fast can we log something to an SD card though?

[remote pond]

i guess you could hop the LO among those frequencies, if you can tolerate discontinuous reception and don't want to have a communication channel to ask the receiver to switch frequencies

[untold needle]

mmmm, what about 3 different oscillators for the 3 different frequencies, I'd prefer a graph

well my group would

because depending on how fast we can store it, it affects the amount of samples. Therefore 27MHz could look closer to 27.5MHz with too few samples couldn't it?

[granite spear]

Pretty fast, at least 1000 measurements/sec without too much trouble. You probably wouldn't need anything like that.

[remote pond]

you want to sample all 3 frequencies simultaneously?

[untold needle]

no

different time points

so like 5 minutes of 27

5 minutes of 27.5

5 minutes of 28

[untold needle]

my friend in my group says you need at least 300,000 a second to reliable discern signals in the 10s of MHz? Is he mistaken

[remote pond]

i think that really needs more context to give a reasonable answer

[untold needle]

he first said, 'samples just means data points - the more data points we can take per second, the more accurate our reading is (27MHz will look like 27.5MHz with too few samples)'

then said we need 0.3MHz of measurements

[remote pond]

you can send a 1kHz bandwidth signal on a 27MHz carrier, and you'd still only need 1kHz of bandwidth to sample the demodulated signal

[untold needle]

I think he must be mistaken

[remote pond]

could you ask him to explain what he means by "27MHz will look like 27.5MHz with too few samples"?

[granite spear]

In this case your oscillator tuning and your filters would be what distinguishes the frequencies from each other, so you don't have to sample fast to tell them apart.

[remote pond]

your transmitter is sending unmodulated carriers at those three frequencies, one at a time, right?

[untold needle]

300,000 measurements/secright right

we don't want to send information on a carrier signal like FM/AM radio does, we want the carrier signal itself basically

[remote pond]

do you expect nonlinear behavior? that might be the source of some of the concern

[untold needle]

he is understanding now

thanks so much guys

and girls

[untold needle]

thanks a lot

[untold needle]

right so specific band-pass filters for each frequency of course

[untold needle]

yes, exactly that

I dont think we expect nonlinear behaviour, should we?

[remote pond]

if you wanted to measure them simultaneously, you'd need a mixer with more than 1MHz bandwidth, and then digitize the demodulated signal (tricky), or feed it into three narrowband filters, one for each possible transmission frequency, and sample those

[untold needle]

nah thankfully not, they are one at a time

[remote pond]

i dunno. the ionosphere does weird things sometimes

[untold needle]

this is very true

So far this seems quite...easy, I say that hesitantly because lots could still go wrong and we are all beginners in our team but your advice and EdKeyes seems great.

[primal warren]

That is a whole different kettle of fish, known as a "spectrum analyzer"

[untold needle]

well thats ideal, just knowing frequency and strength would suffice

but if it isnt much harder or impossible

a spectrum analyzer would be nice

since we do have to present our data

at the end of the project

[primal warren]

That, on the other hand, is a much easier proposition and you may not even need a heterodyne circuit, since you'd basically just looking for the presence or absence of a signal, and don't need to decode modulation. You might be able to do that with a set of tuned circuits/filters, one for each frequency you want to detect.

[untold needle]

thats the plan now, thanks to you and several others clearing it up

[primal warren]

Yeah, the more details we know, the more specific answers we can offer

[untold needle]

you guys are great at this problem solving

must have loads of experience

[untold needle]

the last problem I'm having is with the antenna

using my research, it says optimum length of an antenna is quarter wavelength

that would mean we need a 10m+ antenna

how would that work out

[granite spear]

You should be able to get a more compact one using a helical coil or something similar. There are lots of off-the-shelf options in the 27MHz band you can look at.

[primal warren]

If it's on a a balloon, you can just have a thin wire hanging down.

[untold needle]

Cant find anything like it

I cant find anything lightweight(<200g) that receives 27MHz

[untold needle]

How long would this have to be?

[primal warren]

A quick search for "antenna wavelength calculator" yields a bunch of hits including this one https://www.southwestantennas.com/calculator/antenna-wavelength which says 277cm for 27MHz. I have seen balloon transmitters tailing J-pole antennæ but a simple quarter wave one seems lighter and easier.

[granite spear]

This claims to be 150g... a lighter-weight "tactical" antenna. https://www.amazon.com/ABBREE-Tactical-42-5-Inch-Connector-Compatible/dp/B07VF563FB

[primal warren]

It's just a simple quarter wave wire, just a piece of wire would be lighter and cheaper and work as well.

[untold needle]

hmmmm

[remote pond]

you could use a magnetic antenna, but the receive signal strength calibration might be trickier with those

[untold needle]

https://www.roadtechmarine.com.au/3-foot-27mhz-helical-whip/p/DC2030

I found this one

any thoughts?

[granite spear]

But I think madbodger probably has the best solution. In most terrestrial applications the antenna needs some structure to stand up vertical, but for a balloon, you can have a wire just hanging down, which could be very light.

[untold needle]

riht

right

but how to connect that to my circuit

[granite spear]

...mumble mumble matching-network circuit mumble mumble RF black magic mumble mumble...

[untold needle]

https://www.sotabeams.co.uk/antenna-wire-lightweight-100m/ quickly I found this and its only 3.3g per metre, so 3 metres is 10g

but to connect that to my circuit

I dont know

[untold needle]

Black magic

[untold needle]

I agree but I don't know how I would connect that to my circuit

[granite spear]

To a first approximation, the antenna wire is just where the radio signal enters your circuit, so the wire connection itself would be an input to your receiver/amplifier/mixer. However, there might need to be a little bit of impedance-matching "stuff" or the like involved. <hand-wave>

[untold needle]

yes

the wire connection is just where the radio wave hits it

but I dont know how to connect that to my other components

I will research

[remote pond]

yeah, you probably need an impedance matching network to connect the antenna to your circuit's input. i think often it's a simple hand-wound transformer on a toroid? your RF mixer probably has specs on what impedance it needs to see on its ports

[primal warren]

Are you in Australia? Did you know about this balloon project done by a student there? https://diyodemag.com/features/new_heights_part_1_high_altitude_weather_balloon_artemis_global_tracker

[primal warren]

The impedance of a quarter wave antenna at its feed point is about 36Ω, and most radio gear is designed to operate into a 50Ω load. Since it's a receiver, not a transmitter, that impedance mismatch (which isn't too bad anyway) wouldn't cause a problem.

[untold needle]

my teammate seems to think I need fourier analysis to find out the exact frequency of the incoming wave from the antenna, without using fourier, how would I find out the EXACT frequency of the incoming wave

@granite spear if you can help with this that would be much appreciated

[granite spear]

As I understand your goal, you don't need to precisely measure an unknown frequency, but only need to measure the signal strength at certain known frequencies. So your filters will already zero in to say "Give me only the radio signals between 26.9 and 27.1 MHz", and then you'd have the amplitude of everything in that band, which would encompass your source signal.

It would be more complicated if you needed to measure the frequencies themselves, and in that case you would need something like Fourier analysis, but I don't think you need to do that as you've described the problem.

If you're worried about confusing the signal-strength measurements with background radio noise, then you may want to consider doing some slow modulation of your transmitter, like turning it on and off every second. That way the receiver data logs would show an obvious pattern in the signal strength, and you can distinguish between your transmitter's impact versus background noise.

[remote pond]

could you describe at a higher level the experiment you're going to be doing with this equipment? that might help narrow down the solutions

[remote pond]

asking your teammate why knowing the exact frequency is important might be helpful, if the transmitter will only be one of three fixed frequencies. also, there's no such thing as "exact frequency" in practice. any measurement of frequency will have sources of error, and finite Fourier transforms will trade off frequency resolution vs time resolution

also, if your three fixed transmitter frequencies span 1MHz (which i think is true based on previous postings), you will need to sample the mixer output at more than 2MHz (Nyquist limit) for your FFT, which could be challenging on a microcontroller

[half plover]

Getting into fpga territory

Or, at the very least. PIO territory

[remote pond]

PIO is digital only right?

[granite spear]

I believe so, yeah. It's not impossible that the ADC is also hooked up to the PIO state machine, but I haven't run across anyone using it like that.

[remote pond]

a parallel-output ADC might be one such situation

[half plover]

That’s what I was thinking

I suggested one a couple days back

[remote pond]

but 2MHz ADCs seem likely to be on the expensive side

[half plover]

$10 for a 50MHz 8Bit ADC parallel out

[remote pond]

nice

[half plover]

https://www.digikey.com/en/products/detail/analog-devices-inc/AD9283BRSZ-50/621671

Well, $11

But still not terrible

[remote pond]

wow this stuff got a lot cheaper since the last time i looked at parts in this space

[untold needle]

if we were to do fourier, we would do it at ground level rather than in the balloon itself

[granite spear]

Recording enough data on the balloon to do the analysis later might be challenging. Not impossible, but you'll be pushing the processor and SD Card to keep up.

[primal warren]

You don't have to do Fourier anyway, there are other techniques available to measure incoming frequency (such as sweeping the LO). But if you just need to know which of 3 known frequencies it is, that's a much simpler problem. If you actually need to know the "exact" frequency, it depends on what you mean by "exact" (Fourier is little help for high accuracy here). And I'd be curious as to why (are you trying to measure Doppler shift?).

[wary arrow]

What does DC mean in this, 0Hz? “frequency in the range from DC to whatever MHz”

[granite spear]

Yes, that's correct. DC is often used to mean "a constant signal", which would also be considered 0 Hz.

[wary arrow]

Thanks!

If I am trying to measure the strength of 50 predetermined frequencies between DC and 15MHz, would I be better off using some kind of filter bank, or an FFT (maybe in an FPGA)? Like the poster above, I’m just testing for the strength of a signal not its information content. I’m new to radio

[granite spear]

Hmmm, that's kind of a weird intermediate case. One tool to potentially have in your toolkit is the Goertzel algorithm, which is basically like calculating a single FFT term at a particular frequency.

[primal warren]

There's also the swept LO technique. You dial your LO into each frequency to test, mix it with the incoming signal, then measure the intensity at a fixed frequency out of the mixer

Note that a frequency range like that can be problematic. 1MHz to 15MHz is MUCH easier than 10Hz to 15MHz (let alone down to DC)

[wary arrow]

Intermediate because of the number 50?

[granite spear]

Yeah. Annoyingly large to handle every frequency differently, but annoyingly small versus doing a full FFT spectrogram, heh heh.

[wary arrow]

🙂

[wary arrow]

I think I’ll do the FFT approach. Harder, less specialized, but I’ll learn a lot!

[untold needle]

my teammate seems to think FFT will work perfectly fine since we will be getting the same frequency in and it will just be a sine wave. What are the problems with fourier because I remember there are some inaccuracies and difficulties with running fourier on this sort of input.

also @primal warren how would I sweep the local oscillator if it is a fixed frequency LO

[primal warren]

You would need a different LO if you want to sweep it.

[granite spear]

The main difficulty is that you'll need to digitize the signal at relatively high speed (~1 MHz) and either record all that data for later analysis or have a moderately beefy CPU to run the FFT in real time. That would probably also imply having a more sophisticated amplifier on the receiver, since you'll need to have enough SNR to read the instantaneous radio wave instead of averaging the power over many carrier cycles. Not impossible, but seems like overkill for the type of data you want to collect, assuming I'm understanding your project requirements correctly.

[untold needle]

2 LO's?

[primal warren]

That's a possible approach, a fixed one for receiving one particular frequency, and a controllable one for scanning the band

[untold needle]

mmm, any other way to get the exact frequency in real time/afterwards

[primal warren]

Again, it depends on what you mean by "exact frequency" here. The more exact, the more difficult.

[untold needle]

like within 0.01MHz

so relatively close

[primal warren]

Um, why?

[untold needle]

well my teammates want a pretty close estimation of the exact frequency, so that we can plot if the ionosphere changes frequency as an additional

[primal warren]

You might want to consider measuring relative frequency changes instead of trying to measure absolute frequency at that precision

[untold needle]

would fourier afterwards be the only way for exact frequency

[untold needle]

how would we go about that

[primal warren]

Fourier isn't really practical for that kind of resolution

[granite spear]

Note that there is no "the" frequency, since your receiver will be picking up all frequencies in a band. So you could look for "the frequency of the highest-power signal" or something like that.

[untold needle]

ok well what if i said within the nearest 0.05MHz

just closer than our band pass filter can get

[untold needle]

yes but we are only broadcasting one frequency at a time e.g 27MHz

so we want a pretty close estimation of the exact frequency afterwards or in real time

[granite spear]

Yes, but you aren't the only transmitter in the world. 😉

[untold needle]

thats very true

but it's unlikely any others will hit

[untold needle]

any way to get the exact frequency or pretty close without fourier though?

[primal warren]

You're getting into the realm of crystal ovens and the like here

[untold needle]

right

[remote pond]

could use a PLL to both lock onto the frequency for power measurement, and also get you a frequency measurement? though i'm not sure that gets you the required frequency precision

[untold needle]

how close would that be

I mean I just want to get as close as fourier would get or a bit closer

but without using fourier essentially

if that makes sense

[primal warren]

I was thinking a PLL, because monitoring the correction voltage would give a reading of how much the frequency had changed (which seems to be something they want to measure)

[untold needle]

It is an odd request I know

[remote pond]

i guess it would be a very peculiar sort of FM receiver

[primal warren]

Fourier will not get close. It's a pretty crude approximation. The numbers you're quoting aren't practical with a Fourier approach.

[untold needle]

thats true, my teammate will not believe it

any other way to get a bit closer than fourier?

[remote pond]

among other things, you'd need a huge number of samples to get near that precision with Fourier. i don't remember the relations off the top of my head

[untold needle]

mmmm

[granite spear]

Fourier should be able to do that, I think... Getting 0.01MHz with a 1MHz band is just 100 frequency bins.

[untold needle]

apart from fourier is there a different approach, just looking at all possibilities

Fourier seems like the best one though

[primal warren]

Earlier ranges quoted included "0-15MHz" and around 27MHz. Is there just a 1MHz band of interest? I may have missed that.

[untold needle]

27MHz, 27.5MHz and 28MHz we will broadcast

[remote pond]

i think it was a different poster who wanted 0-15MHz

[untold needle]

local oscillator is fixed at 28MHz

[primal warren]

Yes, there are lots of different approaches (such as the aforementioned PLL, swept LO, banks of tuned filters, SAW filters, etc.)

[remote pond]

how large a frequency shift are you expecting from ionospheric effects?

[untold needle]

not sure, probably not large though

[remote pond]

i mean i do recall reading about weird stuff like intermodulation via nonlinear effects from the aurora, but you haven't talked about this being a polar mission…

[untold needle]

no

[untold needle]

just looking at the best way to measure the frequency coming in without Fourier in real time/afterwards

[untold needle]

thats pretty close, anything without fourier would be nice though

[remote pond]

i think if a frequency shift does happen, you're also going to receive some amount of your unmodified transmitted frequency as well. you might be able to use this to your advantage

[untold needle]

how so?

but what if there is no frequency on the other hand

no frequency change?

i meant

[granite spear]

It's probably worth a little research to figure out what kind of ionosphere effects are happening, to decide whether they're measurable at all (with different kinds of circuits) and how much value you will get from the measurement.

[untold needle]

we are expecting no frequency change at all

but in case

we would like to know the (pretty close to) exact frequency

[granite spear]

Just watch out for scope creep. It's waaaay easy to accidentally make your project 10x as difficult chasing "just in case".

[untold needle]

well fourier seems like it would work

but another method would be nice

[primal warren]

You're likely to receive something like this, where the red line is your transmitted carrier, the small triangular set of peaks to the left of it is the echo from the ionosphere, the peak to the left of that is a terrestrial R/C transmitter, etc.

[untold needle]

wouldnt that require a spectrum analyzer

If we could get data like that

that would be amazing

[primal warren]

Basically what you're describing is similar to a spectrum analyzer.

[remote pond]

spectrum analyzers can typically be swept LO or FFT

[primal warren]

Maybe just buy a HackRF One and use that?

[untold needle]

this seems cool, what does this use

[untold needle]

would be perfect, but we have a fixed oscillator not a variable one

[primal warren]

But you still want to analyze the spectrum of the resulting 1MHz band?

[untold needle]

yes

but how to do that without another oscillator

[untold needle]

too expensive and heavy though

[primal warren]

You have the room and weight for a crystal oven and the battery pack to power it, and you can't loft a 200g HackRF One?

[untold needle]

well we dont

whole thing needs to be 200g

[untold needle]

would a HackRF do everything

200g including antenna

[primal warren]

I don't think you're going to have a crystal oven in that case

[untold needle]

no I dont think so

[remote pond]

hm, HackRF One goes up to 6GHz; i wonder how much extra power that is compared to a fixed LO (or narrow-range swept LO)

[untold needle]

oh yeah

we have a 200mA draw limit too

[primal warren]

Yeah, it's probably overkill (for example, they don't need a transmitter)

[untold needle]

😂

[untold needle]

it seems like the only way is fourier

[untold needle]

seems like fourier is the only way apart from maybe a PLL but I'm not sure if that works

[untold needle]

wait a second, didnt you say a 2nd local oscillator could be used to sweep the first one and act as a spectrum analyzer?

@granite spear can you confirm, since I have found a decent local oscillator that operates 1-160MHz and is relatively cheap

[granite spear]

AFAIK if you use a variable LO, you wouldn't need the original fixed one. But yeah, in combination with a narrowband filter, you should be able to get spectral data that way.

[untold needle]

wait I wouldn't even need the fixed one?

[untold needle]

whats the difference between a narrowband filter and a band pass filter

[granite spear]

Yeah, if you can set your variable filter to 28 MHz, then it can assume the role of the fixed LO.

[untold needle]

well the variable oscillator is on its way

so could I use that instead

I mean it was only £2

so if not thats fine

[granite spear]

Same thing... "narrowband" just implies a relatively tight window, like 1.00 +/- 0.01 MHz if you want to measure frequencies with that kind of precision.

[untold needle]

oh

an IF narrowband?

got it

@granite spear something like this???

ive got mixer, amplifier, filter, what is an envelope detector and video filter???

[granite spear]

You wouldn't need those, instead you'd feed the signal at the end to that power-detector chip you have.

[untold needle]

oh the logger?

[untold needle]

the AD80.... logger

[untold needle]

oh thats easy then, only other part I am confused about

Ramp generator???

[granite spear]

That's to control the frequency of the variable LO... It probably takes a voltage input to set the frequency, so a ramp just sweeps through the frequency settings.

[untold needle]

oh thats completely fine

[untold needle]

and an attentuator?

[granite spear]

Not sure about that. Probably it's to keep the mixed signals at a similar amplitude or something like that.

[primal warren]

No, I had said you could have one fixed LO for your fixed frequency and a second variable LO for sweeping.

[untold needle]

would this not work as a spectrum analyzer

[primal warren]

How you control your variable oscillator depends on the oscillator. Some use analog control, some (like DDS types) use digital control. It's hard to obtain accuracy with an analog controlled one without a PLL setup, but the crystal controlled DDS ones can be fairly accurate.

[untold needle]

its a crystal one

https://uk.rs-online.com/web/p/crystal-oscillators/2177989

this one

[primal warren]

That one ... is not variable.

[untold needle]

it is variable isnt it?

[primal warren]

No.

[untold needle]

right.

https://www.giga.co.za/ocart/index.php?route=product/product&product_id=60

@primal warren this one is definitely variable isnt it?

[primal warren]

It's variable, I'm not sure if it will be accurate enough, and it would take some modification to accept a control voltage (or current)

[untold needle]

cant find any variable local oscillators for my frequency

this is ridiculous

[primal warren]

Since you want accuracy, you might want to take a DDS route instead. There are inexpensive AD9850 / AD9851 modules available that might suit your purpose.

[untold needle]

are there?

are they variable local oscillators

I dont know what DDS is

[primal warren]

DDS is Direct Digital Synthesis, it allows a digitally programmable sine wave generator of high accuracy

[untold needle]

right but I want a variable local oscillator dont I?

[primal warren]

You want a source of a high accuracy variable frequency sine wave to feed into a mixer

A VFO is the simplest approach, but may not provide the accuracy you want

[untold needle]

found this one from digikey

https://www.digikey.co.uk/en/product-highlight/a/analog-devices/ad9850-cmos-complete-direct-digital-synthesizer

does this do the same job as a local oscillator

[primal warren]

Basically, but you'll need a microcontroller or something to tell it what to do. I like to buy modules containing that chip instead of trying to build the circuitry to support the chip myself. The modules are fairly common (here's one on eBay https://www.ebay.com/itm/294988082847)

[untold needle]

I mean we have an arduino?

[primal warren]

Yes, you can control it with an Arduino

[untold needle]

oh thats brilliant thank you so much man

[untold needle]

thats perfect as well, thing is delivery is a month away, but there are different ones probably that are quicker

[primal warren]

Modules like that are available from lots of places.

[untold needle]

yeah I see that

this seems amazing

[untold needle]

so with this acting as a local oscillator, a spectrum analyzer circuit would now work

[primal warren]

Yes, you'd just have the software cycle through a range of frequencies and measure the received strength for each one

[untold needle]

software?

[primal warren]

Yes, the Arduino will need to run some software to control everything

[untold needle]

oh right yes

of course

only problem now is I'm thinking of weight

needs to be <200g

[wary arrow]

What’s a low latency way to collect ADC samples for FFT integration, on a CPU? Dump into a ring buffer?

[half plover]

Use something with a DSP core and use DMA to transfer the ADC data

But you could use SPI based ADC on a board that has SPI DMA capabilities to essentially do the same thing

[viscid vector]

hello

do any of you know of any arduino-esque boards with half decent ble built in? I'm using the Portenta H7 and BLE Nano and talking between them but im finding that the nano is a massive range limitation, cant even reliably communicate when 1m away

i need reliable communication at a maximum range of 3m

and i dont have budget for a second portenta

using a board with BLE built in is ideal since this project is large in scope and this was meant to be the simplest part

and the nicla is gone from shelves

[granite spear]

Adafruit's own Bluefruit boards might be ones to consider. There are both ARM https://www.adafruit.com/product/3406 and AVR https://www.adafruit.com/product/2829 versions, both Arduino-compatible.

[grizzled herald]

Anyone aware of a way to use ESP32 as a BLE packet capture for Wireshark? I've done similar with Nordic's "Connectivity Firmware" and wondered if there was a project for ESPs

[young cove]

If you do a websearch for ESP32 BLE Sniffer, that turns up a bunch of things.

https://github.com/Matheus-Garbelini/esp32_bluetooth_classic_sniffer has wireshark support, it appears

[grizzled herald]

I'll check it out!

[fallow plank]

I got 2 of the RFM69 modules, soldered uFL connectors, got those nice 433mhz antennas and all of that for a glorious 70 meters in direct eyesight.
When there is something in between they do not work at all.
Am I doing something wrong?

It's my first time with radio stuff

I tried to keep the soldering as flat as possible with no pointy ends

Running them at default settings of the circuitpython module

Would it inprove if I desoldered the uFL and put an sma connector directly?

[waxen idol]

Hello I have a rfm9x lora radio 433 mhz and for some reason when I plug it into my computer, it j wont turn on or get detected. I tried putting it in bootloader mode, but it doesnt get detected by the computer all it does is blink a red LED , and when I plug it in, the yellow CHG light blinks once really quickly and doesnt turn on again. Anybody have any ideas on what could be wrong?

[granite spear]

You might want to specify exactly what board you have, as the LoRa radio itself won't have a PC interface. There'll be some other microcontroller on the board which is controlling the radio.

[waxen idol]

Oh it’s a feather 32u4

[azure fox]

Anyone have any RFM95 Arduino examples to learn from?

[worn bridge]

Unlikely, if it works without anything in the way it’s probably just an inability to penetrate the object in question. 433MHz is susceptible to interference by solid objects, and the ability to pass through something does depend on material as well.

[fallow plank]

it can't pass a small tree

Could you then recommend me some modules that got some better range?

[worn bridge]

I unfortunately don’t have enough personal experience with other modules or hardware alternatives for that type of range and application. One possibility is to evaluate different types of antenna for better application-specific transmission?

[fallow plank]

Alright, thank you very much!

[remote pond]

yeah, i think 433MHz is high enough frequency that attenuation due to trees, etc. starts to be noticeable. living things are mostly water, and water can absorb radio waves…

[west perch]

Silly question, you do have the 433MHz modules to go with the 433MHz antennas, right?

[fallow plank]

yep

Everything is 433mhz

[fallow plank]

preamble len seems to help a lot

cranked it to 35

[little forge]

Im wanting to create a number of wearable projects using mainly esp32 s2 QT Py boards for the wearable bits and have a control box in my hand bag besed either on an esp32s2 Feather or a Pi zero W. I think I can create a wifi network without connecting to the internet on the Pi I know I can create wifi on my phone but that auto connects to the internet but if I set up so not connecting to internet on my phone I'd Like to see some examples of multiple ESP32 s2 boards sending Data to and from each other If there is an example to Do this on the Adafruit site or elsewhere on the net using Circuitpython, MicroPython or Arduino then Links would be very helpful. Thanks

[short gyro]

" Like to see some examples of multiple ESP32 s2 boards sending Data to and from each other" I think using something like LoRa is a better option than wifi

[young cove]

Double-check that the boards are 433 MHz. The raw radio chip can operate at multiple frequencies, but there are components on the board that tune the output circuitry to 433 or 868/915 MHz.

Are the antennas vertical?

What is the RSSI reported on both sides? They should be more or less equal when A receives B and vice versa.

For entertainment you could try the higher frequency and see whether you get a better RSSI. That would say they might be mislabeled.

a simple vertical wire cut to the right length is often sufficent as an antenna

[fallow plank]

I had taken a pic

[young cove]

is that our antenna? which part number?

[fallow plank]

nah, local retailer didn't have yours

[young cove]

could you give a link?

a wrong antenna could greatly decrease radiated power

[fallow plank]

https://grobotronics.com/antenna-rf-434mhz-2dbi-117mm.html

same accursed shop I got the bad flux, I however couldn't find anywhere alse

They have no text whatsoever on them

RSSI on both sides is close to 20 when close, +-5 side to side

[young cove]

You could try just tacking a vertical wire to the center antenna output. Then make it straight vertical: https://learn.adafruit.com/adafruit-rfm69hcw-and-rfm96-rfm95-rfm98-lora-packet-padio-breakouts/assembly#wire-antenna-2433240 (pictures there don't show it vertical). Length as specified

[fallow plank]

Honestly at this point, yolo

I will disconnect the antenna, but not desolder the ufl connector, is that fine?

[young cove]

70m is getting toward the far range on these

[young cove]

yes

also make sure the sw is setting the frequency correctly. You could also try the Arduino test programs to see if there's some issue with the CircuitPython libraries.

[fallow plank]

Ah so these modules don't go much further with undirected ants?

[young cove]

I take that back, that should be fine for unobstructed. I was thinking about inside a house

[fallow plank]

Tested 915mhz, at point blank rssi -60

while 433 does -24

[fallow plank]

The library is ancient.
Ack code is broken.
Uses ints instead of bools for enabled properties.
No useful error messages whatsoever.
Docs potato.
I will tidy up once I am done getting harassed by the module.

[young cove]

-24 sounds pretty good.

any diff with the wire antennas?

there are a number of range discussions in the adafruit forums

[fallow plank]

I want to find the perfect cable first, solid core chopper. It will take me a few days to find it.

[worn bridge]

https://learn.adafruit.com/mqtt-in-circuitpython/circuitpython-wifi-usage shows an example of an ESP32-S2 board connecting to Adafruit.io via MQTT. With a bit of modification and some setup on a Pi Zero W, you can have the Pi act as an MQTT broker in place of Adafruit.io (see "Self-Hosted Brokers" under "Connectiong to an MQTT Broker" for details) and send data back and forth that way.

[normal drift]

Have you looked at this guide? https://learn.adafruit.com/adafruit-rfm69hcw-and-rfm96-rfm95-rfm98-lora-packet-padio-breakouts/rfm9x-test and at the examples available in Arduino after you install the RadioHead Library http://www.airspayce.com/mikem/arduino/RadioHead/index.html

[normal drift]

What library are you referring to? This is not "constructive criticism".

[fallow plank]

Oh sorry, I was talking about the CircuitPython RFM69 library.
I will take care of it myself.
It's pure python afterall.

[normal drift]

You are welcome to submit PR's to the repository if you make improvements.

[normal drift]

What is "broken" about the ACK Code? Please file an issue in the repository if something is broken https://github.com/adafruit/Adafruit_CircuitPython_RFM69

[fallow plank]

I am incredibly busy these days. I have noted it down and will file proper issues late next month.

University exams.

[normal drift]

Good luck with your exams!

[fallow plank]

Thanks.

[normal drift]

FYI -- I did just verify that the rfm69_node1_ack and rfm69_node2_ack examples run as expected on my feather_m0_rfm69 boards with the current library and Circuitpyhton 8.0_rc1

[fallow plank]

For me the acks always returned true even though node2 was disconnected from power

perhaps I did something wrong then
My code was based on simpletest
I didn't see the ack examples

[normal drift]

I am not seeing that - mine fail the ack if either is disconnnected.

Wen you have time -- feel free to post your example that is not working (or file an issue).

[fallow plank]

I will do everything in the issue.

[untold needle]

Hi guys, I get how to make a band pass filter for my specific values and everything

but how do I make more than one

so one for 26 --> 27.5 MHz and then another for 27.5 -->29Mhz

2 separate

can I make it so that the signal goes to 2 band pass filters

[untold needle]

@primal warren @granite spear

this is the circuit diagram I have from antenna to filters so far

the top filter is from 1-500KHz

the bottom one is from 500KHz to 2MHz

does this seem right at the moment?

it looks far too simple for me, however that does make sense as right now neither the local oscillator nor antenna circuits are present

[half plover]

Top filter (low band) doesn’t look right

[primal warren]

Yeah, just split it up, run it through the two filters, then recombine. Unless you just want one filter at a time, then you just need some signal switching

[untold needle]

Yeah I'm not sure how to split up the filters

[untold needle]

Was my circuit diagram correct

[primal warren]

I'm still fuzz-brained while getting over COVID, so I'm not sure, but it looks to me like you're using one filter as a ground reference for a second filter, and that doesn't seem right to me.

[half plover]

I feel like it should be a cascading band pass filter

Put either one at the source and then the other in line

[primal warren]

Like, two in series? I think that would filter out everything, which is why I proposed two in parallel

[half plover]

Well, I’m not necessarily saying have two band pass filters in series; I’m trying to remember what I’m trying to convey but my brain is fuzzy as well

Not from Covid, just being up all night with sick child 😵

[young cove]

Since these bands are adjacent, I'm confused. Do you want to take the input and direct it to two filters, and then keep their outputs separate? Or do you want to combine their outputs? If the latter, why not a single filter covering the two bands. Also you said these are 27ish MHz filters, but the diagram implies they are AF filters after a mix-down to audio frequencies. What is the use case here?

[granite spear]

I believe he's trying to measure the signal strength for two different possible transmit frequencies.

[untold needle]

I solved this issue now, but now I'm coming up against a fourier transform issue

my professor has told me it wont work due to the nyquist limit

and me not getting enough samples to do fourier

@granite spear do you know about this and can you suggest any possible solutions

due to the nyquist limit I would have to get something like a sample every nanosecond, is this possible, or is there another way with fourier.

[granite spear]

My understanding was that you were trying to measure frequencies between about 26 and 27MHz, so a 1MHz band? That would ordinarily need sampling on roughly the 1 microsecond scale or a little better for Nyquist, but nanosecond sampling would be a major overkill, I think.

[untold needle]

even with mixing the frequency will still be around 0-2MHz

or even 0-3

also how do I sample this quickly

even 1 microsecond or below

[granite spear]

It's challenging, as you'd need a relatively high-speed ADC and data-processing path. Not impossible, but it's not easy, so it's worth examining whether you really need to go down that path or if you can do something simpler like just signal-strength measurements at a couple of known frequencies.

[untold needle]

but fourier requires nyquist doesn't it?

[granite spear]

Yes. The Nyquist limit is a general statement about how much information about frequencies you can get by sampling in time.

[untold needle]

a high speed ADC is not a problem

[untold needle]

yes I understand this but we would like to run fourier, we can have a very fast ADC but I don't know what a data-processing path is

[granite spear]

I mean, you need to be able to store all the data from the ADC and then run a fast-fourier transform algorithm on it, so that requires a decent chunk of memory and CPU power that might be challenging for microcontrollers.

[untold needle]

we don't need to run FFT on the balloon itself

we just need to store enough voltages

to run FFT at ground level

when the balloon comes back down

[granite spear]

Well, then you need to have a lot of fast data storage instead, so still challenging.

[untold needle]

so fourier is looking unlikely?

[untold needle]

why can't it all just store on the SD card from the very high-speed ADC

[granite spear]

It just depends on the time available and the skill level of the folks involved. The real question is, do you truly need it or are you just doing it because someone said it sounded like a good idea? 😉

[untold needle]

well fourier was our initial plan

actually

I tried to find alternatives

[untold needle]

not immensely skilled

but you also said fourier seems like a decent idea

so

That was our final plan

until we found out about nyquist

[granite spear]

Again, it's not impossible, but the word "just" there hides a lot of complexity, heh heh.

[untold needle]

well we have a month

and I'm willing to work quite hard

I'm pretty determined

I also have holidays this coming week

so I could spend 3-4 hours on it per day

[granite spear]

I probably said it's a decent idea IF you are required to measure exact frequency spectra. But it depends on what your goals are.

[untold needle]

So I'd say researching and skill wouldn't be the issue here

[untold needle]

exact frequency is what we want

band pass filters are a back up

the sampling rate is the only problem

one sample every 150 nano seconds seems viable

[granite spear]

Yep, you're going to want something like a 5-10 million samples/sec rate.

[untold needle]

yes something like that

there are 20MSPS ADCs that dont draw too much current

what else is requires apart from a very high speed ADC

[half plover]

This all playing out like the Digital Spectral Analysis grad class I’m taking

[untold needle]

any suggestions to get like 10-20 million samples/sec

[half plover]

Yeah, one sec

[untold needle]

thanks

[half plover]

https://www.digikey.com/en/products/detail/AD9283BRSZ-50/621671

[untold needle]

jesusa

[half plover]

A 50MSPS pipelined 8-bit parallel out ADC

[untold needle]

50 million a second

what else is required

[half plover]

Probably an FPGA

[untold needle]

there must be a catcg

cxatch

FPGA?

[half plover]

Or something that can access pins in real time

Maybe a pocket beagle

[granite spear]

Heh heh, and here's where that word "just" pops up to bite you. 😂

[half plover]

The PRU on that has single cycle gpio state access

[untold needle]

I need this

[half plover]

Plus 200MHz means you should be able to easily avoid any issues over sampling doing periodogram

[untold needle]

i only need like 20MHz

[half plover]

Yeah

[untold needle]

so I need an ADC

and what else

[half plover]

A microcontroller that’s fast enough

[untold needle]

arduino?

[half plover]

And has the ability to read 8 pins at a time

Like an RP2040

[untold needle]

RP2040 never heard of it

[half plover]

PIO on that gives you single cycle access to up to 8 pins on each PIO register

RP2040 is on the Raspberry Pi Pico

[untold needle]

RP2040 looks pretty good

[half plover]

135MHz, 2x 8 bit PIO registers that can run independently of the CPU and transfer directly from the registers into your FFTs

[untold needle]

thats amazing

but theres always a catch

ADC, RP2040

[half plover]

C programming can be hard

[untold needle]

it cant be just 2

[remote pond]

i think it doesn't have an FPU so you'd have to use an integer FFT?

[half plover]

It has a software FPU

[untold needle]

I need to store the voltages

not run FFT there itself

[untold needle]

isnt it python?

[remote pond]

hm, can you write 20Msamp/sec to an SD card?

[half plover]

Ah, yeah just datalogging you can store the ADC values right to a file

[untold needle]

no need for RP2040?

[half plover]

You can use it but it will be considerably slower

No, rp2040 is necessary

[untold needle]

for what though?

[half plover]

Getting the ADC data

[untold needle]

so it goes through a high-speed adc

[half plover]

Let me draw a diagram

[remote pond]

the ADC can't write to flash storage on its own, much less an actual file on a filesystem

[untold needle]

yes please

this seems pretty difficult but I can work hard

[untold needle]

why is there computer

[half plover]

You can put the data into MATLAB

MATLAB would read the values from a CSV file that the RP2040 would write to

Or a text file, either would work in this case

The PIO data is store in a large array and written to the data file

Then use MATLAB to run all your FFTs

[untold needle]

cant I shove it into an SD card as binary or something

[half plover]

I mean, you could log it to an SD Card but the Pico already comes with 2MB of flash so I’m not sure why you’d want to add additional complexity

[untold needle]

oh yes

just save to SD Card

[half plover]

And it can be loaded as a file system directly accessible by a computer

So unless you expecting to log tens of MB of data, I don’t really see it necessary

Adafruit sells a feather with 8MB of flash as well

[untold needle]

well sampling rate of 20MHz

what coding has to be done on the RP2040

@half plover and is python fast enough

since I'm decent at python but have almost no C expertise though I'm not afraid to learn

[remote pond]

how long do you need to sample for?

[untold needle]

intermittent for like 4-5 minutes at t tiome

time

[remote pond]

um, 5 minutes is 6000MBytes of 8-bit samples at 20MHz (edit: ugh, can't arithmetic today)

or do you mean sampling at intervals of 4-5 minutes, but only for enough samples to get a good FFT?

and this is going to be flying in a balloon, right? or was that someone else's project?

[half plover]

4-5 minutes seems like a really long time

Even systems for like cellular signals only sample for maybe a few hundred milliseconds at most

[young cove]

@untold needle Do you have to sample the whole band at once? Could you sweep the range with a variable oscillator and mixer? Not sure whether you're trying to get detailed info on signals in a whole band at once or just detect presence.

[untold needle]

thats the one

[untold needle]

dont have a VFO

[young cove]

you mean too late to add one to the radiio part? Si5351?

is the point of this the design part, or could you get an off-the-shelf SDR and do what you want?

[untold needle]

plus I think this plan is working

[untold needle]

we could do 10 seconds but we arent 100% sure we will get any data

or even milliseconds

we will do 0.1s samples

yes

but @half plover even 0.1s of 40MHz takes up like 20MB or more

[remote pond]

with the data storage requirements, it might be easier to do the FFT on board. but that means a microcontroller that can compute it fast enough

i guess you could stream a data downlink of the raw samples using a carrier frequency that's not as affected by the ionospheric phenomena you're measuring, but that's also challenging, because now you have a high-bandwidth digital RF link design problem

[granite spear]

Hahaha, at that point you could just upconvert the received signal directly to a different band and retransmit it.

[remote pond]

could be tricky. ionospheric doppler shifts are apparently a thing, though i don't know offhand whether they're large enough to interfere with this particular measurement

[untold needle]

We are going to do 50ms intervals

Or 5ma

5ms

We will do 5ms every 5 minutes

Over 3 hours

Every 2 minutes actually

Which is like 0.45s overall

@remote pond @half plover 0.45s total won't take up that much storage surely

Only problem now is will python FFT be fast enough for all the data points

@granite spear i take it you have some expertise in FFT

Will python fourier be fast enough for 0.45s at 30MHz, so effectively 15 million samples

[granite spear]

Grabbing the data from the ADC will still be tricky, but you should have enough RAM to store milliseconds of data easily, and you can take your time to run an FFT on it or save it to SD Card.

[untold needle]

I think @half plover suggested an RP2040 to grab the data

And I know python FFT is O(n log n)

[granite spear]

Yes, a RP2040 can do it. I'm just saying that it will also take some relatively advanced coding.

[untold needle]

gulp

How advanced are we talking

Is 50 hours enough? I'm not a beginner as duh

Suh

Such