💬 MySensors NRF5 Platform

scalz

@rmtucker said in MySensors NRF5 Platform:

So i am afraid i2c is still a non starter.

maybe try this one https://github.com/felias-fogg/SlowSoftWire
I don't use it, but it compiles fine for nrf52. I already tested it a while ago, and it worked for me.

rmtucker

@scalz
my head is battered with this.
tried the library you mentioned and it compiled fine but still gave spurious readings.
so i plugged the htu21d back in to a nano and used the sparkfun library demo.
the readings were perfect.
so frustrated with this because all my nodes use i2c sensors.

NeverDie

@rmtucker
Stupid question, but are your i2c sensors getting all the voltage that they need? After all, a nano runs at 5v, right? I know that some of the TH i2c sensors include voltage regulators, which would drop the voltage even further. If necessary, those could be removed. That's what I generally do.

rmtucker

@NeverDie
Yes They are 3.3v sensors.
What we need is someone else using nrf51822 and i2c sensors that can shine some light.

Nca78

@rmtucker said in MySensors NRF5 Platform:

@NeverDie
Yes They are 3.3v sensors.
What we need is someone else using nrf51822 and i2c sensors that can shine some light.

I'll try soon, hopefully I'll have time tomorrow, my nrf51 module is soldered on a breadboard. I need to make the promised tests on the nrf52 too ...

rmtucker

@Nca78
Just an observation after many hours of tinkering.
There is no problem reading one byte back from i2c,like when doing an i2c scan.
It comes back with the address of the htu21d no problem.
But if you try to read 3 bytes which is what you do when reading the temp etc,it goes T*TS up.

rmtucker

Well guys after all that
I had not installed the development version from git.
After installling the development version the i2c is working fine.
Suprised no one else has done this??

From git (for core development)

Follow steps from Board Manager section above
cd <SKETCHBOOK>, where <SKETCHBOOK> is your Arduino Sketch folder:
OS X: ~/Documents/Arduino
Linux: ~/Arduino
Windows: ~/Documents/Arduino
Create a folder named hardware, if it does not exist, and change directories to it
Clone this repo: git clone https://github.com/sandeepmistry/arduino-nRF5.git sandeepmistry/nRF5
Restart the Arduino IDE```

rmtucker

@rmtucker

Please can someone update the instructions below to mention installing the development version

Preparing Arduino
Arduino doesn't support the nRF5 platform in a standard setup. You have to add the arduino-nRF5 platform via Arduino's Board Manager like described in https://github.com/sandeepmistry/arduino-nRF5/#installing

Nca78

@rmtucker said in MySensors NRF5 Platform:

Please can someone update the instructions below to mention installing the development version

Preparing Arduino
Arduino doesn't support the nRF5 platform in a standard setup. You have to add the arduino-nRF5 platform via Arduino's Board Manager like described in https://github.com/sandeepmistry/arduino-nRF5/#installing

@d00616 maybe ?

NeverDie

If there are multiple libraries with the same name (as there seem to be), then how does the compiler know which one to pick? Is it just the first one whose name matches in whatever pathway it follows? Since it obviously compiled even though rmtucker was missing the proper libraries... this just sounds like a heartache waiting to happen unless everything is setup just-so. rmtucker's case seems to prove there won't be any warnings if it picks the wrong one.

scalz

looking at d00616 howto it's explained to use Mysensors development branch (or maybe he added it today ). but that makes sense as this port is newer than master branch.

@NeverDie
it uses the core libraries depending on the board used in Arduino.

mfalkvidd

@scalz: I think @rmtucker was discussing using the development version of the sandeepmistry nrf5 Arduino library, it was not related to the MySensors library.

rmtucker

@mfalkvidd
Sorry i should have made it more understandable.
When you follow the instructions below as per sandeeps site it does not install the latest NR5 libraries that you see on github.

Installing

Board Manager

Download and install the Arduino IDE (At least v1.6.12)
Start the Arduino IDE
Go into Preferences
Add https://sandeepmistry.github.io/arduino-nRF5/package_nRF5_boards_index.json as an "Additional Board Manager URL"
Open the Boards Manager from the Tools -> Board menu and install "Nordic Semiconductor nRF5 Boards"
Select your nRF5 board from the Tools -> Board menu

Following this you must install the development version aswell which is what you see on github by doing this.

From git (for core development)

Follow steps from Board Manager section above
cd <SKETCHBOOK>, where <SKETCHBOOK> is your Arduino Sketch folder:
OS X: ~/Documents/Arduino
Linux: ~/Arduino
Windows: ~/Documents/Arduino
Create a folder named hardware, if it does not exist, and change directories to it
Clone this repo: git clone https://github.com/sandeepmistry/arduino-nRF5.git sandeepmistry/nRF5
Restart the Arduino IDE

Otherwise you run into all kinds of problems using the old library.

d00616

I have added a note at the end of the installation documentation. To make things easier, I have opened an issue to update the Arduino-nRF5 release.

https://github.com/sandeepmistry/arduino-nRF5/issues/182

d00616

I2C should be fixed now. Version 0.4.0 is released!

rmtucker

@d00616
So should i be able to delete the hardware folder that the development version created from github and just update from boards manager?.

d00616

@rmtucker said in MySensors NRF5 Platform:

So should i be able to delete the hardware folder that the development version created from github and just update from boards manager?.

If this don't work, remove all folder created by hardware and boards manager, close Arduino IDE and reinstall it via Boards Manager.

Fabien

Hi, I don't us nrf52 yet but as i'm a platformio supporter, There is a new release for nrf52 platform with initial arduino framework support ! Good news, i think I will test nrf52 in few weeks.
https://github.com/platformio/platform-nordicnrf52/releases/tag/v1.2.0

NeverDie

@d00616 said in MySensors NRF5 Platform:

I2C should be fixed now. Version 0.4.0 is released!

To help clarify for others who are reading this: the Arduino IDE board manager let me upgrade to 0.4.0 from 0.3.0 by just clicking on the board entry "Nordic Semiconductor NRF5 Boards by Sandeep Mistry", selecting the 0.4.0 version, and then clicking Install.

d00616

I have updated the radio driver with fixes for some hardware errata.
I have updated the https://github.com/mysensors/ArduinoHwNRF5 with an enhanced definition format. File name and contents are changed! The board description is compatible with the NRF5 variant provided by Arduino, but it's not possible to use this variant at the moment.

NeverDie

@d00616 said in MySensors NRF5 Platform:

I have updated the radio driver with fixes for some hardware errata.

Where do we find the update? i.e. is it baked into the regular MySensors development code thread, or is it elsewhere?

NeverDie

I'm guessing it's this one? https://github.com/mysensors/ArduinoHwNRF5

NeverDie

I refreshed all the libraries. I'm able to compile and upload my sketches, but now it complains a lot about "invalid libraries."

WARNING: Spurious .ci folder in 'MySensors' library
WARNING: Spurious .mystools folder in 'MySensors' library
Sketch uses 3836 bytes (0%) of program storage space. Maximum is 524288 bytes.
Open On-Chip Debugger 0.10.0-dev-00254-g696fc0a (2016-04-10-10:13)
Licensed under GNU GPL v2
For bug reports, read
   http://openocd.org/doc/doxygen/bugs.html
debug_level: 0
0
adapter speed: 10000 kHz
cortex_m reset_config sysresetreq
nrf52.cpu: target state: halted
target halted due to debug-request, current mode: Thread 
xPSR: 0x01000000 pc: 0x000006e8 msp: 0x20010000
** Programming Started **
auto erase enabled
wrote 4096 bytes from file C:\Users\David\AppData\Local\Temp\arduino_build_255419/Interrupt_button_test_v005.ino.hex in 0.800001s (5.000 KiB/s)
** Programming Finished **
** Verify Started **
verified 3964 bytes in 0.060000s (64.518 KiB/s)
** Verified OK **
** Resetting Target **
shutdown command invoked
Invalid library found in C:\Users\David\Documents\Arduino\libraries\arduino-nRF5: C:\Users\David\Documents\Arduino\libraries\arduino-nRF5
Invalid library found in C:\Users\David\Documents\Arduino\libraries\ArduinoHwNRF5: C:\Users\David\Documents\Arduino\libraries\ArduinoHwNRF5
Invalid library found in C:\Users\David\Documents\Arduino\libraries\arduino-nRF5: C:\Users\David\Documents\Arduino\libraries\arduino-nRF5
Invalid library found in C:\Users\David\Documents\Arduino\libraries\ArduinoHwNRF5: C:\Users\David\Documents\Arduino\libraries\ArduinoHwNRF5

d00616

@NeverDie said in MySensors NRF5 Platform:

I refreshed all the libraries. I'm able to compile and upload my sketches, but now it complains a lot about "invalid libraries."

At the moment, I cannot reproduce this with Linux and Arduino 1.8.2. I try to find out what's wrong.

d00616

@d00616 said in MySensors NRF5 Platform:

@NeverDie said in MySensors NRF5 Platform:

I refreshed all the libraries. I'm able to compile and upload my sketches, but now it complains a lot about "invalid libraries."

At the moment, I cannot reproduce this with Linux and Arduino 1.8.2. I try to find out what's wrong.

Can you try to remove the .ci and .mystools folders?

NeverDie

@d00616 said in MySensors NRF5 Platform:

Can you try to remove the .ci and .mystools folders?

Done. So, here it is now:

Archiving built core (caching) in: C:\Users\David\AppData\Local\Temp\arduino_cache_16539\core\core_MySensors_nRF5_MyBoard_nRF52832_bootcode_none,lfclk_lfxo,reset_notenable_2d457965a40837233e636c052c58b359.a
Sketch uses 3872 bytes (0%) of program storage space. Maximum is 524288 bytes.
Open On-Chip Debugger 0.10.0-dev-00254-g696fc0a (2016-04-10-10:13)
Licensed under GNU GPL v2
For bug reports, read
	http://openocd.org/doc/doxygen/bugs.html
debug_level: 0
0
adapter speed: 10000 kHz
cortex_m reset_config sysresetreq
nrf52.cpu: target state: halted
target halted due to debug-request, current mode: Thread 
xPSR: 0x01000000 pc: 0x0000070c msp: 0x20010000
** Programming Started **
auto erase enabled
wrote 4096 bytes from file C:\Users\David\AppData\Local\Temp\arduino_build_610023/Interrupt_button_test_v006.ino.hex in 0.800001s (5.000 KiB/s)
** Programming Finished **
** Verify Started **
verified 4000 bytes in 0.060000s (65.104 KiB/s)
** Verified OK **
** Resetting Target **
shutdown command invoked
Invalid library found in C:\Users\David\Documents\Arduino\libraries\arduino-nRF5: C:\Users\David\Documents\Arduino\libraries\arduino-nRF5
Invalid library found in C:\Users\David\Documents\Arduino\libraries\ArduinoHwNRF5: C:\Users\David\Documents\Arduino\libraries\ArduinoHwNRF5
Invalid library found in C:\Users\David\Documents\Arduino\libraries\arduino-nRF5: C:\Users\David\Documents\Arduino\libraries\arduino-nRF5
Invalid library found in C:\Users\David\Documents\Arduino\libraries\ArduinoHwNRF5: C:\Users\David\Documents\Arduino\libraries\ArduinoHwNRF5

BTW, I'm using the Windows version of Arduino IDE 1.8.5, which is the most current.

d00616

@NeverDie said in MySensors NRF5 Platform:

BTW, I'm using the Windows version of Arduino IDE 1.8.5, which is the most current.

I don't know the correct path on an Windows system. On my system the Sketches are in the ~/Arduino folder in my home directory (~). The arduino-nrf5 is in ~/.arduino15/packages/sandeepmistry/hardware/nRF5/0.4.0 and the ArduinoHwNRF5 files are in ~/.arduino15/packages/MySensors/hardware/nRF5/0.2.1

mfalkvidd

on my Windows machine, all boards are located in c:\Users<username>\AppData\Local\Arduino15\packages
Strange that they've been installed to C:\Users\David\Documents\Arduino\libraries\ on your machine.

Maybe just delete the folders and re-install using the board manager?

NeverDie

@mfalkvidd All I know is that if I put a library in the folder: C:\Users\David\Documents\Arduino\libraries
it always gets detected and used. It also means you can completely uninstall/re-install the arduino IDE, and nothing gets lost. I'm not sure about other locations.

Am I the only one doing it this way?

mfalkvidd

@NeverDie for libraries, doing what you do should result in the same behavior as when using the Library Manager.
But I'm not sure it works with boards (especially given that the Arduino IDE complains when you put boards in the libraries folder).

NeverDie

I use the board manager to manage the boards. I believe it connects directly into github.

Anyhow, I stripped out all the non-library stuff, and now Windows IDE works without complaining.

NeverDie

Has anyone here yet figured out how to have more than one type of interrupt at a time wake-up the MCU from sleep? Based on the examples so far, it seems as though only one type at a time can be active. I'm sure there must be some way to do it. For instance, it would be desirable if the MCU could wake up not just from a timer event every, say, 5 minutes, to take a temperature reading, but also immediately if there is a leak detected. Right now it's just one or the other.

rmtucker

@NeverDie

So we can not use the normal statement

int8_t sleep(int interrupt, int mode, unsigned long ms=0);

I did not realize that was the case as i have never tried it.

NeverDie

I believe hwSleep(..) is the preferred incantation, but beyond that I'm not sure of anything. Maybe @d00616 can comment?

I should be receiving a number of different PCB projects tomorrow for final assembly and test, and if I can't resolve this soon, I'm simply going to release them without full demo code.

scalz

@NeverDie said in MySensors NRF5 Platform:

I believe hwSleep(..) is the preferred incantation, but beyond that I'm not sure of anything.

In MySensors, the regular way to use sleep mode is by using the sleep() functions:
https://www.mysensors.org/download/sensor_api_20#sleeping

But like you said, you can also use the raw hwSleep() from the hw abstraction layer. Or you could also rewrite it!
Which means at each abstraction layer, there are additional logics. sleep() functions have more logics (regarding MySensors states and features) than the raw hwSleep for example (like testing if there is an ongoing ota, smartsleep, heartbeat etc..). I have nodes where i use raw or not, but it's good to know what's behind.

rmtucker

@scalz
So the statement
int8_t sleep(int interrupt, int mode, unsigned long ms=0);
Can be used for timer and pin interrupt at he same time on the NRF5 platform just like the arduino?

NeverDie

I'll try it, but I have strong doubts that it's going to work.

FWIW, here's the pin mapping currently being supplied by digitalPinToInterrupt(..) for the nRF5:
Digital pin 0 = interrupt pin 0
Digital pin 1 = interrupt pin 1
Digital pin 2 = interrupt pin 2
Digital pin 3 = interrupt pin 3
Digital pin 4 = interrupt pin 4
Digital pin 5 = interrupt pin 5
Digital pin 6 = interrupt pin 6
Digital pin 7 = interrupt pin 7
Digital pin 8 = interrupt pin 8
Digital pin 9 = interrupt pin 9
Digital pin 10 = interrupt pin 10
Digital pin 11 = interrupt pin 11
Digital pin 12 = interrupt pin 12
Digital pin 13 = interrupt pin 13
Digital pin 14 = interrupt pin 14
Digital pin 15 = interrupt pin 15
Digital pin 16 = interrupt pin 16
Digital pin 17 = interrupt pin 17
Digital pin 18 = interrupt pin 18
Digital pin 19 = interrupt pin 19
Digital pin 20 = interrupt pin 20
Digital pin 21 = interrupt pin 21
Digital pin 22 = interrupt pin 22
Digital pin 23 = interrupt pin 23
Digital pin 24 = interrupt pin 24
Digital pin 25 = interrupt pin 25
Digital pin 26 = interrupt pin 26
Digital pin 27 = interrupt pin 27
Digital pin 28 = interrupt pin 28
Digital pin 29 = interrupt pin 29
Digital pin 30 = interrupt pin 30

NeverDie

The last I heard, from @d00616 , we had to supply code such as:

    // Enable interrupt
  NVIC_SetPriority(RTC0_IRQn, 15);
  NVIC_ClearPendingIRQ(RTC0_IRQn);
  NVIC_EnableIRQ(RTC0_IRQn);

and

// This must be in one line
extern "C" { void RTC0_IRQHandler(void) {rtcInterruptCounter++; NRF5_RESET_EVENT(NRF_RTC0->EVENTS_OVRFLW); NRF_RTC0->EVENTS_OVRFLW=0; }}

to get interrupts to work. Even with that approach, though, I haven't gotten it to support any interrupts in addition to a timed sleeping, though I have gotten it to support one interrupt that's separate from a timed sleeping (basically, sleeps indefinitely until the interrupt happens).

NeverDie

OK, I just now tried:

  sleep(3,CHANGE,3000);

and, as I suspected, it does nothing but sleep for 3 seconds. It's not responsive to any changes on pin P03 on an nRF52.

@scalz Are you getting a different result? It seems like there's a strong disconnect somewhere between what you're recommending and what I am experiencing.

scalz

@NeverDie said in MySensors NRF5 Platform:

@scalz Are you getting a different result? It seems like there's a strong disconnect somewhere between what you're recommending and what I am experiencing.

I didn't give any recommandation, you misread. I just said the regular way to use sleep feature in mysensors, for users, is with sleep(..). But I agree, I misread you too! when you were asking for the specific nrf52 case I guess

That said, I got this working when I made my recessed node for door (accelerometer and hall effect sensor with sleeping). Maybe things have changed in the lib?? or not. I'm struggling between different version of the lib, and some are different from the dev branch.. I'll recheck this asap (not sure for this evening).

NeverDie

I tried toying around with it a bit, and I got a useful result:

#define MY_CORE_ONLY

#include <nrf.h>
#include <MySensors.h>


const byte ledPin = LED_BUILTIN;
const byte interruptPin = 3;
volatile byte state = LOW;

void blinkityBlink(uint8_t pulses, uint8_t repetitions) {
  for (int x=0;x<repetitions;x++) {
    for (int i=0;i<pulses;i++) {
      digitalWrite(LED_BUILTIN,HIGH);
      wait(20);
      digitalWrite(LED_BUILTIN,LOW);
      wait(100);
    }    
      wait(500);
  }
}

void setup() {
  hwPinMode(LED_BUILTIN,OUTPUT_D0H1);
  hwPinMode(interruptPin, INPUT);
  attachInterrupt(digitalPinToInterrupt(interruptPin), blink, RISING);
  blinkityBlink(2,1);  //signify power-on and start of main loop
}

volatile bool buttonPressed=false;
void loop() {
  state = !state;
  digitalWrite(ledPin, state);
  sleep(digitalPinToInterrupt(interruptPin), RISING, 3000);
  wait(20);  //wait 20 milliseconds for button to debounce if it was pressed
  if (digitalRead(interruptPin)) { //if button is pressed
    blinkityBlink(2,1);
  }
  if (buttonPressed) {
    buttonPressed=false;
    blinkityBlink(20,1);
  }
}

void blink() {
  buttonPressed=true;
}

So, with this approach, pushing the button on pin 3 does wake up the nRF52 from sleep, whereupon the button press can still be detected and serviced, but it also demonstrates that the ISR per se isn't working.

Anyhow, with this I'm able to get the PIR or leak sensor or magnet sensor or light sensor doing useful things in a timely manner, even if it isn't ideal. That puts me further ahead than I was before.

rmtucker

@NeverDie
This is a test program that wakes up from either that i was using some time ago.

/**
   The MySensors Arduino library handles the wireless radio link and protocol
   between your home built sensors/actuators and HA controller of choice.
   The sensors forms a self healing radio network with optional repeaters. Each
   repeater and gateway builds a routing tables in EEPROM which keeps track of the
   network topology allowing messages to be routed to nodes.

   Created by Henrik Ekblad <henrik.ekblad@mysensors.org>
   Copyright (C) 2013-2015 Sensnology AB
   Full contributor list: https://github.com/mysensors/Arduino/graphs/contributors

   Documentation: http://www.mysensors.org
   Support Forum: http://forum.mysensors.org

   This program is free software; you can redistribute it and/or
   modify it under the terms of the GNU General Public License
   version 2 as published by the Free Software Foundation.

 *******************************

   REVISION HISTORY
   Version 1.0 - Henrik EKblad

   DESCRIPTION
   This sketch provides an example how to implement a distance sensor using HC-SR04
   Use this sensor to measure KWH and Watt of your house meeter
   You need to set the correct pulsefactor of your meeter (blinks per KWH).
   The sensor starts by fetching current KWH value from gateway.
   Reports both KWH and Watt back to gateway.

   Unfortunately millis() won't increment when the Arduino is in
   sleepmode. So we cannot make this sensor sleep if we also want
   to calculate/report watt-number.
   http://www.mysensors.org/build/pulse_power
*/

// Enable debug prints
#define MY_DEBUG

// Enable and select radio type attached
//#define MY_RADIO_NRF24
#define MY_RADIO_NRF5_ESB
//#define MY_RADIO_RFM69
//#define MY_RADIO_RFM95

#include <MySensors.h>
#include <Wire.h> // must be included here so that Arduino library object file references work
#include <RtcDS3231.h>
RtcDS3231<TwoWire> Rtc(Wire);

#define DIGITAL_INPUT_SENSOR 2  // The digital input you attached your light sensor.  (Only 2 and 3 generates interrupt!)
#define PULSE_FACTOR 1000       // Nummber of blinks per KWH of your meeter
//#define SLEEP_MODE false        // Watt-value can only be reported when sleep mode is false.
#define MAX_WATT 10000          // Max watt value to report. This filetrs outliers.
#define CHILD_ID 1              // Id of the sensor child

unsigned long SEND_FREQUENCY = 20000; // Minimum time between send (in milliseconds). We don't wnat to spam the gateway.
double ppwh = ((double)PULSE_FACTOR) / 1000; // Pulses per watt hour
//bool pcReceived = false;
volatile unsigned long pulseCount = 0;
volatile unsigned long lastBlink = 0;
volatile unsigned long watt = 0;
volatile unsigned long kwh = 0;
unsigned long oldWatt = 0;
double oldKwh;
unsigned long lastSend;
MyMessage wattMsg(CHILD_ID, V_WATT);
MyMessage kwhMsg(CHILD_ID, V_KWH);


void setup()
{
  Serial.begin(115200);
  Rtc.Begin();
  Rtc.Enable32kHzPin(false);
  Rtc.SetSquareWavePinClockFrequency(DS3231SquareWaveClock_1Hz);
  Rtc.SetSquareWavePin(DS3231SquareWavePin_ModeClock);

  // Use the internal pullup to be able to hook up this sketch directly to an energy meter with S0 output
  // If no pullup is used, the reported usage will be too high because of the floating pin
  hwPinMode(DIGITAL_INPUT_SENSOR, INPUT_PULLUP);

  attachInterrupt(digitalPinToInterrupt(DIGITAL_INPUT_SENSOR), onPulse, FALLING);
  //pcReceived = true;
  lastSend = millis();
}

void presentation()
{
  // Send the sketch version information to the gateway and Controller
  sendSketchInfo("Energy Meter", "1.0");

  // Register this device as power sensor
  present(CHILD_ID, S_POWER);
}

void loop()
{
  unsigned long now = millis();
  // Only send values at a maximum frequency or woken up from sleep
  bool sendTime = now - lastSend > SEND_FREQUENCY;
  if (sendTime) {
    // New watt value has been calculated
    if (watt != oldWatt) {
      // Check that we dont get unresonable large watt value.
      // could hapen when long wraps or false interrupt triggered
      if (watt < ((unsigned long)MAX_WATT)) {
        send(wattMsg.set(watt));  // Send watt value to gw
      }
      Serial.print("Watt:");
      Serial.println(watt);
      oldWatt = watt;
    }

    // Pulse count has changed
    //kwh = pulseCount;
    //double kwh = ((double)pulseCount / ((double)PULSE_FACTOR));
    send(kwhMsg.set(pulseCount));  // Send kwh value to gw
    Serial.print("Wh = ");
    Serial.println(pulseCount);
    pulseCount = 0;
    lastSend = now;
  }
  sleep(SEND_FREQUENCY);
}

void receive(const MyMessage &message)
{
}

void onPulse()
{
  unsigned long newBlink = micros();
  unsigned long interval = newBlink - lastBlink;
  if (interval < 50000L) { // Sometimes we get interrupt on RISING
    return;
  }
  watt = (3600000000.0 / interval) / ppwh;
  lastBlink = newBlink;
  pulseCount++;
}

NeverDie

Well, that's interesting. In the sketch I posted, I invoked sleep with:

  sleep(digitalPinToInterrupt(interruptPin), RISING, 3000);

which caused the mcu to wake up immediately after I press the button, but it didn't process the ISR.

Using an invocation like yours in the sketch you just posted above:

  sleep(3000);

the ISR executes and then terminates when I press the button, but the MCU doesn't wake and continue with the loop as it did with the prior incantation. Instead, it has to wait for the timer cycle to finish.

What I want is for it to wake up, do the ISR, and continue with the loop where it left off until it is explicitly put back to sleep again. How do I do that?

d00616

@NeverDie I have checked the sleep routine in all three variants. It's working with my setup.

There is no API stopping the sleep() by another ISR. Sleep only ends at one of the given conditions.

When you use the MY_CORE_ONLY define, please add "hwInit();" into the setup() routine.

NeverDie

@d00616 said in MySensors NRF5 Platform:

@NeverDie I have checked the sleep routine in all three variants. It's working with my setup.

There is no API stopping the sleep() by another ISR. Sleep only ends at one of the given conditions.

When you use the MY_CORE_ONLY define, please add "hwInit();" into the setup() routine.

Would you please post the three demo code examples that you tested?

NeverDie

@d00616 said in MySensors NRF5 Platform:

add "hwInit();" into the setup() routine.

OK, I just now did that, but I'm not getting any difference in the results.

NeverDie

In any case, I'm sure the question will ultimately turn from "How do I wake up based on a pin change?" to "How do I wake up based on the LPCOMP output, which has that pin as its input?" The reason: as covered earlier in this thread, much lower current consumption while sleeping if doing it via LPCOMP rather than the more straightforward pin monitoring.

d00616

@NeverDie said in MySensors NRF5 Platform:

In any case, I'm sure the question will ultimately turn from "How do I wake up based on a pin change?" to "How do I wake up based on the LPCOMP output, which has that pin as its input?" The reason: as covered earlier in this thread, much lower current consumption while sleeping if doing it via LPCOMP rather than the more straightforward pin monitoring.

It's a good question about, how to design the API to do this. I have no good idea.
Until an API, you can set MY_HW_RTC->CC[0] to (MY_HW_RTC->COUNTER+2). This ends sleep with some latency.

NeverDie

This post is deleted!

NeverDie

This post is deleted!

NeverDie

This post is deleted!

NeverDie

Is it just me, or does the myBoardNrf5 cause I2c to fail during initialization? I have code which worked fine on mNrf5Board but which now hangs during initialization when using myBoardNrf5. Is it working for anyone else?

NeverDie

@NeverDie said in MySensors NRF5 Platform:

Is it just me, or does the myBoardNrf5 cause I2c to fail during initialization? I have code which worked fine on mNrf5Board but which now hangs during initialization when using myBoardNrf5. Is it working for anyone else?

I've re-installed everything and am still getting no joy using I2C currently. The non-I2C stuff seems to be working fine though.
So, before I spin more cycles trying to figure it out, is i2c working for anyone else right now using the latest builds and myBoardNrf5? Or, is I2C currently broken?

NeverDie

It appears that the place where it hangs is this line here in Wire_nRF52.cpp:

  while(!_p_twim->EVENTS_LASTTX && !_p_twim->EVENTS_ERROR);

This is too bad, as I2C seemed to work fine prior to around a couple weeks ago. I think maybe (?) the latest update somehow broke it.

Please advise.

NeverDie

Well, since I'm dead in the water as things stand, I moved the code over to run on an nRF52 DK. Then, usingy Sandeep's code and none of the mysensors code, I was able to get the nRF52 DK to read an attached Si7021 temperature-humidity sensor. i.e. that worked without hanging.

So, it would appear that there's something about the mysensors code that is causing the problems. @d00616 Can we please get it fixed?

d00616

@NeverDie said in MySensors NRF5 Platform:

So, it would appear that there's something about the mysensors code that is causing the problems. @d00616 Can we please get it fixed?

I have no I2C Hardware for testing, but I take a look into the I2C and MySensors code soon.

NeverDie

The sooner the better. I'm pretty much stuck until it gets fixed.

rmtucker

@neverdie
I have been using The NRF5 with mysensors for a number of weeks with the si7021.
So it must be just a recent problem with the mysensors code?

NeverDie

@rmtucker said in MySensors NRF5 Platform:

@neverdie
I have been using The NRF5 with mysensors for a number of weeks with the si7021.
So it must be just a recent problem with the mysensors code?

I think so. Are you running the most current version? si7021 was working for me too on earlier versions. The latest update seems to have broken it. It's the one that uses myBoardNrf5 instead of myNrf5Board.

NeverDie

@rmtucker said in MySensors NRF5 Platform:

@neverdie
I have been using The NRF5 with mysensors for a number of weeks with the si7021.
So it must be just a recent problem with the mysensors code?

Would you mind trying the most recent version also to see if it breaks on you too? That would at least confirm it.

Right now my only alternative is to roll-back to the earlier version and just never upgrade again.

rmtucker

@neverdie
Where is myBoardNrf5 so i can check?

NeverDie

@rmtucker said in MySensors NRF5 Platform:

@neverdie
Where is myBoardNrf5 so i can check?

https://github.com/d00616/ArduinoHwNRF5

rmtucker

@neverdie
That explains it i am not using code from d00616.
I am only using the mysensors library and sandeeps stuff.
Sorry mate.

NeverDie

Well, in general, it's very handy for handling the pin mappings. Up until the latest release, it has worked fine. It still works fine now, except for the I2C pin mappings.

rmtucker

@neverdie
Hmmm at what cost?
Maybe i will stick with what i have got as it just works.

NeverDie

@d00616 Maybe you can just make it like it was before, where there was only one pair of I2c pins, not two as in your upgraded version? Maybe then it would work again.

scalz

@rmtucker said in MySensors NRF5 Platform:

That explains it i am not using code from d00616.
I am only using the mysensors library and sandeeps stuff.
Sorry mate.

If you're using mysensors, then you're using d00616 code (i'm kidding)

@NeverDie I can only confirm, it is related to the ArduinoHwNRF5 code as I have not the problem (I'm not using this part yet. I'm still using my own custom board definitions for the moment).
This should be fixed soon (december can be a busy month..)

d00616

@rmtucker said in MySensors NRF5 Platform:

Where is myBoardNrf5 so i can check?

This is the better repository: https://github.com/mysensors/ArduinoHwNRF5

@neverdie said in MySensors NRF5 Platform:

@d00616 Maybe you can just make it like it was before, where there was only one pair of I2c pins, not two as in your upgraded version? Maybe then it would work again.

Please give me some time. After an debug session with @scalz and @Yveaux, I think this is a tool chain (GCC 5.3) problem. The pin mapping array is part of the binary, but the Wire constructor gets an array with 0 while the Wire instance has pin mapping information.

Yesterday, I had a try to port ArduinoHwNRF5 back to the same tool chain (GCC 4.8) used by Arduino Primo and SAMD boards. I think for compatibility, this is the best option.

If I have no luck, I either remove the option of the pin remapping table or I go back to the old version. I think going back is no good idea, because the NRF52840 has more than one 32 GPIO, which requires advanced parameters.

d00616

@d00616 said in MySensors NRF5 Platform:
Back from the compiler hell. There is no chance getting the GCC 4.8 to work without forking the arduino-nrf5 code. Compiling with GCC 4.8 has the same result.

I have changed the templates a little bit and moved the code for pin mapping generation. Now it works for me. Please give me some time to clear the code and create a new package (0.3.0). With the new package, you have to add '#include <compat_pin_mapping.h>' before '#end" in MyBoardNRF5.cpp. You have to rename the '#ifdefs' and '#define' in both board files to match MYBOARDNRF5 instead of MYNRF5BOARD.

d00616

@d00616 said in MySensors NRF5 Platform:

I have changed the templates a little bit and moved the code for pin mapping generation. Now it works for me. Please give me some time to clear the code and create a new package (0.3.0). With the new package, you have to add '#include <compat_pin_mapping.h>' before '#end" in MyBoardNRF5.cpp. You have to rename the '#ifdefs' and '#define' in both board files to match MYBOARDNRF5 instead of MYNRF5BOARD.

Now, there is a new Version 0.3.0 for MyBoardNRF5 available. It can be installed via the Board Manager. Please change the templates like described above. I think, now the board definition is stable and ready to support other Arduino variants without changing the board definition.

NeverDie

@d00616

I think I made the changes you recommended, but now I'm getting:

Board MyBoard_nRF51822 (platform nRF5, package MySensors) is unknown

Error compiling for board MyBoardNRF5 nRF51822.

d00616

@neverdie said in MySensors NRF5 Platform:

I think I made the changes you recommended, but now I'm getting:
Board MyBoard_nRF51822 (platform nRF5, package MySensors) is unknown

Error compiling for board MyBoardNRF5 nRF51822.

I had the same message. In my case, I have uninstalled arduino-nrf5, cleared files, which are not removed from packages directory (the 0.2.1 MyBoard files) and reinstalled arduino-nrf5. After restarting Arduino it was fine.

NeverDie

@d00616
I'm still getting the error. I tried deleting arduino-nrf5 library and re-installing it fresh, but I don't know how, on Windows, to clear files and remove them from the packages directory. I'm not even aware of their being a "packages directory." Sorry.

NeverDie

@d00616 You mentioned that it works only with a particular version of GCC. Does that mean I need to have a particular version of the Arduino IDE installed? I'm currently running the latest version of Windows Arduino IDE, which is 1.8.5.

NeverDie

The irony of it all is that the earlier setup from 3 or so weeks ago seemed to work just fine. Unfortunately, I no longer have any idea how to get back to that.

scalz

@NeverDie
C:\Users...\AppData\Local\Arduino15\packages\MySensors\

NeverDie

@scalz said in MySensors NRF5 Platform:

@NeverDie
C:\Users...\AppData\Local\Arduino15\packages\MySensors\

Thanks!

The good news is that it compiles now. Also, the pin mapping for an LED is working. However, the bad news is that pin mapping for Serial Rx and Tx doesn't seem to be working, so I can't read any serial debug messages. Therefore, not sure if I2C is working or not.

NeverDie

I did another purge and reinsatall of the boards, and now serial is working.

NeverDie

Unfortunately, it does not appear that I2C is working.

NeverDie

Curiously, I2c does seem to work on a nRF52 DK using the myboard board definition.

NeverDie

However, it doesn't seem to be working on the nrf51.

d00616

@neverdie said in MySensors NRF5 Platform:

@d00616 You mentioned that it works only with a particular version of GCC. Does that mean I need to have a particular version of the Arduino IDE installed? I'm currently running the latest version of Windows Arduino IDE, which is 1.8.5.

The GCC version is part of the arduino-nrf5 distribution.

@neverdie said in MySensors NRF5 Platform:

Unfortunately, it does not appear that I2C is working.

I have I2C and serial running with a NRF52832.
You can try to remove the compat_pin_mapping line and add the 0.1.0 pin map at this place. Then the board definition is nearly the 0.1.0 version.

@neverdie said in MySensors NRF5 Platform:

However, it doesn't seem to be working on the nrf51.

I check the NRF51 board. What do you mean with no working?

NeverDie

I hope that narrows down the likely cause of the problem. The fact that it's working on the nRF52 gives me some hope it can be made to work on the nRF51.

NeverDie

@d00616 said in MySensors NRF5 Platform:

I check the NRF51 board. What do you mean with no working?

The nRF51 can't find the si7021 device, even though it's connected.

NeverDie

@d00616 said in MySensors NRF5 Platform:

I have I2C and serial running with a NRF52832.

Yes, it does seem to work on the nRF52832. Just not presently on the nRF51822.

NeverDie

@d00616 said in MySensors NRF5 Platform:

I check the NRF51 board. What do you mean with no working?

I'm running the sparkfun si7021 code. When I run it on the nrf52832, it yields:

Si7021 Found
Temp:70.51F, Humidity:40.71%

which is correct, but when I run it on the nRF51822, the result is:

No Devices Detected
Temp:-51.85F, Humidity:-5.81%

which is incorrect.

d00616

@neverdie said in MySensors NRF5 Platform:

@d00616 said in MySensors NRF5 Platform:

I have I2C and serial running with a NRF52832.

Yes, it does seem to work on the nRF52832. Just not presently on the nRF51822.

Can you check, if the device works with the "Generic NRF51822" board? You can edit the variant files in the arduino-nrf5\variants directory.
C:\Users...\AppData\Local\Arduino15\packages\sandeepmistry\hardware\nRF5\0.4.0\variants\Generic

NeverDie

@d00616 said in MySensors NRF5 Platform:

@neverdie said in MySensors NRF5 Platform:

@d00616 said in MySensors NRF5 Platform:

I have I2C and serial running with a NRF52832.

Yes, it does seem to work on the nRF52832. Just not presently on the nRF51822.

Can you check, if the device works with the "Generic NRF51822" board? You can edit the variant files in the arduino-nrf5\variants directory.
C:\Users...\AppData\Local\Arduino15\packages\sandeepmistry\hardware\nRF5\0.4.0\variants\Generic

I tried, but I guess I'm not doing it right. I couldn't even get it to blink an LED that way.

NeverDie

@d00616 said in MySensors NRF5 Platform:

You can try to remove the compat_pin_mapping line and add the 0.1.0 pin map at this place. Then the board definition is nearly the 0.1.0 version.

So, I take it you are suggesting that in myboardnrf5.cpp, where it says:

#ifdef MYBOARDNRF5
#include <variant.h>
#include <compat_pin_mapping.h>

I replace

#include <compat_pin_mapping.h>

with what exactly? At this point I'm not even sure where 0.1.0 pin-map is, or even what it's called.

NeverDie

@neverdie said in MySensors NRF5 Platform:

I tried, but I guess I'm not doing it right. I couldn't even get it to blink an LED that way.

After I modify the generic variant board definition, is there some trick to making it go "live"?

scalz

@NeverDie
there is no trick.. it should work.
d00616 said, for a test, to not use his defines, the method that you're used to.
So change the defines you need in the generic variant file, then in arduino:

NeverDie

I think this must be where the disconnect is happening. Where do I find the files or directory for Generic nRF51? The closest I see is a directory called just "Generic":

That's the one I tried modifying:

and it had no effect. So, I suspect there must be a different "generic nRF51" somewhere which is closely tied with the menu item you show the picture of. However, I don't see where.

Nca78

@neverdie did you check there ?

NeverDie

I didn't try that one, but it's a similar dilemma:

i.e. Does "Generic" do double duty for both "Generic nRF51" and "Generic nRF52"?

NeverDie

Aha! That did it. Thanks @NCA78. It's working now, including the I2C.

I had to switch nodes though. It's possible that the mysensors code does work, as the node I was testing it on may have had a bad si7021 module on it. I'll post again after I retry with this different node.

NeverDie

I'm back with the results. The same node where I2C works using the customized "Generic nRF51" does not have working I2c if using the mysensors "MyBoardnRF5 nRF51822".

So, I think I've taken this as far as I can. Hopefully this info helps you solve whatever the problem is.

I'm grateful I can keep moving along now using a modified "Generic nRF51822". Thanks everybody!

NeverDie

Ashes on my head. The mysensors code is working with I2c after all. I confirmed it on the second (alternate) node. I had switched two of the wires in the myboardnrf5 pinout, but hadn't on the generic nRF51 pinout. Correcting for that, it now works.

What threw me completely off was the defective si7021 module on the first node. Well, now everything is sorted.

d00616

@neverdie said in MySensors NRF5 Platform:

Ashes on my head. The mysensors code is working with I2c after all. I confirmed it on the second (alternate) node. I had switched two of the wires in the myboardnrf5 pinout, but hadn't on the generic nRF51 pinout. Correcting for that, it now works.

Thank you for reporting this. Great news to hear this.

NeverDie

@d00616 said in MySensors NRF5 Platform:

Thank you for reporting this. Great news to hear this.

You're welcome, of course. Though it was a little bit of a wild goose chase, at least your code is now pretty well tested and proven to be "known good," so others can start using it without fear.

With this now out of the way, please do continue your excellent work!

Omemanti

This post is deleted!

💬 MySensors NRF5 Platform

Suggested Topics

💬 Teleinfo with parasitic power
OpenHardware.io • mysensors rfm69hw teleinfo • • openhardware.io

💬 EASY PIR multisensors box
OpenHardware.io • mysensors rfm69 temperature light pir si7021 bh1750 dc-dc longlife • • openhardware.io

SCT-013-000 Calibration Value
General Discussion • mysensors • • Junior Ximenes

💬 Capacitive Soil Moisture Sensor
OpenHardware.io • mysensors nrf24l01+ atmega328p soil moisture capacitive soil moisture • • openhardware.io

💬 MySensors equiped Clock Radio
OpenHardware.io • mysensors contest2017 • • openhardware.io

💬 Smart Alarm Clock
OpenHardware.io • mysensors domoticz alarm clock led ring light speaker • • openhardware.io

💬 NRF52 Recessed Vibration+Contact sensor for door/window
OpenHardware.io • nrf52 door accelerometer contact recessed hall effect • • openhardware.io

MySensors Mini Scene Controller
OpenHardware.io • mysensors arduino battery contest2016 • • openhardware.io

34
Online

11.4k
Users

11.1k
Topics

112.6k
Posts

💬 MySensors NRF5 Platform

Suggested Topics

💬 Teleinfo with parasitic power OpenHardware.io • mysensors rfm69hw teleinfo • • openhardware.io

💬 EASY PIR multisensors box OpenHardware.io • mysensors rfm69 temperature light pir si7021 bh1750 dc-dc longlife • • openhardware.io

SCT-013-000 Calibration Value General Discussion • mysensors • • Junior Ximenes

💬 Capacitive Soil Moisture Sensor OpenHardware.io • mysensors nrf24l01+ atmega328p soil moisture capacitive soil moisture • • openhardware.io

💬 MySensors equiped Clock Radio OpenHardware.io • mysensors contest2017 • • openhardware.io

💬 Smart Alarm Clock OpenHardware.io • mysensors domoticz alarm clock led ring light speaker • • openhardware.io

💬 NRF52 Recessed Vibration+Contact sensor for door/window OpenHardware.io • nrf52 door accelerometer contact recessed hall effect • • openhardware.io

MySensors Mini Scene Controller OpenHardware.io • mysensors arduino battery contest2016 • • openhardware.io

34Online

11.4kUsers

11.1kTopics

112.6kPosts

💬 Teleinfo with parasitic power
OpenHardware.io • mysensors rfm69hw teleinfo • • openhardware.io

💬 EASY PIR multisensors box
OpenHardware.io • mysensors rfm69 temperature light pir si7021 bh1750 dc-dc longlife • • openhardware.io

SCT-013-000 Calibration Value
General Discussion • mysensors • • Junior Ximenes

💬 Capacitive Soil Moisture Sensor
OpenHardware.io • mysensors nrf24l01+ atmega328p soil moisture capacitive soil moisture • • openhardware.io

💬 MySensors equiped Clock Radio
OpenHardware.io • mysensors contest2017 • • openhardware.io

💬 Smart Alarm Clock
OpenHardware.io • mysensors domoticz alarm clock led ring light speaker • • openhardware.io

💬 NRF52 Recessed Vibration+Contact sensor for door/window
OpenHardware.io • nrf52 door accelerometer contact recessed hall effect • • openhardware.io

MySensors Mini Scene Controller
OpenHardware.io • mysensors arduino battery contest2016 • • openhardware.io

34
Online

11.4k
Users

11.1k
Topics

112.6k
Posts