💬 Power Meter Pulse Sensor

[gohan]

@sundberg84

Instead of the 10k resistor, do you think I could use a potentiometer to "adjust" sensitivity? I would like to avoid messing up the readings with lights entering when opening the cabinet door where the meter is located

@flopp
Do you mind sharing your wiring? 940nm, isn't it infrared?

[sundberg84]

@gohan - a potentiometer is just a variable resistor so why not? Good idea if you have one laying around.
As i said before though, the try with the transistor is just a try - im not that good with transistors yet and dont know how they work exactly. After this i read somewhere that the base resistor should be bigger to protect the transistor... dont know how to calculate that.

[flopp]

@gohan said in Power Meter Pulse Sensor:

@sundberg84

@flopp
Do you mind sharing your wiring? 940nm, isn't it infrared?

I connected the photo resistor directly to the LM393.
I don't know if that is IR.

[moumout31]

Hello,

First, thanks for this tutorial !
I used the code defined in tutorial but I have the following message for "void receive(const MyMessage &message)" :

'MyMessage' does not name a type

Do you know what can be the cause of this error please ?

Thanks in advance for your help

[mfalkvidd]

@moumout31 do you have void receive(const MyMessage &message) somewhere in your code or where does it come from?
Could you post the full error message?

Have you installed the MySensors library?

[moumout31]

@mfalkvidd void receive(const MyMessage &message) is in the code.
MySensors library is installed, I already use it in other nodes.
It's strange because it works in another computer...

The full error messages, in french beacause I'm french are :

Arduino : 1.8.0 (Windows 7), Carte : "Arduino Pro or Pro Mini, ATmega328 (3.3V, 8 MHz)"

_02_Main_loop:271: error: 'MyMessage' does not name a type

In file included from C:\Users\Anne-Laure\Documents\Arduino\libraries\arduino_759467/MySensors.h:257:0,

                 from C:\Users\Anne-Laure\Dropbox\Maison\A récupérer sur OneDrive\my_teleinfo_light\_02_Main_loop.ino:118:

C:\Users\Anne-Laure\Documents\Arduino\libraries\arduino_759467/core/MyTransport.cpp: In function 'void transportProcessMessage()':

C:\Users\Anne-Laure\Documents\Arduino\libraries\arduino_759467/core/MyTransport.cpp:745:14: error: cannot resolve overloaded function 'receive' based on conversion to type 'bool'

   if (receive) {

              ^

C:\Users\Anne-Laure\Documents\Arduino\libraries\arduino_759467/core/MyTransport.cpp:811:15: error: cannot resolve overloaded function 'receive' based on conversion to type 'bool'

    if (receive) {

               ^

exit status 1
'MyMessage' does not name a type

Bibliothèque non valide trouvée dans C:\Users\Anne-Laure\Documents\Arduino\libraries\MySensors : C:\Users\Anne-Laure\Documents\Arduino\libraries\MySensors

Ce rapport pourrait être plus détaillé avec
l'option "Afficher les résultats détaillés de la compilation"
activée dans Fichier -> Préférences.

Thanks for your help

[mfalkvidd]

@moumout31 My guess is that MyMessage is defined in your sketch, overriding the definition in the MySensors library. Could you post your sketch?

[moumout31]

@mfalkvidd It's strange... It works with an older version of Arduino software... Thus, the problem is solved ! Thanks

[moumout31]

Hello,

I don't understand why, but my power meter sensor gives a higher index than the real power meter index after 2 days.
I think that it counts more pulses than pulses provided by the power meter...
Does anybody encounter this problem ?
Can it be cause to bounce for example ?

Thank you

[gohan]

@moumout31
could it be getting some light from another source?

[flopp]

@moumout31
I had same issue, check here for my solution
https://forum.mysensors.org/topic/4716/two-energy-meter/4

[Fat Fly]

Hei.

How to connect arduino to electricity meter pulse out. Not to led.

[moumout31]

@gohan No, the light is always switched off in this room and I tried, when I switch on the light, there is no pulse from the sensor.

@flopp Thanks for your help. What I see in your topic is that you increase from 10000µs to 40000µs the interval to avoid corrupted interrupts, is that correct ?

Thank you

[flopp]

@moumout31
Yes correct.
When I checked with Serial Monitor and some serial.print in different places I could see that arduino registered double interrupts

[moumout31]

@flopp Thanks a lot, I will try this solution and check if it's better !

[moumout31]

@flopp It seems to work, index on Domoticz is still consistent with the real index... I will see in a few days if it's still the case.
Thanks a lot for your help !

[moumout31]

@flopp After a few days, I confirm you that the index is consistent with power meter index.
However, instant power in watt is not correcly calculated.
For example, when power meter indicates 550W, about 1500W is calculated by the sketch. Do you know how I can resolve this problem please ?

Thanks a lot !

[flopp]

@moumout31
Watt calculation is using time between two interrupts.
How often do you send watt?
Did you measure watt with a measurement tool How did you get 500 watts from power meter?
When you send watt you need to measure exact same second otherwise it will not be correct.

[moumout31]

@flopp Watt is sent every 20 seconds (as done in the original sketch).
500 watts is the power indicated on the screen of the power meter.

[NiklasO]

Just changed from 1-wire counter to MySensors for logging my power consumption. Using the same LED detector as I did when using 1-wire. Also changed watt limit to 20 000 (my heat pump likes to use 9-10kW sometimes). The wires from the old 1-wire net now provide power for the new MySensors-sensor. I have 10 000 led blinks per kWh. Had som crazy counter values with 1-wire that totally messed up the graphs and haven't seen anything like that with this sketch. Let's hope it stays that way. Dumping data to Domoticz. This sensor gives me higher resolution as a bonus. The 1-wire counter was read every minute, this every 20-sec.

My old 1-wire counter with one crazy value messing upp all the graphs:

I know that you can hold shift and click the value that is crazy to make the graphs show up good but you loose the data for the full day. That's not good.

[Samuel235]

I seem to be getting a connection to my gateway, while monitoring the two topics for the gateway i only seem to see any activity regarding this sensor on the gateway-out topic, and it keeps sending out a type 2 message (req). The serial monitor of the sensors is showing:

104198 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
114199 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
124200 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
134201 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
144202 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
154203 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
164204 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
174205 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
184206 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
194207 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:

So it seems to be sending a request package, but i don't see it on the gateway-in topic and i see other sensors sending data to the gateway-in topic so i'm not subscribed to the wrong one. Is this a structural request message or is this the feature that requests the last known message from the gateway? Either way, there doesn't seem to be any sensor data being sent from the node to the gateway.

[sundberg84]

@Samuel235 said in Power Meter Pulse Sensor:

164204 TSF:MSG:SEND,16-16-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:

I dont know if I dont understand you, but it seems like your node is requesting the last pulse counter. I have mentioned this in several threads I have a problem with this (both this thread and rain sensor thread). Im using an ethernet gw and the request reaches the gw but there my error led blinks and it fails to send it back (or recieve it on the node) so it re-request it.

I have never been able to pinpoint it due to repeaters in between.
https://forum.mysensors.org/topic/2116/hard-to-grab-time-and-value-sent-from-controller/ (old).

[Samuel235]

@sundberg84 - You do understand me correctly :). I made a thread on the troubleshooting section of the forum for this simply because i suspected it was normal behaviour, however. It seems that the node isn't sending any other data to the gateway at all and it can not get the previous value as there isn't one. So, my post here and on the troubleshooting section was me aiming to get to this point and then troubleshoot the actual node to see what can be done to get it to send data.

[Samuel235]

Has anyone came across their pulse meter not sending the message for WATT, the other two messages are being sent perfectly fine. My WATT message has been working all day and suddenly stopped.

[NiklasO]

Make sure that you don't reach your wattage limit in the sketch.

[Samuel235]

@NiklasO - Where is this value stored? I've cleared the EEPROM and the MQTT topic too, but it still refuses to send the watt value.

[NiklasO]

@Samuel235
In the sketch, change this
#define MAX_WATT 10000 // Max watt value to report. This filetrs outliers.
I have it set to 20000, using 15 000 easily when it's cold outside. Just make sure that you use the right divider for your pulse/kWh count.

[Samuel235]

Ahh i see what it is it; its set to not send if it hasn't changed.... how much of a silly error that was from me. The wattage is way below 1000 but because i'm using a normal LED on another arduino flashing at a constant rate to set my node up, its not changing the watt value.

Back to solving my initial issue now >.<

Thank you.

[NiklasO]

Ah, I see. Well, I would like the sensor to send data even if it hasn't changed. What did you change?

[gohan]

Watt usage is normally changing all the time, so you should have data reported quite often

[NiklasO]

@gohan But Samuel235 had that problem? Edit: well, ok, he had a different setup. I need to read better.

[Samuel235]

To be honest, i had it running for a while before it stopped sending the WATTS. But i'm guessing this is the issue as if i vary the blinks it sends it. I'de like to see if anyone has this working recently on OpenHAB through means of not touching the node sketch like people have posted before. I have an idea to send the data of the pulse count message onto the request message topic on MQTT and so then when it requests previous count it would have it.

[Samuel235]

This post is deleted!

[mickecarlsson]

EDIT:
This was not the case, I am still debugging the issue, in the mean time I have deleted the code.

After some debugging it seems that Domoticz messes with the V_WATT and V_KWH, and the solution is to have two sensors, one for kWh and one for W.
After I changed it, I have consistent W and kWh.
See also this issue for Home Assistant

[mickecarlsson]

OK, I think I have nailed it. I am now running a beta code pulling S0 from my energy meter and I am getting consistent results.
Before when I checked the domoticz.db I got this for Energy:

sqlite> select Name, sValue from DeviceStatus;
Storage kWh|0.000;4773.700

Now I get this:

sqlite> select Name, sValue from DeviceStatus;
Storage kWh|118.000;11678062.000

Notice that the first field is now with value, previously it was 0.000 (and I have updated the meter count to the actual value)
And in table Meter I now get consistent results:

sqlite> SELECT Value, Usage, Date FROM Meter WHERE DeviceRowID=43;
11678003|1170|2017-07-22 09:25:00
11678012|1170|2017-07-22 09:30:00
11678021|1180|2017-07-22 09:35:00
11678031|1180|2017-07-22 09:40:00
11678043|1170|2017-07-22 09:45:00
11678053|1180|2017-07-22 09:50:00
11678062|1180|2017-07-22 09:55:00
11678072|1180|2017-07-22 10:00:00

The field Usage is now populated as it should.
I will let the code run for a couple of days, then I will post the updated code here.

[mickecarlsson]

Here is a code example that works for me. I have an energy meter with S0 output and I use Domoticz as the controller.

/**
 * 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
 * Version 1.1 - Mikael Carlsson
 *
 * DESCRIPTION
 * This sketch provides an example how to implement a sensor reading from an energy meter
 * 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_RFM69

#include <MySensors.h>

#define DIGITAL_INPUT_SENSOR 3  // 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 meter
#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 filters bad data.
#define CHILD_ID 1              // Id of the sensor child

unsigned long SEND_FREQUENCY =
    20000; // Minimum time between send (in milliseconds). We don't want 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;
unsigned long oldPulseCount = 0;
unsigned long oldWatt = 0;
double oldKwh;
unsigned long lastSend;
MyMessage wattMsg(CHILD_ID,V_WATT);
MyMessage kwhMsg(CHILD_ID,V_KWH);
MyMessage pcMsg(CHILD_ID,V_VAR1);


void setup()
{
    // Fetch last known pulse count value from gw
    request(CHILD_ID, V_VAR1);

    // 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
    pinMode(DIGITAL_INPUT_SENSOR,INPUT_PULLUP);

    attachInterrupt(digitalPinToInterrupt(DIGITAL_INPUT_SENSOR), onPulse, RISING);
    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 (pcReceived && (SLEEP_MODE || sendTime)) {
        // New watt value has been calculated
        if (!SLEEP_MODE && watt != oldWatt) {
            // Check that we dont get unresonable large watt value.
            // could happen when long wraps or false interrupt triggered
            // We need to send in all values at one time to get consistent data
            if (watt<((unsigned long)MAX_WATT)) {
                send(wattMsg.set(watt));  // Send watt value to gw
                Serial.print("Watt: ");
                Serial.println(watt);
                oldWatt = watt;
                send(pcMsg.set(pulseCount));  // Send pulse count value to gw
                Serial.print("pulseCount: ");
                Serial.println(watt);
                double kwh = ((double)pulseCount/((double)PULSE_FACTOR));
                oldPulseCount = pulseCount;
                send(kwhMsg.set(kwh, 4));  // Send kwh value to gw
                Serial.print("kWh: ");
                Serial.println(kwh);
                oldKwh = kwh;
            }
        }
        lastSend = now;
    } else if (sendTime && !pcReceived) {
        // No count received. Try requesting it again
        request(CHILD_ID, V_VAR1);
        lastSend=now;
    }

    if (SLEEP_MODE) {
        sleep(SEND_FREQUENCY);
    }
}

void receive(const MyMessage &message)
{
    if (message.type==V_VAR1) {
        pulseCount = oldPulseCount = message.getLong();
        Serial.print("Received last pulse count from gw:");
        Serial.println(pulseCount);
        pcReceived = true;
    }
}

void onPulse()
{
    if (!SLEEP_MODE) {
        unsigned long newBlink = micros();
        unsigned long interval = newBlink-lastBlink;
        if (interval<10000L) { // Sometimes we get interrupt on RISING
            return;
        }
        watt = (3600000000.0 /interval) / ppwh;
        lastBlink = newBlink;
    }
    pulseCount++;
}```

[dakipro]

One (probably stupid) question about the batter powered option: Since sensor cannot keep track of real time power usage as it sleeps most of the time, would it be possible to do the math on the controller side and get "5min delay real time power usage"?
By dividing number of pulses since last transmission, is that basically what this sensor is doing when not being powered by batteries?

I could experiment with 2min delay f.eks, and see how batteries hold. If dht22 temp sensors works over a year with 5min transmission time, i hope 2min delay in pulse meter can hold 6mnt or something?
Another thing is that the electrical box is all metal, and there is a lot of electricity around the sensor so radio could struggle (I guess a repeater would definitively help)

[gohan]

Is it only me that is getting a compile error?

Error: call of overloaded 'set(volatile long unsigned int&)' is ambiguous
send(wattMsg.set(watt)); \ Send watt value to gw

error: call of overloaded 'set(volatile long unsigned int&)' is ambiguous
send(pcMsg.set(pulseCount)); \ Send pulse count value to gw

followed by all the possible candidates of .set arguments

[mfalkvidd]

@gohan I get the same, both for 2.1.1 and 2.2.0-beta. Strange. I thought there were Jenkins tests that endured the examples compiled.

[mfalkvidd]

Replacing all

unsigned long

with

uint32_t

makes the sketch compile fine, but I don't know if that is the correct solution.

[mfalkvidd]

I created an issue for this, https://github.com/mysensors/MySensors/issues/910

[hek]

The problem with "unsigned long" is that it can have different size based on platform it's compiled on.

Like @mfalkvidd says, use int32_t, uint32_t, int16_t and uint16_t which has a fixed width.

[gohan]

I was suspecting something like that, but I was not sure if changing variable type would affect something else

[gohan]

Ok, I got it working. On wemos it doesn't count pulses though. On domoticz does it have to configured as the value computed or from device? I'm asking because if you disconnect the sensor, on domoticz I keep seeing the same power reading

[gohan]

BTW, is there a way to have the pulse meter running on gateway? I got stuck because at startup it asks the controller the last count value and since the controller is not yet connected it just doesn't work.

[arifuleee]

@gohan for the first time use only the watt value....that time oldcounts will not bother.....

[gohan]

What do you mean? And how do I keep the count of the pulses if I don't get it from controller?

[gohan]

@hek is it me or the WATT readings are never sent? What could it be? Only V_VAR1 and V_KWH are sent.

[rejoe2]

@gohan How about

#define MY_TRANSPORT_WAIT_READY_MS 2000

This will start up your code so you won't miss any pulses until controller is ready.
You may ask for the last known Value in presentation(), then it will be fetched as soon as the controller is really there after startup. This could/should be combined with a bool pcReceived, so you may decide to either add the value received from controller to the counts the node meassured in between or just use the received value (starting from second time value is received from controller).

[gohan]

That actually is not that critical, as I am going to move to MQTT and that should be working anyway. Right now it is working on a UNO with NRF24, but my biggest issue is the WATT reading that it is missing

[gohan]

Any more suggestions on the Watt issue?

[hek]

Did you by any chance change SLEEP_MODE to true?

[gohan]

No, I get every pulse reported as soon as it is detected.

[gohan]

I think I found a problem: the WATT calculation is returning values over 172000 watts and since I put 10000 as the max value, it was not sending anything. I double checked my energy meter and indeed it is 1000 pulses per KWH and this is what I set in sketch

[gohan]

@hek I said I found the problem, I didn't say I know how to solve it

[gohan]

Here is the log I get from node

7230 TSF:MSG:READ,0-0-250,s=1,c=2,t=24,pt=0,l=7,sg=0:1682675
rece

Received last pulse count from gw:1682675

11817 TSF:MSG:SEND,250-250-0-0,s=1,c=1,t=17,pt=5,l=4,sg=0,ft=0,st=OK:528
Watt:528

Watt:197065

21820 TSF:MSG:SEND,250-250-0-0,s=1,c=1,t=24,pt=5,l=4,sg=0,ft=0,st=OK:1682677
21913 TSF:MSG:SEND,250-250-0-0,s=1,c=1,t=18,pt=7,l=5,sg=0,ft=0,st=OK:1682.6770
Watt:196807

36824 TSF:MSG:SEND,250-250-0-0,s=1,c=1,t=24,pt=5,l=4,sg=0,ft=0,st=OK:1682679
36917 TSF:MSG:SEND,250-250-0-0,s=1,c=1,t=18,pt=7,l=5,sg=0,ft=0,st=OK:1682.6790
Watt:197368

51827 TSF:MSG:SEND,250-250-0-0,s=1,c=1,t=24,pt=5,l=4,sg=0,ft=0,st=OK:1682681
51920 TSF:MSG:SEND,250-250-0-0,s=1,c=1,t=18,pt=7,l=5,sg=0,ft=0,st=OK:1682.6810
Watt:196850

Am I the only one using this sketch?

[mickecarlsson]

Please post you sketch.

[gohan]

it is the one in https://www.mysensors.org/build/pulse_power

[mickecarlsson]

How is the pulse counter set up? Is it connected to the S0 port?
Edit:
If you are using S0 it is vital that you use a pull down resistor, if you dont’t have that you get false pulses.

[gohan]

I'm using the lm393 light detector. Domoticz is able to do a calculation from the pulses to give a Watt value and it is correct but it does it every 5 minutes and it is too slow for me

[sundberg84]

@gohan - is Domoticz updating anything (except switches) faster than 5 min? I dont think so...

[gohan]

If you send the watt reading faster it updates like the switch, anyway I need the watt values for node-red also

[gohan]

Could it be the Light sensor is triggering more interrupts from each pulse?

[henninne]

@niklaso

Hi Niklas,

I know this is a very old post, but its worth a try. How did you get out the daily Kwh usage? As default my sensor only gives Watts and Kwh (accumulated from start), I have tried to run a Scene in Vera UI7 where there are a function in the sensor called resetKWh(), but it looks like it doesn't do anything. Any tips would be great, thanks.

Henning

[smilvert]

@gohan I think it is a little tricky to update domoticz every minute but try update the sql-file, see the domoticz forum
https://www.domoticz.com/forum/viewtopic.php?t=10183

[gohan]

The main issue is the fact that the sensor is not calculating the instant power correctly

[DaTim69]

I created this kWh sensor with the TCRT5000 IR Barrier Line Track Sensor.
My powermeter is a ferarrismeter with 375 rotations per kWh.
I also had the problem with spikes in the power usage. But after long...too long investigation I solved the problem.
At my config it was a double issue.

Issue 1.
The problem is the pulse width of the rotation.
The pulse are at night ( low power) much wider that during the day when more power is consumed. At night the spikes in power usage where huge.

Issue 2.
Switching on/off the halogen light causes spikes and related to also strange power measurement.

Solution:
I added a second arduino mini pro as a pulse regulator between the TCRT5000 and the power meter arduino.
this arduino triggers on Rising and will always give a 100ms puls. Als debounce is handled (spikes).
I don't upload the sketch here, I'm a beginning programmer on arduino. It's made on " trail and error" but it functions as it supposed to do.
If someone wants a copy feel free to contact mee

[xydix]

@gohan Did you solve this?
I get very high watt usage as well. Sometimes like 40 000 - 130 000 watt if I check the serial monitor.
Of course, those values will never be sent to the controler.

[gohan]

Nope, I am looking at other things now. Very little time available for too many things. I'm looking at buying CT clamp and do a more direct measurement.

[xydix]

Okay. Really annoying. Check this.

Received last pulse count from gw:2527
Watt:121065
22235 TSF:MSG:SEND,78-78-0-0,s=3,c=1,t=24,pt=5,l=4,sg=0,ft=0,st=OK:2539
22246 TSF:MSG:SEND,78-78-0-0,s=2,c=1,t=18,pt=7,l=5,sg=0,ft=0,st=OK:2.5390
Watt:121621
42238 TSF:MSG:SEND,78-78-0-0,s=3,c=1,t=24,pt=5,l=4,sg=0,ft=0,st=OK:2551
42250 TSF:MSG:SEND,78-78-0-0,s=2,c=1,t=18,pt=7,l=5,sg=0,ft=0,st=OK:2.5510
Watt:121342
62239 TSF:MSG:SEND,78-78-0-0,s=3,c=1,t=24,pt=5,l=4,sg=0,ft=0,st=OK:2563
62251 TSF:MSG:SEND,78-78-0-0,s=2,c=1,t=18,pt=7,l=5,sg=0,ft=0,st=OK:2.5630
Watt:121539
82240 TSF:MSG:SEND,78-78-0-0,s=3,c=1,t=24,pt=5,l=4,sg=0,ft=0,st=OK:2575
82251 TSF:MSG:SEND,78-78-0-0,s=2,c=1,t=18,pt=7,l=5,sg=0,ft=0,st=OK:2.5750
Watt:121951
102239 TSF:MSG:SEND,78-78-0-0,s=3,c=1,t=24,pt=5,l=4,sg=0,ft=0,st=OK:2589
102250 TSF:MSG:SEND,78-78-0-0,s=2,c=1,t=18,pt=7,l=5,sg=0,ft=0,st=OK:2.5890
Watt:122067
122243 TSF:MSG:SEND,78-78-0-0,s=3,c=1,t=24,pt=5,l=4,sg=0,ft=0,st=OK:2601
122254 TSF:MSG:SEND,78-78-0-0,s=2,c=1,t=18,pt=7,l=5,sg=0,ft=0,st=OK:2.6010

The plus count seems OK, right?

My energy meter is 1000 pluses /kwh

Here is my sketch. I use Home assistant so i created three sensors for this as I don't think V_VAR1 is supported with S_POWER.
Otherwise it is the standard example.

/**
 * 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_RFM69
#define MY_NODE_ID 78
#include <MySensors.h>

#define DIGITAL_INPUT_SENSOR 3  // 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 20000          // Max watt value to report. This filetrs outliers.
#define WATT_CHILD_ID 1              // Id of the sensor child
#define KWH_CHILD_ID 2
#define PC_CHILD_ID 3

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;
unsigned long oldPulseCount = 0;
unsigned long oldWatt = 0;
double oldKwh;
unsigned long lastSend;
MyMessage wattMsg(WATT_CHILD_ID,V_WATT);
MyMessage kwhMsg(KWH_CHILD_ID,V_KWH);
MyMessage pcMsg(PC_CHILD_ID,V_VAR1);


void setup()
{
  // Fetch last known pulse count value from gw
  request(PC_CHILD_ID, V_VAR1);

  // 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
  pinMode(DIGITAL_INPUT_SENSOR,INPUT_PULLUP);

  attachInterrupt(digitalPinToInterrupt(DIGITAL_INPUT_SENSOR), onPulse, RISING);
  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(WATT_CHILD_ID, S_POWER);
  present(KWH_CHILD_ID, S_POWER);
  present(PC_CHILD_ID, S_CUSTOM);
}

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 (pcReceived && (SLEEP_MODE || sendTime)) {
    // New watt value has been calculated
    if (!SLEEP_MODE && 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 cout has changed
    if (pulseCount != oldPulseCount) {
      send(pcMsg.set(pulseCount));  // Send pulse count value to gw
      double kwh = ((double)pulseCount/((double)PULSE_FACTOR));
      oldPulseCount = pulseCount;
      if (kwh != oldKwh) {
        send(kwhMsg.set(kwh, 4));  // Send kwh value to gw
        oldKwh = kwh;
      }
    }
    lastSend = now;
  } else if (sendTime && !pcReceived) {
    // No count received. Try requesting it again
    request(PC_CHILD_ID, V_VAR1);
    lastSend=now;
  }

  if (SLEEP_MODE) {
    sleep(SEND_FREQUENCY);
  }
}

void receive(const MyMessage &message)
{
  if (message.type==V_VAR1) {
    pulseCount = oldPulseCount = message.getLong();
    Serial.print("Received last pulse count from gw:");
    Serial.println(pulseCount);
    pcReceived = true;
  }
}

void onPulse()
{
  if (!SLEEP_MODE) {
    unsigned long newBlink = micros();
    unsigned long interval = newBlink-lastBlink;
    if (interval<10000L) { // Sometimes we get interrupt on RISING
      return;
    }
    watt = (3600000000.0 /interval) / ppwh;
    lastBlink = newBlink;
  }
  pulseCount++;
}

Anyone got any ideas?
BTW. First i tried the original example sketch, unmodified, with the same result.

[gohan]

That's pretty much the same problem as I have

[xydix]

I think i solved this.
I don't know if it is a good solution.
Maybe it is possible to solve in the code?
I added a 0,1uf ceramic capacitor between DO and GND on the LM393.
Then i tuned the LM393 until i had it blink as expected.

I think, in my case the LED pulse from my meeter was to short, it gives a very short blink.
Anyway.
I tested it all day and it seems fine now.

[gohan]

Are you getting instant power measurements correct?

[xydix]

Yes. I guess.
I don't have anything to compare with but i think it is accurate.
I have a 1000blink/kwh
Example:
Node send state every 20 sec.
pulsecount: 8 pulses every 20 sec.
Instant power in W is about 1400W

8 pulses x 3 = pulses /minute =24
24 pulses x 60 min = pulses /hour = 1440.

Edit.
Now I get a new value every 20 sec, before I didn't get any value (W) from the node at all because it was too high. I only received pulse count and kWh earlier.

[gohan]

Besides the cap, what exactly did you do?

[xydix]

Nothing more.
I use the same arduino, haven't uploaded anything new to it since i tried it last time.

What I had to do was lower the sensitivity on the LM393 because at first the signal-LED glowed a bit but when i adjusted it to lower sensitivity and closed the door to get it totaly dark the led blinked exactly as the meeter.

I have it on a bredboard and connected the cap between GND and to pin 3 so it is connected close to the arduino but i guess it shouldn't matter.

Have you tried it?

[gohan]

I can't believe it, IT WORKS!!!!! I added the cap directly under the pins of the LM939 board. Thanks man!

[xydix]

I am glad to hear it works for you too.

[gohan]

There is probably a sw way to correct this problem, I'll think about it

[xydix]

It waswhat I thought too.
As I said earlier, when I tested this my conclusion was, short blink, almost a flash will counted as a pulse (kWh) but it does something wrong with the calculation of instant power usage.
A slower blink will correct this.
My meeter gives a very short blink.

[gohan]

Yes, could be that a simple debounce could do the trick but I'll have to look deeper into it. @Yveaux what do you think?

[gohan]

@xydix Could you try to increase the 10000L to 100000L interval in the OnPulse function and see if it works without the capacitor?

[user2684]

I tried to build something similar and it is working nicely, thanks for sharing the idea! However, I've noticed the light sensor (alone) consumes 1mA constantly which is quite a lot when used for a battery powered project. Even if I report to the controller once per hour (summing up the power consumption along the way), still the light sensor needs to be always on, meaning I'd probably need to replace batteries after a month or so. Is there any workaround or alternative sensor which I can use or am I doing something completely wrong? Thanks

[cstewy]

@user2684 I had the same issue when I built this with a standalone atmega328p running on 2x AA's. I set it up with wake on interrupt from the It was constantly drawing 1.6mA to power the tsl257 to detect the flashes. My target was 1+ years on 2 x AA's. My solution to achieve this was to sleep for 125ms, wake up, send power to tsl257 to check light state (on/off), if it changed from previous reading then there was a pulse (well half pulse). By measuring every 125ms I can guarantee to capture fast pulses up to 14.4kW. (3600000/125)/2. It is 20kW max draw for residential in my country. Now:
Average mA Consumption Sleep 0.1
Average mA Consumption Wake 0.064516129
Average mA Consumption Transmit 0.002916667
Battery Life (2xAAs) = 1.7 years

[user2684]

@cstewy many thanks and very interesting. You basically saying it consumes less by waking up every 125ms rather than being asleep but with the sensor eating up current continuously. Definitely worth trying thanks! Meanwhile I've reduced the number of radio transmission and the batteries lasted for a couple of months but still is not ideal. Thanks! I'll try and report back

[artipi]

Hello, i'm stuck with this sensor, because no Interrupt is really working. I've cut down the code to this:

/**
 * 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 LM393 PCB
 * Use this sensor to measure kWh and Watt of your house meter
 * You need to set the correct pulsefactor of your meter (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 value.
 * 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>

#define DIGITAL_INPUT_SENSOR 2  // The digital input you attached your light sensor.  (Only 2 and 3 generates interrupt!)
#define PULSE_FACTOR 500       // Number of blinks per of your meter
#define SLEEP_MODE false        // Watt value can only be reported when sleep mode is false.
#define MAX_WATT 1000000          // Max watt value to report. This filters outliers.
#define CHILD_ID 10             // Id of the sensor child

uint32_t SEND_FREQUENCY =
    15000; // Minimum time between send (in milliseconds). We don't want to spam the gateway.
double ppwh = ((double)PULSE_FACTOR)/1000; // Pulses per watt hour
bool pcReceived = true;
volatile uint32_t pulseCount = 0;
volatile uint32_t lastBlink = 0;
volatile uint32_t watt = 0;
uint32_t oldPulseCount = 0;
uint32_t oldWatt = 0;
double oldkWh;
uint32_t lastSend;
//MyMessage wattMsg(CHILD_ID,V_WATT);
//MyMessage kWhMsg(CHILD_ID,V_KWH);
//MyMessage pcMsg(CHILD_ID,V_VAR1);


void setup()
{
  Serial.begin(9600);
  Serial.println("Start setup");
  // Fetch last known pulse count value from gw
  //request(CHILD_ID, V_VAR1);

  // 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
  pinMode(DIGITAL_INPUT_SENSOR,INPUT_PULLUP);

  attachInterrupt(digitalPinToInterrupt(DIGITAL_INPUT_SENSOR), onPulse, RISING);
  lastSend=millis();
  Serial.println("Start setup: last sent: ");
  pinMode(LED_BUILTIN, OUTPUT);
  digitalWrite(LED_BUILTIN, LOW);
}

void presentation()
{
  // Send the sketch version information to the gateway and Controller
  //sendSketchInfo("Energy Meter", "1.3");

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

void loop()
{
  
}



void onPulse()
{
  Serial.println("onPulse:");
  digitalWrite(LED_BUILTIN, HIGH);   // turn the LED on (HIGH is the voltage level)
  delay(1000);                       // wait for a second
  digitalWrite(LED_BUILTIN, LOW);    // turn the LED off by making the voltage LOW
  delay(1000);      
  pulseCount++;
  Serial.println("pulseCount:");
}

The Problem ist, that only on Start the interrupt ist working once, but no LED blinks.

Can someone help?

Thanks.

[mfalkvidd]

@artipi using Serial.print and delay from within an interrupt service routine will mess things up. Interrupt service routines must be executed quickly.

From https://www.arduino.cc/reference/en/language/functions/external-interrupts/attachinterrupt/:

Since delay() requires interrupts to work, it will not work if called inside an ISR.

You need to do the blinking inside loop().

[artipi]

Ah, this explains many things. Thank you, now its working:)!

[alowhum]

@artipi can you share your code?

[artipi]

@alowhum: Its the original code. Nothing changed. I've only made some tests with my version...

[patrikr76]

Hey guys

So i am trying to use this sketch with the sleep mode set to true 'cause i need to run it on battery.
It works fine running with usb power on it and sleep mode on false.
But with sleep mode true i am getting nothing in myscontroller after the initial startup.
All i changed is the sleep mode from false to true.
Running it on a clone nano with some chinese "flame detector" sensor.
Been trying to figure out if it has something to do with how it reports when in sleep mode but it should still report whatever the send frequency is set to right?
I was using a tv remote to fake impulses, with seems to work on usb power, but not on battery.
Any suggestions?

Best regards
Patrik

[mfalkvidd]

@patrikr76 just to make sure we are on the same page, could you clarify which sketch you are using? This thread has discussed lots of sketches and it would be a pity if we’re looking at different code

In any case, the debug log from the node will give the best information on what is happening.

It could also be useful to only change one thing. Right now you change sleep mode and switch to battery, right? How does the node behave with sleep mode on, but still running on usb power? How does the node behave with sleep mode off, but running on battery?

[patrikr76]

@mfalkvidd , you are absolutely right. Let me clarify.
I am using the example code that is under build and power meter pulse sensor, updated may 1st, 2018.
Running it with "sleep mode = false" it works fine with both usb and battery power, just drains the battery quite fast.
Setting it to "sleep mode = true" it doesn't work either on usb or battery power.

Here's how the serial monitor looks for the startup phase with sleep mode as true.

23:21:24.219 ->  __  __       ____
23:21:24.219 -> |  \/  |_   _/ ___|  ___ _ __  ___  ___  _ __ ___
23:21:24.219 -> | |\/| | | | \___ \ / _ \ `_ \/ __|/ _ \| `__/ __|
23:21:24.219 -> | |  | | |_| |___| |  __/ | | \__ \  _  | |  \__ \
23:21:24.219 -> |_|  |_|\__, |____/ \___|_| |_|___/\___/|_|  |___/
23:21:24.219 ->         |___/                      2.3.0
23:21:24.219 -> 
23:21:24.219 -> 16 MCO:BGN:INIT NODE,CP=RNNNA---,VER=2.3.0
23:21:24.254 -> 25 TSM:INIT
23:21:24.254 -> 26 TSF:WUR:MS=0
23:21:24.254 -> 33 TSM:INIT:TSP OK
23:21:24.254 -> 35 TSF:SID:OK,ID=2
23:21:24.254 -> 37 TSM:FPAR
73 TSF:MSG:SEND,2-2-255-255,s=255,c=3,t=7,pt=0,l=0,sg=0,ft=0,st=OK:
201 TSF:MSG:READ,0-0-2,s=255,c=3,t=8,pt=1,l=1,sg=0:0
23:21:24.428 -> 205 TSF:MSG:FPAR OK,ID=0,D=1
2080 TSM:FPAR:OK
23:21:26.306 -> 2081 TSM:ID
23:21:26.306 -> 2082 TSM:ID:OK
23:21:26.306 -> 2084 TSM:UPL
23:21:26.306 -> 2087 TSF:MSG:SEND,2-2-0-0,s=255,c=3,t=24,pt=1,l=1,sg=0,ft=0,st=OK:1
23:21:26.306 -> 2095 TSF:MSG:READ,0-0-2,s=255,c=3,t=25,pt=1,l=1,sg=0:1
23:21:26.306 -> 2100 TSF:MSG:PONG RECV,HP=1
23:21:26.306 -> 2103 TSM:UPL:OK
23:21:26.306 -> 2104 TSM:READY:ID=2,PAR=0,DIS=1
23:21:26.306 -> 2109 TSF:MSG:SEND,2-2-0-0,s=255,c=3,t=15,pt=6,l=2,sg=0,ft=0,st=OK:0100
23:21:26.341 -> 2118 TSF:MSG:READ,0-0-2,s=255,c=3,t=15,pt=6,l=2,sg=0:0100
23:21:26.341 -> 2125 TSF:MSG:SEND,2-2-0-0,s=255,c=0,t=17,pt=0,l=5,sg=0,ft=0,st=OK:2.3.0
23:21:26.341 -> 2134 TSF:MSG:SEND,2-2-0-0,s=255,c=3,t=6,pt=1,l=1,sg=0,ft=0,st=OK:0
2203 TSF:MSG:READ,0-0-2,s=255,c=3,t=6,pt=0,l=1,sg=0:M
23:21:26.410 -> 2210 TSF:MSG:SEND,2-2-0-0,s=255,c=3,t=11,pt=0,l=12,sg=0,ft=0,st=OK:Energy Meter
23:21:26.445 -> 2220 TSF:MSG:SEND,2-2-0-0,s=255,c=3,t=12,pt=0,l=3,sg=0,ft=0,st=OK:1.0
23:21:26.445 -> 2228 TSF:MSG:SEND,2-2-0-0,s=1,c=0,t=13,pt=0,l=0,sg=0,ft=0,st=OK:
23:21:26.445 -> 2234 MCO:REG:REQ
23:21:26.445 -> 2237 TSF:MSG:SEND,2-2-0-0,s=255,c=3,t=26,pt=1,l=1,sg=0,ft=0,st=OK:2
2252 TSF:MSG:READ,0-0-2,s=255,c=3,t=27,pt=1,l=1,sg=0:1
23:21:26.480 -> 2257 MCO:PIM:NODE REG=1
23:21:26.480 -> 2259 MCO:BGN:STP
23:21:26.480 -> 2263 TSF:MSG:SEND,2-2-0-0,s=1,c=2,t=24,pt=0,l=0,sg=0,ft=0,st=OK:
23:21:26.480 -> 2269 MCO:BGN:INIT OK,TSP=1
23:21:26.480 -> 2272 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
23:21:26.480 -> 2277 TSF:TDI:TSL

As far as i can see it seems fine and also shows up in myscontroller.

And here is an example how it looks after 20 seconds sleep.

2414 MCO:SLP:WUP=-1
23:26:36.346 -> 2416 TSF:TRI:TSB
23:26:36.346 -> 2418 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
23:26:36.346 -> 2423 TSF:TDI:TSL

I don't see any connection attempts with my gateway, or shouldn't there by any if the sensor is not triggered?

Here i can show how it looks when tricking the sensor with the ir on the mouse.

2456 MCO:SLP:WUP=-1
23:27:48.785 -> 2458 TSF:TRI:TSB
23:27:48.785 -> 2460 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
23:27:48.785 -> 2465 TSF:TDI:TSL
2467 MCO:SLP:WUP=-1
23:27:48.924 -> 2468 TSF:TRI:TSB
23:27:48.959 -> 2470 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
23:27:48.959 -> 2476 TSF:TDI:TSL
2478 MCO:SLP:WUP=-1
23:27:49.098 -> 2480 TSF:TRI:TSB
23:27:49.098 -> 2481 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
23:27:49.098 -> 2486 TSF:TDI:TSL
2488 MCO:SLP:WUP=-1
23:27:49.235 -> 2490 TSF:TRI:TSB
23:27:49.235 -> 2492 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
23:27:49.270 -> 2497 TSF:TDI:TSL
2498 MCO:SLP:WUP=-1
23:27:49.408 -> 2500 TSF:TRI:TSB
23:27:49.408 -> 2502 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
23:27:49.408 -> 2507 TSF:TDI:TSL
2509 MCO:SLP:WUP=-1
23:27:49.546 -> 2510 TSF:TRI:TSB
23:27:49.546 -> 2512 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
23:27:49.546 -> 2518 TSF:TDI:TSL

Interrupt is triggered and it wakes up, but nothing is sent to the gateway.

Have i missed something?

[mfalkvidd]

@patrikr76 great info, thanks!

The -1 in Wup=-1 indicates that the node was waken up by timer and not by interrupt. That’s very strange, since the time stamps clearly show that 20 seconds has not passed.

Could you add

Serial.println(pulseCount);

after

sleep(SEND_FREQUENCY);

to see if pulseCount is increased?

[patrikr76]

@mfalkvidd , Here's the serial monitor after adding the serial print.

2269 MCO:SLP:WUP=-1
14:15:15.125 -> 2270 TSF:TRI:TSB
14:15:15.125 -> 0
14:15:15.125 -> 2272 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
14:15:15.125 -> 2277 TSF:TDI:TSL
2279 MCO:SLP:WUP=-1
14:15:37.264 -> 2281 TSF:TRI:TSB
14:15:37.264 -> 0
14:15:37.264 -> 2282 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
14:15:37.264 -> 2287 TSF:TDI:TSL
2289 MCO:SLP:WUP=-1
14:15:59.354 -> 2291 TSF:TRI:TSB
14:15:59.354 -> 0
14:15:59.354 -> 2293 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
14:15:59.389 -> 2298 TSF:TDI:TSL

As you see, it sleeps about 22 secs with no pulse on it.

Moving the mouse over it gives this.

2353 MCO:SLP:WUP=-1
14:18:09.982 -> 2355 TSF:TRI:TSB
14:18:09.982 -> 4
14:18:09.982 -> 2357 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
14:18:09.982 -> 2362 TSF:TDI:TSL
2363 MCO:SLP:WUP=-1
14:18:23.059 -> 2365 TSF:TRI:TSB
14:18:23.059 -> 5
14:18:23.059 -> 2367 MCO:SLP:MS=20000,SMS=0,I1=255,M1=255,I2=255,M2=255
14:18:23.059 -> 2372 TSF:TDI:TSL

Looking through the code, this one raises a flag.

if (pcReceived && (SLEEP_MODE || sendTime))

The pulseCount send is within that if statement which is not fullfilled because it doesn't receive pcReceived.
Or am i wrong?

[patrikr76]

Found a solution, which is not ideal.
I was having similar problem with a motion sensor not being able to send, adding a delay after sleep helped on that one, but it was just a 5ms delay needed.

On this one i had to add a 1 sec delay in the begining of the loop. I tried smaller but then it didn't work.
I am starting to think it is the nano clone that has crappy components.

[mfalkvidd]

@patrikr76 nice work.
Sounds very similar to https://forum.mysensors.org/topic/9655/delay-after-tsf-tri-tsb/14

[StKilda]

I have two meters I would like to measure: One that the LED is off most of the time and briefly pulses on per increment (at 3200 pulses per Kwh - so I adjusted the pulse factor accordingly). This works fine. The second meter, the LED pulses off (meaning the led it lit most of the time, and pulses off briefly). Do I need to adjust the code in any way to deal with this meter? Thanks. Really impressed it works virtually out of the box.

[mfalkvidd]

Great work @stkilda
I think the sketch will work without modification. The easiest way to find out is probably to try it.

[gohan]

you may need to change the interrupt behavior, but try it first