Hi everyone, I started a discussion on the JeeLabs forum (since it is the meeting point of RFM12B users) to see if it is possible to sense the voltage and current on two seperate devices. You can have a look at this page and contribute with your ideas/thoughts...
http://forum.jeelabs.net/node/1006
"Hi everyone, I am trying to make a wireless zero crossing detector with a jeenode and a nanode RF (arduino clone with ENC28J60 and RFM12B). The idea is to detect on the nanode when a voltage sinusoidal waveform starts and directly transmit one bit over the radio to the base (nanode RF). Since the transmission can be made in 160µs (see this post:http://jeelabs.org/2011/01/14/nodes-addresses-and-interference/) this would represent 0.008 of the waveform length (20ms) and can be corrected with software. Any ideas/thoughts/on this? Thanks"
Re: wireless voltage sensor
the conversation is progressing here: http://forum.jeelabs.net/node/1006
Re: wireless voltage sensor
From JeeLabs forums:
Actually the problem is that there is no plug next to the currents sensing node. You are definitely right about the voltage drop, but experience showed that the waveform doesn't really get distorted. And anyway, I think that the load distorts the current waveform and not the voltage right?.
I think measuring the power factor accurately is more decisive than the drop in Vrms.
My idea is to have the current measuring node listen to when the voltage waveform start (no amplitude, a single bit that tells: NOW the voltage waveform started) then it calculates the current's waveform phase shift, and it sends both values to the voltage sensing unit back, which multiplies both current Irms and PF by the Vrms and sends it to a server.
So in this order:
Feasible?
I saw a similar application with Bluetooth modules (3000 measurements/sec sent), is it possible with RFM12B ? Theory says YES
Re: wireless voltage sensor
making slow progress here: http://forum.jeelabs.net/node/1006
Re: wireless voltage sensor
PROGRESS!
I managed to send an empty packet from one node to another to indicate the start of a waveform. I am using two identical setups, one sender (jeenode 6) and one receiver (jeenode usb). Both are hooked to two identical AC-AC adapters with step down voltage dividers and a DC offset. I could manage to measure the time between the start of the waveform and the time of its reception on the receiving node, but the results are quite surprising:
Serial Monitor:
ZeroCross Wireless Receiver From 7 ok | Timer: 62084 uS
From 7 ok | Timer: 82036 uS
From 7 ok | Timer: 42444 uS
From 7 ok | Timer: 102100 uS
From 7 ok | Timer: 2316 uS
From 7 ok | Timer: 2300 uS
From 7 ok | Timer: 22236 uS
From 7 ok | Timer: 2308 uS
From 7 ok | Timer: 22220 uS
From 7 ok | Timer: 2316 uS
From 7 ok | Timer: 42420 uS
From 7 ok | Timer: 2332 uS
From 7 ok | Timer: 2324 uS
From 7 ok | Timer: 22388 uS
From 7 ok | Timer: 42340 uS
From 7 ok | Timer: 102004 uS
From 7 ok | Timer: 82120 uS
From 7 ok | Timer: 42452 uS
From 7 ok | Timer: 82068 uS
From 7 ok | Timer: 101980 uS
From 7 ok | Timer: 142116 uS
From 7 ok | Timer: 181784 uS
From 7 ok | Timer: 201844 uS
From 7 ok | Timer: 2316 uS
From 7 ok | Timer: 82156 uS
From 7 ok | Timer: 42036 uS
From 7 ok | Timer: 62436 uS
From 7 ok | Timer: 2316 uS
From 7 ok | Timer: 22204 uS
From 7 ok | Timer: 41988 uS
From 7 ok | Timer: 61964 uS
From 7 ok | Timer: 82164 uS
From 7 ok | Timer: 62524 uS
From 7 ok | Timer: 42020 uS
From 7 ok | Timer: 102140 uS
From 7 ok | Timer: 2300 uS
From 7 ok | Timer: 42452 uS
From 7 ok | Timer: 42036 uS
From 7 ok | Timer: 41988 uS
From 7 ok | Timer: 61988 uS
From 7 ok | Timer: 21964 uS
From 7 ok | Timer: 82148 uS
From 7 ok | Timer: 22388 uS
From 7 ok | Timer: 42348 uS
From 7 ok | Timer: 62364 uS
From 7 ok | Timer: 122020 uS
From 7 ok | Timer: 2300 uS
From 7 ok | Timer: 22348 uS
From 7 ok | Timer: 42444 uS
From 7 ok | Timer: 82020 uS
From 7 ok | Timer: 41972 uS
Here are the sketches I use:
[Sending]:
#include <JeeLib.h>
Port voltage (4); Port reference (1);
void setup () {
Serial.begin(57600);
Serial.println("ZeroCross Wireless");
rf12_initialize(7, RF12_868MHZ, 7); }
void loop () { word volt = voltage.anaRead(); word ref = reference.anaRead(); int diff1 = volt-ref; delayMicroseconds(500); volt = voltage.anaRead(); ref = reference.anaRead(); int diff2 = volt-ref; // Serial.println(diff2-diff1);
if (diff2 <10 && diff2 >-10 && diff2-diff1>0){ rf12_recvDone(); if (rf12_canSend()){ rf12_sendStart(0); Serial.println("RF OK"); delayMicroseconds (500); } }
}
[Receiving:]
#include <JeeLib.h>
Port voltage (4); Port reference (1);
unsigned long time = 0; unsigned long period;
void setup () {
Serial.begin(57600);
Serial.println("ZeroCross Wireless Receiver");
rf12_initialize(8, RF12_868MHZ, 7);
}
void loop() {
rf12_recvDone(); word volt = voltage.anaRead(); word ref = reference.anaRead(); int diff1 = volt-ref;
volt = voltage.anaRead(); ref = reference.anaRead(); int diff2 = volt-ref; if (diff2 <10 && diff2 >-10 && diff2-diff1>0){ time = micros(); } if (rf12_recvDone() && rf12_crc == 0 && rf12_len == 0 && rf12_hdr == 7) { period = micros()-time; Serial.print ("From 7 ok | Timer: "); Serial.print(period); //time in micros Serial.println(" uS");
}}
Any thoughts on the strange results?
Re: wireless voltage sensor
Progress was made here: http://forum.jeelabs.net/node/1006 I could measure 8uS between the detection of the beginning of a waveform on one node and the reception of that signal on another node. Now I need to measure the period between the beginning of the current waveform and the voltage to deduce the power factor.
Any ideas?
You can see all the code I use on the link: http://forum.jeelabs.net/node/1006