Solar Pannel and LiOn or Lipo

[epierre]

sleep is the problem, it loads at 30s never above...

Batt: 4.06V ; 0.00A 
Solar: 2.56V ; 0.00 A; charge: No
BattPct: 79% 

Batt: 4.05V ; 0.00A 
Solar: 3.38V ; 0.00 A; charge: No
BattPct: 78%

on the other:

2015-07-05 20:32:40 14 10 1 0 38 4.024
2015-07-05 20:32:40 14 10 1 0 39 0.001298
2015-07-05 20:32:41 14 11 1 0 38 3.991
2015-07-05 20:32:42 14 11 1 0 39 0.000000
2015-07-05 20:32:42 14 255 3 0 0 74

2015-07-05 16:24:08 14 10 1 0 38 3.953
2015-07-05 16:24:08 14 10 1 0 39 0.001298
2015-07-05 16:24:09 14 11 1 0 38 4.731
2015-07-05 16:24:10 14 11 1 0 39 0.015577
2015-07-05 16:24:10 14 255 3 0 0 64

2015-07-05 18:55:40 14 10 1 0 38 4.050
2015-07-05 18:55:40 14 10 1 0 39 0.061008
2015-07-05 18:55:40 14 11 1 0 38 4.212
2015-07-05 18:55:42 14 11 1 0 39 0.066200
2015-07-05 18:55:42 14 255 3 0 0 78

One one of the two units I have issues on radio:

Batt: 4.05V ; 0.00A 
Solar: 1.80V ; 0.00 A; charge: No
BattPct: 78% 
send: 15-15-0-0 s=10,c=1,t=38,pt=7,l=5,st=fail:4.050
send: 15-15-0-0 s=10,c=1,t=39,pt=7,l=5,st=fail:0.001298
send: 15-15-0-0 s=11,c=1,t=38,pt=7,l=5,st=ok:1.804
send: 15-15-0-0 s=11,c=1,t=39,pt=7,l=5,st=fail:0.000000
send: 15-15-0-0 s=255,c=3,t=0,pt=1,l=1,st=fail:78

[ceech]

@epierre I'll ask @AWI, he may help.

[AWI]

Your circuit seems to be working correctly. I cannot figure out why the sleep time should influence the metering. I suggest you first test it with a minimal sketch to figure out this sleep issue.

[epierre]

@AWI What should I keep as minimal ? also any idea about the radio hazard ? I've changed several antennae and it is either half wrong or "check wire"

[AWI]

@epierre in a minimal sketch I would remove have it only send one message and give it enough println's to debug via serial. Check the voltages with a meter and let it run with an increasing sleep time. It could be one of those 'bad radios' which are discussed here on the forum.

For the charging part: it only charges a partly empty cell and the charge voltage should be high enough to let it happen.

Debugging is where the fun starts (if you are able to solve it)..

[epierre]

so the same without sensor nor mysensors... 30 secs fine 40 secs dead...

I've removed the radio chipset, same behavior...

/*
 
#include <SPI.h>

#include <Wire.h> 				// I2C


#define LTC4067_CHRG_PIN	A1		//analog input A1 on ATmega 328 is /CHRG signal from LTC4067
#define batteryVoltage_PIN	A0		//analog input A0 on ATmega328 is battery voltage ( /2)
#define solarVoltage_PIN	A2		//analog input A2 is solar cell voltage (/ 2)
#define solarCurrent_PIN	A6		//analog input A6 is input current ( I=V/Rclprog x 1000 )
#define batteryChargeCurrent_PIN	A7	//analog input A7 is battery charge current ( I=V/Rprog x 1000 )
#define LTC4067_SUSPEND_PIN	9		//digital output D9 - drive it high to put LTC4067 in SUSPEND mode

const float VccMin        = 1.0*3.5;  // Minimum expected Vcc level, in Volts. Example for 1 rechargeable lithium-ion.
const float VccMax        = 1.0*4.2;  // Maximum expected Vcc level, in Volts. 

#define LIGHT_SENSOR_ANALOG_PIN 3   // Digital input did you attach your soil sensor.  
#define CHILD_ID_LIGHT 0   // Id of the sensor child
#define BATT_CHILD_ID 10
#define SOLAR_CHILD_ID 11

// PIN Radio
#define RADIO_CE_PIN    7       // radio chip enable
#define RADIO_SS_PIN    8      // CS SS serial select

float lastBattVoltage;
float lastBattCurrent;
float lastSolarVoltage;
float lastSolarCurrent;
int lastBattPct = 0;
uint16_t lastlux;
float VccReference = 3.3 ;				// voltage reference for measurement, definitive init in setup


unsigned long SLEEP_TIME = 40*1000;  // sleep time between reads (seconds * 1000 milliseconds)

int lastSoilValue = -1;

void setup()  
{ 
 
  Serial.begin(115200);
  // use VCC (3.3V) reference
  analogReference(DEFAULT);								// default external reference = 3.3v for Ceech board
  VccReference = 3.323 ;									// measured Vcc input (on board LDO)
  pinMode(LTC4067_SUSPEND_PIN, OUTPUT);					// suspend of Lion charger set
  digitalWrite(LTC4067_SUSPEND_PIN,LOW);   			//  active (non suspend) at start

}
 
void loop()     
{     

  sendVoltage();
  
    
  // Power down the radio
 delay(SLEEP_TIME);
}

void sendVoltage(void)
// battery and charging values
{
	// get Battery Voltage & charge current
	float batteryVoltage = ((float)analogRead(batteryVoltage_PIN)* VccReference/1024) * 2;	// actual voltage is double
	Serial.print("Batt: ");
	Serial.print(batteryVoltage);
	Serial.print("V ; ");
	float batteryChargeCurrent = ((float)analogRead(batteryChargeCurrent_PIN) * VccReference/1024)/ 2.5 ; // current(A) = V/Rprog(kohm)
	Serial.print(batteryChargeCurrent);
	Serial.println("A ");

   
	// get Solar Voltage & charge current
	float solarVoltage = ((float)analogRead(solarVoltage_PIN)/1024 * VccReference) * 2 ;		// actual voltage is double
	Serial.print("Solar: ");
	Serial.print(solarVoltage);
	Serial.print("V ; ");
	// get Solar Current
	float solarCurrent = ((float)analogRead(solarCurrent_PIN)/1024 * VccReference)/ 2.5;		// current(A) = V/Rclprog(kohm)
	Serial.print(solarCurrent);
	Serial.print(" A; charge: ");
	Serial.println(digitalRead(LTC4067_CHRG_PIN)?"No":"Yes");
	
	// send battery percentage for node
	int battPct = 1 ;
	if (batteryVoltage > VccMin){
		battPct = 100.0*(batteryVoltage - VccMin)/(VccMax - VccMin);
	}
	Serial.print("BattPct: ");
	Serial.print(battPct);
	Serial.println("% ");

	
}

[epierre]

I've received a string of 10 radio chipset tonight and changing the component has no effect compared to the other board where radio is ok.

are you sure the problem is softare for what I see here is not dependant of a sophisticated library ?

I've learned recently of IAQ (incomming assurance quality) because Photon (formerly Sark) has been hit by a suppliers problem: https://community.particle.io/t/photon-manufacturing-shipping-update/12275

[AWI]

@epierre this sounds really weird. The delay() you used in this sketch does not sleep anything but just takes an active break. Can you measure the supply voltage on vcc to determine what is happening?

[epierre]

voltage was made through the ftdi...

I'll check again, but an interresting result, the unit with sending problems works well when used with the amplified nrf24l !!! ??? !!!

I've plugged it too with a 4000mAh LiPo battery I just received. I'll try later to play with sleep or delay

2015-07-09 15:14:04 15 10 1 0 38 4.128
2015-07-09 15:14:04 15 10 1 0 39 0.001298
2015-07-09 15:14:05 15 11 1 0 38 4.524
2015-07-09 15:14:06 15 11 1 0 39 0.003894
2015-07-09 15:14:06 15 255 3 0 0 89

[AWI]

@epierre said:

voltage was made through the ftdi...

Are you powering or measuring the unit from ftdi? You should only use the on-board power supply.

[epierre]

Hello,

since I have 2 units, I have one on battery/solar panel and the other with issues on radio through ftdi else I couldn't see what it had to display...

I'm off for a week, I'll do further testing after.

Can you reproduce some part of the delay/speel issue ?

[epierre]

@awi @ceech what test should I do about those units:

• delay test: what kind? what to watch ?
• sleep test: what kind? what to watch ?
• radio issue on one of the boards working only with amplified board, not with small one.

The good news it although they pollute the radio by sending infos every 30s, they are still alive alone with the sun one week after.

[ceech]

@epierre Try comparing the two boards side by side. You'll get an idea where the problem is.

[epierre]

@ceech do you mean on SW or HW side ?

HW:

• I cannot plug the standard radio in one, only amplified works with it
• same solar pannel 1W
• different batteries, one is LiPo, the other is LiOn, but same behavior on delay/sleep seen on both
• same lux sensor

SW: they run the same software

any idea ?

[ceech]

@epierre Not enough power for the amplified radio, probably. Try adding bigger capacitor. In the hundreds of uF region.

[epierre]

@ceech are the radio 3.3V or not ? why should I need a capacitor and between what parts ?

[AWI]

@epierre I quess what @ceech meant is that the on board power supply is not equiped to power the amplified radio. I you put a large capacitor on the power lines (near to the radio) it will handle the power better.

I am still confused as to the problem you are having with the delay. I am able to reproduce your > 30second problem and it is driving me nuts. When I execute the 'original' sketch (including MySensors) it is working fine with gw.sleep() and times longer than 30 secs.

i will do some additional measurements...

[epierre]

@AWI @ceech I have tried this:
-I have removed the mysensor debug mode on the sensor

• I have removed the 4 x gw.send in the sendVoltage

For a sleep time of 31-32 this is fine , start to fail at 33...

Batt: 4.07V ; 0.00A 
Solar: 0.84V ; 0.00 A; charge: No
BattPct: 81% 

@AWI the one with amplified has a 4000mAh Lipo and the radio is perfect. Anyway simple radio does not work with it.

[epierre]

@AWII @ceech any advancement on this ?

[ceech]

@epierre Nope, sorry.