I can always make another circuit in case there is any problem with the current one, even though I have already done several checks.
But only tomorrow. It’s 2:00 in the morning here.
But I liked that someone told me, from their own experience, if they had effectively measured the 10μA in deep sleep.
I dont have an ESP module to test with, only dev boards, but ill knock something up when I get home and post up my results.
Each time I reset the result is expected.
Yup that is the typical serial output of a deepsleep reboot.
@toco Well, “deep sleep” on it 
… but since low power modes have nothing to do with Blynk directly, you may find only limited quantity and subjective advice here. I recommend you also ask around at dedicated ESP8266 forums.
Then share your findings here to increase our knowledge 
and the 190mA - where is this measured?
OK
I’ll take the advice and try elsewhere.
Then, if it’s worth it, I’ll come back here to share.
THANK YOU ALL.
I have this running on my mailbox and it works great for weeks on a 800mah cell phone battery. Ground is tied to reset through a normally closed push button switch. When the mailbox is closed the circuit sleeps at about 10 micro ah. When the mailbox door is opened the switch closes and then opens when it is closed sending the data. Battery voltage is an ESP8266 function so no wires required. Oh…it also sends the temp outside through a dallas onewire sensor on pin 0. This is a picture of the board if it links.
Hope it helps.
#include <ESP8266WiFi.h>
#include <BlynkSimpleEsp8266.h>
#include <DallasTemperature.h>
#include <OneWire.h>
char auth[] = "basdlfj;alkdfjalskdfjals;kdfjaslkdfj"; //mail box
const int sleepTimeS = 1000;
#define ONE_WIRE_BUS 0
OneWire oneWire(ONE_WIRE_BUS);
DallasTemperature DS18B20(&oneWire); // Pass our oneWire reference to Dallas Temperature.
float T1,T2; // Temps in C&F
ADC_MODE(ADC_VCC);
void setup()
{
DS18B20.begin();
Blynk.begin(auth, "SSID", "PW", IPAddress(123.456.1.1));
DS18B20.requestTemperatures(); // Send the command to get temperatures
T1 = (DS18B20.getTempCByIndex(0)); //Centigrade
T2 = int(((T1*1.8)+32)-2); // F
Blynk.virtualWrite(5, T2); // Round F
float voltage = ESP.getVcc()/1072.43; // mess with this denominator to get accurate voltage
Blynk.virtualWrite(2, voltage); //battery voltage
Blynk.email("Mail", "You've got mail!");
//delay(1000);
Blynk.run();
ESP.deepSleep(sleepTimeS * 10000);
}
void loop()
{
}
1 Like
Thanks randobi99 for the help.
Please tell me how much your circuit consumes when it’s in deep sllep.
@toco What @Dave1829 had pointed out 
@randobi99 The mailman(woman) hasn’t freaked and called in any bomb scares yet? 
So you mean 1 week or 2 weeks or more?
@speed57 Google is our friend (unfortunately, so is math 
)
http://www.digikey.ca/en/resources/conversion-calculators/conversion-calculator-battery-life
800 / 1 * 0.7 = 560 Hours aprox. so depending on “stuff” (thus the * 0.7) probably between 3-5 weeks
1 Like
I agree with that google stuff but my problem is math we are not good friends but since blynk I try to understand it 
.
thanks @Gunner short cuts makes me understand more faster than google
Hello
Here I am, the friend of Google.
I’m enjoying noticing that I’ve arranged a reason for everyone to talk.
I would like to tell you that I still can not get those blessed 10μA. Something is wrong but I will continue to search until I realize what it is.
Anyway, thank you for the interventions you have made and I am going to take advantage of some suggestions that have been made here to understand more about this subject.
Hang in there… every hardware device is a little bit (or a lot) different. Something as simple as firmware version or the types of support circuitry can make all the difference to current draw. Let us know how it works out.
If your ESP has a power LED that stays on, your power consumption will be fairly high. You have to remove it from the board to achieve the low micro amp power consumption.
Hello everyone.
To play safe I bought another ESP 8266-12 (actually five) and I came to two conclusions:
First - The ESP I was using is not good. Either because of manufacturing defect or because of my “experiences”, it has a very high consumption.
Second - Consumption is highest when GPIOS are "high"after ESP goes into deep sleep.
So I changed the sketch to work backwards. Instead of the closed door information being “high” became “low”.
And everything is all right.
Each time the door opens or the door is opened I receive a notification informing what happened and the volts of the batteries.
Maybe due to my multimetre defect, the consumption is not stable, but varies between 10 and 20 microA in deep sleep.
I installed the device for two weeks with two 1.5V alkaline batteries and continues to work perfectly with 3.1v information for the BLINK. Let’s see how long they last.
I find the second conclusion a bit strange, but it actually solved the problem.
I also discovered that the circuit I made does not work with an ESP-01 because when the GPIO 2 is low and reset, everything hangs and does not work.
I do not know why.
By the way, for those interested, this is the circuit I’ve done.
The LEDs are optional, and do not have much influence on the consumption because they only light momentarily: the green lights when the door is closed but immediately turns off when the ESP goes into “deep sleep” and the red when the magnetic goes through the reset reed. I just left the red.
The initial idea of the LEDs was to facilitate the tuning of the reeds with the magnetics. The magnetics can not be very strong because as the reeds are very close to each other the desired effect is not achieved. Do not forget that the magnetic door information must be aligned with the respective reed and the reset must be misaligned so as to only activate the reset reed momentarily when the door moves in any direction.
And, that’s all.
If you need any help, just let me know.
Thanks for your contribution.
2 Likes
Just FYI, sharing information.
I am using ESP8266-12E nodemcu dev kit with DHT22 and deepsleep (sleep 5 minutes, measure temp and humidity, sleep 5 minutes, …)
This is how the board exactly looks like:
When I measure current it was 75mA when wake up and around 2,5mA when in deepsleep. It is powered by 4 AA rechargable NiMH batteries (4x 1,25V = 5V). I also expected lower current consumption when in deepsleep, but for now I can live with it.
@helpfinder The drain you see comes from the usb serial adapter and the regulator which are embedded in the nodemcu
2 Likes