Root causes of high deep sleep current [LoPy1, WiPy2 and SiPy], all the new modules **do not have deepsleep issues**

[mcook]

@Fred

Hi,

I received my PyTrack boards today. I have attached my LoPy to one of them. I am powering the PyTrack via USB. The USB is plugged into a USB multimeter. With the WiFi on, I see it is drawing about 125 mA at 5 VDC. After entering deep sleep, I see the current is reduced to ~48 mA. Is there some configuration that needs to be done in the code in order for the PyTrack boards to work properly with Deep Sleep?

Mike

[daniel]

@mcook, entering deepsleep using the Pytrack as the controller of the operation requires a different approach. We have a library for this that will be published in a few days this week. This library also allows to performing several different operations on the PIC microcontroller that's on the Pytrack/Pysense like reading the battery voltage.

Cheers,
Daniel

[daniel]

Hello,

In the meantime while we publish the official library later this week, here is a short version of it which can be used to enter deep sleep mode using the Pytrack/Pysense board and wakeup after a pre-defined number of seconds:

from machine import I2C
from machine import Pin
import machine
import time


I2C_SLAVE_ADDR = const(8)    # This one seems to be an unpopular addresss
LIS2HH12_I2C_ADDR = const(30)   # both Pysense and Pytrack have the accelerometer onboard

CMD_PEEK = const(0x0)
CMD_POKE = const(0x01)
CMD_MAGIC = const(0x02)
CMD_HW_VER = const(0x10)
CMD_FW_VER = const(0x11)
CMD_PROD_ID = const(0x12)
CMD_SETUP_SLEEP = const(0x20)
CMD_GO_SLEEP = const(0x21)
CMD_BAUD_CHANGE = const(0x30)
CMD_DFU = const(0x31)

REG_CMD = const(0)
REG_ADDRL = const(1)
REG_ADDRH = const(2)
REG_AND = const(3)
REG_OR = const(4)
REG_XOR = const(5)

ANSELA_ADDR = const(0x18C)
ANSELB_ADDR = const(0x18D)
ANSELC_ADDR = const(0x18E)

ADCON0_ADDR = const(0x9D)
ADCON1_ADDR = const(0x9E)

_ADCON0_CHS_POSN = const(0x02)
_ADCON0_ADON_MASK = const(0x01)
_ADCON1_ADCS_POSN = const(0x04)
_ADCON0_GO_nDONE_MASK = const(0x02)

ADRESL_ADDR = const(0x09B)
ADRESH_ADDR = const(0x09C)

PORTC_ADDR = const(0x00E)

WPUA_ADDR = const(0x20C)

MEMORY_BANK_ADDR = const(0x620)

class Pysense:

    def __init__(self, i2c=None, sda='P22', scl='P21'):
        if i2c is not None:
            self.i2c = i2c
        else:
            self.i2c = I2C(0, mode=I2C.MASTER, pins=(sda, scl))
        self.reg = bytearray(6)
        scan_r = self.i2c.scan()
        if not scan_r or LIS2HH12_I2C_ADDR not in scan_r or not I2C_SLAVE_ADDR in scan_r:
            raise Exception('Pysense board not detected')

    def _write(self, data, wait=True):
        self.i2c.writeto(I2C_SLAVE_ADDR, data)
        if wait:
            self._wait()

    def _read(self, size):
        return self.i2c.readfrom(I2C_SLAVE_ADDR, size + 1)[1:(size + 1)]

    def _wait(self):
        time.sleep_us(100)
        while self.i2c.readfrom(I2C_SLAVE_ADDR, 1)[0] != 0xFF:
            time.sleep_us(100)

    def _send_cmd(self, cmd):
        self._write(bytes([cmd]))

    def setup_sleep(self, time_s):
        self._write(bytes([CMD_SETUP_SLEEP, time_s & 0xFF, (time_s >> 8) & 0xFF, (time_s >> 16) & 0xFF]))

    def go_to_sleep(self):
        self._write(bytes([CMD_GO_SLEEP]), wait=False)
        # kill the run pin
        Pin('P3', mode=Pin.OUT, value=0)

Example usage:

from pysense import Pysense
py = Pysense()
py.setup_sleep(5)   # setup a 5 seconds sleep time
py.go_to_sleep()

This works for both Pysense and Pytrack.

Cheers,
Daniel

[lj]

@Fred

I didn't received my Deep Sleep Shield , what is the process to check if my request has been registered, plan and hipped to me?

best
Luc

[irvinvp]

@bucknall Did the shields arrive?

[leogaggl]

I am also wondering what the process is if the Deep Sleep Shields don't arrive. I ordered them when the email went around for my order of LoPy's but never received anything ?

[rskoniec]

I've filled the form at https://www.surveymonkey.co.uk/r/PSTPWBB (mid April) and still nothing - no email confirmation, order progress and no DS shields.

@Fred said in Root causes of high deep sleep current:

Dear All,
A quick update on this topic to confirm we are currently testing the Pilot Run samples in Eindhoven. All going well and we expect to have shields to ship circa 2 to 3 weeks e.g. 5/6 June. (...)

@Fred Any news?

[daniel]

Hello,

We have received all the surveys filled up to now and everything is going OK. We are targeting mass production of the deep sleep shields in early July and ship shortly afterwards.

Cheers,
Daniel

[crankshaft]

@Fred - Is it possible to acknowledge receipt of the request for a shield when it is initially made and then notify us when it is dispatched, as at the moment I have no idea whether my submission was successful, or when I am likely to receive the shields ?!

[athtest800]

Hi!
The PIC10F322 needs programming? If yes, where can I find the code?
Thank you!

[irvinvp]

@bucknall Did the shields arrive? is 13th July and i don´t have news.

[rskoniec]

@irvinvp Here is the answer https://forum.pycom.io/topic/1480/deep-sleep-shields-delivery-update

[jalperin]

I wish to run my lopy on a pysense with the lowest possible current draw from a LiPo battery. I assume I would plug the deep sleep shield in between them. Correct?

Could someone please summarize in one place all the appropriate actions to turn off everything but awake periodically to take a lux reading and transmit it via lora? Thanks for your help.

[jcaron]

@jalperin from what I understood, the Pysense (and Pytrack) boards actually incorporate what is present on the deep sleep shields.

However, there is still a lingering issue on the Pysense (not the Pytrack) which still uses more current than it should in deep sleep, still waiting for a solution for this.

[daniel]

@jcaron the reason for the higher current on the Pysense is because the sensors need to be disabled as part of the process of going to deepsleep. We will add that to the official library within the next few days.

Cheers,
Daniel

[jalperin]

Thank you @jcaron.

Unfortunately, I remain where I started. I am unsure whether I am shutting down WiFi and Bluetooth (I read various posts), and I am unsure how to properly get both the pysense and lopy to deep sleep.

Do I understand it correctly that I don't really need the deep sleep shield but I do need new libraries (that will show up in a few days and will hopefully contain examples)? If so, how does one shut off wifi and bluetooth?

Thanks for your help.

[jcaron]

@jalperin With the Pysense board, you need to:

• use the libraries that are available on github:
  https://github.com/pycom/pycom-libraries/tree/master/pysense/lib

• add the following code:

  from pysense import Pysense
  py = Pysense()
  py.setup_sleep(number_of_seconds)
  py.go_to_sleep()
  

This will tell the Pysense to stop the LoPy (including Wi-Fi, BT, LoRa...). It's actually powered off. It will be woken up after the given number of seconds has elapsed. Work is in progress to allow wake up on other conditions (some input going high for instance).

At the moment, this will bring down power consumption to about 2.5 mA when running from a LiPo battery. Ultimately it should go down to a fraction of that once everything is correctly shut down.

Note that when the LoPy comes out of deep sleep, it's as if it just rebooted. There are still issues with LoRa keeping state for instance.

As the firmware turns on Wi-Fi when it starts, it's a good idea to stop it right away if you don't need it during the time the LoPy is awake. Not sure about Bluetooth (I stop it, but I'm not sure that's actually needed).

Hope that helps.

[jalperin]

Thank you, @jcaron. Very helpful.

I have been executing that code, but followed by machine.deepsleep(). I gather that is not necessary. Correct?

So, lastly, exactly how does one turn off wifi and bluetooth (I've read different methods in this forum)?

[jmarcelino]

@jalperin
This works to turn off WiFi

import network
n = network.WLAN()
n.deinit()

Turning off Bluetooth would be similar but for network.Bluetooth() instead.
However removing Bluetooth doesn't change anything in power consumption at the moment - but also it won't be on or consuming anything unless you had previously initialised Bluetooth() in your code. So it's actually better to do nothing at all for the Bluetooth side.

[jalperin]

Thanks jmarcelino.

Does the below earlier comment of yours still apply?

By the way I found that if I do:

from network import WLAN
wlan = WLAN(mode=WLAN.STA)
power saving is enabled, but if I use this instead it doesn't:

from network import WLAN
wlan = WLAN()
wlan.mode(WLAN.STA)