ports/esp32:* Implementation of deep sleep, esp32 module, machine.RTC #3531

MrSurly · 2017-12-31T01:29:27Z

Note 1: Copied over from the old micropython-esp32 repo.
Note 2: Fixed the 1970/2000 bug in RTC.datetime()

Current API:

Basic sleep

import machine
machine.deepsleep() # Sleep until hard reboot
machine.deepsleep(5000) # Deep sleep for 5 seconds

machine.sleep() is implemented, but currently throws an error. If/when the IDF "light sleep" works, this can be removed.

EXT0 wakeups, using the common Micropython idiom

EXT0 wakes are for a single pin, either high or low level.

This is separate from the IRQ handler; the WAKE_LOW and WAKE_HIGH only setup the EXT0 type wake. Using a value other than WAKE_LOW or WAKE_HIGH will clear the EXT0 wake from that pin.

Additionally, any EXT0 wake will fail if esp32.wake_from_touch(True) has been called, since wake from touch and EXT0 are mutually exclusive in the IDF.

>>> import machine
>>> def h(*args): print(args)
...
>>> p1 = machine.Pin(26)
>>> p2 = machine.Pin(27)
>>> p1.irq(trigger = machine.Pin.WAKE_LOW, wake = machine.DEEPSLEEP)
<IRQ>
>>> p2.irq(trigger = machine.Pin.WAKE_LOW, wake = machine.DEEPSLEEP) 
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ValueError: no resources
>>> p1.irq(trigger=machine.Pin.IRQ_RISING, handler = h) # Clears EXT0 wake
>>> p2.irq(trigger = machine.Pin.WAKE_LOW, wake = machine.DEEPSLEEP) # EXT0 now assigned to p2
<IRQ>

Other changes to `machine`

machine.reset_cause() has been implemented, though usually returns machine.WDT_RESET since most of the time the IDF doesn't reset properly. Does properly return machine.DEEPSLEEP_RESET when waking from deep sleep.

machine.wake_reason() has been implemented.

The new top-level `esp32` module

esp32.wake_on_touch(<True|False>) Configures waking on touch. Throws ValueError('no resources') in the event that EXT0 is already configured, since they are not compatible. For this to work, you need to use machine.TouchPad().config() to configure the threshold; an interrupt handler is not required.

esp32.wake_on_ext0(<Pin>, level = <0 | 1>). Configures the EXT0 wake. This will raise ValueError('no resources') if esp32.wake_on_touch(True) has been called. It will not raise this error if another pin has already been configured, it will simply configure the new pin for EXT0 and deconfigure the old pin. If level is not specified, then it is unchanged, and defaults to 0 (low).

esp32.wake_on_ext1(pins = <sequence of Pin object>, level = <esp32.WAKEUP_ALL_LOW | esp32.WAKEUP_ANY_HIGH> ): As described above, wake on multiple pins, either "all low" or "any high".

The new `machine.RTC` module

machine.RTC.datetime() and machine.RTC.init() work as documented in mainline MicroPython.

machine.RTC.memory() Without arguments returns the RTC nonvolatile memory, as bytes. To write memory, pass a bytes, bytearray, or str object.

mactijn · 2017-12-31T15:59:58Z

@MrSurly This patch seems to remove esp_debug stuff, is that intended?
Also, MICROPY_PY_WEBSOCKET is changed to 0, not sure if this is wanted

MrSurly · 2017-12-31T18:00:52Z

@MrSurly This patch seems to remove esp_debug stuff, is that intended?
Also, MICROPY_PY_WEBSOCKET is changed to 0, not sure if this is wanted

Unintended side-effects of the merge. Let me get that fixed up.

MrSurly · 2017-12-31T18:07:46Z

@mactijn Fixed, squashed, pushed

mactijn · 2017-12-31T19:12:56Z

Funnily enough I ran into a similar thing yesterday.

Anyway, seems to work as promised. I'll do some more testing tomorrow, and hopefully release my M5Stack framework.

MrSurly · 2018-01-01T02:23:36Z

I'll do some more testing tomorrow, and hopefully release my M5Stack framework.

Is that a MicroPython framework for the M5Stack? I received my hardware a couple of days ago.

mactijn · 2018-01-01T15:51:01Z

Is that a MicroPython framework for the M5Stack? I received my hardware a couple of days ago.

Yep. Awesome stuff. Great support from the supplier. I've ordered the faces kit too, so I can add support for those modules as well.

Lisa999 · 2018-01-06T01:46:45Z

@MrSurly : Just to let you know, i immediately put your pr to work in my firmware version... ;-)
Wonderful work, just what i needed..

MrSurly · 2018-01-06T06:38:17Z

@Lisa999 Glad it's working well for you. Please let me know if you have any issues.

Lisa999 · 2018-01-13T16:30:45Z

tm = (2018, 1, 13, 0, 16, 18, 31, 0)
machine.RTC().datetime(tm)
print(utime.localtime())
(2018, 1, 12, 16, 18, 39, 4, 12)

I'm missing a day?

MrSurly · 2018-01-13T17:35:08Z

@Lisa999
I think this is an issue with RTC. Under the hood, I'm using C gmtime, and off-setting the year by 1930 (to get to year 2000). Python localtime is using a special function to deal with 2000 vs 1970, and I should probably be using that, because my way likely screws up a leap year somewhere.

I'll take a look tonight (PST).

MrSurly · 2018-01-14T01:10:54Z

@Lisa999
Fixed, squashed, pushed; give it a shot, please?

Lisa999 · 2018-01-17T11:41:22Z

git pull origin pull/3531/head
From https://github.com/micropython/micropython
branch refs/pull/3531/head -> FETCH_HEAD
Auto-merging ports/esp32/machine_rtc.c
CONFLICT (add/add): Merge conflict in ports/esp32/machine_rtc.c
Automatic merge failed; fix conflicts and then commit the result.

?

MrSurly · 2018-01-17T15:11:22Z

@Lisa999
You have a merge conflict. The file in question was changed by two or more different sources, and the changes are not compatible (same lines changed by each source). This usually happens when you edit a file, then attempt to pull or merge code that was modified by someone else.

Lisa999 · 2018-01-20T23:16:03Z

Ok, resolved (i hope)
Now i get this error:

machine_rtc.c:192:1: error: no return statement in function returning non-void [-Werror=return-type]
};
^
cc1: all warnings being treated as errors
../../py/mkrules.mk:47: recept voor doel 'build/machine_rtc.o' is mislukt
make: *** [build/machine_rtc.o] Fout 1

MrSurly · 2018-01-21T01:56:29Z

Ok, resolved (i hope)
Now i get this error:

The errors you're experiencing are unusual if you're simply trying to merge this PR into master. I think you have an XY problem. What are you trying to accomplish?

Lisa999 · 2018-01-21T11:03:50Z

git pull origin master
git pull origin pull/3531/head
And then build so that i have micropython ESP32 master + your PR for using the RTC on ESP32
Dump everything and clone again?

MrSurly · 2018-01-21T22:54:00Z

From scratch:

git clone https://github.com/micropython/micropython.git 
cd micropython/
git remote add mrsurly https://github.com/MrSurly/micropython.git
git fetch mrsurly 
git checkout -b my_new_branch
git merge mrsurly/dev-deepsleep

Lisa999 · 2018-01-24T23:05:18Z

Thx! Worked like a charm!
After a 'git submodule update --init'
Everything is now working as it should be.
Bug solved

dpgeorge · 2018-02-14T13:05:22Z

@MrSurly thanks very much for persisting with this and getting it moved across from the old repo to this one. Much appreciated, and sorry for letting it sit for so long.

Most of it looks good, but the main two items to discuss are:

Licensing and copyright: can you please confirm that you and Tom Manning can release this under an MIT license and you didn't in any way use any other code (apart from referring to the IDF) to write this code.
Can I please bother you to split it up into a few commits, like: addition of machine.deepsleep and machine.sleep; addition of machine.RTC class; addition of esp32 module? Or whatever makes the most sense and whatever order of adding them makes the most sense. Thank you!

dpgeorge · 2018-02-14T13:05:59Z

ports/esp32/sdkconfig.h

@@ -23,7 +23,7 @@
 #define CONFIG_ESP32_DEFAULT_CPU_FREQ_MHZ 240
 #define CONFIG_ESP32_DEFAULT_CPU_FREQ_240 1
 #define CONFIG_ESP32_DEBUG_OCDAWARE 1
-#define CONFIG_ESP32_DEEP_SLEEP_WAKEUP_DELAY 0
+#define CONFIG_ESP32_DEEP_SLEEP_WAKEUP_DELAY 2000


What exactly does this do? What is 2000 measured in, seconds?

What exactly does this do?

Ensures proper wakeup

Discussion here: micropython/micropython-esp32#32 (comment)

2000 is now the default value (instead of 0) in the ESP-IDF as of 9 months ago: espressif/esp-idf@3c583a7

What is 2000 measured in, seconds?

Microseconds

Ok, thanks, understood.

MrSurly · 2018-02-14T16:35:39Z

Can I please bother you to split it up into a few commits, like: addition of machine.deepsleep and machine.sleep; addition of machine.RTC class; addition of esp32 module? Or whatever makes the most sense and whatever order of adding them makes the most sense. Thank you!

Certainly, but n the past, it has been requested (by pfalcon, IIRC) that these sorts of things be not split into multiple commits. I can split this up if desired, but would appreciate consistent policy.

Also not sure if you mean separate commits, or separate PRs for RTC / deepsleep ?

MrSurly · 2018-02-14T16:38:16Z

Licensing and copyright: can you please confirm that you and Tom Manning can release this under an MIT license and you didn't in any way use any other code (apart from referring to the IDF) to write this code.

Yes, this code was written by Mr. Manning and myself. You have my full release uner the MIT license. @manningt, You ok with this?

manningt · 2018-02-14T19:50:32Z

@MrSurly Yes - You have my full release under the MIT license and I didn't in any way use any other code.

dpgeorge · 2018-02-15T04:12:54Z

@MrSurly @manningt thanks for confirming re copyright/license.

I can split this up if desired, but would appreciate consistent policy.

Yes I understand it's not always consistent, sorry. In this case adding a new module (esp32) should be a separate commit. And ideally have a separate commit for a new class (RTC). They can just be separate commits in this PR here (no need at all for a new PR). (I would be happy to do the split myself when merging but then I couldn't retain you as the commit author.)

ports/esp32/modmachine*: RTC class, sleep impiementation ports/esp32/Makefile: Changes to support RTC/sleep ports/esp32/sdkconfig: Changes to support RTC/sleep

ports/esp32/machine_pin.c: esp32 wake-on-pin changes ports/esp32/Makefile: modified to support pin wake ports/esp32/mpconfigport.h: New top-level module for esp32 wake features

MrSurly · 2018-02-15T16:04:52Z

@dpgeorge Split.

dpgeorge · 2018-02-16T13:55:32Z

Thank you very much @MrSurly and @manningt! Merged in 44033a1 and abec47a with some very minor edits to whitespace formatting.

diginfo · 2018-06-01T11:59:07Z

This is also reported here: https://forum.micropython.org/viewtopic.php?f=18&t=4856

This is what I thought would be a simple toggling of the wake pin trigger edge, but it turns out not !

This code should get the current value of door pin, and set the wake from deepsleep edge based on the pin value high or low and then put the device into deepsleep awaiting the trigger edge.

There is a 10K pull-down on pin 34 and the switch pulls it high.

The purpose is to wake up and sleep again each time the interrupt is triggered on a different edge.

The time.sleeps() are just to give time to view the output and should not have any other effect other than de-bouncing, but given that the esp32 startup time is 2 seconds, debouncing should not be needed.

It appears to work the first time deepsleep is fired, but thereafter door.value() always reads 0 even though the switch is closed and pulled high !

boot.py

import machine as mc, time
time.sleep(1)
door = mc.Pin(34,mc.Pin.IN)
edge =  5- door.value()
edgeMc = {
  5:'WAKE_HIGH',
  4:'WAKE_LOW'
}[edge] or 'BAD'

print('door:{}, edge:{}, edgeMc: {}'.format(door.value(), edge, edgeMc))
door.irq(trigger = edge, wake = mc.DEEPSLEEP)

time.sleep(5)
mc.deepsleep(60000)

Initially the switch was low:
door:0, edge:5, edgeMc: WAKE_HIGH

Then switch was then enabled (high):
door:1, edge:4, edgeMc: WAKE_LOW

Then, without changing the switch position, it just continuously loops, apparently unable to read the value of door any more:

door:0, edge:5, edgeMc: WAKE_HIGH
door:0, edge:5, edgeMc: WAKE_HIGH
door:0, edge:5, edgeMc: WAKE_HIGH
door:0, edge:5, edgeMc: WAKE_HIGH
door:0, edge:5, edgeMc: WAKE_HIGH
door:0, edge:5, edgeMc: WAKE_HIGH

MrSurly · 2018-06-02T02:33:01Z

@diginfo Please open a new issue rather than necroposting here. I can take a look next week.

diginfo · 2018-06-02T02:46:04Z

Done: #3835

Singein · 2018-11-19T06:40:40Z

For this to work, you need to use machine.TouchPad().config() to configure the threshold; an interrupt handler is not required.

machine.TouchPad().config() What parameters are needed?
configure the threshold the threshold means what?

I'm a little confused.

MrSurly · 2018-11-19T17:25:22Z

When reading the touchpad, you generally just get a number. When the touch sensor is touched, this value changes. When that change crosses a certain value (the threshold), then it's considered to be triggered.

config() takes a single integer parameter setting the threshold. Generally, you'd read the touchpad value for both touched and not touched, and set the threshold somewhere between. The actual values will vary based on the physical design of your circuit.

Singein · 2018-11-20T03:55:03Z

em...Thank you for your answer.
I'm a newbie to MicroPython..So I still have the following confusion:

>>> import machine
>>> from machine import Pin
>>> t = machine.TouchPad(Pin(32))
>>> t.read()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ValueError: Touch pad error

Could you please help me again about Touc pad error?
I would be grateful if you could show me the sample code below.

MrSurly · 2018-11-20T17:26:31Z

@nickzoic
I'm getting this error as well with touchpads on pins 0, 32, 15, and 33 -- Is there something we're missing?

manningt · 2018-11-20T18:19:35Z

It works for me on all pins mentioned (0,15,32,33) with the latest release:

MicroPython v1.9.4-684-g51482ba92 on 2018-11-20; ESP32 module with ESP32
Type "help()" for more information.
>>> from machine import TouchPad, Pin
>>> t32 = TouchPad(Pin(32))
>>> t32.read()
1269

The ESP32 PCB I'm using is custom and none of the aforementioned pins are connected to anything. The pin read values returned are all different, but within the expected range.
Not sure why you get the error, but I posted this to show it does work.

MrSurly · 2018-11-20T18:51:51Z

Mystery Solved (probably)

Using MP @51482ba92568a9793fea34213ed74be850920a5a (HEAD as of today):

import machine
tp = {}
for i in (4,0,2,15,13,12,14,27,33,32):
 tp[i] = machine.TouchPad(Pin(i))
 
for i,t in tp.items():
 try:
  print (i, end=' ')
  print (t.read(), end = '')
 except:
  print ('error', end = '')
 finally:
  print()

Result:

0 error
32 error
2 290
13 409
14 436
4 260
12 428
27 428
15 error
33 error

Here's why

I've checked the schematic for my board:

GPIO0 is connected to an auto-reset circuit
GPIO32, GPIO33 are connected to pullups for I2C
GPIO15 is connected to a resistor / LED

MrSurly · 2018-11-20T18:52:48Z

@fuermohao

It's likely that the pin you're using on your PCB is already connected to something (likely internal to the PCB), resulting in an error (see above). You should use a different pin.

nickzoic · 2018-11-20T23:24:57Z

@fuermohao thanks for reporting this, although this might not be quite the right place for it ...

What board are you using, or if it is a custom board could you let me see a schematic?

I'll check it out, but the pins work with a free-running oscillator, so just about anything connected to them, even pullups/pulldowns and LEDs, will prevent the oscillation from happening at all. I'm not sure whether this will produce the error you mentioned, but I'll check ASAP & we can add this

EDIT: ... "& we can add this to the documentation", I meant to say :-)

Lisa999 · 2018-11-20T23:53:05Z

MTDO (gpio15) is a strapping pin used for timing of the SDIO slave...
Gpio0 is used for boot control...

Singein · 2018-11-21T03:00:02Z

@MrSurly @nickzoic
I created the new issuse #4313
Thanks for your contribution to the MciroPython community!

Singein · 2018-11-21T07:06:09Z

@MrSurly
final question: how can I wake up esp32 from touchpad?

esp32.wake_on_touch(True)
machine.TouchPad(Pin(32)).config(200)
machine.deepsleep()

These codes don't seem to work

MrSurly force-pushed the dev-deepsleep branch from f9869bd to c7bb301 Compare December 31, 2017 18:06

mactijn mentioned this pull request Dec 31, 2017

backport of MrSurly/micropython-esp32@dev-deepsleep #3529

Closed

MrSurly force-pushed the dev-deepsleep branch from c7bb301 to 668abb2 Compare January 14, 2018 01:10

MrSurly mentioned this pull request Jan 29, 2018

Implemented deepsleep, reset_cause and rtc functions memory, string a… micropython/micropython-esp32#224

Closed

nickzoic mentioned this pull request Jan 30, 2018

is deepsleep() anywhere near to being implemented ? micropython/micropython-esp32#32

Closed

dpgeorge reviewed Feb 14, 2018

View reviewed changes

MrSurly closed this Feb 15, 2018

MrSurly force-pushed the dev-deepsleep branch from 668abb2 to 42c4dd0 Compare February 15, 2018 15:17

MrSurly reopened this Feb 15, 2018

MrSurly added 2 commits February 15, 2018 08:02

ports/esp32/machine_rtc.*: Added RTC class to machine

44867d6

ports/esp32/modmachine*: RTC class, sleep impiementation ports/esp32/Makefile: Changes to support RTC/sleep ports/esp32/sdkconfig: Changes to support RTC/sleep

ports/esp32/modesp32.*: New top-level module for esp32 wake features

ab6544d

ports/esp32/machine_pin.c: esp32 wake-on-pin changes ports/esp32/Makefile: modified to support pin wake ports/esp32/mpconfigport.h: New top-level module for esp32 wake features

MrSurly force-pushed the dev-deepsleep branch from e344396 to ab6544d Compare February 15, 2018 16:03

dpgeorge closed this Feb 16, 2018

diginfo mentioned this pull request Jun 2, 2018

esp32 - Cannot Read Pin Value After Wake on Pin #3835

Closed

Singein mentioned this pull request Nov 21, 2018

Touch Pad Error #4313

Closed

ports/esp32:* Implementation of deep sleep, esp32 module, machine.RTC #3531

ports/esp32:* Implementation of deep sleep, esp32 module, machine.RTC #3531

Conversation

MrSurly commented Dec 31, 2017

Basic sleep

EXT0 wakeups, using the common Micropython idiom

Other changes to machine

The new top-level esp32 module

The new machine.RTC module

mactijn commented Dec 31, 2017

MrSurly commented Dec 31, 2017

MrSurly commented Dec 31, 2017

mactijn commented Dec 31, 2017

MrSurly commented Jan 1, 2018

mactijn commented Jan 1, 2018

Lisa999 commented Jan 6, 2018

MrSurly commented Jan 6, 2018

Lisa999 commented Jan 13, 2018 • edited Loading

MrSurly commented Jan 13, 2018

MrSurly commented Jan 14, 2018

Lisa999 commented Jan 17, 2018 • edited Loading

MrSurly commented Jan 17, 2018

Lisa999 commented Jan 20, 2018

MrSurly commented Jan 21, 2018

Lisa999 commented Jan 21, 2018

MrSurly commented Jan 21, 2018

Lisa999 commented Jan 24, 2018

dpgeorge commented Feb 14, 2018

dpgeorge Feb 14, 2018

Choose a reason for hiding this comment

MrSurly Feb 14, 2018 • edited Loading

Choose a reason for hiding this comment

dpgeorge Feb 15, 2018

Choose a reason for hiding this comment

MrSurly commented Feb 14, 2018

MrSurly commented Feb 14, 2018

manningt commented Feb 14, 2018

dpgeorge commented Feb 15, 2018

MrSurly commented Feb 15, 2018

dpgeorge commented Feb 16, 2018

diginfo commented Jun 1, 2018

MrSurly commented Jun 2, 2018

diginfo commented Jun 2, 2018

Singein commented Nov 19, 2018

MrSurly commented Nov 19, 2018

Singein commented Nov 20, 2018

MrSurly commented Nov 20, 2018

manningt commented Nov 20, 2018

MrSurly commented Nov 20, 2018

Mystery Solved (probably)

Here's why

MrSurly commented Nov 20, 2018

nickzoic commented Nov 20, 2018 • edited Loading

Lisa999 commented Nov 20, 2018

Singein commented Nov 21, 2018

Singein commented Nov 21, 2018

Other changes to `machine`

The new top-level `esp32` module

The new `machine.RTC` module

Lisa999 commented Jan 13, 2018 •

edited

Loading

Lisa999 commented Jan 17, 2018 •

edited

Loading

MrSurly Feb 14, 2018 •

edited

Loading

nickzoic commented Nov 20, 2018 •

edited

Loading