This repository was archived by the owner on Sep 6, 2023. It is now read-only.
Repository navigation
This repository was archived by the owner on Sep 6, 2023. It is now read-only.
APA102 Driver support #121
Description
Hello, I have MicroPython v1.9.1-219-g3580284e installed at my ESP32 NodeMCU development board, but unfortunately I can't find APA102 library included altought it should be in ESP8266 1.9.1 microPython.
from apa102 import APA102
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
ImportError: no module named 'apa102'
how should I add it so I can use it with ESP32?
and one more question - is this implementation of apa102 using real HSPI (hardware SPI) or it just bit-bangs the bits?
Activity
dpgeorge commentedon Jul 5, 2017
The APA102 driver for esp8266 is very specific to that chip (it uses inline assembler and has specific timing). So it's not straightforward to port it to the esp32. It's anyway just a bit-banging implementation.
It would be easier to write a new driver from scratch for the esp32, in Python, using the SPI peripheral (SCK and MOSI lines).
Hokyjack commentedon Jul 5, 2017
Oh, thanks for the answer, I'll try to write my own driver then
mattytrentini commentedon Aug 12, 2017
Also see #159
mattytrentini commentedon Sep 3, 2017
I'm just wondering why the APA102 implementation was implemented with bit-banging on the ESP8266? I've read the conversation on PR1941 but, while there was some solid, animated discussion and excellent collaboration, I'm not sure why an SPI implementation was chosen?
Was it perhaps that hardware SPI wasn't available at the time? Or that bit-banging produced faster updates that the 80MHz SPI limit? I'm sure I'm missing something obvious!
My understanding was that it should be possible to write a pure Python APA102 driver that just takes in a
machine.SPIobject. This will work across all ports and be optimised where hardware SPI is available...mattytrentini commentedon Sep 3, 2017
Oh, and @Hokyjack, are you still looking at creating an APA102 driver? If so, you should also be aware of the end-frame issue in the ESP8266 implementation, documented in PR3037: APA102 LED "n+1" is lighted on write().
mcauser commentedon Sep 5, 2017
Mention me (@mcauser) if you want me to test any code on my ESP32 (lolin32) and APA102 strip.
dpgeorge commentedon Sep 5, 2017
Yes it should work, even at low SCK frequency. But note that the SPI will use up 3 pins on the MCU whereas the APA102 only needs 2.
mattytrentini commentedon Jan 13, 2019
With the micropython-dotstar library we can probably close this ticket? @mcauser if you could test that library I'd really appreciate it!
Also posted on the forum, for reference.
BEMNETADEFIS commentedon Jan 15, 2019
mattytrentini commentedon Jan 28, 2019
@BEMNETADEFIS the current design actually works for Linux right now. The gotcha is that it requires an object (
spi) to handle the SPI comms - to do this it just requires two commands;writeanddeinit. It should be a simple matter to create a class that forwards these methods on to an underlying implementation.Unfortunately, as I understand, SPI comms is non-standard in Linux right now. But if you have a way of sending SPI data in Python this library could use it with minimal fuss.
RobertJBabb commentedon Jan 29, 2019
This is probably only of historical interest now, but I developed a MicroPython APA102/DotStar support library as well over the past few weeks: https://github.com/RobertJBabb/micropython_generic_apa102
The library supports the Adafruit DotStar API, the MicroPython ESP8266 APA102 API, and I added support for Matt's MicroPython SPI DotStar API after I saw his post. I focused on backward compatibilty instead of performance, so I expect that my class will be slower and consume more memory than Matt's. The repository also includes two test files that contain more compact examples of common DotStar demo code than I have seen elsewhere, perhaps that may be of interest to some.
I like Matt's implementation so I will probably use his library going forward.
As a newbie, what is the best way to coordinate development work with other MicroPython contributors?
I enjoyed writing this library but I would like to avoid duplicating effort in the future.
dpgeorge commentedon Jan 31, 2019
It's best to start by discussing on the https://forum.micropython.org
BEMNETADEFIS commentedon Feb 3, 2019
mattytrentini commentedon Feb 4, 2019
@RobertJBabb As Damien mentioned, raising a topic in the forum is the best bet. Sorry I hadn't stumbled on your effort but I'll be sure to take a look at your implementation!
@BEMNETADEFIS To allow the APA102 driver to work on Linux you just need to come up with a Python class that can satisfy the SPI requirements of the APA102 driver. As mentioned, that means you just need a
writeand adeinit(and even the latter is not strictly necessary if you don't ever return the resources). Let me know if you need more assistance.BEMNETADEFIS commentedon Feb 8, 2019
BEMNETADEFIS commentedon Feb 8, 2019