Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Output 9 (GPIO 5) does not work on 0.15.0-b2 #3922

Open
1 task done
bandi13 opened this issue Apr 22, 2024 · 16 comments
Open
1 task done

Output 9 (GPIO 5) does not work on 0.15.0-b2 #3922

bandi13 opened this issue Apr 22, 2024 · 16 comments
Labels
bug confirmed The bug is reproducable and confirmed fixed in source This issue is unsolved in the latest release but fixed in master

Comments

@bandi13
Copy link

bandi13 commented Apr 22, 2024

What happened?

Simply updating to 0.15.0-b2 makes one of my panels no longer light up. I use GPIO 5 for panel 9 in a 2D matrix. Reverting back to 0.14.3 (or even 0.15-0-b1) fixes the issue. I use nine WS2812B panels in a 3x3 matrix. The connections are soldered on to a circuit board, so rewiring won't be the easiest. I use the WLED_N_ESP32.bin type for the firmware.

To Reproduce Bug

Ugrade the firmware to 0.15.0-b2. Completely reproducible by doing a "Manual OTA". The 9th panel stays lit with whatever was drawn on it last.

Expected Behavior

I expect the same configuration to work regardless of the version the firmware is running.

Install Method

Binary from WLED.me

What version of WLED?

WLED 0.15.0-b2 (build 2404100)

Which microcontroller/board are you seeing the problem on?

ESP32

Relevant log/trace output

No response

Anything else?

My suspicion is that some new feature is mapped onto GPIO5, but the release notes don't mention anything.

Code of Conduct

  • I agree to follow this project's Code of Conduct
@bandi13 bandi13 added the bug label Apr 22, 2024
@blazoncek
Copy link
Collaborator

There are no restrictions or default assignments for GPIO5. Please do a factory reset and configure the device from scratch, it may be a configuration issue.
You can backup configuration and presets from "Security & Updates" settings page.

If resetting the device has no effect on behaviour then please get a debug build (from @srg74 repository) and post boot sequence serial output. You will need to attach the device to computer using USB cable and use tool for dumping serial output (like Putty or other terminal tool).

@bandi13
Copy link
Author

bandi13 commented Apr 22, 2024 via email

@blazoncek
Copy link
Collaborator

blazoncek commented Apr 22, 2024

You need a debug build, not sources from his repository.

@bandi13
Copy link
Author

bandi13 commented Apr 22, 2024 via email

@blazoncek
Copy link
Collaborator

You are using Audioreactive usermod which prohibits more than 8 buses/outputs on versions prior to 0.15.0-b3.
Please use (unreleased ATM) 0.15.0-b3.

@blazoncek blazoncek removed the bug label Apr 22, 2024
@bandi13
Copy link
Author

bandi13 commented Apr 22, 2024 via email

@softhack007
Copy link
Collaborator

I just used the version you linked for the Debug version. Previously I was loading https://github.com/Aircoookie/WLED/releases/download/v0.15.0-b2/WLED_0.15.0-b2_ESP32.bin which exhibits the issue. I don't think that's the Audioreactive version.

I agree with the conclusion from @blazoncek - this line in the log

AR: Failed to initialize sound input driver. Please check input PIN

means you're running a firmware that includes the audioreactive usermod. In version WLED_0.15.0-b2 (or older 0.15 pre-releases), WLED with audioreactive will not drive any LEDs on output 9 or 10. Technically its a resource conflict on the I2S driver - outputs 9 & 10 need the I2S unit, however the unit is already "claimed" by audioreactive even when the usermod is disabled.

@softhack007 softhack007 changed the title GPIO 5 does not work on 0.15.0-b2 Output 9 (GPIO 5) does not work on 0.15.0-b2 Apr 22, 2024
@bandi13
Copy link
Author

bandi13 commented Apr 22, 2024 via email

@DedeHai
Copy link

DedeHai commented Apr 22, 2024

I just tried the latest 0.15 build (from source) and it outputs data on GPIO5.

@bandi13
Copy link
Author

bandi13 commented Apr 24, 2024

I tried on the latest 0.15 build from source as well, and the 9th panel does not work. I will try to put an oscilloscope on to the line. My thinking is maybe the timing isn't correct on it? This week I'm pretty busy, but I'll try to get to it one evening.

@blazoncek
Copy link
Collaborator

You can compile custom binary by switching to parallel I2S. Almost all necessary code exists in bus_wrapper.h. You will need to select I2S for the first 8 buses in such case.

@Konos3
Copy link

Konos3 commented Apr 24, 2024

I have the same issue with another data pin GPIO27, downloaded from https://install.wled.me/, not audioreactive and 9 led strips

@blazoncek
Copy link
Collaborator

@Konos3 please provide cfg.json and debug dump.

@bandi13
Copy link
Author

bandi13 commented Apr 26, 2024

I ended up hooking up an oscilloscope, and on my 9th panel (GPIO5) I don't get any output with the official https://github.com/Aircoookie/WLED/releases/download/v0.15.0-b2/WLED_0.15.0-b2_ESP32.bin release.
My cfg.json is here: wled_cfg.json. Though I don't think it's relevant since other versions work fine with the same configuration. I'll try to get you a debug dump when I can.

@blazoncek
Copy link
Collaborator

FYI found the bug. It was well hidden.

@blazoncek blazoncek added bug confirmed The bug is reproducable and confirmed labels Apr 26, 2024
blazoncek added a commit that referenced this issue Apr 26, 2024
- getPixelColor() for analog
- RMT channel (#3922)
@blazoncek blazoncek added the fixed in source This issue is unsolved in the latest release but fixed in master label Apr 26, 2024
softhack007 pushed a commit to MoonModules/WLED that referenced this issue Apr 27, 2024
- getPixelColor() for analog
- RMT channel (Aircoookie#3922)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug confirmed The bug is reproducable and confirmed fixed in source This issue is unsolved in the latest release but fixed in master
Projects
None yet
Development

No branches or pull requests

5 participants