-
-
Notifications
You must be signed in to change notification settings - Fork 234
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
Can't run on webOS 7 LG OLED C2 #13
Comments
Hi |
Resetting (RST) esp32 definitely helps, it starts working only after reset. Enabling/disabling Led device doesn't help. I wonder where's the issue.
|
HyperHDR ESP handshake also resets the board using RST/dtr at the begining and set proper state. Don't know what is causing the issue. |
can reset workflow fail if in the beginning there's some messages from bootloader? |
I’ve managed to reproduce same issue on my laptop running mac os. |
In your logs the reset succeeded, there is an initial message from HyperSerialESP32 when the board with uploaded firmware starts again. Again, check the statistics when it doesn't work as it's described in ESP handshake help: it's important to see if there are any or if the communication is broken. It's a wide shot but if you have SPI led strip connected to GPIO0 change it to other GPIO (e.g. 4) GPIO0 could interfere in the boot process triggering the bootloader mode while restarting, but your HyperHDR logs doesn't confirm such scenario and it's depends on your electrical scheme how it's connected. |
I have quite simple wiring to gpio 4/2. It’s definitely esp initialisation issues. I’ve ordered couple of other boards just in case if current issue be is dodge |
So in both cases the issue with the board affects SPI LEDs but it's handled by Makuna neopixel library (older version because we can't use latest Arduino ESP32 yet due to some performance problem with the serial port in 2.x) and beyond possible diagnostic in this project: unless somebody has other ideas. Bootloader mode can be disabled by putting high (3.3v) on GPIO0 but verify it first because currently I have only mobile access to the internet. |
I don't see a source code now but I think HyperSerialESP32 is using HSPI method for SPI. you can try to change it to other method available for ESP32 and SPI in Makuna Neopixelbus library, compile the firmware to check if it helps |
I hooked up esp to my logic analyser and clearly see there's issue with esp firmware startup. After reset pulse I only can see messages from bootloader. But not And bootloader message doesn't look like it boot mode, usually when it's it boot mode then it prints something like |
Before that initial message completely nothing happens for SPI (for RGBW there is setup of the calibration), even neopixel is initialize later. Try to remove serial.flush in the setup sequence but it does make no sense since sk6812 is not affected. To "wake up" the serial communication after the reset you may need to set proper DTR/RTS but again: the same sequence works for sk6812. Test without LED strip connected to eliminate electrical provlem |
BTW esp32 boot is using 115200 if I remember correctly. Then HyperSerialESP32 switches to 2Mb: you may need to increase the resolution for the logic analizer. Otherwise it may be seen as a single pulse. |
Yep, I have 2 separate decoders enabled for same input. I've found another board with CH9102 chip, it has the same issue |
And what's common with this boards (the're different manufacturers and pcb layouts), that I can't flash them without manually grounding GPIO0. I also found a github issue with similar problem Xinyuan-LilyGO/LilyGo-T-Call-SIM800#139 |
Okay, just in case if someone else is looking:
https://discord.com/channels/407937994037919756/413498235307425802/1054412614396027011 |
With my board mh-et live I had to ground GPIO0 for the first time, to upload firmware with a bootloader. It came without it. Later when you have a bootloader you can upload/upgrade the firmware without using GPIO0, py-esptool can handle it. |
with cp210x driver and chip it's still same problem. Firmware doesn't start after rst pulse, only if I pre RST button manually
|
okay, for the future google people: https://discord.com/channels/407937994037919756/974942819707924511/1023685372905009233
|
I've soldered electrolyte cap between EN/GND. This allowed esp to start from tv off state. The only problem left is that if I manually disable LED device from Remote Control page it doesn't come back to life. |
Using longer reset pulse (150ms) solved issue with esp being in half booted state after reset. Now esp32 comes back to live after reset sequence. awawa-dev/HyperHDR#468 also fixes issue with CH9102 module |
@nurikk-sa could you check if above PR solves also your problem? |
It’s the same person as me. Sorry messing up with personal/work github accounts from different browsers |
Ah, OK then :) So I think this issue can be closed if it works for you and it's specific board/webos problem? I test PR later with my ESP8266/ESP32 boards collection but I think that increasing timing slightly should not be a problem (even if so, we could detect the board using productId/manufacturerId and apply longer pulse, I also consider such solution). Could you post your esp32 IDs because for now I want to just compare it with mine boards? |
Wrong board :/ |
I doubt that even 0.5s reset pulse would be noticed by anyone. Since it’s just during board initialisation. |
Board1: with ch9102x chip
|
Board2: with
|
board3: with cp1204, looks the same as board1
|
all board had reset issue with 100ms pulse and works fine with 150ms pulse |
Ok, thanks. The first looks like mine. It's not directly related (100 or 150ms doesnt make any difference) but with |
Tested two Esp8266 (mini and mini pro) and one generic ESP32 with ch340g: they also work fine with your PR. I have one issue with resetting ESP32 S2 Lolin Mini (refuses to reset itself) so maybe I'll experiment with that pulse time in future. PR can be merged now. Thank you. |
Any thoughts as to why Lolin Mini's aren't resetting properly? I think I'm running into this on my LG CX + Mini S2 setup. Changing the setup of the HyperLED output frequently logs up the RGB output until I sleep the TV and wake it up again. |
Hello! Thanks for a great project! :)
I'm trying to make it work on my rooted lg tv using esp32 dev kit with CH9102F usb2uart chip, which is supported natively by tv os.
But it doesn't work some how. In hyperhdr logs I can see that it successfully opens uart interface and manages to complite handshake, but after that nothing happens, no colour changes on led strip. I'm using latest v19.0.0.0beta2 hypehdr and v8.0.0.0 hyperserial driver.
I've verified that this board and hyperhr works using my laptop running mac os, everything works as expected but when board is hocked to tv it doesn't work. I've checked serial and clock pins using my scope, and there's not signals there
webosbrew/hyperhdr-webos-loader#9
The text was updated successfully, but these errors were encountered: