Please specify "FQBN" (Fully Qualified Board Name). It is the first line in the compiler output in the Arduino IDE 2.x. E.g. "esp32:esp32:featheresp32:FlashFreq=80". The FQBN starts with the core name, which is essential for bugfixing. For example, at least two prominent cores are available for the Bluepill board, namely STMicroelectronics and STM32duino. Without the FQBN we are most likely not able to fix the the issue.
atmega1284
The board manager URL used for exotic cores.
No response
What IDE are you using?
Other - please specify below
What IR protocol are you using?
RC5, RC6
Pin(s) used for IR-receive, if not default.
Arduino pin number: 14
Example(s) you have checked while hunting the bug.
Example(s) to reproduce the issue.
The library version you are working with.
What are the steps to reproduce this issue?
I just attempted to upgrade a sketch that works with 4.5.0 to 4.6.0. I saw the new deprecated warning and made this change:
< IrSender.sendRC5ext(device, obc, false);
---
> IrSender.sendRC5(device, obc, NO_REPEATS, false);
(sendRC5ext() does not appear to repeat commands so NO_REPEATS looked appropriate.)
What happens?
The device (a siriusxm onyx plus) no longer obeys any IR commands after building with 4.6.0 and sendRC5().
I have another sketch that decodes IR commands (atmega1284p @ 16 MHz) that still uses 4.5.0. It correctly decodes both versions (sendRC5ext() and sendRC5()).
The output which indicates the error happened.
What were you expecting to happen?
No response
Additional context.
I updated the FreeBSD arduino-irremote port from 4.5.0 to 4.6.0 today. I build my sketches with make from the command line (no ide).
I am using arduino 1.8.19 (i.e. the most recent 1.X).
I am using an atmega1284 clocked at 20 MHz.
I have a different sketch (atmega328 @ 16 MHz) that sends commands to the onyx plus and works with 4.6.0.
I found that the atmega1284 will work every other attempt if I enable toggle:
IrSender.sendRC5(device, obc, NO_REPEATS, true);
Looking at the code in sendRC5ext(), I see toggleBit defaults to "1" when toggle is false. I thought this might explain why using toggle=1 with sendRC5() works every other time.
I verified (on the atmega1284 @ 20 MHz) that delay(1) takes ~1010 us and delayMicroseconds(1) takes 3-6 us (pretty much the same with/without a delay).
I wonder if the non-default F_CPU factors into this issue?
Final checklist for the bug report.
Please specify "FQBN" (Fully Qualified Board Name). It is the first line in the compiler output in the Arduino IDE 2.x. E.g. "esp32:esp32:featheresp32:FlashFreq=80". The FQBN starts with the core name, which is essential for bugfixing. For example, at least two prominent cores are available for the Bluepill board, namely STMicroelectronics and STM32duino. Without the FQBN we are most likely not able to fix the the issue.
atmega1284
The board manager URL used for exotic cores.
No response
What IDE are you using?
Other - please specify below
What IR protocol are you using?
RC5, RC6
Pin(s) used for IR-receive, if not default.
Arduino pin number: 14
Example(s) you have checked while hunting the bug.
Example(s) to reproduce the issue.
The library version you are working with.
What are the steps to reproduce this issue?
I just attempted to upgrade a sketch that works with 4.5.0 to 4.6.0. I saw the new deprecated warning and made this change:
(sendRC5ext() does not appear to repeat commands so NO_REPEATS looked appropriate.)
What happens?
The device (a siriusxm onyx plus) no longer obeys any IR commands after building with 4.6.0 and sendRC5().
I have another sketch that decodes IR commands (atmega1284p @ 16 MHz) that still uses 4.5.0. It correctly decodes both versions (sendRC5ext() and sendRC5()).
The output which indicates the error happened.
What were you expecting to happen?
No response
Additional context.
I updated the FreeBSD arduino-irremote port from 4.5.0 to 4.6.0 today. I build my sketches with make from the command line (no ide).
I am using arduino 1.8.19 (i.e. the most recent 1.X).
I am using an atmega1284 clocked at 20 MHz.
I have a different sketch (atmega328 @ 16 MHz) that sends commands to the onyx plus and works with 4.6.0.
I found that the atmega1284 will work every other attempt if I enable toggle:
Looking at the code in sendRC5ext(), I see toggleBit defaults to "1" when toggle is false. I thought this might explain why using toggle=1 with sendRC5() works every other time.
I verified (on the atmega1284 @ 20 MHz) that delay(1) takes ~1010 us and delayMicroseconds(1) takes 3-6 us (pretty much the same with/without a delay).
I wonder if the non-default F_CPU factors into this issue?
Final checklist for the bug report.