Re: mode2 shows output, irrecord fails (using the remote out port of an amplifier)

On 2020-03-06 08:54, Robert Jäschke wrote:
> Dear Matthias,
> 
> Am 05.03.20 um 23:00 schrieb Matthias Reichl:
>>> Is there any way to invert the signal on the software side? The gpio-ir
>>> kernel module in 4.19.66-v7+ apparently does not have a parameter for
>>> that
>> I've pushed a PR to make the signal polarity configurable in the
>> gpio-ir device tree overlay:
>>
>> https://github.com/raspberrypi/linux/pull/3490
>>
>> I also posted an answer in the thread on the RPi forum and attached
>> the enhanced gpio-ir.dtbo file there (just copy it to /boot/overlays)
>>
>> https://www.raspberrypi.org/forums/viewtopic.php?p=1622026#p1622026
> 
> Thank's a lot! It works like a charm. Problem solved. :-)

Nice that it works now (I have not tried it myself). My gutfeeling is 
that is would probably be best to turn that driver into a /dev/lirc 
device, but I am not the expert here. Please keep us posted.

@Hias:
nice work. But for the record: I checked the Marantz "remote out" some 
15 years ago, long before the RPi was invented. It *is* active low -- 
there is no "likely" in the equation.

 >> BTW, irrecord is a really not state-of-the art. Please use
 >> IrScrutinizer, which runs natively on the RPi. Although in this case it
 >> would not have helped you, because the driver for reading GPIO inverted
 >> is still not written...

 > Good to know. I have tried irscrutinizer before but I have always
 > received an OddSequenceLengthException.

IrScrutinizer wants signals to start with a pulse and end with a space, 
i.e. be of even number of duration. Lirc tends  to leave out an ending 
space. Anyhow, IrScrutinizer should be more forgiving in the current 
version (2.2.4). If problems occur with the current version, please 
report as bug on GitHub: 
https://github.com/bengtmartensson/IrScrutinizer/issues .

Greetz,

Bengt