TV-B-Gone in MicroPython

[bikeNomad]

As a fun project, I wrote a version of the classic TV-B-Gone tool (which just emits lots of power-toggle IR remote codes for different TVs) yesterday.
The codes came from a universal remote control, so should be more up to date for those of us in North America.
https://github.com/bikeNomad/micropython-tv-b-gone

[Josverl]

Saleae's automation API and Python helped me automate the collection process. Then I had Cody write a Python script that could read the CSV files produced by the Saleae logic analyzer and convert them into a format that could be used by the ESP32

💡Do you think that approach could also be used to record and replay signals with reliable timing ? either on esp32 or rp2.PIO ?

[mattytrentini]

The ESP32 RMT module is designed to support that use case, though MicroPython doesn't have bindings for the receive side. Adding receiver support for RMT would make it a very accurate record/replay device - down to 12nS from memory.

Rp2's PIO is capable of performing similar tasks.

[ned-pcs]

@peterhinch 's micropython_ir library uses a simple scheme (GPIO interrupts) to capture timing. The timing isn't super critical, because typical remotes don't use crystals and can have variable timing (±10% at least). The shortest pulses that I know of are about 160µs long (RCMM, RECS80, LEGO). I will add a simple capture routine (and deep sleep, of course).

[KnorkFox]

Many IR-reciever work on a frequency of 36-40 kHz.

[sosi-deadeye]

Haha, very good. Good. Preventing others to watch TV is a good thing. It could save many brain cells.

For replay attacks this RMT could be helpful, if reading pulses were implemented: https://docs.micropython.org/en/latest/library/esp32.html#esp32-rmt

Maybe someone is willing to do it.
The docs of ESP32: https://docs.espressif.com/projects/esp-idf/en/stable/esp32/api-reference/peripherals/rmt.html#install-rmt-rx-channel

[peterhinch]

My micropython_ir library includes an experimental module which enables an unknown IR stream to be captured for replay.

[Josverl]

Thanks for sharing Peter.
I will take a look - when time allows.
I have been looking for a way to capture a (partial) DMX frame on the wire , and replay that on demand for HIL testing.
For DMX there is a standard - but most consoles I have seen have their own broader timings and interpretations.
adding multiple complete light consoles to the test was ... impractical.
Capturing and replaying different recordings of ~50us -- 22ms would help in testing.

[ned-pcs]

I've added a capture facility to my code that uses the same scheme as @peterhinch used: pin change interrupts collecting time deltas.
Unfortunately, the deepsleep() implementation on the ESP32-C6 is broken, so my device doesn't work (yet). But it should work on other ESP32 chips.

[ned-pcs]

As far as getting low jitter captures of shorter pulses (like the DMX activity), you'd be better off with hardware helping. I'd recommend using an RP2040 because its PIO peripheral is handy for this kind of thing. There is off-the-shelf firmware for the RP2040 for use with sigrok (https://github.com/pico-coder/sigrok-pico); see also sigrok-cli. This may help with your testing.

[Josverl]

Thanks for ideas and inspiration. Now to find some time...