analogRead() returns bogus values when i2s DAC running

[Harvie]

Hello, i have issues with analogRead() when running DMA for I2S DAC. To be honest i don't understand why running external DAC messes up internal ADC. But it surely is problem:

static const i2s_port_t i2s_num = I2S_NUM_0; // i2s port number

  static const i2s_config_t i2s_config = {    
    .mode = (i2s_mode_t)(I2S_MODE_MASTER | I2S_MODE_TX /*| I2S_MODE_DAC_BUILT_IN*/ ),                                                                                                                                                            
    .sample_rate = SAMPLE_RATE,                                                                                             
    .bits_per_sample = I2S_BITS_PER_SAMPLE_16BIT,  // only the top 8 bits will actually be used by the internal DAC, but using 8 bits stra> 
    .channel_format = I2S_CHANNEL_FMT_RIGHT_LEFT,  // always use stereo output. mono seems to be buggy, and the overhead is insignifcant o> 
    .communication_format = (i2s_comm_format_t)(I2S_COMM_FORMAT_I2S | I2S_COMM_FORMAT_I2S_LSB),  // this appears to be the correct setting> 
    .intr_alloc_flags = 0, // default interrupt priority                                                                                    
    .dma_buf_count = 4,    // 8*128 bytes of buffer corresponds to 256 samples (2 channels, see above, 2 bytes per sample per channel)      
    .dma_buf_len = BUF_LEN,    
    .use_apll = false         
  };   

i2s_driver_install(i2s_num, &i2s_config, 0, NULL);   //install and start i2s driver

//i2s_adc_disable(i2s_num);  //Tried this, did not helped.

//So far i've only tested this issue on following three pins:
analogRead(2); //always returns 4095 no matter what signal on pin
analogRead(4); //always returns 4095 no matter what signal on pin
analogRead(35); //returns correct value measured by ADC on pin 35

Is there any chance to get proper ADC readings on all analog pins without shutting down I2S entirely? (this would cause audio glitching, so it is not possible)

[Harvie]

Problematic analog pins 2 and 4 are ADC2, while working pin 35 is ADC1, so i guess this affects whole ADC2.
BUT! When i call adc2_get_raw(...) directly, it does work without problem. So this really has to be a bug in arduino-esp32 core. (I have version 1.0.4 installed)

[stale]

[STALE_SET] This issue has been automatically marked as stale because it has not had recent activity. It will be closed in 14 days if no further activity occurs. Thank you for your contributions.

[stale]

[STALE_DEL] This stale issue has been automatically closed. Thank you for your contributions.

[Harvie]

Not sure if this stale bot thing is useful... "let's just pretend problem doesn't exist because it was reported more than 5 months ago". Also the first notice is considered a spam by gmail, so there wasn't even a warning.

[me-no-dev]

There are many issue reports that are left abandoned by their submitter. The stale bot is there to take care of those. It's enough to just chime-in a message if the bot puts "stale" flag and the issue will be un-staled. We will also get an email that an issue is still active, which let's us know that the problem described persists and that there is someone on the other end.
Unfortunately it not very liked by active folks :) hopefully soon we can turn it off

[stale]

[STALE_CLR] This issue has been removed from the stale queue. Please ensure activity to keep it openin the future.

[stale]

[STALE_CLR] This issue has been removed from the stale queue. Please ensure activity to keep it openin the future.

[stale]

[STALE_SET] This issue has been automatically marked as stale because it has not had recent activity. It will be closed in 14 days if no further activity occurs. Thank you for your contributions.

[Harvie]

bump