Distortion question

[mryndzionek]

I'm attaching a sample wav:
distortion.wav.zip

There is this strange distortion, like wind blowing on a microphone. Is this a sign of something getting overloaded?

[penfold42]

Maybe the AGC trying to track a fading signal?

If it's reproducable, try AGC fast and extra slow and see it that makes a difference.
I might capture some audio tomorrow of an obviously overloaded front end. I have multiple 5kW AM towers 1-1/4 miles from me than can light a blue LED using my EFHW

[dawsonjon]

Does it only happen with very strong signals? I haven't allowed for worst case growth in the FFT, but instead chose a scaling that seemed to avoid overflow even for strong stations (that saturate the op-amps). It's possible that there might still be some overflow with very strong stations. To check, you could try setting extra-bits to zero, this should avoid overflow even for worst case scenario.

[mryndzionek]

AGC - not much difference, extra-bits also. It happens on fairly strong signals (S7-S9), but seems to randomly go away. It's also on frequencies above 10-13MHz. Sometimes it sounds like a different voice "trying" to come through. I wonder if it's some FM 88-108 interference.

[dawsonjon]

Fair enough, I tend to find that interference from the oled display tends to worsen at higher frequencies, does it make any difference when the display is disabled? It couldn't hurt to try a broadcast FM trap if you have one, the Tayloe Detector is sensitive to interference at odd harmonics, and the 7th/9th harmonic would sit in the broadcast FM band at this frequency range.

…

On Mon, 30 Sep 2024, 15:50 mryndzionek, ***@***.***> wrote: AGC - not much difference, extra-bits also. It happens on fairly strong signals (S7-S9), but seems to randomly go away. It's also on frequencies above 10-13MHz. Sometimes it sounds like a different voice "trying" to come through. I wonder if it's some FM 88-108 interference. — Reply to this email directly, view it on GitHub <#102 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAFPFX6W4ZZZXZKP3P5A7LLZZFQIJAVCNFSM6AAAAABPDJTKOWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOBTGQZDQOJTHA> . You are receiving this because you commented.Message ID: ***@***.***>

[mryndzionek]

Fair enough, I tend to find that interference from the oled display tends to worsen at higher frequencies, does it make any difference when the display is disabled? It couldn't hurt to try a broadcast FM trap if you have one, the Tayloe Detector is sensitive to interference at odd harmonics, and the 7th/9th harmonic would sit in the broadcast FM band at this frequency range.
…
On Mon, 30 Sep 2024, 15:50 mryndzionek, @.***> wrote: AGC - not much difference, extra-bits also. It happens on fairly strong signals (S7-S9), but seems to randomly go away. It's also on frequencies above 10-13MHz. Sometimes it sounds like a different voice "trying" to come through. I wonder if it's some FM 88-108 interference

Now it has gone away and the sound is very good. Actually I am using a FM trap, but DIY and unmeasured, so I don't know if it works :)

[penfold42]

Now it has gone away and the sound is very good. Actually I am using a FM trap, but DIY and unmeasured, so I don't know if it works :)

Was that after my PR which disabled I2C writes if the display is off?

Do you have the band LPF on the front end ?

[mryndzionek]

Was that after my PR which disabled I2C writes if the display is off?

It was before. At night some stations sound perfect. I have some strange local interference. I also put my LNA into a metal box.

Do you have the band LPF on the front end ?

No, no LPF. Just a DIY FM trap.

[penfold42]

Many AM transmitters lower their power at night - maybe try an attenuator during the day to see if that helps?
Which reminds me of one of my todo items - control an external attenuator like a pe4302.

I wonder how hard it would be to implement overload detection. At the ADC its easy - count # of samples that are too close to 0 or 3.3v but I'd like to detect clipping that could be introduced by the opamp too...
This would become super useful if i play with an external attenuator to to AGC on an LNA