Hi Frank,
Glad you have found my project helpful. I have not used the MAX4466.
I cannot remember why I choose the MAX9814. But reviewing my project I think it is related to the fixed 2VPP output with 1.25V bias. Using a 2.5V reference means the signal at the micro pin is between 0.25V & 2.25V. This reduces the number of available samples but means the signal will not swing beyond the range of the ADC.
Faced with the problem you are experiencing …
Try the MAX4466 direct to ear plug headphones, like what you get with a mobile phone. Usually they are high impedance and should produce some sound even at low level. This is what I used in developing my project.
I assume you have been able to record the sound. Use Audacity to check how the audio looks. It might give some clues as to what is happening. Clipping of an audio signal can produce some weird effects. Probably you have already done this.
Try the circuit with a MAX9814, it may just be the MAX446 that is not working correctly.
I note:
Nano works of a 5V supply.
3.3V is generated by the FT232R chip and limited to 50mA.
MAX4466 powered from the 3.3V pin on the Nano; audio swing from 0 to 3.3V, offset bias 1.65V.
ADMUX = 0x62, ADC uses Vcc as reference, 0 to 5V with 2.5V mid point.
(nano does not have seperate AVcc pin available, Vcc & AVcc are connected together)
Maybe you could try powering the MAX4466 from 5V, that would put the bias point at 2.5V.
When I was developing the project, bias other than mid point of the ADC produced some strange results. If I remember right the effect you describe I did experience.
Anyway, all the best
Cheers
Jim
