MK-312BT Audio paths are the component values correct?

The MK-312 supports 3 audio modes I’m guessing

  • Audio1: Mic/Mono input sampled through the ADC
  • Audio2: Stereo Input sampled through the ADC
  • Audio3: Stereo input routed to a comparator / schmitt trigger and into INT0 & INT1

I Think!

Has anyone compared a physical MK-312BT to ‘the other one’ in a “side by side” comparison to see if these are functionally and performance wise identical ?
The component values in the Audio chain seem very high in the MK312 and simulations are not working.



AFAIK the LM324 model is correct, Tested it with some basic op-amp circuits and gave expected results.
But I am human and could have misread some stuff in building the audio path

Any input here is welcome.

Some testing: Injecting line level signals into the ‘Line in’ and observing.
Test setup Line IN left & right connected together, sine wave fed into Left & Right

To me this audio processing doesn’t look right, the lower half of the input signal is clipped
The DC blocking cap C26 is OK, Adding a positive or negative offset to the input signal makes no effect, as expected.

Changing the MA knob does change the amplitude and duty cycle of U2.7 and U2.8

can this behavior be reproduced?


The OpAmp is U2

U2. Pin7 ------------------------------------------------------- U2. Pin8
image image
Sine wave: 500hz, 0v Offset, 0.1Vp-p


image image
Sine wave: 500hz, 0v Offset, 0.2Vp-p


image image
Sine wave: 500hz, 0v Offset, 0.3Vp-p


image image
Sine wave: 500hz, 0v Offset, 0.4Vp-p


image image
Sine wave: 500hz, 0v Offset, 0.5Vp-p


image image
Sine wave: 500hz, 0v Offset, 0.6Vp-p


image image
Sine wave: 500hz, 0v Offset, 0.7Vp-p


image image
Sine wave: 500hz, 0v Offset, 0.8Vp-p



Test Setup

These component values match those of the original MK-312BT.
Look here: https://github.com/buttshock/buttshock-et312-schematics/blob/master/hardware/schematics/02_audio_serial.gif

Many thanks

I have doubts around C26 & C21 1uF is a bit high for an audio shunt

20hz Audio in
image

1khz audio in
image

That shunt cap seems to do a good job smoothing out the signal rectified by the high gain opamp.
Channel 2 is the ADC input to the MCU.
I’m confused!

looking at other sources, these values 1uF look correct!

Time to see why the simulation and hardware didn’t match, then try to resolve whats wrong in my head.

Simulation is now working as HW…

Guess I’m just a little thick!


Diode 1N4148 is a bit funny, might need a lower Vf…

Last thought on this.
the Schematic matches the original hand drawn version,
the simulation now matches the physical hardware. The question now, how does this compare with the ET312? Is it possible to check audio from a signal generator and confirm on the real hardware and see how it compares to the simulation and MK-312?

The big question though, is this the expected behavior?
Input a sine wave, output a rectified wave

Sine in = 900mVp-p results in a rectified ripple = of 80mV

Using a 5v ref and 10bit adc => 4.88mV / bit
80mv ADC Signal = 16 steps

seems incorrect.



NG Spice Audio file input to circuit
1st 10 seconds of Wait for It Mono DFAM.wav

6 seconds in the middle of Wait for It Mono DFAM.wav

I think the hand sketches are the circuit of the ET312 revision-d

In my opinion, D1 / RP3.1 / C26 and D2 / RP3.5 / C21 are rectifier circuits.
A DC voltage proportional to the input volume is applied to U1 Analog in.

Hi Sirius,
I think I’m now inclined to agree with you,
based on photos of the PCB, Hand-drawn schematics, simulation and actual behavior, I think this is the intended function.

That last 6 second clip shows the dynamic range of the signal presented to the ADC input. This is probably a non-issue, just my misunderstanding.

SF