AK4556 - DC Coupled IO

Hardware Development
1

I’m working on a prototype Eurorack module using the Daisy Seed. Being only 4hp wide, and constrained for space, I was curious if the CODEC used here could be misused to handle DC coupled CV signals both on the input, and the output (not for V/Oct obviously, but for LFOs, envelopes, etc).

According to the AK4556 datasheet, DC output from the DAC should be possible with the digital LPF, but DC Input on the ADC side is not possible due to the enabled digital HPF. This filter could be disabled by pulling up pin CKS1 on the codec, but it is currently pulled down.

While it might be possible to remove and bridge the coupling caps, and maybe somehow pull CKS1 up, It would certainly be pretty tricky. While I understand the design decision here, I was wondering if:

  • there is another way to achieve this that I didn’t see
  • a future revision might make this an option.
2

How about splitting your signal at the input, sending a lowpass-filtered version to the ADC on the MCU?

3

Even if you do that then the delta-sigma ADC possibly won’t respond well at DC. Why not just use some of the ten available ADC pins ?

4

I know audio codecs don’t handle DC signals accurately. I’ve tested the CS42448 with the Teensy platform extensively in this regard. Despite this, there are a number of use cases (as I mentioned in my original post) where DC accuracy is not terribly important. AC coupling makes these use cases impossible, and my question revolved around the design decision to bake this into the hardware.

Also mentioned in my original post, my little project space constrained. Since I decided to include the two AC coupled codec inputs and outputs in my design. I am left with only two additional DC coupled CV inputs, and zero DC Coupled outputs. This limits the possibilities of what I might want to create upon this foundation.

It doesn’t matter much in this case. The Daisy offers DSP capabilities that eclipse pretty much any DC coupled use cases I can think of. I mean it would be a waste to make an envelope follower, or an LFO (no matter how fancy) with a platform like this. My question was simple curiosity regarding the design decision.

5

I would agree that forced AC coupling is a problem in certain cases. One thing that I can think of is using it for pitch CV in a VCO. Using ADC inputs for this is suboptimal.

I suspect that this might have been just an omission rather than conscious decision. Current codec is the third one among supported by different board revisions, at least one of the older ones had HPF that can be toggled in software (WM8371). So maybe the dev team haven’t considered that this would be an useful thing to add after switching codecs.