@Paradigm5h1f7 don't be surprised, just be ready to dig deap enough. G & H class are tied to power supply which increases efficiency, not directly tied to one or another traditional typology (A, AB or C or D) and can be based atop any mentioned.
You will find it and on power supply circuit in some cases (not directly on the amp circuit for example Hypex modules supplies). You will always recognise it by presence of charge pump. You can find it and on all newer QC Snapdragon SoC's.
Regarding dac's 24 bit integer is still a never reached stratosphere regarding SINAD but you have 64 bit floating point on your disposal (for EQ-ing and effects) on the digital side (and available commercially).
What I like about CS43131 design is it's claimed EMI rejection rate of 100 dB (that's a main reason for so many successful dongle alike designs that come very close to state of art desktop one's, G6 implementation isn't on such level only what should be considered as good from ASR point of view).
Best regards.
Very cool, I didn't know class H was used for Snapdragon, it makes sense. You would think they just go straight class D amplification on an SOC for cellular devices, or devices that need to maintain the utmost efficiency of battery life? I did a quick search and noticed Qualcomm are moving from audio codecs to a dedicated all in one 32bit DAC device as of last year, which is also cool! It is great seeing all this innovation in the Sound space again.
In regards to the G6, I am toying with the idea of designing my own amplifier, and integrated DAC. I like the idea of compact devices (within reason of course) sounding amazing. I would like to employ BASH, for small footprint with big sound, with a small area similar to the G6 for DAC.
In regards to the testing, I am an EE that really really believes simplicity always shines over complexity and am wondering where audio sits in this realm. I think this is why R2R DACs are so popular these days. If the components are selected properly, I don't see why a small footprint DAC couldn't perform equally as well as a larger DAC when shielding, power, LR isolation, and ground planes are added for noise rejection. USB DAC's I can't really get on board with because we do need a little more room for the previously mentioned. This is where the G6 32bit DAC peaked my interest. It is a small footprint, that looks to tik a lot of the boxes, and ASR seemed to back up my ideas.
If you are interested in how I came to my curiosity about the G6. I found the pictures of the guts on reddit and noticed the following.
- I noticed separate voltage conditioners for amplification, processor, and DAC. PSUs are wildly large in expensive DACs for no reason.
- There is isolation for the DAC with dedicated ground planes / components.
- The left and right channels have separate traces, with sufficient trace spacing to each other and the ground planes for a 5V signal, mitigating any EMI issues. You can see them in separate areas around the DAC, and to / from the OPAMPS for pre amplification. Larger PCB designs get carried away for trace spacing.
- The G6 DAC seemingly uses tantalum capacitors rather then the Polyester film stuff used in expensive DACS around the DAC chip, as well as the pre amplification of the OPAMPS. Capacitors around the DAC / OPAMPS, in my opinion, are meant for noise suppression of the internal components and device function such as oscillators, not audio shaping, so technically, the tantalum caps are vastly superior in every way in those areas.
-For audio shaping the G6 employs Polyester film caps into the final output amplification stage, which is where I would "shape" the audio signal if I designed something similar.
It is eerily close to something I would design, which has peaked my curiosity.
I plan on buying a higher end DAC to listen to, this is really just all experimental, and will be an interesting listen to see the real world results without me having to design something myself.
