Cayin N6iii: Nothing is Impossible with Gen2 User-replaceable Audio Motherboard

[Gold Fox]

Just a quick update on the R202 —

Last time I mentioned that

That’s because our goal isn’t simply to copy and paste the R01’s architecture onto the R202 — we’re aiming to push the boundaries and fully take advantage of the 31% larger PCB space offered by the Gen2 audio motherboard.

Just now I received the reference materials — and honestly, I was stunned. I always thought the engineer team was refining the R2R architecture or something along those line. But in reality, the R202 has pushed that extra 31% of PCB space to the absolute limit.

Suddenly, it all made sense — why it’s taken so long, and why it’s gone through so many revisions...

Feel free to take a WILD guess why I’m saying that
If anyone figures it out, I’ll reveal design details early. If not… well, I guess I’ll save the surprise for the official launch!

 

[WDitters]

Feel free to take a WILD guess why I’m saying that
If anyone figures it out, I’ll reveal design details early. If not… well, I guess I’ll save the surprise for the official launch!

R2R + Nutube? (and completely willing to sign an NDA )

 

[Gold Fox]

R2R + Nutube? (and completely willing to sign an NDA )

Nice guess — but nope

In fact, we didn’t go with that combo because of interference issues. I explained a bit of this in RU9's thread:

If there's any RU6/RU7 user, you may knew that both RU6 and RU7 followed discrete architectures — specifically, discrete R-2R and discrete 1-Bit respectively. The challenge with discrete circuitry, however, is its relatively low interference resistance. This is especially significant in compact designs like dongles, where the PCB space is extremely limited, discrete components tend to pick up more noise due to crossover and other factors. To address this issue, both the RU6 and RU7 adopted a vertical discrete structure, with components arranged on two separate PCBs to allow more space and minimize crosstalk. However, since the RU9's height is lower than both of them, this approach simply isn’t feasible here.

Generally speaking, discrete designs require longer signal paths and more open routing compared to integrated chips. In desktop applications, we can space out different boards to minimize crosstalk, but when it comes to dongles, it is inevitable constrained by size.

Now, imagine adding a Nutube 6P1 into the mix — another component known for being sensitive to noise and interference. Combine that with our support for MagSafe (which requires a rather strong internal magnet), it's essentially a 'perfect storm' for interference. It’s like keeping two sensitive components together in a small room——and then throwing an interference in there.

Oh, but you did actually get the one after R202 correctly. Following R202, we do have a tubey motherboard planned.

Now then… anyone else want to take a crack at guessing what makes the R202 so special?

 

[WDitters]

Nice guess — but nope

In fact, we didn’t go with that combo because of interference issues. I explained a bit of this in RU9's thread:

I was getting my inspiration from the upcoming iBasso D17 Atheris that combines both .. But your explanation from the RU9 thread does make sense ...

Then (next) guess: Did you by any chance manage to use all that extra space to actually combine R2R and 1bit on one PCB?

 

[Dron86]

Nice guess — but nope

In fact, we didn’t go with that combo because of interference issues. I explained a bit of this in RU9's thread:


Oh, but you did actually get the one after R202 correctly. Following R202, we do have a tubey motherboard planned.

Now then… anyone else want to take a crack at guessing what makes the R202 so special?

Maybe, R-2R and 1-Bit combo, switching between these 2?
Or r2r with high class amplification.
Don’t know what crazier to guess…

 

[drftr]

Maybe, R-2R and 1-Bit combo, switching between these 2?
Or r2r with high class amplification.
Don’t what crazier to guess…

Any of the above would do!

Never hurts to dream out loud.

drftr

 

[WDitters]

Another thing I noticed on the pictures was the CT7302PL chip that would indicate an i2s connection .. But not sure how much of a compelling feature that would be for the main target customer ....

Furthermore, the presence of a NJW1195A would indicate a digital volume control, something that Cayin originally was not able to implement on the RU6. It also is a 4-in 2-out audio selector .. Could that possibly point to a dual R2R / 1Bit design?

Or rather combined LO / PO capabilities? Between the Hiby RS6 and the Cayin R01 it was the RS6 that included proper LO capabilities ... and that player also included a NJW1195A ...

I guess I need more pictures to investigate @Gold Fox ...

 

[drftr]

Another thing I noticed on the pictures was the CT7302 chip that would indicate an i2s connection .. But not sure how much of a compelling feature that would be for the main target customer ....

Furthermore, the presence of a NJW1195A would indicate a digital volume control, something that Cayin originally was not able to implement on the RU6. It also is a 4-in 2-out audio selector .. Could that possibly point to a dual R2R / 1Bit design?

Or rather combined LO / PO capabilities? Between the Hiby RS6 and the Cayin R01 it was the RS6 that included proper LO capabilities ... and that player also included a NJW1195A ...

I guess I need more pictures to investigate @Gold Fox ...

I certainly wouldn't mind a much better volume control with way more/smaller steps; especially stepping up from 50 to 51, from 60 to 61, et cetera. Those were huge for some IEMs on RU6.

drftr

 

[justsomesonyfan]

Just a quick update on the R202 —

Last time I mentioned that

Just now I received the reference materials — and honestly, I was stunned. I always thought the engineer team was refining the R2R architecture or something along those line. But in reality, the R202 has pushed that extra 31% of PCB space to the absolute limit.

Suddenly, it all made sense — why it’s taken so long, and why it’s gone through so many revisions...

Feel free to take a WILD guess why I’m saying that
If anyone figures it out, I’ll reveal design details early. If not… well, I guess I’ll save the surprise for the official launch!

my bet is that is has 4 channels on the dac. Meaning double the resistors of last time, allowing it to be true balanced.

 

[WDitters]

my bet is that is has 4 channels on the dac. Meaning double the resistors of last time, allowing it to be true balanced.

Not sure .. .On the limited pictures that are available I was able to count 96 resistors which would indicate 2-channel 24bit ... But I agree, that would be awesome ....
However, since the CT7302PL is mostly associated with DSD setups, my bet is on a combined R2R+1bit design. With (at least) DSD512 as a possibility .. An R2R board with a touch of N7 so to speak, not unlike for instance the Gustard R26. If true, the R202 would definitely be the best possible front end to the Cayin C9/C9ii

 

[Gold Fox]

Any other guesses before we lift the curtain?

 

[justsomesonyfan]

Any other guesses before we lift the curtain?

no one take any more guesses, let the man lift the curtain!!

 

[drftr]

Any other guesses before we lift the curtain?

It comes for free if you have the 2 other cards already??

drftr

 

[Sweet2K]

Any other guesses before we lift the curtain?

It has a built-in heat sink and fan to keep it running cooler?

 

[justsomesonyfan]

It mines crypto. Mark my words.