AK4191EQ

Looks like it could be the digital front end for something like AK4499, but apparently with enhanced specifications. Not a complete dac though. Made to be used with AK4498.

From a Japanese website and translated by google:
"AKM's new DAC may be the answer we expect. AK4191EQ and AK4498EQ split the traditional DAC's digital filtering & delta sigma modulation and analog parts to reduce the interference between the digital and analog parts. The general idea is similar to removing the current-voltage conversion with the AK4499EQ. The more scattered the better, moderately reduce the degree of integration to reduce signal interference. As for the indicator, it is inconvenient to directly compare with the AK4499EQ because it is split into two parts."
 

Attachments

  • img158331588837198.jpg
    img158331588837198.jpg
    93.9 KB · Views: 856
Last edited:
These are great parts - I've been introduced to the AK4498 a year ago by AKM japan and at the time the engineers where using it with a FPGA development board - we heard the combination verses there highest end device and I can attest that they sound VERY much better!! Which makes them VERY VERY good!!! :D

If the AK4191 is as good as the FPGA demo solution I heard then the combination its going to be very VERY good - maybe the best off the shelf DAC solution available today!!!

I'd be more then happy to design a PCB around them :)
 
Last edited:
Wow - exciting stuff!

John - you were very privileged getting the private demo.
I was planning to visit AKM in Tokyo this month but decided to postpone due to the Corona Virus.
I just got in-touch with AKM about these new chips & they are not targeted for MP until
Q2 2021, dam.
I have requested samples but I have a feeling it will be a long wait.
Love the concept of separating the digital & analogue sections.
It means you can really optimise the PSU's, layout etc to suit each job the best.
Cheers Clay
GIESELER AUDIO
 

Terry Demol

Member
2002-04-07 3:12 pm
*
Wow - exciting stuff!

John - you were very privileged getting the private demo.
I was planning to visit AKM in Tokyo this month but decided to postpone due to the Corona Virus.
I just got in-touch with AKM about these new chips & they are not targeted for MP until
Q2 2021, dam.
I have requested samples but I have a feeling it will be a long wait.
Love the concept of separating the digital & analogue sections.
It means you can really optimise the PSU's, layout etc to suit each job the best.
Cheers Clay
GIESELER AUDIO

Clay,

Agreed. Also , having the opportunity to tap earlier into the signal at a point of greater DR and lower distortion can be an advantage. (SN / THD = -153dB).

TCD
 

Terry Demol

Member
2002-04-07 3:12 pm
*
That's still in the digital domain, I'm not sure you understand what you are talking about here. By that logic, AK4137 and CS8421 already allow this.

I know exactly what I'm talking about.

The ASRC's you have mentioned a/ OP I2S or in the 4137's case also DSD (1
bit) b/ remove jitter and actually encode it below the rejection CF (depending
on how they are configured).

The 4191 allows access to multi bit OP's after the modulator. We won't know
exactly what features that will include till a more detailed data sheet is
available but suffice to say this chip could drive a multi bit (unity weighted)
discrete DAC OP stage be it resistor or CCS.

It may end up having some advanced features such as RTZ encoding or
other algorithms to linearise the OP DAC. There are white papers (not akm)
kicking around on the various OP coding schemes.

TCD
 
I know exactly what I'm talking about.

The ASRC's you have mentioned a/ OP I2S or in the 4137's case also DSD (1
bit) b/ remove jitter and actually encode it below the rejection CF (depending
on how they are configured).

The 4191 allows access to multi bit OP's after the modulator. We won't know
exactly what features that will include till a more detailed data sheet is
available but suffice to say this chip could drive a multi bit (unity weighted)
discrete DAC OP stage be it resistor or CCS.

It may end up having some advanced features such as RTZ encoding or
other algorithms to linearise the OP DAC. There are white papers (not akm)
kicking around on the various OP coding schemes.

TCD

"Tap into the signal at a point of greater DR and lower distortion". Sure sounds like you were confused and talking about an analog signal.

Surely, you must know that the original digital data before it was passed through ANY of those blocks is "a point of greater DR and lower distortion". It would be possible to develop your own on a PC or in an FPGA and do better than AK4191. Claiming that processing that adds distortion is a point of lower distortion is nonsensical.

Further, without knowing what the contribution of the digital portion of integrated DACs are, how do you know it's lower than what is out there? What if CS43198 or ES9038 performed better?

Maybe you do understand and it's just poor wording / semantics. Either way, to me, the interesting part here is the AK4498, not the AK4191. The AK4191 is virtually useless without the AK4498. The AK4498 can now do what Mark has dreamed of and completely DIY the modulator in a clean fashion.
 
Last edited:

Terry Demol

Member
2002-04-07 3:12 pm
*
"Tap into the signal at a point of greater DR and lower distortion". Sure sounds like you were confused and talking about an analog signal.

No confusion here :)

Surely, you must know that the original digital data before it was passed through ANY of those blocks is "a point of greater DR and lower distortion". It would be possible to develop your own on a PC or in an FPGA and do better than AK4191. Claiming that processing that adds distortion is a point of lower distortion is nonsensical.

Are you just arguing for the sake of it? (probably... :) )

OK - I will spell it out in very simple terms.....
OP of DAC (chip) absolute best = THD -124, DR = -140 (mono)
OP of modulator THD / DR = -153dB

You got it?

Of course you can implement all this in an FPGA (or other), these are all
design choices... development cost, time etc etc.

Further, without knowing what the contribution of the digital portion of integrated DACs are, how do you know it's lower than what is out there? What if CS43198 or ES9038 performed better?

Well, we don't have access to their modulator OP's do we, so it's irrelevant.

Maybe you do understand and it's just poor wording / semantics. Either way, to me, the interesting part here is the AK4498, not the AK4191. The AK4191 is virtually useless without the AK4498. The AK4498 can now do what Mark has dreamed of and completely DIY the modulator in a clean fashion.

Of course.... exactly.

Just the same, I would suggest you study some whitepapers on various DAC
OP architectures, IOW the part after the modulator. Once you are a little
more familiar with them, their various strengths and deficiencies you will
understand a little better why it can be advantageous to 'break into the
chain'.

WRT 4498, interesting in that they have gone back to pre resistor weighted
architecture (4499) to the switched cap / cmos opamp style AKA4497.
It's certainly easier to work with.

TCD
 
Well, we don't have access to their modulator OP's do we, so it's irrelevant.



Of course.... exactly.

Just the same, I would suggest you study some whitepapers on various DAC
OP architectures, IOW the part after the modulator. Once you are a little
more familiar with them, their various strengths and deficiencies you will
understand a little better why it can be advantageous to 'break into the
chain'.

WRT 4498, interesting in that they have gone back to pre resistor weighted
architecture (4499) to the switched cap / cmos opamp style AKA4497.
It's certainly easier to work with.

TCD

I do understand, and I do understand the various architectures out there. Your initial statement is still incorrect and/or misleading.

Ah, so this is a regression back to AK4497's DAC section. No wonder it can't reach the same numbers as AK4499. Seems like a pointless exercise. It's also amusing to see people argue that the digital filter section and modulator is some kind of bottleneck when the factory measured difference between the two parts clearly demonstrates that the less ideal section of the DAC is the converter itself. Pay more, get less?

Don't worry, I'm sure the gurus will jump on it for this feature of questionable importance given the demonstrated performance ("isolation"). Some of the same gurus that denounce op-amps will then be pushing their "best sound ever" DAC with CMOS op-amps - wait I thought those were bad? :confused::p

Give it a year or two and AKM will release their AK4499 based version of the converter section and then that will be the new hot thing and the old one rubbish.
 
It's certainly easier to work with.
TCD

For sure, it is. But then, so is the integrated solution. I know I am coming across as a curmudgeon here, but this has all the hallmarks of a solution to a problem that doesn't exist.

Can you make a better DAC, in theory, if you remove some logic from it? Probably. Is this product a realization of that? I don't see it in the data they have revealed. Despite taking out what are computationally small blocks, they still have high speed digital I/O on the chip, and the interfaces are now running at much higher rates. Can they fab the DAC in a different process? Probably not. Can they fab the AK4191 in a better digital process at TSMC or similar now? Probably, but I don't see a concrete advantage anywhere but in the marketing aspect so far.

I am sure some guys at AKM will tell you this is the most amazing thing ever. Especially in Japan, you will never hear anything else anyway. Proof is in the numbers though, 98 is one less than 99 :D.
 

baten

Member
2019-05-14 10:39 am
I do understand, and I do understand the various architectures out there. Your initial statement is still incorrect and/or misleading.

Ah, so this is a regression back to AK4497's DAC section. No wonder it can't reach the same numbers as AK4499. Seems like a pointless exercise. It's also amusing to see people argue that the digital filter section and modulator is some kind of bottleneck when the factory measured difference between the two parts clearly demonstrates that the less ideal section of the DAC is the converter itself. Pay more, get less?

Don't worry, I'm sure the gurus will jump on it for this feature of questionable importance given the demonstrated performance ("isolation"). Some of the same gurus that denounce op-amps will then be pushing their "best sound ever" DAC with CMOS op-amps - wait I thought those were bad? :confused::p

Give it a year or two and AKM will release their AK4499 based version of the converter section and then that will be the new hot thing and the old one rubbish.

:D
Thumbs up

I am sure some guys at AKM will tell you this is the most amazing thing ever. Especially in Japan, you will never hear anything else anyway. Proof is in the numbers though, 98 is one less than 99 :D.

Proof is also in the pudding ;)
 
I know I am coming across as a curmudgeon here, but this has all the hallmarks of a solution to a problem that doesn't exist.
Amen brother.

It's not like audio DACs haven't basically been a solved problem for like... 20 years?

Mind you, I'm sure the developers of SDR equipment aren't unhappy.
Proof is also in the pudding ;)
Strictly speaking, the proof has never been in the pudding. In its original, uncorrupted form, this proverb said "the proof of the pudding is in the eating".
Do you think AK4191EQ converts pcm signals to 1 bit dsd at output?
The block diagram certainly wouldn't seem to suggest so (all the remains for the AK4498 to do would appear to be synchronous switched capacitors filter and other analog stuff). Why would one do something stupid like that anyway? Lipshitz/Vanderkooy showed that pure 1-bit streams have issues with transparency years ago.
 
Lipshitz/Vanderkooy showed that pure 1-bit streams have issues with transparency years ago[/url].

Then others came along and showed DSD is fine: AES E-Library >> Why Direct Stream Digital (DSD) is the best choice as a digital audio format.

Academic arguments aside, many listeners prefer the sound of their present day dacs when operating in DSD mode at higher sample rates such as DSD256 or DSD512. You saying you don't notice any difference with any of your dacs?
 
Last edited: