Hi arivel,
No AK4499 bare board diy projects have been completed yet, although a few were started and then kind of abandoned.
There are a few commercial AK4499 dac products available at this time. The only one I have heard is Topping D90. IMO it sounds excellent for the price, but its not cheap at a little under $700.
It has been said that Acko dac has an AK4499 board too, but can't recommend it since I don't know enough about it.
Regarding the AK4499 chip itself, its a great sounding dac, better than what I have heard of ESS chips. That's my opinion only, of course.
No AK4499 bare board diy projects have been completed yet, although a few were started and then kind of abandoned.
There are a few commercial AK4499 dac products available at this time. The only one I have heard is Topping D90. IMO it sounds excellent for the price, but its not cheap at a little under $700.
It has been said that Acko dac has an AK4499 board too, but can't recommend it since I don't know enough about it.
Regarding the AK4499 chip itself, its a great sounding dac, better than what I have heard of ESS chips. That's my opinion only, of course.
Last edited:
Oops! Forgot about that.
Guess there are probably a few more such projects around the world, but not that are detailed in a thread here.
May I ask how much it would cost a diy'er like arivel to get something going with the AK4499 board you are using?
Mark,
I got more prompted from the fact that he is from Italy too, and maybe, although small chance, but maybe we could arrange for a listen somewhere, sometime.
Otherwise, I would say the kit is quite a challenge.
The starting price is very honest and good, but not exactly cheap. (it's on the homepage) Then a Bom should be completed, (only active parts are included) which adds up. Delivery from Japan should be organized, maybe with some proxy contacts.
The kit is bare boards, it means one has to manage to hand solder the (double) dac chips. Whew.
But Yanasan is great, precise, quick, professional, prompt. The BOM, if ordered from Japan Akizuki Denshi, is not really painful extra. Looking from Europe, it's a wee bit more complicated.
Manuals are Japanese. Forum too.
But I enjoyed it greatly, and the sound even more.
It can cost more at the end then a Topping, but it's a dual mono, master clock ready, fully galvanically isolated, 8 independent power supply feed, DSD direct-prepared, very nice complete design.
If I should restart, maybe I would chose a more minimalist, compact project, let's say in the direction of an SMSL m400..?
But It had cost me definitely less than that, for a project more feature rich. ~~~...
Where I can (and must) control a lot of vital aspects.
Ciao, George
ヤナソフトのホームページ
I got more prompted from the fact that he is from Italy too, and maybe, although small chance, but maybe we could arrange for a listen somewhere, sometime.
Otherwise, I would say the kit is quite a challenge.
The starting price is very honest and good, but not exactly cheap. (it's on the homepage) Then a Bom should be completed, (only active parts are included) which adds up. Delivery from Japan should be organized, maybe with some proxy contacts.
The kit is bare boards, it means one has to manage to hand solder the (double) dac chips. Whew.
But Yanasan is great, precise, quick, professional, prompt. The BOM, if ordered from Japan Akizuki Denshi, is not really painful extra. Looking from Europe, it's a wee bit more complicated.
Manuals are Japanese. Forum too.
But I enjoyed it greatly, and the sound even more.
It can cost more at the end then a Topping, but it's a dual mono, master clock ready, fully galvanically isolated, 8 independent power supply feed, DSD direct-prepared, very nice complete design.
If I should restart, maybe I would chose a more minimalist, compact project, let's say in the direction of an SMSL m400..?
But It had cost me definitely less than that, for a project more feature rich. ~~~...
Where I can (and must) control a lot of vital aspects.
Ciao, George
ヤナソフトのホームページ
Last edited:
I bought a D90 sounds good but no better then the dac I paid $249 or one i paid $420.
sellers of the D90 don't provide the drivers you need to download them and Zip them. easy enough , I had used up winzip free so had to pay $70 to use it. D90 I am sure is very good, but i can not hear any difference.
sellers of the D90 don't provide the drivers you need to download them and Zip them. easy enough , I had used up winzip free so had to pay $70 to use it. D90 I am sure is very good, but i can not hear any difference.
multi,
If you are looking to understand why you are not hearing differences between your dacs maybe we can help you sort through any problems with your system. Any interest in trying that?
If you are looking to understand why you are not hearing differences between your dacs maybe we can help you sort through any problems with your system. Any interest in trying that?
You don't have to pay to unzip files. There are plenty of free utilities. 7-zip is a good one.
Thought some people might be interested to know that Chinese USB boards are starting to appear that feature Accusilicon clocks (same AS318B model Topping D90 uses, except smaller case size), and that support ASIO up to PCM1536 and up to DSD1024.
USB digital interface AS318B PCM1536 DSD1024 compatible with Amanero Italy XMOS | eBay
However, there is a catch...
In order to use ASIO in Windows it takes a paid version driver, which given how things often tend to work in some places, the board maker did not purchase. What they did do was provide a program (that uses an un-signed driver they have) that can reprogram the USB board to emulate the VID and PID of some other USB board that does use paid drivers.
Moreover, instructions for configuring the boards is only written in Chinese language PDFs. None of the software and documentation described above comes with the board, the seller must be asked to provide it (although they don't tell you that).
Another thing about the boards is that they are USB powered, pretty much modeled after Amanero. Looks like the same method used to run Amanero from clean 5v power could also be used with the Chinese USB boards.
Like many other USB boards, all features of the new Chinese USB boards are available if using Linux or MacOS without playing any questionable driver games. Drivers are only an issue if using Windows.
USB digital interface AS318B PCM1536 DSD1024 compatible with Amanero Italy XMOS | eBay
However, there is a catch...
In order to use ASIO in Windows it takes a paid version driver, which given how things often tend to work in some places, the board maker did not purchase. What they did do was provide a program (that uses an un-signed driver they have) that can reprogram the USB board to emulate the VID and PID of some other USB board that does use paid drivers.
Moreover, instructions for configuring the boards is only written in Chinese language PDFs. None of the software and documentation described above comes with the board, the seller must be asked to provide it (although they don't tell you that).
Another thing about the boards is that they are USB powered, pretty much modeled after Amanero. Looks like the same method used to run Amanero from clean 5v power could also be used with the Chinese USB boards.
Like many other USB boards, all features of the new Chinese USB boards are available if using Linux or MacOS without playing any questionable driver games. Drivers are only an issue if using Windows.
Last edited:
This is the funniest thing i have read this month. And in a month, rich in hilarious international events this is saying something 😎
I don't see anything remotely attractive about these boards apart from the clocks. Is it known what voltage regulator feeds the clocks?
Easy to trace out the power distribution.
The very high sample rates possible with the boards may be of interest to some people.
The very high sample rates possible with the boards may be of interest to some people.
I meant the actual regulator chip.
May be of interest to some people but apparently we don't know any such people, right? 😀
Anyone with an ESS Sabre dac can potentially use the high sample rates. Up until now no USB board was available to support that.
Also, the actual regulator chip can be identified, however, it feeds more than the clocks. It also appears to feed the CPLD (or whatever it is) that divides the clock frequencies by 2 for the MCLK output pin. (that said, the configuration software may be able to change whether the clocks are divided, not sure how it works exactly).
The hot setup would probably be to remove the clocks and put them on the other side of an isolator, then send isolated copies of the clock outputs back to the USB board. Lots of work to do well. Either that, or maybe use a USB isolator box between the USB board and PC, if a good isolator of that type has been identified.
Last edited:
RE very high DSD rates. People should note that the AKM 4499 data sheet clearly shows that this chip performs better at DSD 256 than it does at DSD 512. I suspect the same is true for the ESS 9038 (and other variants as well) With my ESS 90381 DAC I could not discern any improvement with DSD 512 vs. 256, in fact, if anything 256 seemed a little more resolved and cleaner.
I have had some interaction with Andreas Koch re very high rates of DSD, and his opinion is that DSD 256 is about the highest rate it makes any sense to use. He explains that this is because of the limitation of the speed of the switches used for the conversions stage: he feels that at rates above DSD 256, the switches spend too much of the cycle time in the transition zones, and that the dSD 256 rate is the sweet spot where the advantages of the higher sample rate balance with the speed limitations of the switches.
I have had some interaction with Andreas Koch re very high rates of DSD, and his opinion is that DSD 256 is about the highest rate it makes any sense to use. He explains that this is because of the limitation of the speed of the switches used for the conversions stage: he feels that at rates above DSD 256, the switches spend too much of the cycle time in the transition zones, and that the dSD 256 rate is the sweet spot where the advantages of the higher sample rate balance with the speed limitations of the switches.
Which sample rate sounds best still seems to be very dependent on the particular dac architecture. In theory 16/44 or maybe 24/48 PCM should sound as good anything else. It just doesn't work that way in practice with most dacs. DSD1024 is said to give the best sound with Holo Spring dacs. For AKM dacs maybe DSD256 sounds best, but IME it seems to depend somewhat on which algorithms are used to produce the DSD256. Regarding ESS dacs, I prefered DSD512 to DSD256 when I had a chance to compare. Never had a chance to try DSD1024 with ESS, so no idea how it might sound.
Last edited:
Very good question. We know the presumably-real Accusilicon clocks in large cans (like D90 uses) are pretty expensive from European sources. However, I doubt Topping had to pay anywhere near as much as I did to get my hands on a pair.
well...
For the AKM 4499, one can see right in the data sheet that the specified performance of the chip is worse with DSD 512 vs. DSD 256. Now perhaps you are suggesting that some might prefer the sound of more noise and distortion? This appears to support Andreas Koch's position. Considering Andreas' experience level with DSD, and the audible performance of his (Playback Designs, and formerly EMM Labs) DACs, I tend to be more inclined to place more value in his opinion's and thoughts on the matter than those of a few audiophiles who may be impressed by very high numbers.
For the AKM 4499, one can see right in the data sheet that the specified performance of the chip is worse with DSD 512 vs. DSD 256. Now perhaps you are suggesting that some might prefer the sound of more noise and distortion? This appears to support Andreas Koch's position. Considering Andreas' experience level with DSD, and the audible performance of his (Playback Designs, and formerly EMM Labs) DACs, I tend to be more inclined to place more value in his opinion's and thoughts on the matter than those of a few audiophiles who may be impressed by very high numbers.