DAC 2496 (AK4393) DAC KIT With CS8416+AK4393+5532

on stage to provide Volume Control and buffering for a 125ASX2 Icepower module?

I must admit, I'm not sure I understand what the Servo stage will do for me - is it to reduce DC offset and remove the need for coupling capacitors? The input to the 125ASX2 already has an LPF with a coupling capacitor, so I'm not sure if the added complexity of the Servo stage is required/necessary/beneficial in this case, and it may be just as good to have a simpler design - thoughts?
 
I just received my DAC kit from China. It turns out I bought one with SPDIF and USB input (CS8616) and the DAC chip actually is an AK4398. The kit came with LPF values same as for CS4393 but looking at the 4398 data sheet CS suggests different values. Anyone has any experience there? Also the data sheet suggest COG caps and I can get some SMD COG caps to solder onto the board but I can also get THT polystryrene / polypropylene caps. Which way should I go?
 
My first try to get Ak4396 kit work in direct mood was a failure. But I keep trying by working on the volume control as Sandller advice. It is a shunt attenuator with 10k shunt resistor parallel. I replaced it with a 2.4k carbon foil same value from AK4396 datasheet. And it made the trix.
I replaced the 3 elctrolytics in power supply too, thanks to Erin. Three Mundorf 15000uf are playing the bass notes to Patricia Barber’s “Autumn Leaves”.

The sound is totally different vs the same board with a LME49720. I have made two, the only way to get own experience in this endless discussions.
Direct mood sounds very relaxed filling bass slam in stomach. Although rhythm and pace is there, more obvious, intimate yes sexy. Never experienced from a digital source before, maybe 12 years ago from a Linn top player.
The board with opamp sounds good. Every thing is there but fasten, stressed someone called it, hard not enjoyable as direct mood.

Although I did not put money on the coupling caps(suspicious at the beginning). 10uf Nichicon ES (says better than BG N serie) on 0.1uf pio CCCP to the attenuator and 10uf Nichi to the gnd. I guess the PIOs are not suitable in this bypass. Problem with big orchestral! Teflon may works better? Waiting from Russia by doller.
Looking to Jantzen Superior or Silver mkp as well vs Mundorf silver/oil witch costs double! Do not know if the they worth the price (2x €53 for 4.7uf)
Any suggestion that is easy to order from Europe?

Here you go. Good job. Finally someone with courage to go against the traditional engineering. I don't understand how people can be to stubborn
to try this 10 minute no cost tweak. The " I did it once on a dac and din't like
it crowd" don't know what they are missing with the AKM dacs.
.
.
 
I was getting excited that there was a cheap board with an AK4399 and you had made a typo..... oh well... perhaps in a few years :)

Actually, why do we need ready made dacs anyway. As soon as we get them
we start to take off half of the components and substitute them with better
ones. I don't even use the receiver of this AK4396 dac-board. Straight I2S
from my SD player to the dac chip. For better sonics I also bypass the whole
active stage. So what I am left with is just a board that holds the dac chip.
SSOP 28 boards with the AK4396 already soldered on them cost about $ 12
and the better SSOP boards that have near by decoupling possibilities cost
about $2 and the AK4396 from the same seller $5. Simple star earth wiring,
few .1 µf caps and Salas shunt regs will get you into sonic bliss rivalling
commercial offering even up to five digits price tags.
Same can be made for the AK4399 that I have seen selling for $23 on ebay.
Extracting the I2S signals from any gear is not rocket science. If the wires are
kept short which is easy to achieve with those little boards all is fine.
I am about to hook them up to a Envy24ht soundcard for pc crossover tests.
What I want to say is “why bother with complex if simple is way better?

Greets,
Klaus
 
What I want to say is “why bother with complex if simple is way better?

Sometimes it is nice to have a ready made board that you can get running in an hour, and then tinker with components until you get the sound you like. We are discussing a stand alone DAC here, not mods and additions to sound cards and CD players!

Your comments are somewhat valid if modding a CD player with pre existing power supply and i2s lines. If building a stand alone DAC the "simple" suggestion as you put it involves point to point wiring, SSOP to DIL mounted on veroboard connected to power supplies with more point to wiring etc. etc. Then you need to find a way to mount the jungle of wires and boards to a case. IMO not so fun. But whatever does it for you. It actually takes more time and effort than a PCB we can buy from ebay.

Sound wise, probably not a lot of difference between PCB or point to point. Depends on your wire length and layout. PTP is more hit and miss. PCB from ebay works well - guaranteed. Either way gets you better sound at a cheaper price than buying a brand name DAC from the shop. Its still DIY Audio!

Thanks for your comments, but possibly in the wrong thread. We are discussing a stand alone DAC. Not sure, but perhaps your comments are better in the Vendors forum if you are trying to spruik your wares???
 
Last edited:
Hi erin, I think you took my post the wrong way. I was not badmouthing this
dac as I bougt one myself to see what it's all about.
I just gave my logical conclusion based on what I did to get it singing.
Think about it. The active stage is just crap and can't compete with two
simple caps. That's not just a matter of taste, it's flat out missing on detail
separation of instruments and not close to the bass precision of the cap out
mod. So everything on the right side of the dac is not needed.
Then there is the regulators. Crap as well. So getting clean power to the
chips will take some effort. Not only that but the receiver needs clean
power too. But why do we need it ? So we can feed the dac I2S that the
source has converted to Spdiff.
Why in the world do we need too format conversions if we are dealing with a
board from the size of a wallet that can be put about one inch close to the
I2S source. No matter how much we tweak it, this Spdiff will not sound as
good as simple I2S transmission.
So taking all this into account. The SSOP 28 board with the dac on it will
sonically be better and it will not be more work to do it.

If one thinks he really needs Spdiff and already has this dac, then he can put
an AK4399 on an adapter board and connected it to the four resistors
feeding the AK4396.

Greets,
Klaus
 
Thanks for your clarification. I now read your comments sort of like musings on the meaning of life. Now I get you. :) Please forgive my heavy handed reply.

I have been back and forth with i2s. I used to like it but always had incredible interference with clicks and pops even when using very short wires. (6cm)
I don't understand why I got the interference when other people don't have this problem.
FWIW I was even running my source off a battery. I think it was RF interference. Mostly when the washing machine motor turned off.

i2s is in theory the best way to run digital signal to the DAC. I just think in practice, from personal experience, that it is problematic, and system and location dependent, as to the sort of sound you will get. In an ideal setup, results will certainly be better than a SPDIF receiver. In an unlucky setup (like mine) it will cause frustration.
 
Hi all,
I give a try with the TAMradio transfo i've bought from Ebay ...and nothing..only noise (As Nosian). Just to be sure the dac is working, i want to use a capacitor output (+R/+L) to 100R resistors , 3.3uF mkp caps ; the two rca grounds to ground somewhere on the board.
is this OK ? Where can i find a suitable gnd on the board ?
Thanks.
 
Hi all,
I give a try with the TAMradio transfo i've bought from Ebay ...and nothing..only noise (As Nosian). Just to be sure the dac is working, i want to use a capacitor output (+R/+L) to 100R resistors , 3.3uF mkp caps ; the two rca grounds to ground somewhere on the board.
is this OK ? Where can i find a suitable gnd on the board ?
Thanks.

For a trial just use any ground you can find on the board.
.
 
Thanks for your clarification. I now read your comments sort of like musings on the meaning of life. Now I get you. :) .

I am glad that my English isn't good enough to get the real gist of
this phrase.

I had also a few issues with clicks and static noise at first but after using
twisted pairs and close grounding of each line it was just fine. Better than
spdiff for sure. If you see how far I2S is routed on some boards the challenge is just to get the grounding right.
One of the best purchases I did for DIY electronics was a very fine silver wire.
A touch of a second with the iron and it sticks easily to any small smt resistor or cap. If I place my board close to where I want to extract the I2S lines from, I just stick the fine silver wire with a tweezers through the pins on my board and pull them tightly up to solder. After that they can be painted with a fine brush a few times so the paint will be an insulator. Then liquite silver can be applied to go from the ground on one board via I2S silver wires to the ground of the other board. That will do it just fine.
On the first pic you can see that my longest I2S wire to my sound-card is only
3cm. It was just a quick and dirty job with hot glue and my old AK4396 boards,
so don't look to close. The paint an the liquid silver has yet to be applied.

The second pic is my QA550 player with our dac in discussion.

Greets,
Klaus
 

Attachments

  • I2S out.jpg
    I2S out.jpg
    149.3 KB · Views: 874
  • QA550-AK4396.jpg
    QA550-AK4396.jpg
    148.1 KB · Views: 844
Last edited: