ackoDAC based on ES9018

[acko]

Thanks everyone for all the feedback and comments. I will be updating with more details on how to put things together to get a complete system. Lots more assembly and testing being done at the moment to get this ready for release.
Special thanks to Jkenny (John) who helped to co-design the DAC module. John will help with support for nearby regions and is looking at battery psus and usb audio receiver for this DAC. Also appreciate the contributions of 'Qusp' who will be assisting with IV stages as well as detailed tests/comparisons with other DACs. Hope to hear his objective comments. So this project has become a truly collaborative effort and we plan to take this right up to a finished 'product' that would rival the best of commercial offerings!

Now on the main DAC module the design is modular from ground-up with various options. The DAC board has only the Sabre chip and we decided not to crowd it psu, mcu etc, leaving the chip to roam around freely. The layout is therefore simple 2 layer with all straight connection for signals and power. Also, all the power lines to the chip are brought out for independent supply feeds. So you have:

AVCC_L (3.3V)
AVCC_R (3.3V)
AVDD_L (1.2V)
AVDD_R (1.2V)

DVDD (1.2V)
DVCC_T (3.3V)
DVCC-B (3.3V)

VOSC (3.3V)

All supply lines are decoupled close to the chip with COG ceramics with additional Polymer filter caps for noise suppression.

The Analog supplies requires ultra low noise regulators for best performance.
I have shown a number of regulator modules that has been specially designed for this DAC. They will plug in directly. Will provide details on these soon. If you prefer you could also use others like Paul Hynes super shunts:
High performance regulators for Sabre DACs

Note: only the Analog supplies (and VOSC) are critical. The digital ones will do alright with LT1763 IC regs or similar, so there can be some cost savings here. You will need matching PSUs to feed the regulators.

There is an on-board mounting for an smd XO like Crystek 950 series and the frequency option will be based on source. If feed is from standard 44KHz 16 bit CD audio then a 40-50Hz on is sufficient. For 192Kz, requires 80MHz. If you are planning to play around with DXD audio then consider the maximum 100MHz option to give up to 500KHz capacity. There is an option to pipe in an external Master Clock instead. The smd coax connection (marked CLK under X1) and will take the Hirose connector/cable assembly (PNCL331-0452-5).

Hope this helps. More updates to come.

Regard, Acko

[qusp]

excellent mate, looking forward to running it through its paces. buying some O-Netics trannies too from another member by the looks of it.

also this looks rather interesting
VIA Vinyl EnvyUSB 2.0 Audio Controller - VIA Technologies, Inc.

the Via 'Envy' USB 2.0 audio receiver/controller. will do 8 channels of 24/192; has i2c, i2s, spdif and too many other functions to name. its an audio/visual product from the ground up, designed for hidef production systems. I have emailed them requesting the datasheet and spec.

[labjr]

Originally Posted by qusp

excellent mate, looking forward to running it through its paces. buying some O-Netics trannies too from another member by the looks of it.

also this looks rather interesting
VIA Vinyl EnvyUSB 2.0 Audio Controller - VIA Technologies, Inc.

the Via 'Envy' USB 2.0 audio receiver/controller. will do 8 channels of 24/192; has i2c, i2s, spdif and too many other functions to name. its an audio/visual product from the ground up, designed for hidef production systems. I have emailed them requesting the datasheet and spec.

I think that's made more for home theater receiver, external sound card or other consumer products than audiophile DACs Since it uses PLLs etc. No mention of asynchronous transfer.

[qusp]

and that matters how? it has the capabilities we need for multichannel hardware with the interfaces/formats all covered. who gives a toss if it has asyncronous transfer or not is they can get all the bits there on time without any missing the bus. as far as i'm concerned async is simply another buzz word. provided jitter is under control (yet another buzzword) and there is a means of communicating this clock to the dac well, I really dont mind how its done. async has its place yes, it sells units quite well and it does allow a high quality external clock to be substituted, but the need for this is largely to do with the fact that the clock in the signal is bad. if the embedded clock is of high quality what does it matter if we have to use it? sure i'll wait till I know more about it before I make a judgement call, but so far it sure sounds very interesting

[qusp]

apologies for some snippy comment above, was not meant to be personal, but I just get a little annoyed when things are discounted because they arent in keeping with the latest audiophile trends. just as the mass market is manipulated by the large scale marketing strategies, as is here in many cases where we are led to believe that if something doesnt have all the magic boxes ticked its dismissed. personally I find the biggest ut off with this is that it will most likely end up being a heavily controlled device due to having blueray support

[Doctorcilantro]

There is an option to pipe in an external Master Clock instead. The smd coax connection (marked CLK under X1) and will take the Hirose connector/cable assembly (PNCL331-0452-5).

Interesting! I have been looking for a DAC with AES input and BNC word clock input; why the Highrose? Is it 75ohm? Yes, I'm a luddite.

[qusp]

probably the Hirose because it fits and is the right part for the job. i'm sure its as close to 75ohms as possible, but the connector has to be soldered to the board, so it wont be perfect. it never is. perhaps you could look up the datasheet for the part number? i havent used it before O I dont know for sure

[pidesd]

Originally Posted by qusp

for sure, well the 1v2 analogue reference for the chip could indeed be quite important.

just bypassed the decoupling cap before vdd1.2 (after the 1.2v reg) on the buffalo 1.2 with a teflon ft-3 and it made a noticeable difference

[qusp]

cool, enjoy

[pidesd]

i dont think i m going to keep it like this though, i just wanted to mention the 1.2v analogue is probably indeed well worth the trouble of a good supply as the little reg on the buffalo seems not what the sabre really deserves.