DIY ES9018 Hi-end USB DAC

Thanks for link.Its good that it has asynchronous USB.Also Windows driver supported.
So all I need to do is connect USB streamer to PC,then I2S out to BuffaloIII input and rest is done automatically.That is channels will be input to corresponding DAC channel(F/C/Surr) and get 5.1 analog out.Any more to consider?

You will need need 3 stereo I/V stages for 5.1.

As far as software is concerned have a look a Jriver Media Centre.
Free 30 day trial.

It will connect to the MiniDSP Streamer via ASIO and provides room correction and loads of other useful stuff.
 
Oh yes ! I2S

Remove LCD+ES9018 High-end USB DAC CPU Coaxial fiber 192K-24BIT KIT WLX | eBay

2013-08-19 15.58.53.jpg
 
where you got that pcb picture from? I don't see it in the actual ebay listing...it seems like the I2S traces to the DAC are on the backside of the PCB; maybe someone who already bought this PCB version could confirm

Could you confirm that this link
Remove LCD+ES9018 High-end USB DAC CPU Coaxial fiber 192K-24BIT KIT WLX | eBay
Does NOT work for you.
It appears thet Ebay links may not work internationally ?
The item number on ebay.co.uk is 200948482004

I purchased the board and took the picture.
 
thx for your reply, I could see the ebay listing but the picture with the bare pcb doesn't appear in the ebay listing that's what I said, I didn't know that you purchased the board and made the picture yourself... but this is great that you can confirm the I2S traces on this pretty cheap design which I think it's worth the money... I would be also interested of the sound quality impressions after you finish the assembling job - good luck!

p.s.
it seems like that pin from I2S connector which is not connected is the MCLK, so this device cannot be run in synchronous mode due to that onboard 100 mhz XO, unless one can inject the synchronous MCLK instead the onboard XO which would have to be disconnected then; so this design is only for asynchronous mode using the 80 or 100 mhz onboard XO; I would also attach my I2S source directly to those little 3 pins on the top side of the dac to keep the traces as short as possible ;
too bad that this pcb isn't available as bare pcb with the dac only soldered on it so that I could try a tube output, because I think that it's opamp output isn't a great performer...or at least I would try joe rasmussen's approach from this thread here

http://www.diyaudio.com/forums/digi...o-s-bdp105-discussions-upgrading-mods-65.html
 
Last edited:
Well my weiliang 9018 board is about 2 months old and has started making a very odd noise on one channel. It's a buzzing, starts quiet, gets louder, stops with a popp and then repeats over and over every 2 seconds.

If I touch the casing on one of the stereo opamps it gets quieter, but it doesn't go away. The USB board I got from the same supplier blew a track on the PCB which i had to repair, and now it sometimes doesnt get detected by the OS.

Overall not very impressed with this DAC. Support from the supplier was good until I mentioned the fault, now they are refusing to refund me, but I can get a replacement if I pay shipping back to China. They cannot however credit what I spent against anything else. Should have bought a buffalo IIIse.

AVOID. I've effectively lost £200.
 
everyone knows that the best place for the termination resistors is on the source side (ex. at the USB-I2S converter) and not at the DAC side; as for the long traces that's why I said that I would solder the 3 I2S signals directly to those tiny 3 little pads on the top side of the dac and cut the traces to the big connector to avoid making "antennas"; anyway the soldering job weillang has made to the ES chip is not one of the best ones, especially there where the I2S traces are going - looks like a mess there, you could check with a multimeter before powering up for any shorts on the pins, to be on the safe side
 
everyone knows that the best place for the termination resistors is on the source side (ex. at the USB-I2S converter) and not at the DAC side; as for the long traces that's why I said that I would solder the 3 I2S signals directly to those tiny 3 little pads on the top side of the dac and cut the traces to the big connector to avoid making "antennas"; anyway the soldering job weillang has made to the ES chip is not one of the best ones, especially there where the I2S traces are going - looks like a mess there, you could check with a multimeter before powering up for any shorts on the pins, to be on the safe side

Placing source (not termination) resistors at the DAC end is and attempt to match impedance of the line, to do this we need to know Zo of the line.

Experimenting with split termination at the 9018 end may be of benefit here.

Sure, I mentioned the 3 vias in post #122.

The soldering of the 9018 leaves a little to be desired but looks OK under the microscope.
 
Hi deanoUK, a little offtopic though, I saw on another thread that you received an XMOS board from a seller which didn't provide any windows drivers for it; so I wonder what driver are you using? is it the evaluation one from THESYCON with periodical beeps after every 30 min.? have you also ordered the PCM5102 daughter-board for it? I have also oreder one cheap XMOS+PCM5102 board from ebay (70$ with free shipping) and I received a full functional working driver together with ASIO feature but as of the PCM5102 I remember someone reported heavy clipping on it's output in case you fed it with very high level digital material and it's powered from 3,3V (and not 3,6V) so that's why I wonder about implementation regarding this eroneous functionality
 
Last edited:
Hi deanoUK, a little offtopic though, I saw on another thread that you received an XMOS board from a seller which didn't provide any windows drivers for it; so I wonder what driver are you using? is it the evaluation one from THESYCON with periodical beeps after every 30 min.? have you also ordered the PCM5102 daughter-board for it? I have also oreder one cheap XMOS+PCM5102 board from ebay (70$ with free shipping) and I received a full functional working driver together with ASIO feature but as of the PCM5102 I remember someone reported heavy clipping on it's output in case you fed it with very high level digital material and it's powered from 3,3V (and not 3,6V) so that's why I wonder about implementation regarding this eroneous functionality

I have replied here -
http://www.diyaudio.com/forums/digital-source/236028-xmos-dsd-384-khz-32bit-usb-6.html#post3608015
 
Some mods done and some in the pipe

Hi!
I did some tweaking to the board using parts already lying around. Sadly it was an all in once tweak (due to lack of time) so i have no way of knowing what effect each modification has.

There is also not a great deal of thought going in to this mod. I simply assumed that the components i had lying around is of better quality and the dac would benefit from a change.

This is what i have done:

- Replaced all 103 and 104 caps in output stage and analogue psu with wima mkp.
- Replaced all 100pf caps in output stage with wima Fkp2.

- Replaced all 10uf output stage and analogue psu with silmic II.

- Replaced the 4700uf Sanwas with 8200uf panasonic TSUP.

- Replaced the Xo with noname TCXO from the board supplier.

The change in SQ was quite remarkable. It was like a thin veil that i did not noticed before was lifted. There is also a more 'analogue' presentation, simply more pleasant to listen to. It also added a little more body to the sound, not as thin. I was hoping for an improvement of the bass, but it is still i little thin in my system (may not be the dac). Before the change to the TCXO the dac kind of hesitated when locking, the led flashed a few times before locking. This is complete gone with the new TCXO. I wished i had the time to do the mods one at a time so i could learn and draw some conclusions, but it was a well spent two hours and the parts was already paid for.

Next step:

Obviously better components pay of sound wise, but further upgrades has to wait until i the next order from mouser (expensive to ship to Sweden). But i do have a bunch of arduinos and 16x2 lcd:s. I´m thinking of hooking up one of them running the HiFiDuino software. I have looked through the HiFiDuino blogs. A wonderful project with wonderful documentation. I have a couple of questions though:

1 - I assume that the registers in the es9018 is non persistent and simply removing the arduino and restarting the dac will bring the 9018 to its default values, is this correct?

2 - Is there any risk of bricking the dac by manipulating its registers?

3 - Is the any obvious reason why hifidiuno would not work with this dac 'out of the box (other than changing the code to match 16x2 lcd)?

4 - The primary goal of this project is to learn more about dacs and have fun, but i´m also thinking of adding more SPDIF inputs and have them selectable from the hifiduino. Assuming that i manage to connect to any of the data 2-7 of the 9018, would that be feasible?

Best Regards // Christian