ackoDAC based on ES9018

[labjr]

Originally Posted by realmadrid

Look at
TeraDak Teralink X2 (USB to SPDIF/TOSLINK/I2S Converter @ 24bit 96k) - AVSite

It contains the same information as the ebay page. Most of the imported DACs are crude designs or they copy the application notes. Either way don't necessarily perform well.

[jkeny]

As labjr said, this doesn't look technically up to the best - not asynchronous, uses a single 12MHz clock from which audio clock is synthesised.
The HiFace is far superior technically & with a couple of PS mods sings wonderfully.

[acko]

Originally Posted by labjr

Cool Project. Any plans to do a USB or firewire input board for the system?

The DAC system is modular so you can attach external USB-I2S or USB-Spdif to the Digital Inputs. As JKenny pointed out the HiFace USB-Spdif is an excellent choice and this can be connected via coax cable to the Spdif Receiver as shown in the preview. The DAC has 8 Spdif channels (D1 to D8) so could be used to easily switch from PC to CD player (with spdif/AES out)

What about similar USB to I2S as this would be more direct in the conversion?

Regards, Acko

[labjr]

Asynchronous USB to I2S would be the obvious choice. I don't know why SPDIF even exists anymore.

It would be nice not to have modify another product. Modifying the Hiface is risky because it's so small and delicate.

Having it's own USB input module, the power supplies and such could be designed with the module in mind.

[darkfury]

Originally Posted by labjr

Asynchronous USB to I2S would be the obvious choice. I don't know why SPDIF even exists anymore.

It would be nice not to have modify another product. Modifying the Hiface is risky because it's so small and delicate.

Having it's own USB input module, the power supplies and such could be designed with the module in mind.

For the Teralink X2 convertor in my earlier post, designer plan to add a FIFO buffer in that converter to get lower jitter
Did you try Musiland Monitor 01 USD;Asynchronous and 24/192 (using xilinx chip)
I am wondering performance of it comparing to HiFace USB-Spdif

[labjr]

Jkeny has both the Hiface and Musiland. He says there is no comparison. The Hiface is much better.

[Magsy]

Looks great and it is for sure going to mate well with a hacked up hiface. I think given the state of USB Async there is no point in trying to include it unless it comes licensed from Hiface/Wavelength Audio, all the other solutions seem to have issues and at some point you need to just get the product out.

How is the ETA looking Acko?

[jkeny]

Originally Posted by labjr

Jkeny has both the Hiface and Musiland. He says there is no comparison. The Hiface is much better.

Yep, labjr, but just to clarify - The MODDED HiFace is WAY better than the MODDED Musiland. The stock HiFace is about the same as the Musiland AFAI can remember . But when the HiFace is modded with a clean supply to the clocks & a separate clean supply to the IO part of the Xilinx CPLD chip, then it really sings

[glt]

Originally Posted by labjr

Jkeny has both the Hiface and Musiland. He says there is no comparison. The Hiface is much better.

Probably implementation dependent.

For many, hiface vs musiland will depend on soldering skills (I myself would destroy the hifice if I attempt to mod it)

In addition, the ess chip will throw away the master clock and resample the data, so the difference in clock implementation between musiland and hiface may or may not matter. I think jkenny was going to repeat his experiment with the asrc enabled with a proper clock.

There was comment in my blog about an implementation of using the i2s from the musiland to a BurrBrown DAC and it was an improvement over his previous implementation using an resampler with a tent clock.

[jkeny]

Originally Posted by glt

Probably implementation dependent.

Yes, glt, I forgot about the Musiland's clock putting out 128fs & this not being optimal for your Opus WM DAC or my ESS DAC. Whereas the HiFace puts out 256fs from 32 to 96KHz, which is fine! BTW, Does the Musiland use 128fs at all samplerates? This might be another reason though to favour the HiFace & I2S or use SPDIF as you have done?

For many, hiface vs musiland will depend on soldering skills (I myself would destroy the hifice if I attempt to mod it)

Yes, the HiFace is not very DIY friendly - it's a very tight fit but I have to say that all the critical supplies are exposed & accessible so they can be optimised given a good magnifier& steady hand.

In addition, the ess chip will throw away the master clock and resample the data, so the difference in clock implementation between musiland and hiface may or may not matter. I think jkenny was going to repeat his experiment with the asrc enabled with a proper clock.

Yes, my report is based on using the I2S MCLK which comes from the transport so the transport clock's jitter & noise is important. I haven't tried the 50MHz clock in asynchronous mode on the ESS DAC. I'm not sure what Valeriano is using?

There was comment in my blog about an implementation of using the i2s from the musiland to a BurrBrown DAC and it was an improvement over his previous implementation using an resampler with a tent clock.