• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

The Asynchronous USB Audio class 2 module.

Well Folks,

It has been a long time coming, but it is almost here. I spent many months of prototyping and research before ending up here.

I am just finalizing the layout of the Asynchronous USB 2.0 input module.

I wanted to show you folks a quick preview of the module. We are just about to order the prototypes.

Some features:

Fully asynchronous operation based on XMOS technology utilizing custom firmware.

All audio signals and power including the clocks are fully isolated from the USB side of things - this can be overridden with jumpers if you want to power the entire device from the USB bus. External power for the audio side is very flexible. 5 - 12VDC input should be fine.

I2S (TTL and LVDS for teleporter) and SPDIF output.

Access to the master clock signal.

All digital audio signals (clocks and data) are re-clocked to the audio master clock.

Two extremely low phase noise Crystek clocks for the audio master clocks.

Separate TPS7A47 LDO regulators used on both the USB side and the audio side.

Meticulous 4 layer layout with very careful attention to all of the important details like signal termination and bypass/decoupling. There is even ESD protection for the USB input.

uFL connectors for LVTTL signals to/from DAC/ADC.

The board is designed to be easily panel mounted using brackets.

Here is a preview of the board. Some details will likely change a bit as I finish it up (and if I find anything I want to change after the next prototypes), and certain design features are intentionally obscured. But you should get a good idea what is going on. :cool:

More details to come. Feedback welcome.

Cheers!
Russ
 

Attachments

  • USB_Alpha.PNG
    USB_Alpha.PNG
    82.1 KB · Views: 2,519
Last edited:
This will do simultaneous i2s input and output, right?

Chris

Yes - it is planned to support that - but I need to actually test it with the target DAC and ADC at the same time.

I have it designed to work with the ESS ADC with the ADC in asynchronous slave mode.

With my current firmware the USB module must always be master for both the DAC and the ADC.

The initial focus is output - but I have plans for input too. :)
 
I was intentionally vague about some key features, as I have some firmware coding to do on the prototype. :)

Still what I can say for sure is that it will at least support the following:

44.1 - 192Khz-24bit playback via I2S and SPDIF.

Features still in development that are expected (but not promised):

2 channel Audio input *and* output up to 384Khz 32-bit.

DSD support (to what degree TBD)

ADC input (because I need to test integration with the ADC)

The great thing is that It is a DIY module, and as such we expect to provide updates and even possibly custom firmware as we can to support the community. To facilitate that I plan on installing the DFU firmware. All of this still needs to be tested. :)

Cheers!
Russ
 

Bunpei

Member
2008-08-30 10:31 am
Oh, it's a long-awaited good news!

May I ask a few questions? Please feel free to ignore any of them.

1. "Asynchronous USB 2.0 input module"
Does this mean "USB Audio Class 2.0" compatible or simply "USB 2.0 (High-speed)" compatible?
If it is "USB Audio Class 2.0" compatible, it does require a driver software in Windows environment and does not in Mac OS environment. Is it correct?
If we use it with ASIO compatible players, we can use ASIO4ALL for that purpose?

2. I2S (TTL and LVDS for teleporter)
Does the TTL mean "+5V TTL level" or "+3.3V C-MOS level"?

3. Two extremely low phase noise Crystek clocks for the audio master clocks
Are these clocks of 45.1584/49.152MHz or 90.3168/98.304MHz?
We expected your strong buying power would make them supply 90s MHz pair to you.

4. All digital audio signals (clocks and data) are re-clocked to the audio master clock
I assume you have overcome the 400MHz re-clocking issue of XMOS architecture that a Canadian vendor sometimes blames. I have believed that you would achieve the true re-clocking.

5. I2C control for DAC
Is the USB board configurable to send commands for setting registers in DAC over I2C?

Anyway, congratulations!
 
Oh, it's a long-awaited good news!

May I ask a few questions? Please feel free to ignore any of them.

1. "Asynchronous USB 2.0 input module"
Does this mean "USB Audio Class 2.0" compatible or simply "USB 2.0 (High-speed)" compatible?
If it is "USB Audio Class 2.0" compatible, it does require a driver software in Windows environment and does not in Mac OS environment. Is it correct?
If we use it with ASIO compatible players, we can use ASIO4ALL for that purpose?

2. I2S (TTL and LVDS for teleporter)
Does the TTL mean "+5V TTL level" or "+3.3V C-MOS level"?

3. Two extremely low phase noise Crystek clocks for the audio master clocks
Are these clocks of 45.1584/49.152MHz or 90.3168/98.304MHz?
We expected your strong buying power would make them supply 90s MHz pair to you.

4. All digital audio signals (clocks and data) are re-clocked to the audio master clock
I assume you have overcome the 400MHz re-clocking issue of XMOS architecture that a Canadian vendor sometimes blames. I have believed that you would achieve the true re-clocking.

5. I2C control for DAC
Is the USB board configurable to send commands for setting registers in DAC over I2C?

Anyway, congratulations!

Thanks! I am happy to answer.

1) USB Audio Class 2 :)

2) 3.3V based LVTTL same as ESS DAC and ADC accept/produce.

3) Exploring several possibilities here. More prototyping will be required.

4) This technique I am using should ensure that signals are always aligned with the master clock. I would be interested in reading the experience of the Canadian vendor. :) If you can please send me more info.

5) Not currently in scope, but I will think it over.

Cheers!
Russ
 
Hi Bunpei, of course I meant you to respond privately, because I simply do not agree in many respects with the people you refer to. :)

I understand very well how both the XMOS and ESS I/O schemes work.

https://www.xmos.com/download/final...1718eee0a56586d0e502ff9855&px-time=1362678408

With the ES9018 the effects of the 400Mhz skew on I2S would not matter as everything is asynchronously re-sampled (not merely re-clocked), but even if you are of the mind that it would matter, re-clocking each signal to the audio master clock solves this issue very nicely. :)

I will say that if someone credible can definitively show a direct analog effect (at the DAC output) of the 4ns jitter on anything other than a master clock on the analog audio signal (not digital) I will gladly change my view, until then that idea holds no water for me. :)

Anyway - for this source, it's a complete non-issue. The output is not coming directly from the XMOS chip. And the master clock is definitely not generated by it.

Cheers!
Russ
 
Last edited: