Hey guys,
Has anyone looked at this thing: https://www.xmos.com/products/development-kits/usbaudio2
It's a development board capable of 192Khz/24 bit audio. It can do both SPDIF as well as I2S (although I have not found this explained in the docs). You could even do DSP stuff, and also connect a ADC. It also uses an asynchronous clock
I asked for the price of the hiFace, and I think it is far to high for a little device not doing a lot: $199,-
The xmos board is $149,-, and then you can even do fun stuff with it
Has anyone looked at this thing: https://www.xmos.com/products/development-kits/usbaudio2
An externally hosted image should be here but it was not working when we last tested it.
It's a development board capable of 192Khz/24 bit audio. It can do both SPDIF as well as I2S (although I have not found this explained in the docs). You could even do DSP stuff, and also connect a ADC. It also uses an asynchronous clock
I asked for the price of the hiFace, and I think it is far to high for a little device not doing a lot: $199,-
The xmos board is $149,-, and then you can even do fun stuff with it
Hey guys,
Has anyone looked at this thing: https://www.xmos.com/products/development-kits/usbaudio2
An externally hosted image should be here but it was not working when we last tested it.
It's a development board capable of 192Khz/24 bit audio. It can do both SPDIF as well as I2S (although I have not found this explained in the docs). You could even do DSP stuff, and also connect a ADC. It also uses an asynchronous clock
I asked for the price of the hiFace, and I think it is far to high for a little device not doing a lot: $199,-
The xmos board is $149,-, and then you can even do fun stuff with it
Hi 4real,
it's an interesting device, now finally available for sale.. a few months ago it was jkeny who first sent me info regarding this interesting device, back then it was more a concept .. now available good.
a few remarks.. usb audio classes 2.0 on windows are not available, they claim to have windows driver available.. is it working and free? (they say ASK representative..means $$?)
onboard local oscillators freqs are 11.2896 and 24.576 for 44.1k 48k and multiples.. too bad my wolfson doesn't support 176.4 at 64fs
by looking at the BOM.. xtals are common citizens -+50ppm...jitter unknown. it seems to me not very hi performance.cheap solution
anyway it's very interesting device.. especially this new event driven processor (how it works dunno) .who first gonna try it and report back?
PS: hiface sells at 99EU+shipping from m2tech and 150USD+shipping from tweekgeek.. where did you get 199USD?
cheers
Vale
Last edited:
Looks like you get the driver is you get the board.. I don't know if they plan to release it for everyone. It even looks like Windows 7 supports it out of the box, but I'm not sure. OSX does, at least.
That depends on what you put behind it.. Im my case, it will be a Ess Sabre, and then these clocks are fairly irrelevant.
Well, if the hiface is 99 euro's, then that's another story.. I mailed m2tech, and they told me that I can get it in Australia: M2Tech hiFace USB Digital Audio Interface
According to themselves, you can only buy it there...
Anyway: if I look at the prices of the distributors, I can't find the 99 euro's anywhere...
Last edited:
I would think you need to use crystal clocks with better tolerances than 50ppm and not plain ordinary crystals. Thus the jitter would probably have to be measured in the circuit. I wouldn't have great expectations.
I suspect there will be more and more USB receivers and other DAC boards when USB Class 2 drivers for Windows become available and drivers don't have to be written.
Maybe a good group project to do here on DIY to come up with a very high precision Universal Asynch USB receiver for DAC projects that doesn't require modding some tiny board with tweezers and a microscope.
Hi John,
PS: USB isolation at hi speed (480Mbs) is not available on any known chip (at least to me).. maybe we can focus on I2S isolation using ADUM
You have to be very careful with the ADUMs. The ADUMs work by sending a few pulses across the transformer for each edge it sees. These pulses come from an on chip free running oscillator (NOT a crystal). The result is every edge on the input is quantized to the timing of this on chip oscillator. You are essentially running them through a high speed asynchronous reclocker. This effectively adds several nanoseconds (yes nanoseconds) of jitter to the signals.
So if you are designing a circuit that reclocks the I2S after the ADUM with your low jitter local clock it should not be a problem. But if you are feeding I2S through the ADUM and then directly into a DAC chip it will not be good.
John S.
Maybe in the future Intels Light Peak interface could be used to isolate the computer from the DAC.
Could do but I wonder what the jitter will be?
I think the only current way to address this is to use a low power Music playback PC & run on batteries or a good linear supply. I know there are still switching regs within these that will pollute the PS but I seem to remember John S saying that the FIT PC had a small number of these & they could possibly be replaced with other regs?
Well, I wonder how Wavelength, Ayre etc. will solve this problem when they all upgrade to 192K ? I know Ayre is currently using something in the QB-9 but that's only USB 1.1
Hey guys,
Has anyone looked at this thing: https://www.xmos.com/products/development-kits/usbaudio2
An externally hosted image should be here but it was not working when we last tested it.
It's a development board capable of 192Khz/24 bit audio. It can do both SPDIF as well as I2S (although I have not found this explained in the docs). You could even do DSP stuff, and also connect a ADC. It also uses an asynchronous clock
I asked for the price of the hiFace, and I think it is far to high for a little device not doing a lot: $199,-
The xmos board is $149,-, and then you can even do fun stuff with it
I have this board. Sound quality through my Buffalo32 is quite good, noticeably better than the Musiland USD 01 that we compared it to.
what about NVE ultra fast digital couplers ?
found this note "Isolated I2S Minimizes Distortion Reaching Isolated Audio Heaven (well, nearly)"
http://www.nve.com/Downloads/ab17.pdf
NVE Illustrative Isolator Product Applications
anybody used them?
thanks
Vale
Why would there be a difference? I thought the Sabre DAC eliminated jitter? I guess not literally huh?
Well, accourding to the white paper on the ESS website, it should indeed literally eliminate jitter
I don't if this is true in practice however...Reports from the field indicate that different transports sound different using the Sabre32 top of the range DAC so it doesn't eliminate jitter (quelle surpriseWell, accourding to the white paper on the ESS website, it should indeed literally eliminate jitter
).There is nothing that eliminates jitter without sonic penalty, AFAIK.
AFAIK, MCLK can either be supplied from the source as one of the I2S lines OR this MCLK can be ignored & the DAC runs happily without it! I believe the ESS DAC senses the speed of the incoming data & reclocks it with it's own local clock. So there is an ASRC built into this DAC, it just sounds much better than other ASRCs or gets in the way less!
Here's what Dustin (the Chief Engineer) says about it:
Last edited:
Logically it would seem that not using MLCK is asking for higher jitter. Traditionally an ASRC does not totally eliminate jitter. Though with Sabre may be lower jitter than previous methods. I wouldn't expect the Sabre to lock onto anything and eliminate all jitter. Maybe alright for consumer devices but if you can hear the difference with transports then it doesn't completely eliminate jitter.
I'm not so sure that's true if the local clock is a good low jitter implementation?
Here's some further light from Dustin
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.