DIYHiFi.org

What do you guys think of the technology & it's implementation in these devices?

The Monitor 01 USD is a 24/192 USB to SPDIF - optical, coaxial & BNC SPDIF out. The Monitor 01 US has a PCM1793 DAC inside so analog & 2 headphone outs.

These sound to be too good to be true for $70 & $80 respectively!

According to Musiland, the USB 2.0 Cypress chip interfaces to a Xilinx Spartan FPGA which is the heart of the system. This FPGA creates the SPDIF signal, does clock management (apparently it only adds 1 ppm of jitter in generating the audio clock from the 24MHz crystal), controls the asynchronous USB communication with the PC & can be upgraded in the future through software. It seems to be the first offering in a suite of models all based on this hardware platform (multi-channel recording & playback is mentioned for the future) Early feedback on the sound are very favourable

Edit: There is a third model US 02 - this has a PS in-built & 2 headphone outs - otherwise it's the same as US 01 - Pics here: http://www.soomal.com/search/201.pic/musiland.htm.

_________________
There is a crack in everything, that's how the light gets in!

Just discovered that the same Cypress & Xilinx chips are also used in the Lilo V, which is only supports 24/96. Wasn't there a thread here about this DAC already? In it I seem to remember that it was claimed to use Bulk mode USB but now they claim asynchronous mode is being used. Different drivers & firmware, I guess? I wonder if the Monitor US software could be used with the Lilo ?

_________________
There is a crack in everything, that's how the light gets in!

and you think this is better for you than an ASRC ?

This has nothing to do with ASRC

I'm talking about playback of audio files at their recorded speed not upsampling.

_________________
There is a crack in everything, that's how the light gets in!

http://www.pacificvalve.us/MusilandMonitor01US.html

"Rocketing off the successful LILO DAC, Musiland makes another bold step with their Monitor 01 US,
which uses the same asynchronous topology. Except this time, they extended the sampling rate to 192K!"

Why get all excited over a Xilinx chip? It doesn't speak for the quality of the code.

_________________
'I am... everything is... changed... they're calling... your face... interwoven... who is...

"......the Monitor 01 US is powered by the USB port on your PC."

if want a clean supply, to make a really good clock, yep, that is the place to snag it from.

Jocko

_________________
"Because, I hate you guys. I hate you guys so very, very much.

Yours,
General Cartman Lee"

Have a look at http://www.soomal.com/doc/20100000701.htm for a review. The US01 was tested up to 2496 only and there are some ludicrous results obtained with RMAA. The review actually doubts these. The card can be ordered directly from Hong kong for about $80 delivered. The review of the 02 in the same portal is less informative about hirez capabilities and shows pictures of a switcher built into the dac.!

Sure, mods such as an external clean PS (and not the internal smps of the 02 model) & better clock are a given & possibly tapping I2S is another possibility (for a different internally implemented DAC) - maybe even using a different IV stage if staying with the internal DAC (what's the opinion on the PCM1793?).

But this is jumping the gun - how good a base for modding might this be? Early users have reported that it does 24/192 KHz and shown pics of real devices in use!

I'm specifically interested in their low jitter clock synthesis from the 24MHz crystal - anybody know about doing this via FPGA - avoiding the standard jitter adding PLL approach seems worthwhile. Might replacing the crystal with 24MHz low jitter clock (Clocko Jocko?) bring benefits?

Downsides that I see are no planned support for Linux & not sure about MAC (not that I care about this one) so stuck with windows (it works on windows 7, by all accounts)

Review is hard to read in Google translate form - pidgen english does my head in!

_________________
There is a crack in everything, that's how the light gets in!

fractional frequency divider ? Somehow I doubt you goin to get your bitperfect nonjitter multichannel interface from these guys especially at that price point, seems like the usual 3000 ps jitter business.

I know the price point seems too good to be true (I think I said that) but it's not your usual PLL derivation of the audio clock from a crystal - this is done in FPGA - that's why I'm asking here - has anybody got some knowledge in the FPGA area & can say whether their low jitter claims are warranted/possible. I know the crystal quality will set the baseline jitter (I presume a low jitter clock could be substituted) but they claim the derivation of the audio clock using FPGA only adds 1 ppm extra jitter!

And I know it's all about implementation!!

_________________
There is a crack in everything, that's how the light gets in!

Look up the specs of the IC and you might get an idea what it can do. I dont' see why you should presume it's not working as part of a PLL either.

My EMU Audiodock uses a set of Xilinx chips as well and it IS part of a PLL.

What they did to the clock lines after that at least from the purists perspective is an fn disgrace, but it's also been made to sound surprisingly decent.

My guess is as I think others have already stated, these guys are just copying what they've seen a dozen others do.

_________________
'I am... everything is... changed... they're calling... your face... interwoven... who is...

Which IC are you talking about the Cypress is USB controller/streamer pretty much like a TAS1020B with internal 8051 mcu and all. The FPGA ic specs - I don't have a clue. I'm not particularly interested in their analog out but it would be a bonus if it was any good. PCM1793 DAC, not stellar but probably can be good. IV stage would have to be changed, as usual.

You're right, I should have said that it doesn't use analog PLL. I don't know about digital PLL in FPGA. How good can this be?

Copied who? Who has asynch USB running on FPGA at 24/192KHz? It all really boils down to their clock & how low the added jitter is in the system.

Edit: I saw reference to sarin clock management in the translated text for this DAC but didn't know what, if anything, this referring to. Based on your probing, TD, I researched and found reference to this http://www.actel.com/documents/DigitalPLL_AN.pdf which mentions "Jitter-Bounded Digital PLL" using Successive Approximation Register (SAR) - could this be the FPGA PLL technique used?

_________________
There is a crack in everything, that's how the light gets in!

Well, in my case, and I still need to take a much closer look, there's one of those clock slicer chips used in conjuction with the FPGA. I believe it's reclocked at that point.

You'd have to check but I believe some FPGA's and likely this Xilinx one have internal PLL functionality of some sort at least on a few pins.

What in the hell could they be using an 8051 for lol...

_________________
'I am... everything is... changed... they're calling... your face... interwoven... who is...

There are no other chips than the ones I mentioned & OP275 chip used as IV in the model with the PCM1793 DAC (US)!

There is a Digital Clock Management function within the Xilinx FPGA

The 8051 within the Cypress USB streamer chip is used in exactly the same way as the 8051 inside a TAS1020B USB streamer chip - to control it's functionality!

_________________
There is a crack in everything, that's how the light gets in!

I got one of those (the model USD). It does playback 192/24 and with fairly low jitter measured on an AP SYS-2722. *I know* it is not the best tool to do the measurement so please don't start on that.

The jitter spectrum is purely white noise in the audio frequency range, although the "noise floor" is about 20dB higher than the AP digital generator looped-back.

It is overall quite well designed, using DCM inside the FPGA to generate audio clock from the 24MHz EZUSB system clock. The down side - the coaxial outputs is a bust - transformers not really rated for 192k and an output level only about 57% of what it supposed to be when loaded with a 75 ohm load.