[tosehee]

I thought the bit perfect only worked in KS mode. What modes are currently supporting bit perfect under Windows Vista/7? up to 24/192?

 

[tosehee]

Quote:

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.

Anyone know what this means? I am referring to the last sentence.

 

[glt]

Means it works but not "audiophile perfect"

 

[mmerrill99]

glt is correct about this & if you read the whole thread simmconn later corrects his statement that the noise floor is 20dB higher - he was measuring incorrectly - he later posts a graph which shows about 2dB higher. I believe he also estimates jitter to be <100ps.

I have an issue with my 01 US which you guys might be able to help me with - I don't know if it's the unit or the track or a interaction of the two.

I noticed a muddiness on a 24/96 track "Sister Rosetta goes before U" from the album Raising Sand. This showed up in analog & digital Toslink out but not on headphone. I don't have another track from that album but all other high res tracks (from other albums) are fine.

I had a theory that this track was very hotly cut & might be over-driving the amp's input stage.

So, I took some measurements today - at full volume with the track, I record max 1600mV on the RCA out - any other track I get about 800mV max!

2000mV is the spec on analag out - is this rms? Does this mean +/-1000mV? Does this explain the muddiness? Does this explain the toslink muddiness also?

 

[glt]

jkenny, some dacs have de-emphasis switch. Also the analog is running with 5V supply, so maybe clipping? (just speculating)

 

[macrog]

I purchased a monitor 01usd and monitor 02 which arrived on Wednesday. The sound quality from the monitor 02 used as a coaxial digital input for my Cambridge 840c is much better than the sound quality I get from the monitor 01usd. The monitor 01usd sounds a bit bright in the treble and not as natural pitch wise. I guess the monitor 01usd might take longer to run in due to usb power rather than mains ?

The sound quality from the monitor 02's analogue outputs is a noticeable step up from my Dacmagic when fed a usb source even on 44.1khz recordings with more detail apparent.

The coaxial digital output of the monitor 02 compares well to the coaxial digital output of my Asus Xonar stx which really suprises me. It is probably just a fraction behind and as it is still not fully run in may well equal or beat the stx for digital output.

I am very impressed with musiland and will probably use the monitor 01usd for usb sources to the dacmagic. If it doesnt improve to be closer to the quality of the monitor 02 I will grab another monitor 02.

Hope this helps someone.

 

[TheShaman]

Perhaps 01USD will sound more like (or better) than the 02 when fed from a separate power supply rather than rely on USB power.

 

[mmerrill99]

I agree, TheShaman, I believe the difference in sound between them is down to the internal SMPS powering the DAC & output stage of the monitor 02. I think I remember reading that it runs the output stage at a higher voltage than the +5V available from USB.

My first experiment will be to run the whole unit off external 5V supply - the next step will require a mod inside the unit to split the supply to the output stage & feed it off a higher voltage external supply - can't remember what value supply - will have to check out the datasheets of PCM1793 & OP275

 

[rmask]

Quote:

Originally Posted by macrog /img/forum/go_quote.gif I purchased a monitor 01usd and monitor 02 which arrived on Wednesday. The sound quality from the monitor 02 used as a coaxial digital input for my Cambridge 840c is much better than the sound quality I get from the monitor 01usd. The monitor 01usd sounds a bit bright in the treble and not as natural pitch wise. I guess the monitor 01usd might take longer to run in due to usb power rather than mains ?

Please note any future changes in the USD's sound signature (if any). Thanks

 

[mmerrill99]

OK, here's what I discovered:

I did some experiments & I supplied my monitor 01 US (the PCM1973 DAC based one) with external (SMPS) 5V supply & it greatly improves the sound from the headphone out. I had heard a tiny bit of headphone distortion (on the "Sister Rosetta" track - see my post above) but assumed it was the cheap Sony headphones I'm using - this is now gone. I won't get a chance to try the analog out in the recording studio for a while.

So maybe it was current starved operating off the USB supply? Which would explain why the digital out was also distorted! Maybe the USD model (transport only - no DAC) isn't as badly affected by this as it has no DAC or op-amp drawing current away from the crystal, FPGA & USB controller chips?

I'll next try some batteries running this! I don't know what the current consumption of the unit is as to whether batteries are practical & still portable.

Edit: glt, I recommend trying this - here's what I did - quick & dirty - just tap into the USB leads; cut the +5V (Red wire in my case - I think this is the standard) and connect the DAC facing red to your external 5V supply; expose the gnd wire (black) and connect the external ground to this.

Edit again: glt, first of all try the track & see if you get similar distortion - email me for details!

 

[danny_66]

Has someone tried to get I2S out of these ?

When looking at the pictures of the PCM1793 in the 01 US I see that pins 26 and 28 are tied together and pin 27 not:
Google Translate

According to the datasheet (page 15) of the PCM1793 that corresponds probably with 24-bit I2S format.
http://focus.ti.com/lit/ds/symlink/pcm1793.pdf
Can someone check if pin 26,28 are high?

I2S can than be taken from pins 1,2,3 from the PCM1793 of probably pins 3,4,5 from the FPGA.

 

[mmerrill99]

Hi Danny,
I haven't measured these pins yet mainly because of their proximity - this thing is about 1.5 times the size of a cigarette box. Also I've destroyed too many pieces with a slip of the measuring probe but your analysis is exactly the same as mine & I bought the 01 US specifically for the I2S signals

BTW, I analysed the track & it turns out to be badly over-recorded & distorted - it's not the DAC after-all!!! I'm writing to the on-line shop I bought it from!

 

[danny_66]

Hello jkeny,
it looked bigger on the detailed pictures
Yes, with such small lines an error is easily made.
Maybe you can check with the power off if pin 26,28 is connected with a powerline somewhere?

 

[mmerrill99]

I know, you have to have your magnifier glass cleaned up for this work

I have an idea how to tap into these lines - scratch the varnish off trace 1 & 3 and a bit further along off 2 & 4 - this will allow me to flux & solder these exposed copper tracks & there will be a bit of separation between them now (not a lot

). I was going to then solder fine wires to each of the copper traces & bring them to the edge of the board, ensuring that all the line distance from the FPGA pins to the board edge are the same (recommended for I2S).

So I might do this on one of the lines first and measure it but I'm not ready to do it yet.

I can confirm it by checking PCM1793 pins 26 & 28 are tied to High (+3.3V) & 27 to Low (gnd) so we have 24 bit I2S format !

 

[mmerrill99]

Just in case I wasn't clear in the last post - the pins are high & low as I stated - I measured them - so we have I2S from the FPGA running in those 4 traces to the PCM1793.

Can you think of any better way to tap into those signals other than the method I outlined?