[tiberian]

http://photo.head-fi.org/showphoto.php?photo=615

 

[S_Dedalus]

Well, now we know that modding it would be a nightmare.

 

[gaboo]

If you could take a picture in which we can read what's printed on the chips it would be much more useful. Trust me, nobody is going to clone that ASIC.

 

[tiberian]

http://photo.head-fi.org/showphoto.php?photo=616
was trying to trim the output volume and bypass the volume pot...resulted in horrible clipping, i am still working on it, seems like the output is VERY strong.

 

[gaboo]

Hm, the Xilix chip is a FPGA. Looks similar to the one on 1010m. I don't suppose the minidac is capable of any DSP-like stuff... The clock above it surely looks huge, but we can't read it.

AD1955 is the DAC. What's interesting is that's a Multibit Sigma-Delta (i.e. SACD decoding capable). Probably you'll never see that functionality working.
AD1955 also decodes PCM, it's probably wired to do just that. The numbers look a bit better than 1212m's DAC: 120dB SNR and 120dB dynamic range, -110dB THD+N (vs 120/120/-107 on CS4398). It also has better stopband attenuation. So it's safe to say it is better that 1212m.

Btw, CS4398 and also supports DSD, not that I ever hope to see 1212m decoding SACD.

OPA275 in the lower left corner has 9Mhz bandwitdth and 22 V/µs slew rate. Not exactly something you'd want to DYI.

Is the minidac bus powered? Or is the upside down "daughtercard" PCB the psu?

 

[tiberian]

the card is the usb add-on...the mini dac doesn't use an internal transformer.
the clock says 'apogee 24.576Mhz M2550 0301'.
there a a bunch of OP275s and 2 OP176s at the back.

i will pull away the card and take another pic...

 

[gaboo]

If the lower left opamp is on the output (the lower left seems to go to the output front jack, err, I hope that's a jack, can't quite tell from the pic) you could replace it with AD8610. It will give you twice the slew rate (40-50 V/µs). The bandwitdh of 8610 is only 25Mhz, so it will probably be stable in that circuit.

 

[tiberian]

 

[gaboo]

The USB board has an AD1896 which is an ASRC (async sample rate converter) and it's own clock. The $300 question is what is the square chip that's the USB interface. It's pitch black in the picture.

Anyway, the presence of an ASRC makes me suspect that the USB transfer is not asyncrhonous/buffered. So, despite the USB interface, the design is not very different from what we've seen so far (Benchmark DAC-1). Except that you only get the ASRC for the USB, and not the S/PDIF inputs.

 

[gaboo]

The balanced output path towards the XLR jacks is too complex for me to guess how it's done without being able to see the traces. It seems there are at least two or three stages there... The bad news is that all the caps on that path are surface mounted. You probably don't have many options in replacing those, but I'm no expert on caps, let alone the surface mounted ones.

 

[tiberian]

Quote:

Originally Posted by gaboo The USB board has an AD1896 which is an ASRC (async sample rate converter) and it's own clock. The $300 question is what is the square chip that's the USB interface. It's pitch black in the picture. Anyway, the presence of an ASRC makes me suspect that the USB transfer is not asyncrhonous/buffered. So, despite the USB interface, the design is not very different from what we've seen so far (Benchmark DAC-1). Except that you only get the ASRC for the USB, and not the S/PDIF inputs.

strange i can read the chip from the pic...(symbol) 2AD4LJT TUSB3200AC

 

[gaboo]

Quote:

Originally Posted by tiberian strange i can read the chip from the pic...(symbol) 2AD4LJT TUSB3200AC

Heh, I increased luminosity to max on my lcd, and I too can read the print on the chip.

TUSB3200 is pretty standard 1.1 USB to I2S codec. In theory it supports asynchronous transfers, but I'm not sure it is used that way in the minidac. One way to be sure is to snoop the usb traffic. In Linux it would be trivial to this: the driver will basiacally tell you what kind of pipe the device requested. I don't know what tools are available to this in Windows...

 

[tiberian]

Quote:

Originally Posted by gaboo Heh, my cheap Dell LCD probably doesn't have many shades of black. TUSB3200 is pretty standard 1.1 USB codec. In theory it supports asynchronous transfers, but I don't think it is used that way in the minidac. One way to be sure is to snoop the usb traffic. In Linux it would be trivial. Heh, I increased luminosity to max and I can now read the print on the chip

i think its pretty amazing that apogee didn't put a huge ass power supply in the dac and the dac still has lots of power.

 

[ooheadsoo]

Wow, if I turn my monitor up enough to read it, black isn't 100% pure black any more, as I have it normally.

 

[lan]

Quote:

Originally Posted by tiberian i think its pretty amazing that apogee didn't put a huge ass power supply in the dac and the dac still has lots of power.

What does it use a wallwart AC/DC adaptor?