Sound very interesting the Phidecca dac..i am going to read all about it,,,but where can i
see how it is build in output,,,is it a mix of opamp and discrete ?
is it easy to wire my JLsound usb to the Phidacca ?
are there another tread where there maybe are more information of this dac type ?
Best Bjarne
see how it is build in output,,,is it a mix of opamp and discrete ?
is it easy to wire my JLsound usb to the Phidacca ?
are there another tread where there maybe are more information of this dac type ?
Best Bjarne
You'll find the schematic in post #189 of this thread. Its a mixture of low-noise opamps and discrete - the output stages are discrete, classA single-ended.
I'm not aware of this DAC being a particular 'type' - although its using opamp for I/V it also has an LC filter prior to the opamp which I've not seen in other designs.
@PJN - thanks very much for your comments on how it sounds, very much appreciated!
I'm not aware of this DAC being a particular 'type' - although its using opamp for I/V it also has an LC filter prior to the opamp which I've not seen in other designs.
@PJN - thanks very much for your comments on how it sounds, very much appreciated!
I hope to hear your comments soon, Cuchu
(I can't resist posting and seeing two spanish flags together, we are so few here)
Haha! You will know my mpressions shortly. (Maybe I will have to make some questions to you
) We have to encourage our spanish friends to start making this kind of projects!
24bit word won't squeeze into a 16bit DAC, which is what TDA1387 is. The bottom 8 bits will just get ignored. As for 192kHz it might play, the DAC chip itself accepts it but I haven't tested my logic (level shifter before the DAC chips) at that rate.
Pop on start seems to depend on the USB interface - the CM6631A does occasionally give a mild clunk, another I've tried using STM32F4 gave more undesirable noises. On power up there will be a pop yes because the circuitry is single-ended, not using two power supplies. So the output caps need to charge initially.
Pop on start seems to depend on the USB interface - the CM6631A does occasionally give a mild clunk, another I've tried using STM32F4 gave more undesirable noises. On power up there will be a pop yes because the circuitry is single-ended, not using two power supplies. So the output caps need to charge initially.
Hires (88k and 96k) play for sure, I have tried but not all the bits will make it to influence the output. Above 96k I won't give any guarantees but if you want to play rates higher than 44k1 I'd advise you to tweak the frequency response of the output stage filter to give a flatter characteristic. At present its optimized to correct for NOS droop at 44.1kHz.
As for taking the output to a transformer, that's what I do all the time I would suggest increasing the value of the 4.7uF output coupling caps when feeding a trafo, they should be larger to keep the LF resonance well controlled.
As for taking the output to a transformer, that's what I do all the time
I would suggest increasing the value of the 4.7uF output coupling caps when feeding a trafo, they should be larger to keep the LF resonance well controlled.
I'd not advise connecting a trafo before the output in this case as the output stage provides NOS droop correction. Also there's the slim possibility that you might overload the LT1028 (earlier) stage if you connect the trafo to that. Taking the signal straight after the LC filter will most likely destroy the nice flat FR I've designed so I can't recommend that. It will also very likely make the DAC noisier.
The internal architecture of this DAC isn't similar to the AK4493 - as I understand it that DAC is balanced. You'd likely want to tap off the balanced output of the AK4493 before the external opamp filter in that case.
Yes questions are welcome here
The internal architecture of this DAC isn't similar to the AK4493 - as I understand it that DAC is balanced. You'd likely want to tap off the balanced output of the AK4493 before the external opamp filter in that case.
Yes questions are welcome here
Hi Richard!
You wrote somewhere that you did not prefer Xmos-Usb to i2s conversion. For what reason if I may ask? Is the quality of the clocks important here? What about ground isolation?
I saw at diyinhk an affordable one for 39 USD.(384kHz Asynchronous USB to I2S/SPDIF CM6631A PCB - DIYINHK) Any thoughts?
Lots of q's, hope you can help..
Ed
You wrote somewhere that you did not prefer Xmos-Usb to i2s conversion. For what reason if I may ask? Is the quality of the clocks important here? What about ground isolation?
I saw at diyinhk an affordable one for 39 USD.(384kHz Asynchronous USB to I2S/SPDIF CM6631A PCB - DIYINHK) Any thoughts?
Lots of q's, hope you can help..
Ed
Last edited:
Ed - in the old days XMOS was very power hungry and hence made more RFI than other solutions. They might have improved the EMC profile more recently though, I haven't kept track. I must admit I've hardly paid any attention to clock quality on my DAC design journey, that hasn't been the low-hanging fruit. But in future perhaps I'll realize what I've been missing out on, who knows?
CM6631A is a solid choice for USB-> I2S. Ground isolation is very desirable but often problematic with USB.
@beardman - at present I'm using Wima but this being DIY you can change. As for the bass, see post #196
CM6631A is a solid choice for USB-> I2S. Ground isolation is very desirable but often problematic with USB.
@beardman - at present I'm using Wima but this being DIY you can change. As for the bass, see post #196
That's right - to use an off-board cap, fit those resistors (and don't fit C20,C27)
You can't go direct to a trafo no. A trafo will always need an output cap - in this case the output cap will need to be bigger than 4.7uF so best use an off-board one.
Yes it is something to do with the impedance of the trafo. Trafos always like to be driven by a low impedance source but the output cap makes the source increasingly high impedance at the lowest frequencies. A bigger cap reduces the LF impedance of the source. The precise value needed depends on the trafo's primary inductance - a measure of its impedance at LF.