03-27-2016, 07:58 PM
(Modification du message : 03-27-2016, 07:59 PM par a supprimer merci.)
Bonne question. Il y a un fil sur ComputerAudiophile ou j'avais posé ces mêmes questions. Voici les réponses de Swenson (page 3 dans le fil):
"The microRendu uses the iMX6 DualLite processor. It has many different power supply domains which makes it particularly suited for audiophile use.
Each supply domain gets an appropriate regulation scheme for that function. For example the processor domain uses a high quality switching regulator since it needs low voltage at high current. The on chip PLLs which generate the clocks for all the different system are run off an ultra low noise regulator. The USB subsystem takes three supplies, all of which are fed by ultra low noise regulators.
In addition the chip can be fed an external clock that bypasses the internal oscillator circuit which is not particularly low jitter. The reference clock of the PLLs can changed to point at this external clock input into which I feed a very low jitter main clock.
The result of this is MUCH lower noise and jitter at the USB subsystem."
Et concernant le soft:
"The software for the Sonicorbiter SE and the microRendu are very similar, but not exactly the same. Some obvious things are the SoSE contains an S/PDIF optical output which does not exist in the microRendu, the software has to take that into account. In addition the microRendu includes software to make the clock change mentioned previously, without that software to set the PLL muxes the low jitter clock has no effect.
What you are talking about, using different distributions in the microRendu hardware has not been tested at all, but my gut feeling is that software changes will make less of a difference on SQ with the microRendu VS other hardware platforms (such as the Cubox) due to the much greater power supply isolation between power domains. "
"The microRendu uses the iMX6 DualLite processor. It has many different power supply domains which makes it particularly suited for audiophile use.
Each supply domain gets an appropriate regulation scheme for that function. For example the processor domain uses a high quality switching regulator since it needs low voltage at high current. The on chip PLLs which generate the clocks for all the different system are run off an ultra low noise regulator. The USB subsystem takes three supplies, all of which are fed by ultra low noise regulators.
In addition the chip can be fed an external clock that bypasses the internal oscillator circuit which is not particularly low jitter. The reference clock of the PLLs can changed to point at this external clock input into which I feed a very low jitter main clock.
The result of this is MUCH lower noise and jitter at the USB subsystem."
Et concernant le soft:
"The software for the Sonicorbiter SE and the microRendu are very similar, but not exactly the same. Some obvious things are the SoSE contains an S/PDIF optical output which does not exist in the microRendu, the software has to take that into account. In addition the microRendu includes software to make the clock change mentioned previously, without that software to set the PLL muxes the low jitter clock has no effect.
What you are talking about, using different distributions in the microRendu hardware has not been tested at all, but my gut feeling is that software changes will make less of a difference on SQ with the microRendu VS other hardware platforms (such as the Cubox) due to the much greater power supply isolation between power domains. "