Quote:
Originally Posted by Masterlu
This is off a Trendnet Gigbit switch.
|
Keep us appraised of your findings, Ivan. More data for the database!
The whole Ethernet switch thing is yet another "can of worms" with respect to digital streaming that I've been wrasslin' with of late. A number of companies (e.g. AQVox, SOtM, and soon, Uptone Audio) are releasing "audio-grade" Ethernet switchese and, it turns out, there are very good reasons for doing so: I've been doing a LOT of reseach on this (the impact of switches, Ethernet cables, FMCs, etc) over the last week or so.
It turns out that a lot of the issues from using "consumer-grade" Ethernet switches, cable modems, routers, FMCs, comes down to these devices' power supplies and clocks: 1) passing low-impedance leakage currents, 2) passing high-impedance leakage currents from switch mode power supplies 3) and cheap clocks causing
clock phase noise. The impact of all these of these is quite audible. BTW, I've read that for most Ethernet switches you want put your cables into alternate RJ45 jacks, e.g. 1, 3, 5, 7, etc, but not 1&2 or 3 & 4 or 5&6, etc. Also, I've read that because of the ground arrangement, the Netgear FS105A or FS108 Ethernet switches do not pass these high-impedance leakage currents "down the line" to your clocks, DACs, etc. (they still need to be powered with linear power supplies, though, NOT SMPS).
I'm sure its less of an issue when using really nice Rubidium clocks, e.g. the G01, on DACs and CD spinners, but for us that are not using these "super-clocks", its a source of a real and very audible problems with respect to streaming configurations (the use of switches, that is).
BTW, the reasons we can very clearly hear differnces in Ethernet and USB cables is for the 3 prinicipal reasons I've cited above.
Very interesting but lots of factors to explore.