OK, this is going to be a thread about the new setup, or rather, a new build of an old setup we've had working at a colleague's in Germany for several years. However, some details are different, as there won't be a Sonore SOtM (well, not yet... maybe soon, maybe something else like an NAD M50) music server.
Core component for clock and digital path re-clocking is a Brainstorm DCD-8.
This little beastie is found in a lot of studios and video sound stages and post production studios, because of it's clock distribution and sample rate synchronization capabilities, including with HD video sources. On top of that, a key differentiating feature is the ability to use an external 10MHz rubidium atomic clock as it's internal clocking/reference source for digital re-clocking. Rubidium clocks are very accurate, have very low phase noise and jitter (they're widely used in the telecom industry for these reasons, and that's how I got a used one on eBay- taken out of gear. The LPFRS module was originally made in Switzerland, and requires up to 4 A at 24V during startup, tapering off to about 3/4A during normal operation as the oscillator is up to temperature and stabilized.
This morning I finished making power cabled for the input power and and DC power connection from the OMRON SMPS I also bought used on eBay- this is a 24V 4A supply made in Japan, with good quality on the output and some nice control features- a little overkill, I suppose, but what the heck. My friend powers his off of batteries- most of his system runs off Lithium batteries with Paul Hynes regulators. See, I'm really just a pretty normal guy in comparison. :B
I've yet to mount the LPFRS on the heatsink I bought for it; gotta get the thermal interface material first. With an average power dissipation of about 18W, the heatsink is quite necessary. Eventually I'll probably replace this with a new Stanford Research PRS-10, which is what my colleague was using until he fried it (power supply misconnection) and had to send it back for repairs; then he got a used one on eBay! I think he has the SRS one back. By the specs, it has lower phase noise and better longer term performance. Will it make a difference? Don't know...
But this one is pretty good...
For the inquiring minds, here's the output of the oscillator terminated into 50 ohms.
Yeah, make fun of my antique HP scope, but it was cheap, and it is 2GSa/s sampling rate (2 Giga samples per second). And yeah, that's a CRT, but it's a digital display, not a conventional scope tube. No data export, which limits the usefulness, but it's OK for troubleshooting around the house, you know, when your toaster goes on the fritz!
Next step is mounting the heatsink, then it will be programming the DCD8 and testing with an iMac for now.
The DCD8 will be working in dual domain mode, with one domain coming from the server (what ever it is), and the other from the disk playback transport. The DCD8 also includes a high performance Firewire input, and I'll be using that for interfacing to Mac's for testing with my MBP 17, and interface with the late model 21" iMac.
more to come...
Core component for clock and digital path re-clocking is a Brainstorm DCD-8.
This little beastie is found in a lot of studios and video sound stages and post production studios, because of it's clock distribution and sample rate synchronization capabilities, including with HD video sources. On top of that, a key differentiating feature is the ability to use an external 10MHz rubidium atomic clock as it's internal clocking/reference source for digital re-clocking. Rubidium clocks are very accurate, have very low phase noise and jitter (they're widely used in the telecom industry for these reasons, and that's how I got a used one on eBay- taken out of gear. The LPFRS module was originally made in Switzerland, and requires up to 4 A at 24V during startup, tapering off to about 3/4A during normal operation as the oscillator is up to temperature and stabilized.
This morning I finished making power cabled for the input power and and DC power connection from the OMRON SMPS I also bought used on eBay- this is a 24V 4A supply made in Japan, with good quality on the output and some nice control features- a little overkill, I suppose, but what the heck. My friend powers his off of batteries- most of his system runs off Lithium batteries with Paul Hynes regulators. See, I'm really just a pretty normal guy in comparison. :B
I've yet to mount the LPFRS on the heatsink I bought for it; gotta get the thermal interface material first. With an average power dissipation of about 18W, the heatsink is quite necessary. Eventually I'll probably replace this with a new Stanford Research PRS-10, which is what my colleague was using until he fried it (power supply misconnection) and had to send it back for repairs; then he got a used one on eBay! I think he has the SRS one back. By the specs, it has lower phase noise and better longer term performance. Will it make a difference? Don't know...
But this one is pretty good...
For the inquiring minds, here's the output of the oscillator terminated into 50 ohms.
Yeah, make fun of my antique HP scope, but it was cheap, and it is 2GSa/s sampling rate (2 Giga samples per second). And yeah, that's a CRT, but it's a digital display, not a conventional scope tube. No data export, which limits the usefulness, but it's OK for troubleshooting around the house, you know, when your toaster goes on the fritz!
Next step is mounting the heatsink, then it will be programming the DCD8 and testing with an iMac for now.
The DCD8 will be working in dual domain mode, with one domain coming from the server (what ever it is), and the other from the disk playback transport. The DCD8 also includes a high performance Firewire input, and I'll be using that for interfacing to Mac's for testing with my MBP 17, and interface with the late model 21" iMac.
more to come...
Comment