|
Computer Audio Asylum Music servers and other computer based digital audio technologies. |
|
|
|
Computer Audio Asylum Music servers and other computer based digital audio technologies. |
|
In Reply to: Re: I don't think any USB audio devices have a master clock posted by John Swenson on September 7, 2005 at 23:54:14:
Hi John;
Thanks for the explanation. My understanding of syncronous mode problem as described by the BB designer is this - the USB receiver is phase locked directly to the USB data, with a very short PLL time constant. It needs to be to lock ok, within 2 - 1ms packets.
The drawback, as described, is the USB receiver will have an instantaneous PLL clock change when the number of samples per 1ms packets changes (for redbook the USB transmitter needs to put out precisely 441 samples every 10ms). So you need a short PLL integration time to lock, but are then penalized with it by high jitter. If I understand his example, without the 2706 fancy circuit, the USB receiver pll clock literally runs at 44,000 Hz for 9 ms, then speeds up to 45,000 Hz for 1ms. This is where the horrendous 100Hz jitter comes from.My understanding of the BB solution was to "switch gears" so to speak, and once locked, (using the short analog time constant PLL), go to a long integration time constant, implemented in digital logic, with such a time constant, that it eliminates jitter in the audio range.
You could do this, for example, if the digital logic PLL integration time was 1 second. This would correctly count the data samples/sec provided by the USB transmitter, and only very slowly change the local sample rate to allow for drift in the USB transmit clock, etc.
This is interesting to me, as the same technique is used in Yamaha CSL FM tuners, where the RF local oscillator feedback tunes the normal synthesizer PLL to start, but the switches over to a long time contant PLL implemented in logic by an onboard processor. It seems good ideas are reborn elsewhere by clever people.On the SB2 - I agree using a direct I2S or similar (without the shifted bit...) is the best way to drive a DAC chip from a low jitter receiver, whatever that may be. No reason to bring S/PDIF back in to screw up a good thing, but for "standardization" it frequently happens. I just wish there had evolved a better version of the standard S/PDIF, with re-clocked differential data over twisted pair. Or providing a 2nd line coax line with the master clock became more popular. There are lots of ways to improve it, (too bad the I2S connector movement sorta died) instead of staying with the 1980 RCA or toslink connector implementation...
Bob
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Re: I don't think any USB audio devices have a master clock - BFitz 09/8/0509:10:31 09/8/05 (3)
- Re: I don't think any USB audio devices have a master clock - John Swenson 16:56:08 09/8/05 (0)
- Two stage PLL - Monkey Bones 09:45:09 09/8/05 (1)
- Re: Two stage PLL - BFitz 14:12:21 09/8/05 (0)
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]