Wow, I'm impressed!
The two things I know is that perfect phase is HARD to hear, like was said here, and in the studio, unless it is David Chesky, they are not recorded minimum phase anyway, because of the overdubbing etc. If I could correct one thing that modern engineers mess up, it would not be phase, it would be all of that COMPRESSION that they use. For instance, IMO, Mandy Moore turned out to be a WONDERFUL singer, but her, "Coverage" C.D. sounds absolutely TERRIBLE.
The two things I know is that perfect phase is HARD to hear, like was said here, and in the studio, unless it is David Chesky, they are not recorded minimum phase anyway, because of the overdubbing etc. If I could correct one thing that modern engineers mess up, it would not be phase, it would be all of that COMPRESSION that they use. For instance, IMO, Mandy Moore turned out to be a WONDERFUL singer, but her, "Coverage" C.D. sounds absolutely TERRIBLE.
Jan posted some files from Hawksford- you might do some searching for these. Frankly, it's so easy to create your own that I'd recommend putting your effort in that direction. That way, you can play with frequency, amount of shift, and the like to see what things you're sensitive to and what you're not. In conjunction with foobar's ABX, it can be a very... ear-opening experience.
I just completed the SPDIF input mod on my DCX - replacing the AES/EBU xformer with a couple of MKT caps and a 75 ohm resistor. Removed the 110 ohm resistor across RXP and RXN, and connected a digital ground to the coax shield.
Sadly, it's still regularly going into the weird 'bug' mode. Should I have put the 75ohm resistor on the CS8420 side of the caps, or am I right in putting it on the coax side?
Sadly, it's still regularly going into the weird 'bug' mode. Should I have put the 75ohm resistor on the CS8420 side of the caps, or am I right in putting it on the coax side?
The 'weird bug' mode is not caused by the input config. It is an issue caused by the sensitivity of the receiver chip to some R-C values on the loop filter.
If you Google 'CS8420' you can find a bug report that details the changes to fix it.
Jan
Looking at the recommened filter component tables on pages 88 - 89 of the datasheet pdf, I can see that the DCX uses some quite different parts, so these might be worth changing. I'm not sure how I'll go obtaining the required surface mount caps and resistors, but I will report back if I have any success!
I took a different route to solving this problem. I'm currently feeding the unit from a modded USB-SPDIF converter, on which I've removed the voltage divider that would normally lower the 5v digital level to SPDIF voltage levels. So it's getting the full 5vpp(?) from the 74hc04 chip that follows the main converter IC.
So far it has worked perfectly even at 96k where it is normally really buggy. Fingers crossed!
So far it has worked perfectly even at 96k where it is normally really buggy. Fingers crossed!
Very nice,
Could you please elaborate this little bit more - complete path from USB converter to receiver, as i try to work 96khz for YEARS without success.
many thanks
It's a Tenor TE7022 ebay usb-spdif board, an LJM design. The 74hc04 on the board (I assume) buffers the stream output from the TE7022 chip. There's a couple of surface mount resistors next to the spdif transformer - the 91 ohm to ground is removed and the series 150 ohm is jumpered. This sends the full ttl level to the spdif output.
(This step may be unnecessary with the high output spdif mod above) On the DCX side I have removed the AES digital input transformer on the DSP board, and replaced it with a couple of .01uf MKT caps with a 75ohm resistor across these on the input side. Digital earth is taken from a nearby board point and connected to the capacitor on the ground side of the input. A length of coax is connected directly to the capacitors and becomes the new spdif input connection.
On the underside of the DCX DSP board, the traces connecting the 110 ohm resistor across the input are broken to disconnect the resistor. The old input traces from the board header are also broken. This direct input mod wasn't my idea - I borrowed it from a few other sites.
Some computer motherboards have a spdif header which actually outputs 5vpp ttl level. This is what I first fed the DCX with when I bought it years ago, not realising it was not standard spdif voltage, and it always worked well. When I moved to a different spdif card the troubles started.
The other thought I had was to install a toslink receiver module in the DCX and feed the TTL output directly into the CS8420 chip.
Personally, stability at 96k isn't a huge concern as I doubt there would be any difference to feeding 44.1 or 48k, sound wise. It's a good test though, as this is where it usually struggles.
(This step may be unnecessary with the high output spdif mod above) On the DCX side I have removed the AES digital input transformer on the DSP board, and replaced it with a couple of .01uf MKT caps with a 75ohm resistor across these on the input side. Digital earth is taken from a nearby board point and connected to the capacitor on the ground side of the input. A length of coax is connected directly to the capacitors and becomes the new spdif input connection.
On the underside of the DCX DSP board, the traces connecting the 110 ohm resistor across the input are broken to disconnect the resistor. The old input traces from the board header are also broken. This direct input mod wasn't my idea - I borrowed it from a few other sites.
Some computer motherboards have a spdif header which actually outputs 5vpp ttl level. This is what I first fed the DCX with when I bought it years ago, not realising it was not standard spdif voltage, and it always worked well. When I moved to a different spdif card the troubles started.
The other thought I had was to install a toslink receiver module in the DCX and feed the TTL output directly into the CS8420 chip.
Personally, stability at 96k isn't a huge concern as I doubt there would be any difference to feeding 44.1 or 48k, sound wise. It's a good test though, as this is where it usually struggles.
Is DCX2496 improved during the years?
How is the sound quality compared with f ex DEQX, Groundsound and MiniDSP HD4x10?
Will it be any gain/impedance/noise/quality issues to drive 2 x Pass Labs XA30.5?
I have a decent turntable (Pro-Ject Extension Evolution, Pro-Ject balanced Reference DAC and Ortofon Cadenza Black) , how is the quality of the AD input?
At last, i have SONOS With SPDIF output, how do i Connect it to the AES/EBU input?
How is the sound quality compared with f ex DEQX, Groundsound and MiniDSP HD4x10?
Will it be any gain/impedance/noise/quality issues to drive 2 x Pass Labs XA30.5?
I have a decent turntable (Pro-Ject Extension Evolution, Pro-Ject balanced Reference DAC and Ortofon Cadenza Black) , how is the quality of the AD input?
At last, i have SONOS With SPDIF output, how do i Connect it to the AES/EBU input?
No, it is definitely not.
A DEQX beats it hands down IMHO, a NAJDA does also.
There are at least gain issues.
There is a lot about that in this thread.
So, I got one of these units to tinker with. Does anyone have any of the images that used to be here: passive output stage for DCX2496 tweakanalog5.php and all the rest of the pages seem to be missing. I sent an e-mail to ask, but that was a few days ago. I'm pretty sure it's abandoned.
I'm pretty sure I have the board for the passive output built properly, but the information that's available now is so sketchy, the yahoo group is gone, and most of the pages are at least half gone. I'd really love to see (without digging through 368 pages of mostly dead links) any images of the ribbon cable tapped into or really anything from their passive modded unit.
I'm pretty sure I have the board for the passive output built properly, but the information that's available now is so sketchy, the yahoo group is gone, and most of the pages are at least half gone. I'd really love to see (without digging through 368 pages of mostly dead links) any images of the ribbon cable tapped into or really anything from their passive modded unit.