Hand Made SATA Cable for CAT

Status
Not open for further replies.
AGREE totaly on the clock misunderstanding:)




Sorry to shatter your delusions. but the whole point of digital is the 1s and 0s, and getting the signal from the transmitter to the receiver.
:D


Sorry to shatter your perception. You haven't shattered a thing. I am talking about noise pickup and its passing out to the DAC and analog stages, you are talking about 1s and 0s. :D
 
Last edited:
Sata cable provides a higher bandwidth, more dynamic, more clarity and better musical

Thats just ridicules on so many levels. Does your magic SATA cable also make pictures clearer, spreadsheets have data with more decimal places, and word documents have a conclusion that makes sense? No it dosnt, and it dosnt make any difference with audio. Its all just data. The cable dosnt care if its audio or text. (Timing is immaterial, the data is buffered (in the ram) and timed by the computer). This is equivalent to buffering the whole song turning the SATA drive off and then playing the song out of RAM. How much will the SATA cable effect the sound now.
 
Last edited:
I can't believe I'm falling for this thread again.
If anybody observes that a SATA cable makes a difference to their sound, then their is something seriously wrong with their system.
If you worry about noise radiating from your SATA cable, then sort your DAC PSU and shielding out.
There is no way that the clocking on a SATA interface is at all linked to the master clock of a DAC.

Does anybody worry about other digital transfer mechanisms? I've been thinking about Wifi over nitrogen enriched air. Thing is, it's a sod to keep in the house, and subtly alters the character of my speakers. Mind you, after all the effort, the difference is subtle, because I'm close to asphixiation during listening sessions.
 
Raj, we are talking about a digital interface.
The information was given freely, there is no cost or implication of a cost. The links are to the world rennowed experts in the field of EMC and signal integrity. If the noise is affecting other circuitry it is an EMC issue. Both Henry Ott and Keith Armstrong are your main sources of information here, read some of the stuff, it covers all frequencys of electronics.
Doing PCB's for communication systems and working as part of a team thats spent the last 4 years desiging the next generation system for a very big customer, suprisingly has given us all (in the team) an insight into how digial and analogue systems interact and how the noise from one can effect the other.
But in the digital domain 1s and 0s are king, and the SATA interface is digital.
Digital circuitry is noisy, to an extent, but most of the noise is very high frequency, way outside the audio range. This allows it to be filtered, inter stage CLC combinations are very good. But the only way to totaly isolate your PC form your DAC is galvonic isolation. But that still leaves the DAC where digital and analogue have to co-exist in the same chip, never mind the same board. Isolating the PC is a ggod move and the best way to electronicly isolate it is by using a optical fibre interface, or use a squeezebox wirelessly, youve then isolated the very noisy PC from your DAC, then there is only the local noise to worry about, and that will be a lot less than a PC generates. Another way is to isolate the digital using an ADUM device or a custom planar pulse transformer with shield layers to stop the noise coupling capacitivley.
But playing around with esoteric SATA cables isn't gonna have the magical effect claimed on sound repreduction...
Most PC boards are designed with a minimum number of layers for cost purposes, the interface chips are designed to allow for routing on 6 layers in some cases. That is why you see the signal traces on the top of the board, especially running to the DDR memory. That is not best practice, all the signals are better run on inner layers as stripline, with a contigous unbroken ground plane next to them, and all power planes with a closely coupled ground plane (sub 0.1mm dialectric between the planes). So getting the best layed out motherboard you can with a well speced power supply is the best bet. HDI (high density interconnect) PCB motherboards would be best, I dont know if any mainstreem PC's are designed this way, but a lot of smaller mobile devices and critical electronics (mil/med/aero) use this way of doing digital PCB's to minimise the noise generated by the digital electronics, maximise EMC compatability and improve signal integrity.
So instead of using sticking plasters cure the problem.
 
Raj, we are talking about a digital interface.
The information was given freely, there is no cost or implication of a cost. The links are to the world rennowed experts in the field of EMC and signal integrity. If the noise is affecting other circuitry it is an EMC issue. Both Henry Ott and Keith Armstrong are your main sources of information here, read some of the stuff, it covers all frequencys of electronics.
Doing PCB's for communication systems and working as part of a team thats spent the last 4 years desiging the next generation system for a very big customer, suprisingly has given us all (in the team) an insight into how digial and analogue systems interact and how the noise from one can effect the other.
But in the digital domain 1s and 0s are king, and the SATA interface is digital.
Digital circuitry is noisy, to an extent, but most of the noise is very high frequency, way outside the audio range. This allows it to be filtered, inter stage CLC combinations are very good. But the only way to totaly isolate your PC form your DAC is galvonic isolation. But that still leaves the DAC where digital and analogue have to co-exist in the same chip, never mind the same board. Isolating the PC is a ggod move and the best way to electronicly isolate it is by using a optical fibre interface, or use a squeezebox wirelessly, youve then isolated the very noisy PC from your DAC, then there is only the local noise to worry about, and that will be a lot less than a PC generates. Another way is to isolate the digital using an ADUM device or a custom planar pulse transformer with shield layers to stop the noise coupling capacitivley.
But playing around with esoteric SATA cables isn't gonna have the magical effect claimed on sound repreduction...
Most PC boards are designed with a minimum number of layers for cost purposes, the interface chips are designed to allow for routing on 6 layers in some cases. That is why you see the signal traces on the top of the board, especially running to the DDR memory. That is not best practice, all the signals are better run on inner layers as stripline, with a contigous unbroken ground plane next to them, and all power planes with a closely coupled ground plane (sub 0.1mm dialectric between the planes). So getting the best layed out motherboard you can with a well speced power supply is the best bet. HDI (high density interconnect) PCB motherboards would be best, I dont know if any mainstreem PC's are designed this way, but a lot of smaller mobile devices and critical electronics (mil/med/aero) use this way of doing digital PCB's to minimise the noise generated by the digital electronics, maximise EMC compatability and improve signal integrity.
So instead of using sticking plasters cure the problem.



Marce, my posts were elduing to this in the first place. I work for ASUS so I know a thing or two about what can and cannot be done (depends on how well the fly-by and MRC can be adopted and R&D time). FWIW we don't get control access to all of the signal skew mechanisms any more or some chipsets provide an offset mechanism with an automated base. I think you'll understand what I mean if you know memory systems well.

The DQ pairs for memory are separated on the top and lower sides of the boards (some of the CMD and CTRL lines also). Top end boards are now 8 layer(have been for quite some time). The chipsets have limitations and base layout recommendations plus tolerance margins are often well cited in development only white papers :)

Anyway sorry for not stating my inital post clearly - my point was there are infinite variables that add up to why some people can make out a change. And many people choose to mod as once they have the eqpmnt, resources are limited, we have to accept that. Plus the fact that some love tinkering...


Bear in mind that I am not an advocate for silver SATA cables, but I do understand that noise can be an issue in PCs (for various reasons) and that not all outboard solutions feature galvanic isolation. :)

-Raja
 
Last edited:
:D
We are arguing the same points from different directions I believe.
Having done some PC boards with 6 layers it does make me weep, similar high reliability and quite boards (DDR interface, PC cored FPGA's ethernet etc) we used 14 layers and even that was a compromise, 16 would have been better, or HDI, the next generation will be HDI.
My appologies for going overboard, but usually when 1s and 0s are mentioned in that way a wierd theory follows, with some horendous mods to a digital circuitry, quite often with leaded decoupling audiophile caps involved.:)
 
It's ok beating these things out, I come here because I learn from others, and sometimes that means having to put in effort to get to the point. It takes all sorts to make up a forum community, and it's often where the disagreements or misundertandings start that one learns the most.

-Raja
 
86570d6f2678b17d25ccc986fc838e7e_7b0.jpg

Second generation SATA cable have the conductor of the newly developed
Design for High End Music server


e2d0306057deff69be90ebe45627f425_7b0.jpg

The unique structure to create unparalleled dynamic & liveness than first generation


735f8fbc6005ccee931480c9addf3b83_7b0.jpg

Rugged connectors to improve reliability
 
Status
Not open for further replies.