I'm building my first TDA1543 dac and I wonder if it should be better to use AES/EBU transmission between transport and dac. If I understand I could use 110 ohm CAT5 twisted pair instead of 75 ohm coaxial cable just replacing 75 ohm resistor on receiver side with a 110 ohm resistor.
From Wikipedia:
"The digital audio standard frequently called AES/EBU, officially known as AES3, is used for carrying digital audio signals between various devices. (...) A related system, S/PDIF, was developed essentially as a consumer version of AES/EBU, using connectors more commonly found in the consumer market. (...) The AES3 standard parallels part 4 of the international standard IEC 60958. Of the physical interconnection types defined by IEC 60958, three are in common use:
* IEC 60958 Type I Balanced – 3-conductor, 110-ohm twisted pair cabling with an XLR connector, used in professional installations (AES3 standard)
* IEC 60958 Type II Unbalanced – 2-conductor, 75-ohm coaxial cable with an RCA connector, used in consumer audio
* IEC 60958 Type II Optical – optical fiber, usually plastic but occasionally glass, with an F05 connector, also used in consumer audio
From Wikipedia:
"The digital audio standard frequently called AES/EBU, officially known as AES3, is used for carrying digital audio signals between various devices. (...) A related system, S/PDIF, was developed essentially as a consumer version of AES/EBU, using connectors more commonly found in the consumer market. (...) The AES3 standard parallels part 4 of the international standard IEC 60958. Of the physical interconnection types defined by IEC 60958, three are in common use:
* IEC 60958 Type I Balanced – 3-conductor, 110-ohm twisted pair cabling with an XLR connector, used in professional installations (AES3 standard)
* IEC 60958 Type II Unbalanced – 2-conductor, 75-ohm coaxial cable with an RCA connector, used in consumer audio
* IEC 60958 Type II Optical – optical fiber, usually plastic but occasionally glass, with an F05 connector, also used in consumer audio
AES/EBU is only really going to show its advantages if you're sending it over long distances, via cables that aren't optimised for digital audio, and are susceptible to noise pickup.
Okay, thank you. So I'll have:
cdplayer digital out (RCA connector) --> digital cable 75ohm --> DAC input connector (BNC 75ohm)
But how to connect BNC to CS8414 (inside the DAC)? Do I need a piece of 75 ohm coaxial or could I use a simple copper wire?
cdplayer digital out (RCA connector) --> digital cable 75ohm --> DAC input connector (BNC 75ohm)
But how to connect BNC to CS8414 (inside the DAC)? Do I need a piece of 75 ohm coaxial or could I use a simple copper wire?
If your cable is short - say a few tens of cm, ordinary co-axial cable - audio line level, or microphone screened cable, will work fine. The frequencies used for SPDIF aren't high enough that characteristing impedance of cable is important. You may need terminating resistors though, because the line driver of the source wants to 'see' 75R.
does the line driver see 75ohms or 150ohms? The receive end of the cable is terminated with 75ohms to ground and the transmit end is fed from a 75ohm source impedance. The driver sees this as a total of 150ohms.
If the output of the DAC is balanced, then I would preserve the balanced signal as long as possible.
If the output of the DAC is balanced, then I would preserve the balanced signal as long as possible.
Your description sounds right, although the 'driver' has an inherent source impedance of 75R from the sound of it, in which case it is correct to terminate it with 75R at the 'far end'.
So, yes, ultimately if you consider the source as an ideal Thevenin source with a series resistor of 75R, then the ideal source 'sees' 150R in total.
The OUTPUT of the DAC is a completely different signal and system to the SPDIF part - I'm not sure if your two queries are linked in thought.
The balanced analogue link is most useful if your preamp has a balanced line in. Otherwise you will need a balun somewhere, and it is best put as close to the pre-amp input as possible (an audio grade transformer here would work well). Balanced lines are useful if you have electrically noisy environments or long cables; otherwise the added conversion of bal-to-unbal may degrade the sound more than just using the unbal out on the DAC.
I have a parallel 75 ohm resistor + series 10nF capacitors just before CS8414. Between input BNC connector and CS8414 I need just 10 cm wire so any coaxial will works (also a 50 ohms one?).
Please let me ask another question:
do digital interconnect quality affects sound quality in the same way this happens with analogue interconnects?
Not in the same way, no. But they sure can effect it! Since the receiver chip has to derive the clock from the spdif or AES signal, bad cables can make a mess of it. I keep a few around here just for demo purposes. Both coax and optical.
The most expensive cables are not always the best, either. Radio Shack actually sells some decent ones.
The most expensive cables are not always the best, either. Radio Shack actually sells some decent ones.
So how select the right one? Electrical properties? Conductor and/or shield material? ...?
If you believe the crazy people, digital interconnects affect sound quality just as much as power cords and magic rocks and crystals do. What I wonder is, how come these marvelously subtle influences are almost always positive. How long before some crazed audiophile begins hanging human ears around the room to improve the listening experience or electrocutes himself putting Cramolin on the pole transformer on the street. I also wonder how I can cash in on these wonderfully gullible people.
Anyway...
If you're DIYing the whole chain, consider using something like I2S which includes a separate line for clock as well as data. But if the DAC uses an ASRC or something else that removes jitter, that's probably overkill.
I agree completely, many do sound different and it's not very subtle sometimes. I make my own now, tried Monster, Dayton audio, Radio Shack, Philips,and a couple others. I'm too much of a cheapskate to buy a $200 boutique cable but I assure you cables make a difference. Buy a few cheap ones to try and trust your ears, that's what matters anyway. Check out www.bluejeanscables.com .
Best, Bill
... or get a decent DAC. A DAC that lets you hear all sorts of different cables isn't that good, is it?
jd
Well, yes, I was trying to get some people to realize that the search for the inaudible cable is over once you have a decent DAC. You can also diy a DAC, no?
I hoped it was helpfull, but if not, I apologize.
jd
No apology is necessary, and I agree with you for the most part. Apparently the reciever chips and their implementation are improving so much these days they are more tolerant of imperfect signals. I don't know beans about digital but my ears know audio pretty well, and I still hear differences with the cheaper cables. I'm sure there is science behind it but I'm not well versed in digital technology.
Best, Bill
I think it's just logic. If the cable causes audible differences in your DAC, obviously your DAC is sensitive to RF or jitter or what. Might be worthwhile to try out a different DAC, where those cables don't make any or much less differences. Unless of course you sell cables 😉
jd
Well, yes, I was trying to get some people to realize that the search for the inaudible cable is over once you have a decent DAC.
I wish... same goes for USB cables. It is sort hierarchical. Is the DAC front end good enough not to spit stuff back into the interface causing EMI and jitter. Is the DAC terminated with a decent resistive load impedance of the correct value. Is the cable the right impedance and direction.
I spent several years designing (and listening was a BIG part) SPDIF cables, pulse transformers, differential analog buffers to reject common noise and minimize signal reflections. The better you get the interface the more you can hear the difference between good cables. You know we went through all this before with large data FIFOs in some of the Mark Levinson equipment and Genesis. You could still hear cable differences. Hell it took years before people stopped making 1/2 meter cables ( the kiss of death for the refections to show up at the logic transistion of HCMOS drivers. The devil in the design details.
The SPDIF interface is like the weather... every one complains but nobody does anything about it. Or they put 50 ohm BNC connectors with 9 inches of twisted pair wire to the unbuffered SPDIF receiver and brag that they don't use those nasty RCA connectors.
I wish... same goes for USB cables. It is sort hierarchical. Is the DAC front end good enough not to spit stuff back into the interface causing EMI and jitter. Is the DAC terminated with a decent resistive load impedance of the correct value. Is the cable the right impedance and direction.
I spent several years designing (and listening was a BIG part) SPDIF cables, pulse transformers, differential analog buffers to reject common noise and minimize signal reflections. The better you get the interface the more you can hear the difference between good cables. You know we went through all this before with large data FIFOs in some of the Mark Levinson equipment and Genesis. You could still hear cable differences. Hell it took years before people stopped making 1/2 meter cables ( the kiss of death for the refections to show up at the logic transistion of HCMOS drivers. The devil in the design details.
The SPDIF interface is like the weather... every one complains but nobody does anything about it. Or they put 50 ohm BNC connectors with 9 inches of twisted pair wire to the unbuffered SPDIF receiver and brag that they don't use those nasty RCA connectors.
Try a Benchmark DAC. No difference between boutique cables and anything else that maintains a galvanic connection.
But hey, I'm not in this stuff, so I'll shut up.
jd
yes, if the cable changes the sound from the transmitter or into the receiver then it's not the cable that is at fault. Find and cure the problem with the electronics either side of the cable.
This applies just as much to analogue as to digital transmission.
Unless the cable is so far from the norm that it is deliberately EQing the signal transfer.
This depends on how the DAC deals with imperfection of the signal or transmission-induced noise. If the DAC recovers the embedded clock of AES/EBU or S/PDIF, then it is highly likely that the reconstructed analog waveforms will be different, hence sonically discernable.
Airforce rader-watching personnels are trained to detect any suspecious objects on the radar screen. Similarly, Navy sonar-monitoring personnels can distinguish whale's speech from military artifacts.
This is about human's signal-noise detection performance.
"The detection threshold is defined as "the signal minus the noise level required" "there are any underwater noise sources in the environment,but nothing can replace the human operator." "A human interpretation is also required with minimum visible signal (MVS) and tangential sensitivity. Invariably, the received signal power is competing with some interfering signal in order to be detected or recognised."
Trained human ears are so smart. Be proud guys ! But performance also increases through effective training.
PS Your dog can detect your foot-steps in the dark. 😀
Last edited:
- Status
- Not open for further replies.