Algar_emi said:
Good experience. It seems the essential stuff is impedance matching that makes the most difference. Which really brings up the question what the impedance standard for analog interfaces of audio equipment is.
Hi soongsc,
For audio, there generally is no standard. For XLR connections, there is but it is ignored for the most part. Even the balanced part is not executed properly many times.
-Chris
For audio, there generally is no standard. For XLR connections, there is but it is ignored for the most part. Even the balanced part is not executed properly many times.
-Chris
SPDIF interface impedance is specified at 75 ohms, 0.5V amplitude.
XLR is specified at 110 ohms. You're right Anatech. These interfaces are not implemented correctly most of the time. Take note that most manufacturers output the SPDIF on 50 ohms BNC connectors to start with, or even worst RCA connector. The ML37 is using real 75 ohms connectors, as do my Belden cable. My two DAC input BNC are also using real 75 ohms connectors.
The inputs are then correctly matched pulse transformers.
In interesting detail is that Mark Levinson output the SPDIF at 2V instead of 0.5V. They say that by increasing the pulse risetime amplitude, it helps to reduce the jitter and improve the SPDIF signal recovery... Looking at the ML31.5 Reference Transport (looking like the ML37) outputs, they seem to have did it correctly, or more correctly that most CD players 😉
XLR is specified at 110 ohms. You're right Anatech. These interfaces are not implemented correctly most of the time. Take note that most manufacturers output the SPDIF on 50 ohms BNC connectors to start with, or even worst RCA connector. The ML37 is using real 75 ohms connectors, as do my Belden cable. My two DAC input BNC are also using real 75 ohms connectors.
The inputs are then correctly matched pulse transformers.
In interesting detail is that Mark Levinson output the SPDIF at 2V instead of 0.5V. They say that by increasing the pulse risetime amplitude, it helps to reduce the jitter and improve the SPDIF signal recovery... Looking at the ML31.5 Reference Transport (looking like the ML37) outputs, they seem to have did it correctly, or more correctly that most CD players 😉
While audio frequencies do suffer the same reflections and discontinuities from impedance mis-matches, due the low frequencies involved it takes miles of tranmission line length before it impacts at audio frequencies. My understanding ( as told by Bill Whitlock) is that is why 600 ohm audio transmission lines were adopted. The telephone company adopted 600 ohm lines because that was the average impedance of their transmission lines on the telephone poles. Since the phone company did have to deal with miles of transmission line length, it was important to match source, line and termination impedances to minimize VSWR/ reflections, and or echo. This might be a little tagential for this thread. This is more a topic for high speed edges (digital)
Hi Algar_emi,
I purposely answered soongsc's question. He asked about analog interfaces.
The standards for digital interfaces are adhered to more closely as you've pointed out. 😉
A standard should always be followed within it's tolerance, or it is something else.
-Chris
I purposely answered soongsc's question. He asked about analog interfaces.
The standards for digital interfaces are adhered to more closely as you've pointed out. 😉
I wonder if an interface will be designed that will be damaged or saturated by this non-conforming signal. A transformer that steps this up and possibly overdrives the circuit, or just plain saturates (cheap, but it should still work).
A standard should always be followed within it's tolerance, or it is something else.
-Chris
Hi mrshow4u,
Actually, our standards and circuits have more to do with telephone transmission than you want to know! 😉
-Chris
Actually, our standards and circuits have more to do with telephone transmission than you want to know! 😉
No, it's right on the money - every time someone uses a balanced line. Those basics are fundamental to how it all works and the reasons for doing things this way.
-Chris
Hi mrshow4u,
LOL
Okay, I'm guessing you've never seen a large recording studio!
In reality, the telephone system works amazingly well. The quality cut is for bandwidth reasons. To conserve bandwidth and keep your conversation clear, they cut the low and high frequencies that are only noise anyway (relating to voice). They also sample the audio at lower frequencies for a 56 K data rate before mixing it into a T1 (or higher data rate) data stream, 64 K if it's a PRI. Look!, digital audio. VoIP was already here, they just encapsulated the data differently. Frame relay was first.
I digress. Sorry. My point was that CD audio was a collection of pre existing technologies.
-Chris
LOL
Okay, I'm guessing you've never seen a large recording studio!
In reality, the telephone system works amazingly well. The quality cut is for bandwidth reasons. To conserve bandwidth and keep your conversation clear, they cut the low and high frequencies that are only noise anyway (relating to voice). They also sample the audio at lower frequencies for a 56 K data rate before mixing it into a T1 (or higher data rate) data stream, 64 K if it's a PRI. Look!, digital audio. VoIP was already here, they just encapsulated the data differently. Frame relay was first.
I digress. Sorry. My point was that CD audio was a collection of pre existing technologies.
-Chris
hahaha I was kidding. I used to work for a sound reinforcement company. DI boxes, Lo-z mics, and Jensen transformers were part of my world. Just having a laugh. It's a good topic. Before balanced lines were fashionable, I had suggested (as many others) to my H-Fi store boss using balanced line interfaces for Hi-Fi as they did in pro. He snidely said that "the pro market didn't require quality", and that "balanced lines ruin the sound". So the topic of cable sound goes way back. I wish Blumlein was around to join the topic.
Hey you guys! don't diss the phone company... they're called deci BELLS for a reason you know...
mrshow4u said:
The characteristic impedance of interoffice telephone cable is actually 150 ohms. Even in phone companies there are few who remember that any longer, just the old timers. The 600 ohm end user interface was chosen for different historical reasons which I can't quite recall at the moment.
I'm just newbie here but based on my experience with reliability/efficiency on my R&D days these are my observation between High-end IC & Low cost IC's:
1. Material - High-end normally have better materials (e.g. silver, gold plated, 2N/3N/4N purity etch) of course we need better materials especially coating like Silver & Gold for contact efficiency (corrosion) for a long time use not to mention your location condition like tropical countries with very humidity.
2. Insulation - Generally most of Hign-end IC's have better insulation not to mention shieldings (Tri/Quad shielding add the 80/90/99% braid)
- This not so effective since we are only runs IC mostly 1 to 1.5 meters
3. Build - Most of high-end parts particularly to RCA connectors have locking capability (of course for better connections) the other is how they attached the cables from the RCA jack. Commonly are using silver solder (which I think very important other than connectivity of RCA), others are using compression type (no weld) and the other one is using high current welding for the connections.
For me the most improtant is system matching especially for Silver or pure copper IC's. It's not the question of how much & how good and how well known is your IC, other thing than "Pride of ownership" Contentment or other words SATISFACTION is the word to follow, without it your US$12K IC's is nothing (IMHO)
1. Material - High-end normally have better materials (e.g. silver, gold plated, 2N/3N/4N purity etch) of course we need better materials especially coating like Silver & Gold for contact efficiency (corrosion) for a long time use not to mention your location condition like tropical countries with very humidity.
2. Insulation - Generally most of Hign-end IC's have better insulation not to mention shieldings (Tri/Quad shielding add the 80/90/99% braid)
- This not so effective since we are only runs IC mostly 1 to 1.5 meters
3. Build - Most of high-end parts particularly to RCA connectors have locking capability (of course for better connections) the other is how they attached the cables from the RCA jack. Commonly are using silver solder (which I think very important other than connectivity of RCA), others are using compression type (no weld) and the other one is using high current welding for the connections.
For me the most improtant is system matching especially for Silver or pure copper IC's. It's not the question of how much & how good and how well known is your IC, other thing than "Pride of ownership" Contentment or other words SATISFACTION is the word to follow, without it your US$12K IC's is nothing (IMHO)
anatech said:
This is why there is some much room for interconnect cables to create business. Just tune the cable enough to make a difference, if the customer likes it, then is sells. If something works better for a specific setup, buy new interconnects. Seems like lots of business as long as a difference can be heard, and people buy just to try.
Hi poobah,
Never would I diss the phone company. That is my current business. I think they did it right.
Hi rdf,
I can't remember the moment either, and I don't feel like looking it up. The important contributions were balanced lines and feedback controlled gain and distortion. Also the idea of DC current with audio superimposed to minimize contact issues. They packaged that up with line supervision. Cool. 😎
150 ohm impedance would be better from a noise pick up point of view.
Hi octaver,
Point #1 - I guess I shouldn't mention the interface between the jack and the wire to the board then the component. Or the plug to wire, or the base material to the plating. If it makes you feel better, then great. The rest is audio jewellery. If the materials are the same then so much the better. These things didn't seem to bother people in the 50's and 60's. (or the 70's for that matter)
Point #2 - not really. The rear of your gear (hey! That ryhmes) is where all the noisy AC power cables are. You just need one switchmode power suppy back there on top of everything else.
Point #3 - no. Those are simply destructive. You can get a gas tight seal with far lower tension. Then the jacks will stay inside the equipment. Jacks are getting harder to source with parts stock cut backs. I'm talking high density consumer grade.
-Chris
Never would I diss the phone company. That is my current business. I think they did it right.
Hi rdf,
I can't remember the moment either, and I don't feel like looking it up. The important contributions were balanced lines and feedback controlled gain and distortion. Also the idea of DC current with audio superimposed to minimize contact issues. They packaged that up with line supervision. Cool. 😎
150 ohm impedance would be better from a noise pick up point of view.
Hi octaver,
Point #1 - I guess I shouldn't mention the interface between the jack and the wire to the board then the component. Or the plug to wire, or the base material to the plating. If it makes you feel better, then great. The rest is audio jewellery. If the materials are the same then so much the better. These things didn't seem to bother people in the 50's and 60's. (or the 70's for that matter)
Point #2 - not really. The rear of your gear (hey! That ryhmes) is where all the noisy AC power cables are. You just need one switchmode power suppy back there on top of everything else.
Point #3 - no. Those are simply destructive. You can get a gas tight seal with far lower tension. Then the jacks will stay inside the equipment. Jacks are getting harder to source with parts stock cut backs. I'm talking high density consumer grade.
I can't see myself arguing with you on that one! 😉
-Chris
anatech said:
I never knew that and always assumed the DC was required for the ring tone circuit or housekeeping controls. Makes perfect sense. Surprising performance is achievable from a couple audio transformers and a line amp. We're feeding a transmitter over 2 km of dry copper house pair with a transformer strapped 600:150 on either end and a line amp pounding signal at the source. The run is bundled with phone and data yet still achieves around -85 dB signal to noise and a frequecy response 1 dB down at around 17 kHz.
The DC was also used to energize the relay that disconnected the ring generator... phone never rings off-hook!
The evolution of the phone system is a marvelous read. Yeah, they had a few happy accidents (dumbass luck)... a lot of brilliant engineering though.
🙂
The evolution of the phone system is a marvelous read. Yeah, they had a few happy accidents (dumbass luck)... a lot of brilliant engineering though.
🙂
poobah said:
Can you recommend any books in particular? I might as well get something useful out of this thread... 😉
- Status
- Not open for further replies.