| Home |
| Search |
| Today's Posts |
|
#1
|
|||
|
|||
Is the impedance of a balanced cable less than the impedance of an unbalanced one?
Sorry to waste time on such a basic question - but I'm just not understanding.
I tend to run balanced cables for just about everything - unbalanced only occasionally for very short cables. But I thought the prime issue was external noise coupling into the cable. Do balanced cables inherently have a lower impedance than unbalanced? Thanks MUCH, -lee- |
|
#3
|
|||
|
|||
|
In article ,
Leoaw3 wrote: Sorry to waste time on such a basic question - but I'm just not understanding. I tend to run balanced cables for just about everything - unbalanced only occasionally for very short cables. But I thought the prime issue was external noise coupling into the cable. Do balanced cables inherently have a lower impedance than unbalanced? No. This is discussed in the FAQ... the idea with balanced cables is basically that there are two opposing signals being sent down the line, and what the input looks at is the difference between them. When noise is induced, it is induced into both lines, and so it gets cancelled out. Impedance has nothing to do with it... you can have high-Z balanced lines and low-Z unbalanced lines (although they are not as common). --scott -- "C'est un Nagra. C'est suisse, et tres, tres precis." |
|
#4
|
|||
|
|||
|
The "characteristic impedance" of a cable -- balanced or unbalanced -- is
determined by a number of factors that are under control of the cable designer. One type of cable is not inherently higher- or lower-impedance than the other. (In fact, I've never seen a balanced cable with a spec'd characteristic impedance.) The characteristic impedance is not generally seen as a significant performance factor at audio frequencies. Do not confuse the cable's characteristic impedance with the resistance of its conductors. They have nothing to do with each other. |
|
#5
|
|||
|
|||
|
Leoaw3 wrote:
Sorry to waste time on such a basic question - but I'm just not understanding. I tend to run balanced cables for just about everything - unbalanced only occasionally for very short cables. But I thought the prime issue was external noise coupling into the cable. Do balanced cables inherently have a lower impedance than unbalanced? The notion of a cable's impedance is meaningless until its length is a significant fraction of the wavelengths of the signals it carries. At audio frequencies this length is on the order of 10's of kilometers. -- ================================================== ====================== Michael Kesti | "And like, one and one don't make | two, one and one make one." | - The Who, Bargain |
|
#6
|
|||
|
|||
|
"Scott Dorsey" wrote ...
Jay Kadis wrote: I attended Bill Whitlock's tutorial at the AES and this is at odds with his definition of a balanced system. He says that it is the balance between source and load impedances that determines whether a system is balanced or not and that the differential nature of the signals is irrelevent. That's very much true... as long as the impedances are the same, the noise will be induced in the two lines in the same way. And it is the differential_input_ subtracting the noise common to the two lines that makes the system balanced. Jay seems to be talking about impedance matching between OUTput and INput, and Scott seems to be talking about impedance matching between the two differential input nodes. My study and experience agrees more with Scott's explanation than with Jay's (and his interperetation of Whitlock). In modern solid-state equipment where output impedances are quite low, why does the output impedance match with the input impedance have anything to do with noise rejection? It seems to be almost entirely up to the input to reject whatever inappropriate signal. And isn't Whitlock's POV from a world of equipment which uses proper transformers at each input and output? Alas not likely to be found in the real world. |
|
#7
|
|||
|
|||
|
In modern solid-state equipment where output impedances
are quite low, why does the output impedance match with the input impedance have anything to do with noise rejection? It seems to be almost entirely up to the input to reject whatever inappropriate signal. For ease of analysis, imagine a 50-ohm transformer output feeding a 200-ohm transformer input through a balanced cable. It should be "obvious" (???) that "going around the loop," the total impedance is the same for each line of the cable. The current induced by the noise field should be identical in each line. |
|
#8
|
|||
|
|||
|
In article ,
"Richard Crowley" wrote: "Scott Dorsey" wrote ... Jay Kadis wrote: I attended Bill Whitlock's tutorial at the AES and this is at odds with his definition of a balanced system. He says that it is the balance between source and load impedances that determines whether a system is balanced or not and that the differential nature of the signals is irrelevent. That's very much true... as long as the impedances are the same, the noise will be induced in the two lines in the same way. And it is the differential_input_ subtracting the noise common to the two lines that makes the system balanced. Jay seems to be talking about impedance matching between OUTput and INput, and Scott seems to be talking about impedance matching between the two differential input nodes. My study and experience agrees more with Scott's explanation than with Jay's (and his interperetation of Whitlock). No, I meant both the source/load balance AND the matching of the inverting/noninverting impedances. In modern solid-state equipment where output impedances are quite low, why does the output impedance match with the input impedance have anything to do with noise rejection? It seems to be almost entirely up to the input to reject whatever inappropriate signal. The balance between the source and load impedances IS critical in maintaing the CMRR of the differential amplifier, but in the sense of matching both the inverting and non-inverting impedances of the Wheatstone bridge that is created. And isn't Whitlock's POV from a world of equipment which uses proper transformers at each input and output? Alas not likely to be found in the real world. Bill also has a patent on a bootstrapped balanced amplifier circuit, the IC implementation of which was announced at the AES. -Jay -- x------- Jay Kadis ------- x---- Jay's Attic Studio ------x x Lecturer, Audio Engineer x Dexter Records x x CCRMA, Stanford University x http://www.offbeats.com/ x x---------- http://ccrma.stanford.edu/~jay/ ------------x |
|
#9
|
|||
|
|||
|
On Fri, 05 Nov 2004 14:16:16 GMT, (Don
Pearce) wrote: On 05 Nov 2004 13:57:18 GMT, ospam (Leoaw3) wrote: Sorry to waste time on such a basic question - but I'm just not understanding. I tend to run balanced cables for just about everything - unbalanced only occasionally for very short cables. But I thought the prime issue was external noise coupling into the cable. That's an important issue, but that also assumes the inputs and outputs are balanced, as well as the cable. A lot of the 'semi-pro' gear such as the Delta 44 and 66 cards are not actually balanced, though the marketing may lead one to believe they are. But even so, long runs of unbalanced cables (or more correctly, unbalanced systems/connections, which may use "balanced" cables) often don't pick up much noise in many circumstances. YMMV. Do balanced cables inherently have a lower impedance than unbalanced? Thanks MUCH, -lee- In general the impedance of a balanced cable will be higher than that of an unbalanced one. For practical purposes, audio cables themselves have infinite impedance (perhaps the only notable exception is an guitar without an internal preamp driving a long cable). It's the impedance of the source and destination that count (and where the above generality is true), not the cable. d Pearce Consulting http://www.pearce.uk.com ----- http://mindspring.com/~benbradley |
|
#10
|
|||
|
|||
|
Don Pearce wrote:
On 05 Nov 2004 13:57:18 GMT, ospam (Leoaw3) wrote: Sorry to waste time on such a basic question - but I'm just not understanding. I tend to run balanced cables for just about everything - unbalanced only occasionally for very short cables. But I thought the prime issue was external noise coupling into the cable. Do balanced cables inherently have a lower impedance than unbalanced? Thanks MUCH, -lee- In general the impedance of a balanced cable will be higher than that of an unbalanced one. What drugs are you on ? At audio frequencies a cable has no 'characteristic impedance'. The wavelength at audio frequencies is far too long relative to typical cable lengths. What's more, many balanced circuits operate at *lower* impedances than unbalanced ones ! Graham |
|
#11
|
|||
|
|||
|
Jay Kadis wrote:
I attended Bill Whitlock's tutorial at the AES and this is at odds with his definition of a balanced system. He says that it is the balance between source and load impedances that determines whether a system is balanced or not and that the differential nature of the signals is irrelevent. The differential nature of the signals is hardly *irrelevant* ! Regarding 'balance' between source and load, I'm not sure what you're referring to. Certainly 'matched impedance' ( load and source Z equal ) is *very* old hat and unknown in typical pro-audio gear. It *is* important for good balance that the load impedance is symmetrical about ground, which isn't the case sadly with most 'balanced' inputs using a single op-amp and equal value resistors for the inverting and non-inverting legs. Graham |
|
#12
|
|||
|
|||
|
Richard Crowley wrote:
"Scott Dorsey" wrote ... Jay Kadis wrote: I attended Bill Whitlock's tutorial at the AES and this is at odds with his definition of a balanced system. He says that it is the balance between source and load impedances that determines whether a system is balanced or not and that the differential nature of the signals is irrelevent. That's very much true... as long as the impedances are the same, the noise will be induced in the two lines in the same way. And it is the differential_input_ subtracting the noise common to the two lines that makes the system balanced. Jay seems to be talking about impedance matching between OUTput and INput, and Scott seems to be talking about impedance matching between the two differential input nodes. My study and experience agrees more with Scott's explanation than with Jay's (and his interperetation of Whitlock). In modern solid-state equipment where output impedances are quite low, why does the output impedance match with the input impedance have anything to do with noise rejection? It seems to be almost entirely up to the input to reject whatever inappropriate signal. And isn't Whitlock's POV from a world of equipment which uses proper transformers at each input and output? Alas not likely to be found in the real world. Alas ? You *want* transformers messing up the signal every time it leaves or enters equipment ? Graham |
|
#13
|
|||
|
|||
|
In article ,
Pooh Bear wrote: Jay Kadis wrote: I attended Bill Whitlock's tutorial at the AES and this is at odds with his definition of a balanced system. He says that it is the balance between source and load impedances that determines whether a system is balanced or not and that the differential nature of the signals is irrelevent. The differential nature of the signals is hardly *irrelevant* ! Regarding 'balance' between source and load, I'm not sure what you're referring to. Certainly 'matched impedance' ( load and source Z equal ) is *very* old hat and unknown in typical pro-audio gear. It *is* important for good balance that the load impedance is symmetrical about ground, which isn't the case sadly with most 'balanced' inputs using a single op-amp and equal value resistors for the inverting and non-inverting legs. Graham I think that was his point, that the load impedance must be symmetrical. As far as the signal being differential, what he meant was that you need not use an inverted signal as well as a non-inverted one, as you would get the noise rejection even if only one of the lines carried signal. -Jay -- x------- Jay Kadis ------- x---- Jay's Attic Studio ------x x Lecturer, Audio Engineer x Dexter Records x x CCRMA, Stanford University x http://www.offbeats.com/ x x---------- http://ccrma.stanford.edu/~jay/ ------------x |
|
#14
|
|||
|
|||
|
On Fri, 05 Nov 2004 21:17:31 +0000, Pooh Bear
wrote: Don Pearce wrote: On 05 Nov 2004 13:57:18 GMT, ospam (Leoaw3) wrote: Sorry to waste time on such a basic question - but I'm just not understanding. I tend to run balanced cables for just about everything - unbalanced only occasionally for very short cables. But I thought the prime issue was external noise coupling into the cable. Do balanced cables inherently have a lower impedance than unbalanced? Thanks MUCH, -lee- In general the impedance of a balanced cable will be higher than that of an unbalanced one. What drugs are you on ? At audio frequencies a cable has no 'characteristic impedance'. The wavelength at audio frequencies is far too long relative to typical cable lengths. What's more, many balanced circuits operate at *lower* impedances than unbalanced ones ! Graham 1. Cables have characteristic impedances at all frequencies. They merely have little or no relevance at audio frequencies. 2. Given otherwise similar geometries, the impedance of a balanced cable is higher than that of an unbalanced cable. 3. The question is about cables, not circuits. d Pearce Consulting http://www.pearce.uk.com |
|
#15
|
|||
|
|||
|
Jay Kadis wrote:
In article , Pooh Bear wrote: Jay Kadis wrote: I attended Bill Whitlock's tutorial at the AES and this is at odds with his definition of a balanced system. He says that it is the balance between source and load impedances that determines whether a system is balanced or not and that the differential nature of the signals is irrelevent. The differential nature of the signals is hardly *irrelevant* ! Regarding 'balance' between source and load, I'm not sure what you're referring to. Certainly 'matched impedance' ( load and source Z equal ) is *very* old hat and unknown in typical pro-audio gear. It *is* important for good balance that the load impedance is symmetrical about ground, which isn't the case sadly with most 'balanced' inputs using a single op-amp and equal value resistors for the inverting and non-inverting legs. Graham I think that was his point, that the load impedance must be symmetrical. Right. Makes sense. I bet most ppl don't realise that the typical x1 differential op-amp configuration with 4x10k resistors places an asymmetric load on a balanced line. I've even had ppl query my use of the correct values ( use 2x3k3 instead on the non-inverting side ). Having said that, since the source Z is typically in the 100 ohm region for many outputs, the unbalance thus created is less than that caused by the possible mismatch of the typically used 5% resistors ! As far as the signal being differential, what he meant was that you need not use an inverted signal as well as a non-inverted one, as you would get the noise rejection even if only one of the lines carried signal. Yes, that's true as long as the 'non-signal' line is correctly referred. Graham |
|
#16
|
|||
|
|||
|
Pooh Bear wrote:
At audio frequencies a cable has no 'characteristic impedance'. The wavelength at audio frequencies is far too long relative to typical cable lengths. Why do I get echos when I call long distance, then? What's more, many balanced circuits operate at *lower* impedances than unbalanced ones ! Absolutely. --scott -- "C'est un Nagra. C'est suisse, et tres, tres precis." |
|
#17
|
|||
|
|||
|
|
|
#18
|
|||
|
|||
|
"William Sommerwerck" wrote in message
The "characteristic impedance" of a cable -- balanced or unbalanced -- is determined by a number of factors that are under control of the cable designer. One type of cable is not inherently higher- or lower-impedance than the other. (In fact, I've never seen a balanced cable with a spec'd characteristic impedance.) The characteristic impedance is not generally seen as a significant performance factor at audio frequencies. Let's not forget digital audio. AES/EBU does require 110 ohm cable. SPDIF requires 75 ohm cable. "Characteristic imedance" that is. Do not confuse the cable's characteristic impedance with the resistance of its conductors. They have nothing to do with each other. |
|
#19
|
|||
|
|||
|
"Yuri T." wrote: "William Sommerwerck" wrote in message The "characteristic impedance" of a cable -- balanced or unbalanced -- is determined by a number of factors that are under control of the cable designer. One type of cable is not inherently higher- or lower-impedance than the other. (In fact, I've never seen a balanced cable with a spec'd characteristic impedance.) The characteristic impedance is not generally seen as a significant performance factor at audio frequencies. Let's not forget digital audio. AES/EBU does require 110 ohm cable. SPDIF requires 75 ohm cable. "Characteristic imedance" that is. That's on account of the frequency of digital audio signals being high enough for characteristic impedance to have a meaning. The curious 110 ohm spec derives from the idea that it was originally envisaged that ordinary mic cable might be used for AES/EBU wiring and the characteristic impedance when measured was typically around that figure IIRC. Since then it has been found that a specialised cable type works rather better. Graham |
|
#20
|
|||
|
|||
|
Don Pearce wrote:
On Fri, 05 Nov 2004 21:17:31 +0000, Pooh Bear wrote: Don Pearce wrote: On 05 Nov 2004 13:57:18 GMT, ospam (Leoaw3) wrote: Sorry to waste time on such a basic question - but I'm just not understanding. I tend to run balanced cables for just about everything - unbalanced only occasionally for very short cables. But I thought the prime issue was external noise coupling into the cable. Do balanced cables inherently have a lower impedance than unbalanced? Thanks MUCH, -lee- In general the impedance of a balanced cable will be higher than that of an unbalanced one. What drugs are you on ? At audio frequencies a cable has no 'characteristic impedance'. The wavelength at audio frequencies is far too long relative to typical cable lengths. What's more, many balanced circuits operate at *lower* impedances than unbalanced ones ! Graham 1. Cables have characteristic impedances at all frequencies. They merely have little or no relevance at audio frequencies. 2. Given otherwise similar geometries, the impedance of a balanced cable is higher than that of an unbalanced cable. 3. The question is about cables, not circuits. The correct *answer* is that cables play no important role in this debate. It's entirely the circuits that determine operation. Graham |
|
#21
|
|||
|
|||
|
Scott Dorsey wrote: Pooh Bear wrote: At audio frequencies a cable has no 'characteristic impedance'. The wavelength at audio frequencies is far too long relative to typical cable lengths. Why do I get echos when I call long distance, then? You still do ? Been ages since I heard that stuff. Is the US not yet converted throughout to digital exchanges ? When I referred to typical cable lengths I had a studio / live sound application in mind rather than telephony which may involve somewhat large cable lengths. I appreciate that long 'landlines' operate differently to a 30ft mic cable. I've seen them being EQ'd for quality audio links. What's more, many balanced circuits operate at *lower* impedances than unbalanced ones ! Absolutely. Graham |
|
#22
|
|||
|
|||
|
"Pooh Bear" At audio frequencies a cable has no 'characteristic impedance'. The wavelength at audio frequencies is far too long relative to typical cable lengths. ** 1 km of standard twisted pair mic cable, when unterminated, has about 220nF of capacitance - enough to very seriously attenuate the HF output from typical mics. When terminated with its characteristic impedance ( ie a 100 ohm load ) the response comes back to nearly flat. Standing waves are simply not the issue - that is a silly ham radio operator's idea. ............ Phil |
|
#23
|
|||
|
|||
|
On Sat, 06 Nov 2004 06:57:25 +0000, Pooh Bear
wrote: Don Pearce wrote: On Fri, 05 Nov 2004 21:17:31 +0000, Pooh Bear wrote: Don Pearce wrote: On 05 Nov 2004 13:57:18 GMT, ospam (Leoaw3) wrote: Sorry to waste time on such a basic question - but I'm just not understanding. I tend to run balanced cables for just about everything - unbalanced only occasionally for very short cables. But I thought the prime issue was external noise coupling into the cable. Do balanced cables inherently have a lower impedance than unbalanced? Thanks MUCH, -lee- In general the impedance of a balanced cable will be higher than that of an unbalanced one. What drugs are you on ? At audio frequencies a cable has no 'characteristic impedance'. The wavelength at audio frequencies is far too long relative to typical cable lengths. What's more, many balanced circuits operate at *lower* impedances than unbalanced ones ! Graham 1. Cables have characteristic impedances at all frequencies. They merely have little or no relevance at audio frequencies. 2. Given otherwise similar geometries, the impedance of a balanced cable is higher than that of an unbalanced cable. 3. The question is about cables, not circuits. The correct *answer* is that cables play no important role in this debate. It's entirely the circuits that determine operation. Graham See my answer number 1. And thank you for your retraction. d Pearce Consulting http://www.pearce.uk.com |
|
#24
|
|||
|
|||
|
Phil Allison wrote: "Pooh Bear" At audio frequencies a cable has no 'characteristic impedance'. The wavelength at audio frequencies is far too long relative to typical cable lengths. ** 1 km of standard twisted pair mic cable, when unterminated, has about 220nF of capacitance - enough to very seriously attenuate the HF output from typical mics. When terminated with its characteristic impedance ( ie a 100 ohm load ) the response comes back to nearly flat. I haven't seen many 1 km cables used in recording or live sound though. I'm not sure many mics would appreciate being terminated with 100 ohms either. Graham |
|
#25
|
|||
|
|||
|
"Pooh Bear"
Phil Allison At audio frequencies a cable has no 'characteristic impedance'. The wavelength at audio frequencies is far too long relative to typical cable lengths. ** 1 km of standard twisted pair mic cable, when unterminated, has about 220nF of capacitance - enough to very seriously attenuate the HF output from typical mics. When terminated with its characteristic impedance ( ie a 100 ohm load ) the response comes back to nearly flat. I haven't seen many 1 km cables used in recording or live sound though. ** What you have not seen - ****head - is beside the point entirely. I'm not sure many mics would appreciate being terminated with 100 ohms either. ** What a posturing prick like you is not sure about would fill all the world's libraries. ................ Phil |
| Reply |
| Thread Tools | |
| Display Modes | |
|
|
Similar Threads
|
||||
| Thread | Forum | |||
| on topic: we need a rec.audio.pro.ot newsgroup! | Pro Audio | |||
| DNC Schedule of Events | Pro Audio | |||
Linear Mode
