I MADE SOME ERRORS. But I feel many of my points remain valid. What I didn't understand is how common mode interference is removed. I do now, THANKS to Dave's patience. You can read in the description. Hopefully he'll do another video on it where one doesn't get confused as I did.
In the picture of "Balanced Cables Explained" at time 1:22, the received amplifier (the triangle) is a difference amplifier. The top input should be labeled "+" and the bottom input should be labeled "-". The output of the amplifier is the top input minus the bottom input. The source amplifier produces the signal "signal" and its inverse "-signal". The noise picked up by the cable is the same on both wires. Therefore, the top wire contains "signal + noise") and the bottom wire contains "-signal +noise". The receiver amplifier subracts the bottom wire from the top wire: (signal + noise) - (-signal +noise). This equation simplifies to "signal +signal +noise -noise" which we results in "2 * signal". The cables are called "balanced audio cables" because they carry balanced audio where the normal and inverted audio signals are transmitted on two wires.
Thanks for comment! I recognize I sound nit-picky, but I feel the "balance" part doesn't exist in the cables but in the circuitry at each end. I'm working on a 2nd video where I try to explain what Dave explains in a way that makes more sense to me. He's taken pains to explain to me the importance of common mode rejection, which I didn't understand when I made this video.
The inverted signal is a red herring and nothing to do with being balanced. There are plenty of single ended balanced outputs in use, probably more than double ended now due to the proliferation of cheap mixers. The only thing that makes an output balanced is having closely matched output impedances. If you have the signal on one side and just the signal common 0V on the other you still cancel noise picked up on the cable and end up with just the signal after a differential receiver. The aim is not to double the signal and in fact this can cause clipping with high level signals.
Very cool Max and thank you. I enjoy the nuances of the audio adventure. With a balanced line there are two separate circuits that we are looking at. With an unbalanced line we are looking at one circuit. With an unbalanced line the desired signal and any induced noise is amplified by the mic preamp or instrument preamp or whatever it's plugged into. The signal path for the desired signal is down the hot wire and returning via ground or shield or the second wire. Cables tend to act like antennas so they pick up RF interference and other noise and the signal path for the noise is once it's induced on the cable it travels to ground of the console. It's easy to demonstrate this by just plugging in one end of a guitar cord and touching the other and inducing noise and you hear the buzz or hum coming out of the amp. The induced RF noise is a similar situation. With a balance line there are two separate circuits. One circuit is the loop typically traveling down pin 2 and returning via pin 3 to the sound source or microphone or DI box or whatever the signal source is. The ground or shield or pin 1 of the cable is not involved. Or preferably not involved but may have some involvement if the microphone has a preamp in it or there's an electronic drive for it but for all practical purposes we can look at the pin 2 3 circuit as a loop with the signal. The second circuit is just like the unbalanced cable or noise is induced except with a balanced cable it's induced on both wires of the loop. So it's induced on pin 3 and pin 2 at the same time and hopefully equally if they're twisted well and very close to each other. Sucks at the noises identical or near identical induced on both pin two and three. Any signal that's identical on pin 2 and 3 will not be amplified. With a transformer input this naturally occurs because the same signals on both sides of the primary and therefore as no potential and no signal flow so the noise just cancels itself out or never really goes anywhere to get amplified. With an electronic balanced input it's similar except their are two inputs to the op amp one of them inverting and one of them non-inverting and the inverting signal flips a polarity of the noise to combine with itself canceling it out. But unlike the transformer input they're actually is a path to ground as very few if any electronic inputs are truly 100% ground isolated. We hear this happen when we have a single ended microphone where pin 2 or pin 3 is not connected in the cable or in the mic. Even though there's no loop anymore because one wire is not connected we will still hear sound on a high gain preamp even though the mic will sound very thin and lack low end. If there was absolutely no signal path to ground then we would hear silence. So some fairly simple tests can show that wires act like antennas as well as we can demonstrate the cancellation of noise impact by shorting pin 2 to ground or pin 3 to ground on a balanced line cable for three conductor cable or twisted pair cable with a shield around it or whatever you want to call it. And then putting that cable in a high noise RF environment and then unshorting pin two or three to ground and listening to the difference in noise Very cool Max, let me what ya think
The diagrams are correct though they provide little explanation. Yes a differential amplifier used in a balanced input will reject common mode noise which is the main problem. Yes a twisted pair with a shield (balanced cable) both actually exists and has higher noise rejection than a coaxial cable (single ended). The balanced cable is better suited for differential inputs and single ended cables are better suited for single ended inputs.
Thanks. My focus is only on microphones to pre-amps. Does anyone use coax for microphones? I don't think that diagram is showing coax. I think it's just badly drawn.
@@MaxoticsTV Hi Z microphones use coax. Most young people won't have heard of them. Back in the day you could get cheap 'Hi Z' microphones for use with cheap equipment that didn't have differential inputs. Capacitance is an issue with coaxial cables causing roll of high frequency signal as cable length gets longer.. You could terminate your coax with its characteristic impedance as is done with RF electronics but that is not convenient.
Thanks! I'm not young, 63, so I can't play the too-young card ;) I still feel any diagram showing COAX should show the "return" path/ line because it has one, right? All circuits must. Thanks again for explaining it!
@@MaxoticsTVThe balanced system has a huge benefit over an unbalanced system. It's not rocket science and it's not magic. One of the main problems sending signals is that the ground level of each connected piece of equipment is different and changing over time. That is the ground differential between connected components is a source of noise. A single ended system compares the single ended input to it's own ground and not ground of the component that sends it and that is a major source of noise. That is why the diagram for the single ended system shows only one line because only one line is used. Yes there is a return line and if it worked like you want it to then the grounds of all connected components would agree at all times. In the real world the ground line has both resistance and inductance and cannot bring the grounds of connected components to exactly the same level. The tubular shield of a coax has lower resistance and inductance than a wire and that is why a coax is usually better for single ended systems.
Balanced systems on the other hand use a differential amplifier to receive the signal from the previous component. The differential amplifier only amplifies the difference between the two inputs. Anything the two inputs have in common is rejected. These are known as common mode signals and differential mode signals. A ground differential between the two components will appear the same at both inputs and be rejected by the differential amplifier. That is if the ground of the source suddenly jumps to 1 volt that 1 volt will appear at both inputs and be rejected 1-1=0. if signal however is sent of 0.5 volts on the positive input and -0.5 volts on the negative input then 1 volt will be received. 0.5 - (-0.5)=1. The single ended system cannot reject the common mode noise. Using a transmission line that is as symmetrical as possible with differential (balanced) inputs will insure that other forms of interference affect both lines equally and become common mode signals which will be rejected by the balanced input. That is the reason for the twisted pair used in balanced cables. Twisted pairs are extremely symetrical and interference becomes a common mode signal which will be rejected.
@ I wouldn’t say it’s an issue, but it is a consideration. Not likely to show in a short run, but hundreds of feet could make a difference. Side note-it counts a lot in a twisted pair in an Ethernet cable. In a pro piece of gear where the driving impedance is 50 ohms or so, imbalance is minimized. The type of circuit makes a huge difference. The Bell labs were undoubtedly using transformers. There exist single drive circuits that are impedance balanced. (Mackie) see also the InGenius balanced chip notes from ThatCorp. The cable is usually shielded twisted pair. The impedance of the pair is usually 60 ohms for audio and 110 for AES. Look at a Belsen 9451 datasheet and you will see info on the tolerances for impedance in the pair and pair to ground. A differential circuit may be sensitive to imbalance (simple opamp diff) or quite insensitive (good transformer). I will agree that the term balanced cable that has become standard parlance is sloppy. You certainly could plug an Xlr cable in to two units that are wired in an unbalanced fashion. The old classic White equalizer had Xlr in, but no balancing unless you bought the transformer option. (Unbal was also pin 3 hot - another issue) If you ever have to wire a Neve or SSL patch bay, or run half mile lines, this subject is a deep rabbit hole.
IMHO your explanation with the bike chains and diaphragm is long and convoluted. I would have just whipped out the equivalent circuit diagram and performed nodal analysis or a SPICE simulation to prove my point in under 10 minutes.
Hi. Nodal analysis or a SPICE simulation proves what? I'm not after "proofs", I'm trying to understand, trying to visualize in my head how things work. My fundamental question is: how can the pre-amp know which voltage is coming from the microphone and which is coming from lightning, let's say. How do diagrams answer that? Most diagrams don't. They focus on the circuit of interest. They ASSUME there is no interference ("noise"). But I invite you to prove (haha) me wrong. Thanks for comment!!!
@@MaxoticsTV You can add a common mode voltage source and stray capacitance to the circuit to model common mode currents. Biricha has a video that explains this in 5 minutes called "EMC Filter Design Part 1: Understanding Common Mode and Differential Mode Noise".
@@fablearchitect7645 I don't think I would have understood it better watching that video first. I enjoyed though--THANKS! My core problem with Dave's explanation was that it treated the mic voltage as a circuit but the interference ("noise") as an orphaned voltage. I'm always thinking conservation of energy--there can be no orphans. In the video you pointed out, he indeed shows that the common mode current goes into Earth Ground. So he separates the primary circuit from that secondary circuit. I know no one thinks of anything Earth Ground as a "circuit" but we have to have it, I believe, to keep our explanations obeying the Conservation of Energy.
I MADE SOME ERRORS. But I feel many of my points remain valid. What I didn't understand is how common mode interference is removed. I do now, THANKS to Dave's patience. You can read in the description. Hopefully he'll do another video on it where one doesn't get confused as I did.
In the picture of "Balanced Cables Explained" at time 1:22, the received amplifier (the triangle) is a difference amplifier. The top input should be labeled "+" and the bottom input should be labeled "-". The output of the amplifier is the top input minus the bottom input. The source amplifier produces the signal "signal" and its inverse "-signal". The noise picked up by the cable is the same on both wires. Therefore, the top wire contains "signal + noise") and the bottom wire contains "-signal +noise". The receiver amplifier subracts the bottom wire from the top wire: (signal + noise) - (-signal +noise). This equation simplifies to "signal +signal +noise -noise" which we results in "2 * signal".
The cables are called "balanced audio cables" because they carry balanced audio where the normal and inverted audio signals are transmitted on two wires.
Thanks for comment! I recognize I sound nit-picky, but I feel the "balance" part doesn't exist in the cables but in the circuitry at each end. I'm working on a 2nd video where I try to explain what Dave explains in a way that makes more sense to me. He's taken pains to explain to me the importance of common mode rejection, which I didn't understand when I made this video.
The inverted signal is a red herring and nothing to do with being balanced. There are plenty of single ended balanced outputs in use, probably more than double ended now due to the proliferation of cheap mixers. The only thing that makes an output balanced is having closely matched output impedances. If you have the signal on one side and just the signal common 0V on the other you still cancel noise picked up on the cable and end up with just the signal after a differential receiver. The aim is not to double the signal and in fact this can cause clipping with high level signals.
Very cool Max and thank you. I enjoy the nuances of the audio adventure.
With a balanced line there are two separate circuits that we are looking at. With an unbalanced line we are looking at one circuit.
With an unbalanced line the desired signal and any induced noise is amplified by the mic preamp or instrument preamp or whatever it's plugged into. The signal path for the desired signal is down the hot wire and returning via ground or shield or the second wire.
Cables tend to act like antennas so they pick up RF interference and other noise and the signal path for the noise is once it's induced on the cable it travels to ground of the console.
It's easy to demonstrate this by just plugging in one end of a guitar cord and touching the other and inducing noise and you hear the buzz or hum coming out of the amp. The induced RF noise is a similar situation.
With a balance line there are two separate circuits.
One circuit is the loop typically traveling down pin 2 and returning via pin 3 to the sound source or microphone or DI box or whatever the signal source is. The ground or shield or pin 1 of the cable is not involved. Or preferably not involved but may have some involvement if the microphone has a preamp in it or there's an electronic drive for it but for all practical purposes we can look at the pin 2 3 circuit as a loop with the signal.
The second circuit is just like the unbalanced cable or noise is induced except with a balanced cable it's induced on both wires of the loop. So it's induced on pin 3 and pin 2 at the same time and hopefully equally if they're twisted well and very close to each other. Sucks at the noises identical or near identical induced on both pin two and three.
Any signal that's identical on pin 2 and 3 will not be amplified.
With a transformer input this naturally occurs because the same signals on both sides of the primary and therefore as no potential and no signal flow so the noise just cancels itself out or never really goes anywhere to get amplified.
With an electronic balanced input it's similar except their are two inputs to the op amp one of them inverting and one of them non-inverting and the inverting signal flips a polarity of the noise to combine with itself canceling it out. But unlike the transformer input they're actually is a path to ground as very few if any electronic inputs are truly 100% ground isolated.
We hear this happen when we have a single ended microphone where pin 2 or pin 3 is not connected in the cable or in the mic.
Even though there's no loop anymore because one wire is not connected we will still hear sound on a high gain preamp even though the mic will sound very thin and lack low end.
If there was absolutely no signal path to ground then we would hear silence.
So some fairly simple tests can show that wires act like antennas as well as we can demonstrate the cancellation of noise impact by shorting pin 2 to ground or pin 3 to ground on a balanced line cable for three conductor cable or twisted pair cable with a shield around it or whatever you want to call it.
And then putting that cable in a high noise RF environment and then unshorting pin two or three to ground and listening to the difference in noise
Very cool Max, let me what ya think
Thanks for taking the time to comment!!!
The diagrams are correct though they provide little explanation. Yes a differential amplifier used in a balanced input will reject common mode noise which is the main problem. Yes a twisted pair with a shield (balanced cable) both actually exists and has higher noise rejection than a coaxial cable (single ended). The balanced cable is better suited for differential inputs and single ended cables are better suited for single ended inputs.
Thanks. My focus is only on microphones to pre-amps. Does anyone use coax for microphones? I don't think that diagram is showing coax. I think it's just badly drawn.
@@MaxoticsTV Hi Z microphones use coax. Most young people won't have heard of them. Back in the day you could get cheap 'Hi Z' microphones for use with cheap equipment that didn't have differential inputs. Capacitance is an issue with coaxial cables causing roll of high frequency signal as cable length gets longer.. You could terminate your coax with its characteristic impedance as is done with RF electronics but that is not convenient.
Thanks! I'm not young, 63, so I can't play the too-young card ;) I still feel any diagram showing COAX should show the "return" path/ line because it has one, right? All circuits must. Thanks again for explaining it!
@@MaxoticsTVThe balanced system has a huge benefit over an unbalanced system. It's not rocket science and it's not magic. One of the main problems sending signals is that the ground level of each connected piece of equipment is different and changing over time. That is the ground differential between connected components is a source of noise. A single ended system compares the single ended input to it's own ground and not ground of the component that sends it and that is a major source of noise. That is why the diagram for the single ended system shows only one line because only one line is used. Yes there is a return line and if it worked like you want it to then the grounds of all connected components would agree at all times. In the real world the ground line has both resistance and inductance and cannot bring the grounds of connected components to exactly the same level. The tubular shield of a coax has lower resistance and inductance than a wire and that is why a coax is usually better for single ended systems.
Balanced systems on the other hand use a differential amplifier to receive the signal from the previous component. The differential amplifier only amplifies the difference between the two inputs. Anything the two inputs have in common is rejected. These are known as common mode signals and differential mode signals. A ground differential between the two components will appear the same at both inputs and be rejected by the differential amplifier. That is if the ground of the source suddenly jumps to 1 volt that 1 volt will appear at both inputs and be rejected 1-1=0. if signal however is sent of 0.5 volts on the positive input and -0.5 volts on the negative input then 1 volt will be received. 0.5 - (-0.5)=1. The single ended system cannot reject the common mode noise. Using a transmission line that is as symmetrical as possible with differential (balanced) inputs will insure that other forms of interference affect both lines equally and become common mode signals which will be rejected by the balanced input. That is the reason for the twisted pair used in balanced cables. Twisted pairs are extremely symetrical and interference becomes a common mode signal which will be rejected.
Balanced impedance is the key.
Yes, but is that an issue for modern cables, wire?
@ I wouldn’t say it’s an issue, but it is a consideration. Not likely to show in a short run, but hundreds of feet could make a difference. Side note-it counts a lot in a twisted pair in an Ethernet cable.
In a pro piece of gear where the driving impedance is 50 ohms or so, imbalance is minimized.
The type of circuit makes a huge difference. The Bell labs were undoubtedly using transformers. There exist single drive circuits that are impedance balanced. (Mackie) see also the InGenius balanced chip notes from ThatCorp.
The cable is usually shielded twisted pair. The impedance of the pair is usually 60 ohms for audio and 110 for AES. Look at a Belsen 9451 datasheet and you will see info on the tolerances for impedance in the pair and pair to ground.
A differential circuit may be sensitive to imbalance (simple opamp diff) or quite insensitive (good transformer).
I will agree that the term balanced cable that has become standard parlance is sloppy. You certainly could plug an Xlr cable in to two units that are wired in an unbalanced fashion. The old classic White equalizer had Xlr in, but no balancing unless you bought the transformer option. (Unbal was also pin 3 hot - another issue)
If you ever have to wire a Neve or SSL patch bay, or run half mile lines, this subject is a deep rabbit hole.
IMHO your explanation with the bike chains and diaphragm is long and convoluted. I would have just whipped out the equivalent circuit diagram and performed nodal analysis or a SPICE simulation to prove my point in under 10 minutes.
Hi. Nodal analysis or a SPICE simulation proves what? I'm not after "proofs", I'm trying to understand, trying to visualize in my head how things work. My fundamental question is: how can the pre-amp know which voltage is coming from the microphone and which is coming from lightning, let's say. How do diagrams answer that?
Most diagrams don't. They focus on the circuit of interest. They ASSUME there is no interference ("noise"). But I invite you to prove (haha) me wrong. Thanks for comment!!!
@@MaxoticsTV You can add a common mode voltage source and stray capacitance to the circuit to model common mode currents. Biricha has a video that explains this in 5 minutes called "EMC Filter Design Part 1: Understanding Common Mode and Differential Mode Noise".
@@fablearchitect7645 I don't think I would have understood it better watching that video first. I enjoyed though--THANKS! My core problem with Dave's explanation was that it treated the mic voltage as a circuit but the interference ("noise") as an orphaned voltage. I'm always thinking conservation of energy--there can be no orphans. In the video you pointed out, he indeed shows that the common mode current goes into Earth Ground. So he separates the primary circuit from that secondary circuit. I know no one thinks of anything Earth Ground as a "circuit" but we have to have it, I believe, to keep our explanations obeying the Conservation of Energy.