Is bridged / balanced better than single ended?
A limited number of "wire" SE-SE pcbs needing rework became available. Several fanatics have allowed their imaginations to run loose with those boards. One especially promising avenue was opened by agdr.
The rework pcb allows the LME49860 to replace the LME49990 with about the same disruption as would be required to reroute the supply rails except that the bypass caps remain correctly positioned. Details can be found in the thread where the above quote was posted.
I believe 'the wire' is relevant to this thread because it allows testing several configurations of bridged headphone cabling (or the same configuration with several headamps). The sonic purity of the different versions is quite high in both balanced and single ended configurations.
The HigherWire provides clues about whether it is the higher voltages across the cans that accounts for the perception that 'balanced is better'. With 22v rails it could provide peak voltages across each can of nearly 20 volts whereas the stock SE-SE would peak at around 14 volts and the BAL-BAL at around 27 volts. Keep in mind that sustained voltages in those ranges would ruin both cans and ears of listener.
By comparison, the balanced JC-80 can provide peak voltages above 40 volts and other esoterics, e.g. beta22, even higher. Readers who have followed my thinking will realize that I am really attempting to understand whether the push-pull of balanced sources is actually better or whether the one side grounded of single ended, properly cabled, can be as good.
A limited number of "wire" SE-SE pcbs needing rework became available. Several fanatics have allowed their imaginations to run loose with those boards. One especially promising avenue was opened by agdr.
The rework pcb allows the LME49860 to replace the LME49990 with about the same disruption as would be required to reroute the supply rails except that the bypass caps remain correctly positioned. Details can be found in the thread where the above quote was posted.
I believe 'the wire' is relevant to this thread because it allows testing several configurations of bridged headphone cabling (or the same configuration with several headamps). The sonic purity of the different versions is quite high in both balanced and single ended configurations.
The HigherWire provides clues about whether it is the higher voltages across the cans that accounts for the perception that 'balanced is better'. With 22v rails it could provide peak voltages across each can of nearly 20 volts whereas the stock SE-SE would peak at around 14 volts and the BAL-BAL at around 27 volts. Keep in mind that sustained voltages in those ranges would ruin both cans and ears of listener.
By comparison, the balanced JC-80 can provide peak voltages above 40 volts and other esoterics, e.g. beta22, even higher. Readers who have followed my thinking will realize that I am really attempting to understand whether the push-pull of balanced sources is actually better or whether the one side grounded of single ended, properly cabled, can be as good.
From the point of view of the headphone drivers, what is the difference between the two? I totally understand that the quality of sound of a bridged amplifier can be quite different from a true balanced amplifier, but for the moment I can't see how the headphones would know which is which.
Herez the what without the why
That is the sense in which I have used bridged. Two amplified channels with a common ground, carrying identical but inverted signals, the output of each wired across one can. If the circuitry of the two amplifiers supports common mode noise rejection it would be called balanced AFAIK. Note that all of that is about the amplifiers.
In what cases will common noise be cancelled? Let's start with two identical amplifiers powered by a single power supply. Power supply noise, because it is common, would be amplified but largely cancelled at the can assuming the cabling was twisted pair. The advantage of balanced amplifiers would be cancellation of the common noise in the circuitry, well before transmission to the cans. Those are the only cases: SE has no chance for common noise cancellation because there is no differential signal.
Since both cases of interest include an inverted signal there may be a temptation to consider generating both the signal and an inverted copy from the bits. This would allow common noise cancellation and differential transmission at every step in the chain but many users balk at a complete new start to their system.
For those users, converting their headphone cable to dual (ideally twisted) pairs would yield some reduction of crosstalk whose importance would vary by headset. In order to achieve any level of common mode noise reduction either bridged or balanced (which requires wiring equivalent to bridged) amplification is required. The cheapest alternative is to drive each L-R channel with a stereo amp wired ACROSS a single can. Noise reduction occurs at the can for the typically common power supply of each stereo amp. In addition, because the outputs are out of phase, i.e. the voltage differential is 2x that of one single ended connection and ground.
For now we are ignoring 1)getting an inverted signal and 2)assuring equal gain on both sides. The second of these may be a plus for those with unbalanced ears.
Are there advantages to bridged vis-a-vis single ended? I think there are. Are there additional advantages to balanced vis-a-vis bridged? I think there are. The question of whether the cost of amplifying four separate signals over and above the cost of 1)recabling + 2)obtaining an inverted copy of each of two signals + 3)amplifying just two additional signals is justified. In my case the sum of 1+2+3 was much higher and required several years of fiddling but when I started it was the case that balanced appeared quite costly and the benefits of balanced vis-a-vis were not known to me.
What do I know now that I did not know then? HD-600s sounded better by providing four conductors to the cans. They also sounded better bridged than single ended but that required an inverted signal which was not cheap in my case and so many things changed at once the importance of bridged v. se was clouded. For example, were the Senns merely power hungry, since the inverted signal came from a different DAC . . .
I recently completed and deployed a BAL-BAL "wire" and initially connected it to the same power supply used for the bridged pair of amps. Fortunately qusp spotted that as a potential issue.
The BAL-BAL is great at driving HD-600s for both balanced and se inputs (the inverted signal occurs "automagically" but the output level is lower). I would have saved both time and money had I been able to start with the wire. Alas, it was unavailable then and is again now :-(
Merely recabling won't do much for Senns. Recabling plus spending $300-400 on a balanced headamp / DAC could make quite a difference. Should the BAL-BAL become available again and you are ready to solder surface mount you could jolt yourself all the way to nirvana for $100 plus the cost of a power supply.
That is the sense in which I have used bridged. Two amplified channels with a common ground, carrying identical but inverted signals, the output of each wired across one can. If the circuitry of the two amplifiers supports common mode noise rejection it would be called balanced AFAIK. Note that all of that is about the amplifiers.
In what cases will common noise be cancelled? Let's start with two identical amplifiers powered by a single power supply. Power supply noise, because it is common, would be amplified but largely cancelled at the can assuming the cabling was twisted pair. The advantage of balanced amplifiers would be cancellation of the common noise in the circuitry, well before transmission to the cans. Those are the only cases: SE has no chance for common noise cancellation because there is no differential signal.
Since both cases of interest include an inverted signal there may be a temptation to consider generating both the signal and an inverted copy from the bits. This would allow common noise cancellation and differential transmission at every step in the chain but many users balk at a complete new start to their system.
For those users, converting their headphone cable to dual (ideally twisted) pairs would yield some reduction of crosstalk whose importance would vary by headset. In order to achieve any level of common mode noise reduction either bridged or balanced (which requires wiring equivalent to bridged) amplification is required. The cheapest alternative is to drive each L-R channel with a stereo amp wired ACROSS a single can. Noise reduction occurs at the can for the typically common power supply of each stereo amp. In addition, because the outputs are out of phase, i.e. the voltage differential is 2x that of one single ended connection and ground.
For now we are ignoring 1)getting an inverted signal and 2)assuring equal gain on both sides. The second of these may be a plus for those with unbalanced ears.
Are there advantages to bridged vis-a-vis single ended? I think there are. Are there additional advantages to balanced vis-a-vis bridged? I think there are. The question of whether the cost of amplifying four separate signals over and above the cost of 1)recabling + 2)obtaining an inverted copy of each of two signals + 3)amplifying just two additional signals is justified. In my case the sum of 1+2+3 was much higher and required several years of fiddling but when I started it was the case that balanced appeared quite costly and the benefits of balanced vis-a-vis were not known to me.
What do I know now that I did not know then? HD-600s sounded better by providing four conductors to the cans. They also sounded better bridged than single ended but that required an inverted signal which was not cheap in my case and so many things changed at once the importance of bridged v. se was clouded. For example, were the Senns merely power hungry, since the inverted signal came from a different DAC . . .
I recently completed and deployed a BAL-BAL "wire" and initially connected it to the same power supply used for the bridged pair of amps. Fortunately qusp spotted that as a potential issue.
The BAL-BAL is great at driving HD-600s for both balanced and se inputs (the inverted signal occurs "automagically" but the output level is lower). I would have saved both time and money had I been able to start with the wire. Alas, it was unavailable then and is again now :-(
Merely recabling won't do much for Senns. Recabling plus spending $300-400 on a balanced headamp / DAC could make quite a difference. Should the BAL-BAL become available again and you are ready to solder surface mount you could jolt yourself all the way to nirvana for $100 plus the cost of a power supply.
Twisted pair does not cancel existing noise. Instead, what it does is ensure that any additional noise picked up by the cable is as close to identical in each wire as possible, because the geometric center of a twisted pair places each wire of the pair in the same 'average' location. Note that EM noise is highly dependent upon position, so both wires in a pair must be in the same position to pick up the same noise. The purpose of keeping the noise similar in each wire is so that it can be canceled more completely in the next balanced stage.
This has nothing to do with power supply noise. Power supply noise comes into the system separately from the headphone cable, and thus it makes no difference how the cable is configured. Zip cord is just a good for balanced power supply rejection as twisted pair.
The only exception would be if you had a power supply that was designed so poorly that it emitted EM noise in a huge field around your amplifier, thus causing additional noise to be picked up by the headphone cable. In that bizarre case, twisted pair might help.
By the way, a speaker driver is differential by nature. It responds to the difference in voltage between terminals (actually, it responds to the current that flows through the voice coil as a result of the voltage differential). If power supply noise appears as equal voltage on both terminals, then the speaker driver will basically ignore the power supply noise.
Some of the best DAC chips have both a normal and inverted output. In fact, a by product of certain D/A conversion circuits is that they also produce an inverted output. Obviously, such a DAC would have better matching between + and - than two separate chips. Thus, it may not be so great an effort to include the inverted signal, although you are correct that it might require a whole new design.
Again, twisted pair is not necessarily ideal. Zip cord might perform practically the same. Given the power levels, coaxial is probably a bad idea, and even twisted pair might have some disadvantages.
Hi again rsdio. Other than your love for zip over tp I don't think we disagree about anything. Since the currents for signal and inverted signal must always be equal but opposite, so the magnetic fields they generate must be equal and opposite and thus self canceling. Any stray magnetic field applied to a tp gets canceled at the can. In home runs as short as headphone cables they ought never be subjected to such fields but I admit to never having wired headphones with zip to see what happens.
I should have mentioned that there may be BAL-SE pcbs only remaining. If your cans could be adequately powered by one buffer per can then these might be the next best option but I have not tried it. My preference is definitely for the BAL-BAL.
I will be turning my attention to the digital and DAC ends of the chain and plan no further experiments. Luciens WaveIO, a couple of Opus DACs and assorted stuff is awaiting my attention. Happy New Year to all.
I should have mentioned that there may be BAL-SE pcbs only remaining. If your cans could be adequately powered by one buffer per can then these might be the next best option but I have not tried it. My preference is definitely for the BAL-BAL.
I will be turning my attention to the digital and DAC ends of the chain and plan no further experiments. Luciens WaveIO, a couple of Opus DACs and assorted stuff is awaiting my attention. Happy New Year to all.
Just to be clear, I don't love zip. I was merely pointing out that zip and tp are equivalent when it comes to power supply noise rejection. Only EM noise picked up by the cable will be affected by your choice of zip or tp.
By the way, I'm sure you've heard it recommended that speakers be wired with zip cord, and by no means should they ever use coaxial. When it comes to headphones, I've heard some people say they should be wired exactly like speakers, while other people say they are much lower in power and thus should be wired differently, even going so far as to say that coaxial is just fine for headphones. To be safe, I won't go on record with a preference for either one when it comes to headphone cables. As I explained above, my only point was that the power supply noise rejection ignores the cable topology.
Then we are in total accord. I hope you will consider 'the wire' if it becomes available again. It is good enuff for me to relax about that end of the chain and turn my attention to turning bitz to volts. Thanks for following along with this thread.
I tried to scan this entire thread for a link to whatever it is you're talking about when you say 'the wire' or BAL-BAL, but I could not find anything. Is it some kind of headphone amplifier? At first I thought you were talking about some sort of expensive wire that was for sale (and after a long AES presentation from a Belkin representative, I really do not believe in Transparent or any of the other 'magic' wire brands), but a little digging hints that you're talking about some kind of PCB.
Anyway, reply with a link, and I'll take a look. Hopefully there is sufficient information even if 'the wire' is not available at the moment.
I believe this is the thread you are looking for - The Wire
Also opc now has a build wiki that has schematics and bom attached which might be quicker reading for you
On the topic of a balanced connector standard, I believe I have finally made up my mind.
I believe the 5-pin Mini-DIN is the way to go. It allows for shielding if anybody cares, but it has very unique distinction in that it can plug into a 6-pin socket (it may require shaving the center post a bit). The 6-pin socket can be used for things like inline remote or mic. It can also provide power for things like active NC, loudness circuits, and more.
So my proposed standard is the 5-pin Mini-DIN to the headphones and a 5 or 6 pin socket depending on the designer's objectives.
Thoughts?
I believe the 5-pin Mini-DIN is the way to go. It allows for shielding if anybody cares, but it has very unique distinction in that it can plug into a 6-pin socket (it may require shaving the center post a bit). The 6-pin socket can be used for things like inline remote or mic. It can also provide power for things like active NC, loudness circuits, and more.
So my proposed standard is the 5-pin Mini-DIN to the headphones and a 5 or 6 pin socket depending on the designer's objectives.
Thoughts?
Last edited:
If . . .
If you are wanting to standardize I think you may need to fully describe the pinout and to which pins you plan to apply power. If someone misinterpreted or miswired a plug or socket could the power damage either their amplifier or earphones? Do you forsee a need for power at the headphones? Are there advantages of a DIN over a mini-din? Are there disadvantages? Those are my immediate thoughts, none of which would be relevant if you just want to wire your own headset.
If you are wanting to standardize I think you may need to fully describe the pinout and to which pins you plan to apply power. If someone misinterpreted or miswired a plug or socket could the power damage either their amplifier or earphones? Do you forsee a need for power at the headphones? Are there advantages of a DIN over a mini-din? Are there disadvantages? Those are my immediate thoughts, none of which would be relevant if you just want to wire your own headset.
with din, space is the only advantage and only in one plane, in the others its larger and there are only disadvantages other than that. in hifi, honestly who gives a toss about active noise cancelling when i can use brilliant quality in ears that block 26db passively in situations where i need it? i dont see any reason for having power, i have a remote already and yes with din there is nothing stopping it being plugged in wrong if you have more pins than needed. it also isnt locking and they arent what i would call rugged. the lemo already takes up less space, cannot be plugged in wrong but noone will use that, its too expensive but is otherwise 100% perfect. what would you be using for power ground anyway? youde need a 7pin if you want to maintain a shield, because shield and power ground should not be the same. cmmr of a flat connector is inferior
and come on, another balanced standard? give me a break. have you actually used a din in this manner, or are you proposing a standard without having ever used it?
and come on, another balanced standard? give me a break. have you actually used a din in this manner, or are you proposing a standard without having ever used it?
Last edited:
@Wolfsin, I figured any standard worth spit would have the pinout explained. If the task were handed to me, I'd work it over. Lots of demo builds and external consultations would ensue.
@qusp, all good info. It makes me wonder what we'd be saying if TRS came up in conversation as a new idea for SE standardization (as if the world today were different).
@qusp, all good info. It makes me wonder what we'd be saying if TRS came up in conversation as a new idea for SE standardization (as if the world today were different).
my thoughts on the headphone connector 'standards' are well known, both here and over at the other forum; the trs is idiotic as are 3 of the 5 (yes 5) balanced headphone 'standards' and all of the portable balanced ones except maybe the 4 pin XLR for home (still too big) and the hirose HR10 ibasso and one other asian company are using on their portables; but tbh i dont see yours as being any better and worse than a couple of them at least, it has its own problems and like it or not people are stupid sometimes and dont always have a manual to refer to, or even be the owner, so using something that needs explanation or it fries your headphones is not a great idea IMO and certainly not a good reason to make it 'standard' 6
i think its good that someone is thinking about such an idea, but this is a bit of a pet cause for me, if ANOTHER one is to be introduced it must be compared to those in use favourably before anyone even puts it on something
Last edited:
Until we have a LeMo as small as a TRS and as cheap, we're just going to have to deal with no standards for a standard. My thoughts were certainly aimed at home use. Portable is a different beast.
Also, I don't see how a sixth power pin unconnected to a 5-pin plug would fry the headphones. Anyway, I've been hunting around for your comments and your thoughts are more mature than mine. Very good thinking. The DIY bug bit me before I read much on it, so I was going my own way on this.
Also, I don't see how a sixth power pin unconnected to a 5-pin plug would fry the headphones. Anyway, I've been hunting around for your comments and your thoughts are more mature than mine. Very good thinking. The DIY bug bit me before I read much on it, so I was going my own way on this.
maybe you are right, if everything is correctly in place, but do you really see it as that much of a stretch that someone would plug it in incorrectly when there is no mechanism in the din standard to stop that from happening? if you are going to put DC power enough to kill headphones on the same connector as the headphone audio; you better make it impossible to connect by accident.
also no you see i think this no standard thing is just pure laziness and also people and companies wanting to have something original they can advertise. if it was actually decided that a standard was to be introduced, the manufacturing/purchase cost would come down just like that. the industry is still small enough to make this happen, but commercial interests stop it from happening.
hell i would be happy enough if it was the HR10 or probably lemo which if bought in bulk would probably come down to 10-15 and if you are really going to spend a few hundred (or in some cases 10x that) on headphones and then whine about a connector costing 10-15, then I think there are some priorities out of place. having repaired a number of amps for people because of minijacks failing, i guess i have a different idea of what constitutes value.
balanced audio is not for the daytripper, by the time someone is considering that theyve spent a fair bit of money on their setup, the cost of proper balanced connectors would make little difference to the overall cost
also no you see i think this no standard thing is just pure laziness and also people and companies wanting to have something original they can advertise. if it was actually decided that a standard was to be introduced, the manufacturing/purchase cost would come down just like that. the industry is still small enough to make this happen, but commercial interests stop it from happening.
hell i would be happy enough if it was the HR10 or probably lemo which if bought in bulk would probably come down to 10-15 and if you are really going to spend a few hundred (or in some cases 10x that) on headphones and then whine about a connector costing 10-15, then I think there are some priorities out of place. having repaired a number of amps for people because of minijacks failing, i guess i have a different idea of what constitutes value.
balanced audio is not for the daytripper, by the time someone is considering that theyve spent a fair bit of money on their setup, the cost of proper balanced connectors would make little difference to the overall cost
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.