Personal computing discussed
Moderators: renee, Captain Ned
G8torbyte wrote:I'm using an XLR adapter that takes the left and right XLR outputs into one 4-pin XLR connector direct to my headphones.
Ifalna wrote:G8torbyte wrote:I'm using an XLR adapter that takes the left and right XLR outputs into one 4-pin XLR connector direct to my headphones.
Ah, so you are using the XLRs on the back. Interesting.
Does it sound better/different compared to the single ended front output?
I've read that going balanced for cans is rather useless b/c cable runs aren't long enough for common mode rejection to be needed and that balanced doubles THD and output impedance, thus actually degrades sonic quality.
jihadjoe wrote:One amp per channel is just dual-mono. It's great, but not yet 'fully balanced'.
A true fully balanced design will actually have two amps per channel, with separate amps handling the normal and inverted outputs (for a total of four amps driving two channels). Basically take the dual-mono concept and apply it again to the normal and inverted outputs of each channel.
[IMG SNIP]
sauce
morphine wrote:Balanced vs unbalanced is all about dat noise control. I've yet to hear about that ever being a problem with headphones. To me it just sounds like a feature that exists for the sake of selling something.
Waco wrote:I can see where slew rate and dynamics could be better with quad amps, but two slightly larger amps in dual mono are easier to calibrate (and less likely to have a DC bias in one direction or the other).
DancinJack wrote:Yeah, unless you're using, like, onboard realtek from 2007 it's probably not an issue most of the time now.
Ifalna wrote:Waco wrote:I can see where slew rate and dynamics could be better with quad amps, but two slightly larger amps in dual mono are easier to calibrate (and less likely to have a DC bias in one direction or the other).
I seriously doubt that modern opamps would be slew rate constrained on sth. like a 22KHz audio signal.
Esp not at the low voltages normal headphones operate on. Oo
Ifalna wrote:DancinJack wrote:Yeah, unless you're using, like, onboard realtek from 2007 it's probably not an issue most of the time now.
A balanced setup would not help against that. The distortion is already in the signal before it hits the amp.
Differential amping is only useful to filter out distortion that is added between amp and target because any injected noise will affect both, + and - and subsequently the difference between the two will be the same. The target thus never "sees" the injected noise because only the difference contains the actual information. IIRC that is called "common mode rejection".
Say you have to run your analog cable past a HF light on a stage. No matter how well you shield, that light will induce noise. Going balanced eliminates that noise from the light and you can feed your speaker a "clean" signal.
sluggo wrote:In a previous life I was working on interesting applications for my company's switching amplifier technology. Since the output stage operated in differential mode and we had a stereo pair on the chip, I modified a set of AKG 240 headphones for 4-wire drive to see how it sounded. It sounded great. These were about 8 of the cleanest watts available from a single 12-volt supply, with negligible offset and over 100 dB of supply rejection.
Unfortunately, nearly no one puts a differential connector on consumer electronics and nearly every headphone sold uses a common ground. It was a good idea, but there was literally no market for it at the time (which, yeah, meant it was not really a good idea). Looking through head-fi and similar sites it appears that some people are running balanced drivers, but it's still a small fraction of users.
superjawes wrote:DancinJack wrote:Pet peeve: using impedance alone to determine how hard it is to drive a set of headphones.Those HE-4XX's are only 35 ohms so it shouldn't be that hard to drive them. The Jubilee's on the other hand are 150. Might need an amp to get real quality of those.
Sure, it's part of the equation--high impedance models generally need higher voltages to drive properly--but you really need to consider efficiency/sensitivity as well. The 58X Jubilee is listed at 104 dB (at 1V, 1 kHz), and the HE-4XX just lists 93 dB (also, the AKG K7XX is shown at 105 dB/V).
Specs can be wonky and non-standardized, but wrt to the HE-4XX, you have a less efficient planar driver that might need some extra power. The 58X Jubilee is more efficient, but that higher impedance indicates a desire for higher input voltage to get the best drive.
And G8torbyte...why aren't you using the headphone output on the ADI-2? I'd imagine that output is probably better at driving both sets of headphones without distortion...
cynan wrote:sluggo wrote:In a previous life I was working on interesting applications for my company's switching amplifier technology. Since the output stage operated in differential mode and we had a stereo pair on the chip, I modified a set of AKG 240 headphones for 4-wire drive to see how it sounded. It sounded great. These were about 8 of the cleanest watts available from a single 12-volt supply, with negligible offset and over 100 dB of supply rejection.
Unfortunately, nearly no one puts a differential connector on consumer electronics and nearly every headphone sold uses a common ground. It was a good idea, but there was literally no market for it at the time (which, yeah, meant it was not really a good idea). Looking through head-fi and similar sites it appears that some people are running balanced drivers, but it's still a small fraction of users.
Many (most?) higher end consumer headphones have "balanced" drivers, just like all speakers. The connection to the driver (or crossover) itself has a +ve and -ve connection. It's the cable that combines the +ve into a common grnd. So you can do what you describe with many consumer products. It's just easier to do with headphones that have cables (or detachable cable connectors) with with the +ve and -ve separate (eg, HD 650, and plenty of others). I also have an amp with old school switching technology (not class D, but a Tripath chip). This also sounds pretty good as a balanced headphone amp (eg, if you terminate the speaker terminals as XLR), even though it is only SE. Again, for headphones, I agree that the benefit in perceived sound is not as much from noise defeating but because you have a "push-pull" more articulated control of the drivers (and often more power).
just brew it! wrote:I thought Tripath was essentially Class D with some minor tweaks?
just brew it! wrote:Since the ground return in the headphone cord has non-zero resistance, a common return allows the signal for one channel to affect the ground reference for the other channel, which causes crosstalk. 4-wire connection with separate ground returns fixes this...
G8torbyte wrote:True, the RME’s single ended output is better configured for powering the headphones since the output impedance is 0.1 Ohms and the XLR jacks have 200 Ohms output impedance according to the specs. I’m going back to the headphone output since I’m reading more that the inverted output/input impedance difference is not the correct way. I read somewhere that a general rule should be that output impedance should be 10X less than the input impedance for headphones.
Ifalna wrote:I wonder if crosstalk is an actual problem in headphone applications.
When I am sitting in front of a stereo speaker setup, my right ear does not just hear the right speaker, it hears both + a gazillion reflections of the room wall.
sluggo wrote:
BTW, the amplifier I was talking about was the Tripath TA1101, which operates in bridge mode in each channel (each channel has high- and low-side drivers). Tripath amplifiers are not SE.
cynan wrote:sluggo wrote:
BTW, the amplifier I was talking about was the Tripath TA1101, which operates in bridge mode in each channel (each channel has high- and low-side drivers). Tripath amplifiers are not SE.
I meant the input/preamp portion of the unit was SE. But yes, the amplification chip (TA1101 chip) section itself is balanced. Even more interesting, you worked for Tripath? The Tripath based amps I have are branded as Virtue Audio (now essentially defunct). Their business model was to buy up a lot of Tripath chips (probably the TA1101, but they would never say) and run them at up to 30V (when they were designed for 12V, I think). Essentially, an "overclocked" Tripath chip with fancy input capacitors. Being interested in overclocking CPUs at the time, this naturally caught my interest. The result was that the AMP could deliver something like 50+ watts per channel into 4 ohms (considerably more than any other Tripath design I've seen). Heat dissipation was accomplished by attaching a CPU-like heat-pipe cooler attached to the aluminum enclosure. They were a bit overpriced, but the build quality was good, components were of high quality, and (like most class D) were extremely efficient (i.e., easy to power on batteries within reason), and I still think they sound great.
sluggo wrote:cynan wrote:sluggo wrote:
BTW, the amplifier I was talking about was the Tripath TA1101, which operates in bridge mode in each channel (each channel has high- and low-side drivers). Tripath amplifiers are not SE.
If the Virtue product was running at 30V, then it was not using the TA1101. The 1101 had two die inside - the control/logic die, which ran at 5V, and the power die (sourced from ST), which was rated at no higher than 14.4V max operating, iirc. Go higher than about 19V under any conditions and smoke leaked out.
The Virtue amp was likely using the TK2050, which was a two-package design with a max operating point of 36V, but only with optimal layout. Safer to stay at 30-32V, which provided about 40W per channel into an 8 ohm load at nothing-ish THD. I think a dual TK2050 design with paralleled outputs per channel would get you to 70w per channel into 4 ohms, so I'm guessing that's how your amp is set up. The power dies are current limited and so are not rated under 6 ohms in stereo mode for a single TK2050.
I worked at Tripath from '98 up until things went kablooey. I really liked the 2050 and thought it was one of the best things we did. It had the sound quality of the Overture parts with a far less expensive bill of materials. Put it together with a spare laptop supply and you have a cheap 20W DIY amp with fantastic sound quality.
cynan wrote:sluggo wrote:cynan wrote:
If the Virtue product was running at 30V, then it was not using the TA1101. The 1101 had two die inside - the control/logic die, which ran at 5V, and the power die (sourced from ST), which was rated at no higher than 14.4V max operating, iirc. Go higher than about 19V under any conditions and smoke leaked out.
The Virtue amp was likely using the TK2050, which was a two-package design with a max operating point of 36V, but only with optimal layout. Safer to stay at 30-32V, which provided about 40W per channel into an 8 ohm load at nothing-ish THD. I think a dual TK2050 design with paralleled outputs per channel would get you to 70w per channel into 4 ohms, so I'm guessing that's how your amp is set up. The power dies are current limited and so are not rated under 6 ohms in stereo mode for a single TK2050.
I worked at Tripath from '98 up until things went kablooey. I really liked the 2050 and thought it was one of the best things we did. It had the sound quality of the Overture parts with a far less expensive bill of materials. Put it together with a spare laptop supply and you have a cheap 20W DIY amp with fantastic sound quality.
I'm sure you're right. Great to finally know what the chip in the amps are.
But 20w, at least with a parallel 2050 implementation, seems to discernably limit performance. I started with 24V/30W switching PSU, then went up to a 30V/100W switching PSU and the improvement was definitely more than just a bit of increased volume. I found that running them on batteries (higher current than the switching supplies - 2 5AH lead acid in series producing about 27V on full charge) provided the best results so its interesting that you say there's a hard current limit.
BTW, do you know how widely the 2050 chips were adopted? As stated, aside from the virtue AMPS, I've not seen Tripath designs rated for that much power..All the Chinese versions seem to be using the TA1101...
Mr Bill wrote:Back when I was playing WOW I was listening through Grado SR60's and later SR80's. It kept the noise down for the rest of the house and made the whole experience much better. I thought Blizzard did a fine job of making sound cues and ambient sounds very useful. I know some people liked to play with sound off but I really liked how sounds added to the immersive realism. I hung a cheal plantronics headset around my neck for its noise cancelling mike for talking to other players.