Building balanced headphone cables?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Please disregard my ramblings re IVY III. It bears almost no resemblance to the SuSy circuit commented on by Nelson Pass. It was not merely the names that changed :-(
You may have been right initially. The IVY III is based around four OPA1632D op-amps that are fully differential, allowing a circuit configuration that is "super symmetric." The IVY III appears to have the topology necessary to eek SuSy from the OPA1632D. Nelson Pass has commented that he hasn't worked with the OPA1632, but after the part was brought up in discussion on diyAudio, Texas Instruments has subsequently licensed the SUSY patent.

Trivia aside, if you're not afraid of op-amps, and if you prefer not to be forced to tweak biasing pots while looking at a distortion analyzer (being consigned to forever wonder whether temperature drift has allowed distortion to creep back in to your hand-made amplifier), then the IVY III or a similar design based on the OPA1632 might be just the ticket.
 
By the way, thanks for the links, Wolfsin.

Regarding your quad DAC boards, is that really necessary? It seems like it would be simpler to have dual DAC chips that each have + and - balanced outputs. That way, the +/- balanced outputs should be more closely matched and thus be better at rejecting noise. Then again, I haven't been shopping for DAC chips lately, so the fancy balanced outputs DACs that I've used may not be available in your preferred flavors. Note: I assume all of this is on-topic for the "balanced headphone" part of the topic, if not the "cable" part.
 
The quads may not be necessary in the longrun but for now I am toying with the idea of using the extra flexibility to experiment with interpolation of redbook DAC outputs -- something along the lines of Onkyo's VLSC. Hardware is so available and I am living out my childhood dreams but I am actually a software kinda guy so would rather write a simple (read freeware) filter driver to massage bits on the fly.

If you are right re the Ivy I will be very pleased. I have the kit inhouse but am (re)wiring balanced cables as I await delivery of 'the wire' PCBs due shortly. In addition to the GAC3 I am seriously contemplating cannibalizing an unshielded dual twisted pair by Rockford Fosgate and soldering on Cardas connectors.

I am listening to JMR 'Music for a Glass Bead Game' very carefully for differences among the three cables I have now. Whether I am right at the edge of what remains of my hearing or whether there is little or no difference I cannot tell. Perhaps 'the wire' will reveal differences. Changing cables is painful even with connectors.

Is nobody interested in putting forward a candidate cable to test my offer?
 
When izEnuff ENOUGH?

I like the GAC-3 very much but three of four of my Cardas solder joints failed. Fearing that I might melt the plastic with excessive heat, I failed to apply enough.

I tried unshielded dual twisted pair. Rockford-Fosgate produces assembled cables ending in RCA plugs. One of these was sacrificed and the split used to attach a mini carabiner to anchor the cable to the headset frame. Inside the cable I found a black-white pair twisted separately from the black-gray pair. The former got wired to the black (left) HD-600 connector and the black-gray pair to the red. In each case I chose to connect the black conductor to the larger pin. That pin is in some Senn-docs called gnd, but for floating balanced signals that is meaningless. In my case I will connect it to the inverse phase signal, i.e. the one sometimes marked as -.

On the amp end, the TA4M, male panel connector is the most readily available. Its pins are numbered 1,2,3,4. #1 is at 2 o'clock, #2 at 6, and #3 at 10. #4 is the middle position. This is looking into the open end of a male connector. There is no standard for which wire goes to which pin. Keep track of how you wire these for reference in wiring stereo plugs, extension cables, etc., so you do not mix L-R conductors and lose noise immunity.

The convenience of having the pairs split out at each end can be used to advantage. You might wish to connect TA4? miniXLR to a pair of female XLR if you wish to interconnect with much of the commercial bridged headphone community.

Yes, the GAC-3 sounded better without the cold solder joints and the dual twisted pair also sounds good to me. My final take is that I have wired up several cables whose sonic properties all exceed my ability to distinguish among them. I have built my final headphone cable. Allow me to thank everyone for passing up my offer to build another.

I have gone thru similar experiments with multiple preamps and multiple power supplies driving them. The absence of DC offset in a JC-2 (because of careful component matching?) first allowed me to glimpse sound so crisp I could scarcely believe it. Will "the wire" surpass it or is my hearing the limiting factor? Will the JC-80?

The first step into balanced headphone listening involves many changes to one's system. Determining what is essential and what is ephemeral to the wondrous world of balanced audio can be long and arduous. How important is full balance from bits to cans? Can everything before the final bridged amps be single ended? How important is low noise in the power supply for those amps when virtually all common noise gets cancelled by the cans?
 
Where qusp et. al. spill their prefs

2) What do I do about converting my HD-580's to balanced? DIY? Where do I get plugs? What type of plug/s are commonly used for balanced headphones? (I'm looking at YOU qusp) ;)

The above post and several that follow it explore a larger range of connector possibilities by a couple of fanatics. It makes interesting reading on the topic of this thread.
 
Hi people,

I'm about to rewire my Grado for my new balanced amp project.
I understand that the only "commercial standard" for headphone connectors is from AKG (1000).
Is it a (male) Rean Tiny XLR ? Or a "mini XLR (what's this ? a smaller shell with the same pins ?).
On the amp side, what is the "standard" ?
I'd prefer to stick to the 4-pole (female) XLR, but maybe it's better to standardize on Rean Tiny XLR ?

Confused...
 
The eBay search '(ta3f,ta4f,minixlr)' will turn up quite a number of AKG compatible parts. The selection of panel mount types is limited, however. I use ta3m inputs and ta4m output (to headphones) and a minixlr to xlr converter to provide compatibility with commercial products thru standard xlr microphone cables. Unfortunately pinouts even for xlr vary. qusp et all have other favorites, primarily due to limited panel mount selection. See post just prior to your question for a pointer.
 
FWIW

I just recabled Sony MDR V6 to bridged and the difference is so dramatic I am convinced I cannot be unbiased. I was ready to donate them and now would not part with them. I had used them with Creek OBH11 and Oppo after I stopped listening to those with HD600s. They sounded OK at best.

I have noticed that I am quite sensitive to rail voltage so perhaps it is the increased range as much as getting rid of the common ground. BTW, they require no connectors so cannot be swapped as with the Senns but a recable with GAC-4 is one of the simplest ways to get LOTS of bang for very little bucks.
 
Accidents and failures present some of the best learning opportunities. I will follow up on Steve's contention that SE can be made to sound as good if we just get the wiring and connections good and clean. I just built the Bal-Bal "wire" and it took things up a notch. opc also made the wire in SE-SE and I will hook a ta4f to it so everything will be identical except the version of that preamp. It will be at least a month because I am short of parts and my surface mount skills are not up to snuff.
 
I will follow up on Steve's contention that SE can be made to sound as good if we just get the wiring and connections good and clean. I just built the Bal-Bal "wire" and it took things up a notch. opc also made the wire in SE-SE and I will hook a ta4f to it so everything will be identical except the version of that preamp. It will be at least a month because I am short of parts and my surface mount skills are not up to snuff.
Steve's claims make sense in light of the popularity of bi-wiring as a huge step towards bi-amping. With speaker crossovers that can be separated into high-pass and low-pass, there is a significant improvement to running separate wires (bi-wiring). This is because the speaker wire participates in the crossover currents, and the woofer and tweeter sound clearer when current to one does not starve the current to the other.

Headphones cannot be bi-wired in the same sense because there is no separate woofer and tweeter, however it still makes sense to separate the ground wires for left and right so that one is not starving the current for the other. Electricity always flows in a circuit, and you cannot have full current on the left and right + wires if they share a single - return wire.

I look forward to reading about actual results. At the moment, I have nice headphones but no nice headphone amplifiers, so I probably won't have the time to experiment for myself. But this is all very educational and I hope to learn more by following this thread.
 
Holiday Slowdown

Nice to see you back rsdio. If you are considering a headamp you might find the other thread I started of interest -- building a JC-80 with the help of Mr. JC himself. Following is from that thread.

My trepidation regarding surface mount construction was entirely justified. . .After assembling a Bal-Bal 'wire' I realized the JC-80 servos were not a good first (or even second) surface mount project. . . . and I hope the JC-80 will be even better

I will not get to build and test an Se-Se wire until I get one channel of the JC-80 up and running. It has been a rather more demanding project because of the special power supply requirements for that design.
 
I am short of parts and my surface mount skills are not up to snuff.
You might consider looking around near where you live for assembly shops. I found one that I drive by nearly every day - just a 10 minute walk from where I work. They charge as little as $30 for some SMD work, but that was after I brought a lot of business to them. The more I ask around, the more I learn that there are many assembly shops in the Seattle area, and some are hungry for business. In other words, you don't necessarily need to build your own Toast-R-Oven for SMD or invest in a hot-air rework station. You might be able to just hire someone.
 
Hiring out the soldering izWhut I used to do. Then HP laid off the techs I knew there and my buddy accepted a position at Google and . . .

When my ancient iron died I purchased a knockoff of the Hakko and am just learning to use it. I really think flux was the issue and I was a bit overanxious about how the Bal-Bal would sound. Ovens and hot air are NOT in my future (I hope). If I get to anything needing ovens and rework I will find a shop.

I really hope you can avoid a few of the iterations I have gone through building headamps. It has helped me to understand the problems better but the amazing thing is that I am still getting reasonably big improvements in detail. Failure will come when the improvements halt and I decide that the headset is the limiting factor :-(
 
A chance to relax?

Serendipity took took pity on Wolfsin. qusp demanded pix of the first wire of the second run and he recognized the power supply I was using. This led to the realization that marginal current handling has led to many of my equivocal calls in past tests. I mention this because it is exactly those who might be tempted to build their own cables who might also head down the seemingly never ending pathway Wolfsin has embarked upon.

Combinatorix practically guarantee that experiments will be forgotten, misinterpreted or worst. It was the power sources that tripped up Wolfsin. Knowing that LME49600 was a shining star, Wolfsin 'doubled down' on that IC and bridged eight of those suckers to make sure the current hungry Senns would be satisfied. Well, the Senns sucked the (thought to be adequate) power sources too hard so in some tests glimpses of nirvana would appear only to be lost again.

Yesterday the mysteries were resolved by putting a beefier powsupp to work driving the left can with a partially depopulated 49710-49600 board and the right can with one channel of a Bal-Bal wire. The reduced current requirements allowed both to shine BRIGHTLY. When you find nirvana Wolfsin thinks you will recognize it!
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.