What kind of wire do you use to connect your equipment?

[Matthew J Poes]

I'm looking to do an article on DIY speaker cables, interconnects, etc. It will cover a range of ways to build your own cables at modest prices. I'll readily admit that material costs for DIY will always exceed the absolute cheapest cables one could buy (or the free ones often included) but the quality is still a good bit better. I'll include measurements of the resistance, capacitance, and inductance of these cables.

I'm curious the range of cable's people use. Did you make your own? Free stuff included? High-end?

In my own system I have a mix of cables from a variety of manufacturers and of my own creation. I have pure silver tri-braid cables I made with bullet RCA connectors as I was curious if silver really sounded any different. The cost to make them was less than $40. I have a bunch of BlueJeans Cables' RCA LC1 with Canare RCA connectors for a few connections as well as longer subwoofer runs. I have Zu Audio cables sent to me as samples to test out years ago.

For speaker cables I currently use OFC copper in wall, standard grade, high strand count, and 12 gauge. It's finished with copper BFA connectors I had, not sure where I got them. I've built probably over 100 different audiophile speaker cables in the past, though none of them were/are long enough to run to my speakers in the current system. I can share how to create some of these cables for anyone interested.

I'm not going to delve deeply into any sonic differences among cables as the goal of this and the upcoming article is to focus on getting the job done. If for no other reason, I like to have robust cables in my system to prevent problems due to damage. I've had cheap cables fail quite a bit, even untouched ones. Given the frequency with which I disconnect equipment in my system, having a decent cable makes sense. I also like to tinker and RCA, power, and speaker cables were all an easy target for the DIYer.

My current favorite line level connection is balanced due to its inherently better shielding and noise cancelation along with the more robust locking connectors that are available cheaply. My current favorite speaker connector is speakon. Nothing I've used is as robust, and in discussion with a number of amplifier engineers and an interesting incident with the measurements of a very low noise amplifier in stereophile, I've come to learn that it is by far the lowest noise and most reliable connection on the market (Phoenix is also very good but more uncommon, and harder to use). Starting a few years back I started buying Speakon in bulk and building all my projects with speakon connectors. They are not expensive, handle a ton of current, are easy to connect and lock in place, are durable if pulled on, and won't add any noise ever.

I also like the look of a well finished cable and so I like to add heat shrink and techflex. While it has no real performance advantage, it just looks more professional. I'll upload some pics later when I get a chance.

 

[ddude003]

I use BlueJean 12 White Speaker Cable, terminated with locking banana plugs... Audioquest copper for all power and interconnects... RCA 14 gauge for surrounds... All things in or near the signal path matter... Just ask Tesla, Maxwell and Faraday... ;^)

 

[Tony V.]

I just buy most of my cables from Monoprice. I made all my speaker interconnects with just bulk oxygen free stranded 12awg copper speaker wire. and then I soldered some nice gold plated banana plugs on the one end and "U"crimp on connectors on the other for the speaker ends. I put the wires in a black braided jacket and shrink rapped the ends.

 

[Matthew J Poes]

I just buy most of my cables from Monoprice. I made all my speaker interconnects with just bulk oxygen free stranded 12awg copper speaker wire. and then I soldered some nice gold plated banana plugs on the one end and "U"crimp on connectors on the other for the speaker ends. I put the wires in a black braided jacket and shrink rapped the ends.
View attachment 6927

Lookin' Good Tony!

I'll be sharing some "secrets" for top quality cheap connectors in the article.

 

[Wayne A. Pflughaupt]

I just go totally basic. My favorite connector is the Neutrik Rean NYS373 with Canare’s thin L-2ES. I’ve made component video cables with the same connector and Canare LV-61S and even plain-vanilla RG-59 (albeit with a copper shield). I don’t see the point of heat shrink and techflex for something that’s going behind the gear, never to be seen by anyone anyway.

I have some oxygen-free 12-ga. zip cord that I may resurrect into duty when we buy a house here in Corpus Christi. In our Katy house I just ran enough of the in-wall speaker wire to reach the front three speakers. I can’t tell a difference in sound either way.

BTW, balanced doesn’t necessarily equal better shielding. Take a look at this stuff. That copper spiral is the shield, no kidding. Even the cheap RCAs that come in the box with audio or components have a better shield than that!

By the way, have you seen my DIY cable article @ HT Shack? I did some other tech articles there as well.

http://www.hometheatershack.com/for...-guide-making-your-own-cables.html#post116023


Regards,
Wayne

 

[Matthew J Poes]

I just go totally basic. My favorite connector is the Neutrik Rean NYS373 with Canare’s thin L-2ES. I’ve made component video cables with the same connector and Canare LV-61S and even plain-vanilla RG-59 (albeit with a copper shield). I don’t see the point of heat shrink and techflex for something that’s going behind the gear, never to be seen by anyone anyway.

I have some oxygen-free 12-ga. zip cord that I may resurrect into duty when we buy a house here in Corpus Christi. In our Katy house I just ran enough of the in-wall speaker wire to reach the front three speakers. I can’t tell a difference in sound either way.

BTW, balanced doesn’t necessarily equal better shielding. Take a look at this stuff. That copper spiral is the shield, no kidding. Even the cheap RCAs that come in the box with audio or components have a better shield than that!

By the way, have you seen my DIY cable article @ HT Shack? I did some other tech articles there as well.

http://www.hometheatershack.com/for...-guide-making-your-own-cables.html#post116023


Regards,
Wayne

Hi Wayne,

I have read your article. This won't be a replacement for that at all. It will be a very brief 300-500 words with some pics. Additionally, I will try to focus on cheap but good and primarily using crimp connectors. I want to make it so anyone can do this as simply as possible.

As you probably know, there are a bunch of connectors on the market now that provide a very good connection using nothing by high pressure crimping. I believe the concept was invented by Canare, but at this point a lot of clones exist. They use a crimp pin for the center conductor and a crimp sleeve for the shield. It used to be that the special tools were 100's of dollars, but now you can get a crimper that will do the job for under $20 and the connectors cost a few dollars each. They are as well made as anything I've used. They are pretty much standard in studio/video RCA/BNC connectors.

I like balanced systems for the circuits active noise cancelation, not because the cable is inherently better shielded (I didn't say that right earlier). Like you say, it isn't true that a balanced cable is better shielded. However true balanced circuits are dual differential and I prefer dual differential circuits all else being equal. While there is no rule that says these are always lower in noise, the reality is that in todays cutting edge tech, they do seem to be quieter, eeking out that last few bits of dynamic range. The advantage of canceling induced noise on the line is a plus, the inherently better approach of having +/- and ground as separate legs, all of this is better to me. Finally of course is the superior XLR connector. It's just my preference. Mind you almost none of my gear meets this criteria because of the price point of my equipment, but if I had my druthers it would. At the low end of the price spectrum, a lot of cheap gear is not truly dual differential and in fact uses cheap and noisy differential conversion chips. I don't like that.

But I digress, this is about cables!

 

[Wayne A. Pflughaupt]

Wow, didn’t know about the new options for crimped connectors. I wrote the DIY article partially as an answer to the expensive Canare system, which was all the rage at the time. People raved about them all the time on numerous forums.

But by the time I got the article finished, you weren’t hearing about the Canare’s as much. From there it wasn’t much longer to you didn’t hear about them at all, anywhere. Maybe I managed to shut them down with the cheaper soldering option? LOL (Actually, it was probably the advent of HDMI.)

Crimped connections have the potential advantage of keeping the shield intact all the way up to where the center conductor connects (at least that’s the way it was with the Canare system). That may account for the noise improvement you’re talking about. I look forward to your article and learning about those options. Shoot, I think I’d invest in a $20 crimper that would allow me to make a cable in a fraction of the time soldering takes!

Hey, your article might shut down the DIY soldering method!

Regards,
Wayne

 

[Matthew J Poes]

Wow, didn’t know about the new options for crimped connectors. I wrote the DIY article partially as an answer to the expensive Canare system, which was all the rage at the time. People raved about them all the time on numerous forums.

But by the time I got the article finished, you weren’t hearing about the Canare’s as much. From there it wasn’t much longer to you didn’t hear about them at all, anywhere. Maybe I managed to shut them down with the cheaper soldering option? LOL (Actually, it was probably the advent of HDMI.)

Crimped connections have the potential advantage of keeping the shield intact all the way up to where the center conductor connects (at least that’s the way it was with the Canare system). That may account for the noise improvement you’re talking about. I look forward to your article and learning about those options. Shoot, I think I’d invest in a $20 crimper that would allow me to make a cable in a fraction of the time soldering takes!

Hey, your article might shut down the DIY soldering method!

Regards,
Wayne

Hah, I don't really see it as either or. Each has its advantages and disadvantages. In most of my DIY cables (since I can solder) I tend to do both. Having said that, solder joints aren't used in high current assemblies for the most part, so I prefer to follow suit and use clamp or crimp connectors there. With RCA or high level, it just depends. Realistically, an unshielded cable can work fine with no real induced noise for most people. The silver tri-braid I did (I know technically that is a shield, but not much of one) didn't show any real signs of noise induction as compared with a Canare cable I had. I didn't try to induce noise as I figured my goal was to keep out the noise that is there, not fictitious serious noise that might show up.

So is this the same Canare system you are talking about:

The Canare Rcap? If so then yes, all the cheaper copies have the same method. The only differences between this and the cheapest of copies is the center pin. Some don't use a crimp pin and instead use a compression method. I prefer the crimp pin myself.

If so, then what I will show is that you can crimp this with a cheap RG-6 crimper as its basically the same as the Canare. The center pin can be crimped with a standard crimper.

There is also the compression style connector, which is super easy to use, but must be used with a solid center conductor cable:


Both allow full shield coverage and provide a strong connection. I like crimp better because the "tons" of pressure can make an air tight sealed connection, and if done with enough pressure, actually cold welds the metals together. Compression doesn't do that, but in practice, works fine, noise free.

I like heat shrink because it adds a little strain relief and looks nice. I also tend to use it for channel identification. As for tech flex, I know, it serves no purpose, but it looks nice. It's so cheap. I think I just got a 50 foot roll for like $5 surplus (I'd have to double check, but it was a cheap enough to be an impulse buy). It can also be used for channel identification or to provide abrasion resistance if the cables are coming out of something sharp.

 

[Mark C Flick]

I also have a mix, including some Monster, Phillips, and Monoprice but, most of it is generic. I've never been into the fancy cables.

 

[Matthew J Poes]

I also have a mix, including some Monster, Phillips, and Monoprice but, most of it is generic. I've never been into the fancy cables.

I think what you have done is very common. I started the same way.

Cables can be a fun diy project but it’s not a smart use of funds if you have a limited budget. The actual components, speakers, most life expenses all matter far more.

My interest in high end cables was curiosity and the fact that as a diy project they were easy to whip up. I’ve not bought any really expensive cables but did have a few given to me as samples over the years. I have cables from Kimber Kable, Zu Audio, Cardas, and wireworld. I also have a ton of cheap stuff from the free cables included with gear to AR, Monster, Monoprice, random eBay special. I used to make my own using telephone cable I had that was pulled from a business and cheap Radio shack connectors. Eventually tried better wire and better connectors. My last venture before I stopped building cables was to buy pure single crystal uncoated silver wire and hand feed it into Teflon tubing. Then making my own braids and twists to construct cables. I connected them with pure silver or pure copper connectors. All sounds far more expensive than it was as the wire was raw jewelry wire from Japan, and the connectors were sent as samples. A friend I knew got a few as samples for cables he was manufacturing and he didn’t need them, sent them my way. I though it was worth a shot. They were fun to build and look very fancy. Gave my stereo a little bawler bling.

 

[Sonnie Parker]

After spending quite a bit of time talking with (mostly listening) to Roger Sanders (formerly with MartinLogan and now Sanders Sound Systems)... and having owned a lot of ESL speakers, I am sold on using speaker cables that offer low inductance, low capacitance and some resistance. Even with other types of speakers, I believe this cable gives me the best option of cables NOT interfering with the sound of the speakers. These are not difficult to make, just buy Canare 4s11... order some colorful techflex from Hong Kong in your choice of color, with some good heat shrink tubing... connectors if you like, and you'll have good looking and good quality speaker cable.

For power cables, I make most of my own to the length I need. I use very flexible power cord and purchase some decent connectors off Amazon, with my techflex and heat shrink... make a good looking and good quality power cord. I could probably sell them for 100's if I'd put them in the freezer for a few days.

 

[dc2bluelight]

My current favorite line level connection is balanced due to its inherently better shielding and noise cancelation along with the more robust locking connectors that are available cheaply.

I like balanced systems for the circuits active noise cancelation, not because the cable is inherently better shielded (I didn't say that right earlier). Like you say, it isn't true that a balanced cable is better shielded. 1. However true balanced circuits are dual differential and I prefer dual differential circuits all else being equal. 2. While there is no rule that says these are always lower in noise, the reality is that in todays cutting edge tech, they do seem to be quieter, eeking out that last few bits of dynamic range. The advantage of canceling induced noise on the line is a plus, the inherently better approach of having +/- and ground as separate legs, all of this is better to me. Finally of course is the superior XLR connector. It's just my preference. Mind you almost none of my gear meets this criteria because of the price point of my equipment, but if I had my druthers it would. At the low end of the price spectrum, 3. a lot of cheap gear is not truly dual differential and in fact uses cheap and noisy differential conversion chips. I don't like that.

But I digress, this is about cables!

Well, "balanced interface" actually IS about cables (and the interface too), so it's not a digression at all.

1. There is a prevalent misconception here, it's actually more about terminology, and understanding what means what. "True Balanced circuits", in a discussion of cables and interfaces, needs a bit of clarification. From this link, ""A balanced circuit is a two-conductor circuit in which both conductors and all circuits connected to them have the same impedance with respect to ground and to all other conductors. The purpose of balancing is to make the noise pickup equal in both conductors, in which case it will be a common-mode signal which can be made to cancel out in the load." Notice that what's not mentioned is anything about differential signals or a dual-differential amplifier topology (which is something else entirely). What's important is equal noise pickup in both conductors, and to make that possible, the impedance to ground of both conductors must be equal. That means some cables actually work better than others, and some shielded cable actually is worse than UTP. The other topic here, "dual-differential" amplifier topology is often confused with, or blended with balanced signaling. In fact, a dual-differential circuit is essential two identical circuits, amplifiers, DACs, whatever, that handle the same signal but with inverse polarity. The result may in some cases be a lowering of some types of amplifier induced distortions, but generally this topology does not improve noise performance because of the doubling of uncorrelated noise sources which places both as differential noise sources. And, the lowering of possible nonlinear distortion products is entirely dependent on other design concerns. So we need to eliminate "dual-differential" from the discussion of cables and interface.

2. The entire point of balanced signaling is the cancellation of common mode noise. However, for that to be an actual real benefit there must be some common mode noise to cancel. Unbalanced interconnects do not always have common mode noise, and in fact, usually work very well. Common mode noise can enter an interconnect capacitively, inductively, or electromagnetically. One source of capacitive and inductive noise coupling is actually the shield, if it connects two chassis grounds that have enough difference of potential to cause an AC current to flow through the shield. To be clear, the shield does nothing to prevent that coupling, in fact, it's the cause. Lifting the shield/ground connection at one end breaks the current flow, and reduces that noise coupling. The shield is an electromagnetic shield (if grounded well) and an electrostatic shield blocking capacitive coupling. It is transparent, however, to magnetic coupling, which is not an insignificant noise source. However, there are conditions where none of this is an issue. Shields between two devices with similar ground potential, systems within buildings that provide any form of RF attenuation, cables routed correctly away from high voltage and high magnetic fields, and most importantly, cables driven by low impedance circuits. That last one is important because the lower the circuit impedance is with respect to ground, the less stray noise can couple to it.

There's a down side to balanced signaling: it takes more circuit elements to accomplish, and fairly precision ones to achieve high common mode rejection. More circuit elements mean more chances for uncorrelated noise. And, to achieve high common mode rejection attention must be paid to the line driver as well as the line receiver. Neither is trivial, yet neither is performance tested in specs or reviews. Some interfaces in equipment are quite shoddy.

3. Frankly, not a lot of gear, cheap or expensive, is dual differential because mostly it doesn't need to be. Single ended topology works just fine, is less than half the cost, and has half the components. However, again, dual-differential device topology has little to do with balanced signal interfacing. To drive a balanced cable there must be some means of developing a balanced drive signal. That means each conductor of the pair is driven with an equal but opposite polarity signal, and most importantly, from a source with equal (balanced) impedance from each conductor to ground. The driver can be active or passive (with a transformer), but balanced impedance to ground is the key, and to some extent, lower is better. The line receiver must also present a balanced impedance from each conductor to ground, and preferably fairly high impedance. It musts then have a high CMRR across the audio band, and up through problematic RF. That's a pretty tough task, because RF proofing an input can, and usually does, through its input impedance balance off. Again, the proper execution of a balanced system is not trivial. However, there are inexpensive monolithic line driver and receiver chips that can make the job easier, and more effective. These are not noisy by definition, though. And because all of the critical trimming is done at the chip level, the performance of some of these "chips" will simply trounce discrete designs, often by 30-40dB of CMRR and quite a bit lower noise too. For example, the THAT Corp. 1200 series (there's a dual channel for $3.80 in singles) offers these performance metrics:
• High CMRR:
— typ. 90 dB at 60 Hz
— typ. 85dB @ 20 kHz
— Wide bandwidth: > 22 MHz
— High slew rate: 12 V/us
— Low distortion: 0.0005 % THD
— Low noise: -106 dBu

And the THAT Corp 1606 ($4 singles, $2.50, quantity) line driver offers:
• High-current differential output driver
• Balanced, transformer-like floating output
• +6.0dB gain, balanced or single-ended termination
• OutSmarts technology tames clipping
into single-ended loads
• Stable driving long cables and capacitive loads
• High output: +28dBu into 600 Ω
• Low noise: -101 dBu
• Low distortion: 0.0007% @ 1kHz

Are those cheap? As ICs go, not really, but not so expensive when you consider the alternatives.

I don't think global generalization about the benefits or detriments of balanced cables or interfaces is really productive. It's one of those subjects where the details and specifics are really necessary to understand when to apply the technology, and what products incorporate interfaces that are actually good. Wouldn't it be nice to see balanced I/O tested in reviews?

So what does all of this mean in a discussion of DIY interconnects? Determine, first, if you actually do have noise on an interconnect, and figure out why. It's likely the issue may be solved without using a balanced signal system, which will save significant cost, and possibly offer slightly better total performance in the end. If it actually is a common-mode issue and solving it with a balanced interface and cables is dictated (actually, a pretty rare situation), be aware that not all balanced interfaces or cables are identical, results may vary. There is no electrical reason a balanced interface/wire system should sound better, apart from common-mode noise cancellation, and several reasons why, if noise immunity is not required, it may sound worse.

 

[dc2bluelight]

I use compression RCA connectors on RG59 cable for custom audio wiring because they let me make a reliable cable very, very quickly. But mostly I use pre-made cables because I don't feel right about charging customers for a custom cable when a Monoprice cable will work just fine.

I do make my own XLR cables using Neutrik NC3MXB and NC3FXB (favorites of mine since their introduction) because their strain relief is excellent, assembly requires no tools, contact pressure and material make for a very reliable connection (as long as they're mated with others of their ilk). Wire is dictated by the application.

I stopped using heat shrink years ago because it just slowed down the process and offers no advantage other than visual.

A caution about Monoprice: some of their cables are not shielded, and all of the RCA and 3.5mm TRS have very tiny inner conductors. It's not a performance issue, but makes it challenging to do any custom changes on their pre-made wire.

 

[Matthew J Poes]

I use compression RCA connectors on RG59 cable for custom audio wiring because they let me make a reliable cable very, very quickly. But mostly I use pre-made cables because I don't feel right about charging customers for a custom cable when a Monoprice cable will work just fine.

I do make my own XLR cables using Neutrik NC3MXB and NC3FXB (favorites of mine since their introduction) because their strain relief is excellent, assembly requires no tools, contact pressure and material make for a very reliable connection (as long as they're mated with others of their ilk). Wire is dictated by the application.

I stopped using heat shrink years ago because it just slowed down the process and offers no advantage other than visual.

A caution about Monoprice: some of their cables are not shielded, and all of the RCA and 3.5mm TRS have very tiny inner conductors. It's not a performance issue, but makes it challenging to do any custom changes on their pre-made wire.

My upcoming article will include a compression RCA cable for just this reason. Works as well as anything and is very quick and easy.

I like dressing up cables for myself or if they are on display at all. Otherwise I agree, not a huge issue. If the cables will be removed a lot, like in someone's audio shop, I use adhesive heat shrink. I also use colored for channel idenrification.

 

[dc2bluelight]

My default cable marking is a labeler using the flexible wire making tape. The labels indicate signal source and destination. No confusion. I tag power cords at the plug too so you know what you're about to yank.

The one down side to compression connector RCA cables is RG59 ain't the flexi-est.

 

[Matthew J Poes]

My default cable marking is a labeler using the flexible wire making tape. The labels indicate signal source and destination. No confusion. I tag power cords at the plug too so you know what you're about to yank.

The one down side to compression connector RCA cables is RG59 ain't the flexi-est.

It's not. I have some samples in for the article. One is a flexible version of rg59 with a twisted center conductor. I was curious to try putting a compression connector on it. The Canare connector would work for sure but are not as fast to connect. If it works I'll let you know the specific parts. Cost is still pretty low. If you do a lot of install work, the cable was around a dollar a foot I believe in 1000 ft spools.

 

[dc2bluelight]

I tried the Canare system, got the tool, etc. when it was $100. Might be less now. Nice, if you use their wire, but no real advantage in speed or quality of connector. The tool now sits in my tool cabinet along with a connector or two. I should just sell them.

I have used compression BNCs with stranded-inner RG59. If you use the right connectors its fine. Not all work, the inner tends to fold and not seat. I like the Ideal InSite BNCs but there's no equivalent RCA. The window is really helpful, but still not great for stranded inner RG59. I'll trade the fuss for lack of flex.

 

[Matthew J Poes]

I tried the Canare system, got the tool, etc. when it was $100. Might be less now. Nice, if you use their wire, but no real advantage in speed or quality of connector. The tool now sits in my tool cabinet along with a connector or two. I should just sell them.

I have used compression BNCs with stranded-inner RG59. If you use the right connectors its fine. Not all work, the inner tends to fold and not seat. I like the Ideal InSite BNCs but there's no equivalent RCA. The window is really helpful, but still not great for stranded inner RG59. I'll trade the fuss for lack of flex.

The Belden Rg6 I got is really not bad flexibility wise. For most people once a connection is made it's done never to be touched again. For something like a phono arm it makes sense to use a more flexible low noise like a litz or other flexible thing cable, but for anything else (modern) it's just not a concern right?

 

[dc2bluelight]

The Belden Rg6 I got is really not bad flexibility wise. For most people once a connection is made it's done never to be touched again. For something like a phono arm it makes sense to use a more flexible low noise like a litz or other flexible thing cable, but for anything else (modern) it's just not a concern right?

RG6 is actually fairly stiff compared to RG59. The concern is that stiff wire causes angular force on connectors that are often soldered to boards inside the device. Some are fairly fragile, particularly HDMI connectors, which I've seen snapped off their boards simply by the angular force of a thick HDMI cable, and that's a very expensive repair. While I've never seen an RCA connector break this way, some are not really solidly mounted, so it's a concern. I like flexible cables for cable dress, otherwise you end up with a sort of thick spaghetti looking installation.

There's absolutely nothing gained by using RG6 for audio. Even quad-shield RG59 is overkill, but handy, and a bit more flexible at least. I'll do a flex force measurement of RG6, RG59 and generic audio wire later when I get the chance.

 

[dc2bluelight]

I tested for the amount of force required to bend a 1' section of cable into a natural 90 degree bend.

RG6 quad shield, solid inner: 2.9oz
RG59 quad shield, solid inner: 2.2oz
Generic RCA cable (dual/stereo) 0.3oz
Gepco "Digi-Aud" E131674 (shielded, stranded pair) .5oz

So the difference between the bend force of RG6 and RG59 is minimal, which I didn't expect. However, there are some other factors. Both RG cables slowly formed to the bend, and gradually the bending force reduced to below 1.5oz. Other RG cables have different amounts of flexibility in the jacket and inner insulator, so they may measure differently. Temperature makes a significant difference.

Both the actual audio cables needed far less force to bend, but will have to be purchased at standard lengths or customized with solder-on connectors.

I guess none of it matters really. Probably a good idea to pre-form bends so they don't exert extra force on connectors if you're using RG cable. And make sure to use the right compression connector for the wire you're using. There is a difference in connector for the RG6, RG6 Quad, and RG59! Not interchangeable. I checked, apparently I have the Platinum connectors in stock (wasn't sure, I do so few any more), but there are lots of others.

This side note: the only RCA connectors I've ever had fail were very cheap ones from Radio Shack that were purchased for an emergency need, and that was a physical failure, not an electrical one. I've also used solder-on RCA connectors by Neutrik (nice!) Switchcraft, and several sourced from China. None have ever failed as a connector, though some are more difficult to deal with. The usual RCA jacks and connectors end up with fairly high contact pressure which results in little need for exotic platings. Gold/gold makes you feel good, but honestly, I've yet to have a tin or nickel plated RCA fail as a connector (but I've only been using them for 50 years or so). Not true of 3.5 mm TS or TRS, and 1/4" TS/TRS jacks/connectors which have lighter contact pressure, and there are tons of badly plated 1/4" jacks in the world. My favorite 1/4" plug is a Neutrik (getting a theme yet?), and I have had some cheap 1/4" and 3.5mm connectors literally fall apart.

 

[Matthew J Poes]

I tested for the amount of force required to bend a 1' section of cable into a natural 90 degree bend.

RG6 quad shield, solid inner: 2.9oz
RG59 quad shield, solid inner: 2.2oz
Generic RCA cable (dual/stereo) 0.3oz
Gepco "Digi-Aud" E131674 (shielded, stranded pair) .5oz

So the difference between the bend force of RG6 and RG59 is minimal, which I didn't expect. However, there are some other factors. Both RG cables slowly formed to the bend, and gradually the bending force reduced to below 1.5oz. Other RG cables have different amounts of flexibility in the jacket and inner insulator, so they may measure differently. Temperature makes a significant difference.

Both the actual audio cables needed far less force to bend, but will have to be purchased at standard lengths or customized with solder-on connectors.

I guess none of it matters really. Probably a good idea to pre-form bends so they don't exert extra force on connectors if you're using RG cable. And make sure to use the right compression connector for the wire you're using. There is a difference in connector for the RG6, RG6 Quad, and RG59! Not interchangeable. I checked, apparently I have the Platinum connectors in stock (wasn't sure, I do so few any more), but there are lots of others.

This side note: the only RCA connectors I've ever had fail were very cheap ones from Radio Shack that were purchased for an emergency need, and that was a physical failure, not an electrical one. I've also used solder-on RCA connectors by Neutrik (nice!) Switchcraft, and several sourced from China. None have ever failed as a connector, though some are more difficult to deal with. The usual RCA jacks and connectors end up with fairly high contact pressure which results in little need for exotic platings. Gold/gold makes you feel good, but honestly, I've yet to have a tin or nickel plated RCA fail as a connector (but I've only been using them for 50 years or so). Not true of 3.5 mm TS or TRS, and 1/4" TS/TRS jacks/connectors which have lighter contact pressure, and there are tons of badly plated 1/4" jacks in the world. My favorite 1/4" plug is a Neutrik (getting a theme yet?), and I have had some cheap 1/4" and 3.5mm connectors literally fall apart.

I've had some cheap RCA's fail myself, they likely were of the same ilk as the radio shack type you mention. While it also was a mechanical failure, it led to the cable being unusable.

I too have used a ton of different RCA connectors and most are fine. I consider building cables a great first DIY project for people, especially interconnectors, as its an easy project to learn to solder on. If you damage something, its cheap. I was probably 10 years old when I first soldered a pair of RCA connectors up. I would say I have about half your experience, 25 years, but certainly plenty to experience the scope of the market. I was an installer in HS and part of college but once I college got more serious I didn't have time to continue. I didn't revisit this until later and only as a side job, so I don't likely install anywhere near as many cables as you. Having said all that, I've rarely had an RCA connector fail, and I certainly have plenty of stiff cables in my system (and have over the years). At one time cables were an easy review product and I would get free samples long before I ever had much of an outlet to review. Know a guy at audioquest, phoenix gold, cardas, etc and you get a box of cables.

The cheap connectors that fail most for me are binding posts on receivers and HDMI connectors (like you mention). I believe the problem with HDMI connectors is that most are not bolted to the board or to the frame, and they do not have a plastic sub-structure like RCA connectors. Instead they are wave soldered to the board with some adhesive and are nothing more than thin metal with SMD pads. Still, failures aren't all that common. Receivers have crummy binding posts and these seem to fail a lot. In my 25+ years of using audio electronics, the classic binding posts never failed, but spring clips did. They often broke. Then cheap binding posts became common and they have been just as bad. They can't hand the weight of heavy cables and connectors and if you aren't careful and torque them, they snap off. RCA's are much more robust. Doesn't change that I would prefer a locking XLR. Bruno Putzey has talked about the inherent superiority of the design and small but measurable differences in the performance of the connector itself. More important, as I mentioned earlier, a true balanced connection using a dual differential circuit has superior noise and distortion performance. As you mentioned earlier, this isn't true of converted circuits that are single ended by nature with a single ended to balanced converter chip, and in fact, with the exception of some like the THAT chip, tend to end up noisier than their SE output. However, true balanced designs (which are very common today and at affordable prices) do seem to be superior nearly every time. Emotiva and Teac are two brands with true balanced circuits and superior measurements on the balanced outputs (and among the lowest noise of any product on the market). Same is true of the pro-sumer brands like Benchmark and SPL. The many modern amplifies operate in BTL for greater power and lower rail voltages, which means they are inherently balanced as well. From that stand point, I see no reason not to use real balanced connections since it is the SE circuit that is actually adding parts, not the balanced. It isn't really about the common mode noise reduction of the connection, that is minor (as you say) and in short distances, a non-issue. It's greatest benefit would be for subwoofers. It's really the circuit difference that I am after. Even high end receivers often use DAC's in differential mode, meaning they too are balanced, and they are adding a balanced to single ended converter chip after the DAC to then use it as an SE design. Most receivers are still using Class AB amps and are not balanced in the analogue domain, but I actually don't think it would raise the price to go balanced on these high end designs, especially if they switched to the Class D designs. Any of the good ones are inherently balanced already.

 

[dc2bluelight]

I too have used a ton of different RCA connectors and most are fine. I consider building cables a great first DIY project for people, especially interconnectors, as its an easy project to learn to solder on. If you damage something, its cheap. I was probably 10 years old when I first soldered a pair of RCA connectors up.....<snip>...... At one time cables were an easy review product and I would get free samples long before I ever had much of an outlet to review. Know a guy at audioquest, phoenix gold, cardas, etc and you get a box of cables.

Ever see a "TIFF" connector? I used to have a full color brochure (yes, for an RCA connector). Machined, gold plated, jade sleeve. A gorgeous thing. I don't recall the price, I think in the 100s per unit. Obviously I never got a free sample. I see some evidence that Tiff Electronics is still around, but only see a digital cable with a directional flow indicator on it. Good grief. They had to change their original name (Tiffany) because a certain jewelry company looked at their product and thought they were wearable .

The cheap connectors that fail most for me are binding posts on receivers and HDMI connectors (like you mention). I believe the problem with HDMI connectors is that most are not bolted to the board or to the frame, and they do not have a plastic sub-structure like RCA connectors. Instead they are wave soldered to the board with some adhesive and are nothing more than thin metal with SMD pads.

Ah yes. HDMI "Horribly Designed Mechanical Interface". <rag mode on> Lets see...how about 19 contacts in a half inch non-locking, non-customizable, non-field-terminating connector with a cable the diameter of your big toe, with the flexibility of PVC pipe, plugged into a mating connector mechanically mounted with SMD pads? Yeah, let's do that! And lets make signaling and hand-shaking fragile. And then we'll keep changing the specs as we discover all the features we should have known about but didn't think would ever happen. All that and you can do the same job with RG59 and compression BNC connectors (HDSDI) and still retain the only real motivation for HDMI in the first place: Copy Protection. What a FUBAR. The Consortium should be ashamed, but I doubt they have that capacity. <rag mode off> Yes, I know we have smaller diameter HDMI cable now, and there's even a locking version. Just not what most people use or buy at Menards.

Still, failures aren't all that common.

It only takes a few to mess up a business model. I had a client with a mid-level AVR with a failed HDMI monitor output jack rendering the AVR useless. By the time I dealt with it as a warranty repair issue I should have just bought him a new unit out of pocket since that would have cost me less. And that was just one case. The manufacturer was Denon, in case anyone cares.

Doesn't change that I would prefer a locking XLR. Bruno Putzey has talked about the inherent superiority of the design and small but measurable differences in the performance of the connector itself. More important, as I mentioned earlier, a true balanced connection using a dual differential circuit has superior noise and distortion performance. As you mentioned earlier, this isn't true of converted circuits that are single ended by nature with a single ended to balanced converter chip, and in fact, with the exception of some like the THAT chip, tend to end up noisier than their SE output. However, true balanced designs (which are very common today and at affordable prices) do seem to be superior nearly every time. Emotiva and Teac are two brands with true balanced circuits and superior measurements on the balanced outputs (and among the lowest noise of any product on the market). Same is true of the pro-sumer brands like Benchmark and SPL.

Are there any comparative measurements for a single-ended equivalent though? I suspect not, because if you're going to make a fully balanced device you have to do the best noise design, then there's no purpose in doing half of it for fun. Remember, though, more than twice the noise sources...and it's uncorrelated, so it doesn't cancel....just sayin. The THAT chips are not the only exception, there are others that are also good, just not as smart or flexible.

The many modern amplifies operate in BTL for greater power and lower rail voltages, which means they are inherently balanced as well.

I have not done an exhaustive search on this, because it doesn't make any difference in the end result, but many amps run as differential have degraded noise figures. I don't know if that is any more universal than saying that fully balanced is always quieter. I suspect neither generalization is true.

From that stand point, I see no reason not to use real balanced connections since it is the SE circuit that is actually adding parts, not the balanced.

I don't quite see how that can be true. I've never designed a single ended circuit with more parts than the equivalend balanced one. Balancing, even using conversion to SE in and out, always has the higher parts count. Fully balanced/differential is more than double the equivalent SE.

It isn't really about the common mode noise reduction of the connection, that is minor (as you say) and in short distances, a non-issue. It's greatest benefit would be for subwoofers.

CMRR is the only advantage, though, in balanced interconnection. Yes, it's great for subs and that sub ground loop issue.

It's really the circuit difference that I am after. Even high end receivers often use DAC's in differential mode, meaning they too are balanced, and they are adding a balanced to single ended converter chip after the DAC to then use it as an SE design.

Those are two different topologies. I know there's a point to a differential DAC, I just don't recognize it as audible. Might be a noise advantage over conversion, also might not be. It gets down to the specific design, really not something that can be assumed universally.

Most receivers are still using Class AB amps and are not balanced in the analogue domain, but I actually don't think it would raise the price to go balanced on these high end designs, especially if they switched to the Class D designs. Any of the good ones are inherently balanced already.

I don't know, I'm not sure. Balanced is always more expensive vs the equivalent SE. If you save cost using Class D, then there's budget for balanced. But with single part cost reflecting 5X in the final price, just the cost of all those XLRs might change the prince point.

Don't get me wrong, I've been in Pro audio for all these decades, used balanced circuits since the actual 600 ohm days, designed many devices with balanced I/O, even using the (gasp!) converter chips. I know the point, and the cost, and when you have to have a balanced interface, but also when you don't. When a system cost reaches a certain threshold, balanced interfacing and perhaps balanced topology can just come along for the ride, and that makes sense. But the actual audible impact in small, short-interconnect systems is near or below zero due to other factors...big ones...like acoustic noise floors, speaker distortion, and blah blah. I don't want to give anyone the idea that balanced always means "sounds better", though. Sometimes yes, sometimes no, sometimes its a draw. What I find most disturbing is that designing a truly balanced and effective balanced signally system is not trivial, even in the most critical areas of CMRR and noise. There are so many rudimentary so-called balanced systems in the world, and no product reviews properly done wide-band CMRR testing. It's never spec'd either. And that takes the teeth out of the whole balanced thing.

The other thing I recently tripped over regarding a balanced system was discrepancy in level standardization. The devices were a Trinnov processor and a set of balanced input power amps with too much gain. How about 32 balanced XLR U pads? Ugh. And a bit "common, guys!" from me. You know what a generic U pad or H pad does to your CMRR?

I'll just end with this: My favorite audio connector is also the (Neutrik) XLR on real balanced audio cable plugged into a real balanced interface. I also live quite happily without it in my personal system, hearing and measuring no limitations caused by generic RCA connectors, cables, and SE interfaces. All studios I build use balanced wiring throughout with whatever connectors are dictated by the equipment selection.

 

[Matthew J Poes]

Ever see a "TIFF" connector? I used to have a full color brochure (yes, for an RCA connector). Machined, gold plated, jade sleeve. A gorgeous thing. I don't recall the price, I think in the 100s per unit. Obviously I never got a free sample. I see some evidence that Tiff Electronics is still around, but only see a digital cable with a directional flow indicator on it. Good grief. They had to change their original name (Tiffany) because a certain jewelry company looked at their product and thought they were wearable .

Ah yes. HDMI "Horribly Designed Mechanical Interface". <rag mode on> Lets see...how about 19 contacts in a half inch non-locking, non-customizable, non-field-terminating connector with a cable the diameter of your big toe, with the flexibility of PVC pipe, plugged into a mating connector mechanically mounted with SMD pads? Yeah, let's do that! And lets make signaling and hand-shaking fragile. And then we'll keep changing the specs as we discover all the features we should have known about but didn't think would ever happen. All that and you can do the same job with RG59 and compression BNC connectors (HDSDI) and still retain the only real motivation for HDMI in the first place: Copy Protection. What a FUBAR. The Consortium should be ashamed, but I doubt they have that capacity. <rag mode off> Yes, I know we have smaller diameter HDMI cable now, and there's even a locking version. Just not what most people use or buy at Menards.
It only takes a few to mess up a business model. I had a client with a mid-level AVR with a failed HDMI monitor output jack rendering the AVR useless. By the time I dealt with it as a warranty repair issue I should have just bought him a new unit out of pocket since that would have cost me less. And that was just one case. The manufacturer was Denon, in case anyone cares.
Are there any comparative measurements for a single-ended equivalent though? I suspect not, because if you're going to make a fully balanced device you have to do the best noise design, then there's no purpose in doing half of it for fun. Remember, though, more than twice the noise sources...and it's uncorrelated, so it doesn't cancel....just sayin. The THAT chips are not the only exception, there are others that are also good, just not as smart or flexible.
I have not done an exhaustive search on this, because it doesn't make any difference in the end result, but many amps run as differential have degraded noise figures. I don't know if that is any more universal than saying that fully balanced is always quieter. I suspect neither generalization is true.
I don't quite see how that can be true. I've never designed a single ended circuit with more parts than the equivalend balanced one. Balancing, even using conversion to SE in and out, always has the higher parts count. Fully balanced/differential is more than double the equivalent SE.
CMRR is the only advantage, though, in balanced interconnection. Yes, it's great for subs and that sub ground loop issue.
Those are two different topologies. I know there's a point to a differential DAC, I just don't recognize it as audible. Might be a noise advantage over conversion, also might not be. It gets down to the specific design, really not something that can be assumed universally.

I don't know, I'm not sure. Balanced is always more expensive vs the equivalent SE. If you save cost using Class D, then there's budget for balanced. But with single part cost reflecting 5X in the final price, just the cost of all those XLRs might change the prince point.

Don't get me wrong, I've been in Pro audio for all these decades, used balanced circuits since the actual 600 ohm days, designed many devices with balanced I/O, even using the (gasp!) converter chips. I know the point, and the cost, and when you have to have a balanced interface, but also when you don't. When a system cost reaches a certain threshold, balanced interfacing and perhaps balanced topology can just come along for the ride, and that makes sense. But the actual audible impact in small, short-interconnect systems is near or below zero due to other factors...big ones...like acoustic noise floors, speaker distortion, and blah blah. I don't want to give anyone the idea that balanced always means "sounds better", though. Sometimes yes, sometimes no, sometimes its a draw. What I find most disturbing is that designing a truly balanced and effective balanced signally system is not trivial, even in the most critical areas of CMRR and noise. There are so many rudimentary so-called balanced systems in the world, and no product reviews properly done wide-band CMRR testing. It's never spec'd either. And that takes the teeth out of the whole balanced thing.

The other thing I recently tripped over regarding a balanced system was discrepancy in level standardization. The devices were a Trinnov processor and a set of balanced input power amps with too much gain. How about 32 balanced XLR U pads? Ugh. And a bit "common, guys!" from me. You know what a generic U pad or H pad does to your CMRR?

I'll just end with this: My favorite audio connector is also the (Neutrik) XLR on real balanced audio cable plugged into a real balanced interface. I also live quite happily without it in my personal system, hearing and measuring no limitations caused by generic RCA connectors, cables, and SE interfaces. All studios I build use balanced wiring throughout with whatever connectors are dictated by the equipment selection.

I agree with all of this. As for the last bit on balanced, I agree. I don't see major (or any) audible flaws. It's more the measurable benefit and the why not if its there. I like the connectors better, its a theoretically better connection with some possible benefits, and in this day and age it might not have a cost consequence.

My point about the SE having more parts is that if the circuit is inherently a balanced design that must be converted to single ended, then you have more parts because you added an SE conversion stage. I know that a number of very good chips exist for this today, but then, why not just leave it balanced?

As for comparing Se to balanced, you are right that there aren't much in the way of practical examples of this. The Balanced and SE comparisons made on a same unit are inherently balanced so the degraded SE performance may just be the conversion chip adding noise. I've also looked at the ATI and Emotiva amplifiers though as some of the 2 channel models and monoblocks are the same amplifiers, one operating as a differential pair and the other not (such as is the case with a bridge tied load). The Monoblocks have lower noise in most if not all cases. Its exactly the same noise benefit as with a differential dac design in terms of the analogue portion of the benefit. I too have read the discussions of these balanced amps being noisier because there is now more in the circuit, but it doesn't seem to be true of these products. Even Benchmark, who says this about balanced headphone amps (I agree with their assertions there) use an amplifier whose performance shows lower noise in bridged mode. I suppose another possibility (which I should probably check) is that the noise isn't changing but the peak voltage is and that is the reason for the lower distortion and greater dynamic range and/or signal to noise ratio. The AP reports, those provided by folks online such as Archimego, or John Atkinson do not provide sufficient detail to know that. Bruno Putzey has been good about accommodating these requests, so if people care enough, I can ask him if he could provide some data from his amps. His amps (Hypex) are also inherently balanced designs.

 

[dc2bluelight]

I agree with all of this. As for the last bit on balanced, I agree. I don't see major (or any) audible flaws. It's more the measurable benefit and the why not if its there. I like the connectors better, its a theoretically better connection with some possible benefits, and in this day and age it might not have a cost consequence.

A board-mount XLR connector is about $2 in quantity, a board-mount RCA is about $.20 in quantity. I'm not convinced an XLR makes a better electrical connection, but it is more robust in construction. If repeated insert/extract cycles are important or the connector is exposed to wear (like in a pro application), clearly it's the better choice. Once inserted and not extracted, I don't think there's any performance advantage of the XLR given similar materials and construction quality. Looking at the contact resistance for a Neutrik RCA vs XLR, both gold plated, the RCA is slightly higher, both are so ridiculously low as to be zero in practice.

My point about the SE having more parts is that if the circuit is inherently a balanced design that must be converted to single ended, then you have more parts because you added an SE conversion stage. I know that a number of very good chips exist for this today, but then, why not just leave it balanced?

But there's no need to convert anything either way. A DAC with differential output can drive SE directly without conversion, and often does. It depends on the desired configuration of the filter/buffer amp that follows it. I checked a few data sheets, sorry no time to delve into this in depth, but they do show both methods, and no "conversion" is required either way. The parts count for fully balanced output buffering is nearly double that of a buffer that takes the differential DAC outputs and sums them. And what comes after that? More balanced circuits with twice the parts.

As for comparing Se to balanced, you are right that there aren't much in the way of practical examples of this. The Balanced and SE comparisons made on a same unit are inherently balanced so the degraded SE performance may just be the conversion chip adding noise. I've also looked at the ATI and Emotiva amplifiers though as some of the 2 channel models and monoblocks are the same amplifiers, one operating as a differential pair and the other not (such as is the case with a bridge tied load). The Monoblocks have lower noise in most if not all cases.

Do we know with absolute certainty (like a published schematic) that this is the case?

Its exactly the same noise benefit as with a differential dac design in terms of the analogue portion of the benefit. I too have read the discussions of these balanced amps being noisier because there is now more in the circuit, but it doesn't seem to be true of these products. Even Benchmark, who says this about balanced headphone amps (I agree with their assertions there) use an amplifier whose performance shows lower noise in bridged mode. I suppose another possibility (which I should probably check) is that the noise isn't changing but the peak voltage is and that is the reason for the lower distortion and greater dynamic range and/or signal to noise ratio.

Or, likely if the device is made to be switched from either SE or bridged, it's the same noise sources in the circuit all the time, so there's a 3dB increase in uncorrelated noise, but also a 6dB signal voltage gain...hence the slightly better noise in bridged. However, take away that feature and just compare what's required for SE vs balanced, I don't think you'd see the same improvement because you won't have the 3dB noise gain from the balancing circuit in place in SE mode. Again, almost impossible to compare, and not possible to test for without circuit modification, so kind of silly to discuss.

The AP reports, those provided by folks online such as Archimego, or John Atkinson do not provide sufficient detail to know that. Bruno Putzey has been good about accommodating these requests, so if people care enough, I can ask him if he could provide some data from his amps. His amps (Hypex) are also inherently balanced designs.

If his designs are balanced, then he's not in a position to verify this anyway. Balanced topology increases the number of noise sources, so noise must go up, but then there's a 6dB voltage gain too. The comparison I'm talking about would show the difference between a purely SE design and the equivalent (if there is such a thing) balanced design. That's where I think the differences would start to become very small, with SE standing a good chance at a win. The whole discussion gets very, very device and circuit specific, which doesn't really serve much purpose. For example, there are some distortion mechanisms that could be cancelled in a balanced design (once again converted to SE), but there are other ways do deal with distortion mechanisms besides differential cancellation. I just don't want to mislead anyone into thinking that balanced or fully diff is by definition somehow better, while ignoring the main point which is common mode rejection. We have to be specific about what the advantages actually are, because "balanced" is such a nice warm fuzzy word, and just sounds like it must be better, while "unbalanced" is a negative word biasing the perception of that system, when in reality those biases may be entirely imaginary.

 

[dc2bluelight]

Thought I might contribute something possibly useful for a change, so I listed some of my favorite connectors and wire.

My connector choices (most are links):
Rean/Neutrik RCA solder on
Monoprice Premium RCA solder on, actually quite nice, good for larger cable diameters
Ideal Compression RCA for RG59 They also make a version for RG6 (be careful to check the stated wire dimensions for a match to your wire!)
Platinum Compression RCA (various models for specific wire types - these are my preferred compression seal)
Neutrik NC3FXB Female gold plated XLR - this and the male version have been my standard XLR for 30 years.
Neutrik NC3MXB Make gold plated XLR - Gold version is nice if mating with another gold plated connector, otherwise silver contacts are good too.
Neutrik NP3C-BAG - 1/4" TRS - I moved to these from Switchcraft and never looked back.

Wires:
Installed balanced wiring for studios:
Gepco 61801EZ - except for the time when Gepco was owned by a weird company that made them impossible to deal with, (they are now thankfully owned by Belden) this has been my standard single-pair installed audio wire since the 1980s 100% foil shield, easy strip, multiple colors. I've also used the D61801EZGF, which is two pair, similar to 60801EZ, but it's not as much an advantage as you'd think. Easier to stock one wire type where possible. I like it in purple!

Mics and Line cables that will be handled:
Mogami W2893 and W2534 (I like the first a bit better) High flex, low microphonics, holds up to physical use.
Gepco DS401M - actually a 110 Ohm AES/EBU cable, but works for analog just fine - again, good to stock one product

Unbalanced audio, video depends on several other factors, but I stock RG59 quad and RG6 quad in 1Kft put-ups for other purposes, and use the appropriate compression connectors. But also use pre-made Monoprice cables when appropriate, and not their cheap stuff.

For flexible use, there are "instrument" cables from Mogami and Canare that are very flexible, but otherwise more painful to deal with, and expensive. Also Gepco VHD2000M is in between there somewhere. Pre-made is often a better choice considering termination time.

Since I'm involved in custom A/V install and pro studio construction many cable and connector choices are influenced by speed of termination with reliability of termination which offsets the cost of connectors and tooling.

I have not found a favorite solder-on 3.5mm TRS plug! They all have issues, usually that the shell/handle is too big to work in today's smart phones, and the smaller ones don't fit wire well. Open to suggestions, but that's one I buy pre-made. You won't see me with Monster-level pre-made cables though, just not worth the cost. I can barely justify Bluejeans.

For Cat 5/Cat 6, the Platinum EZ series changed my life. And speaking of Cat cable - it's great for analog audio too! I've wired entire broadcast facilities with it, but the system must be designed considering that.