XLR Interconnects: Is there a difference?

[Labjr]

The video purports to compare their StarQuad to Canare’s StarQuad. Start at 7:35.

Except that Benchmark doesn't manufacture the cable (Canare) or the connectors. They do assemble them and sell them for a reasonable price on their website, but Benchmark specifically discusses why they have selected that cable over others based on TESTING. I use XLR's with Mogami wire, which doesn't have as good of shielding as the Carnare and hence rejects about 10-15db less noise. In my case it makes no difference, as my system has a SN ratio of probably about 90db (and my room about 50-55 db). In well designed studio equipment, they are pushing 120db+ sn ratios and it can make a difference.

I am in the minority here that I believe I can hear very minor differences in speaker cables. Can I tell you which is more accurate, nope. Can I tell you which I prefer, yes. Would I spend $500 on a pair, no (but have spent $200 for a 15foot biwire pair of cables and just built a set of speaker cables for my desktop that set me back a whopping $20). Am I convinced that the differences I am hearing are anything more than placebo, nope. I actually wish I had an easy way to do an ABX of speaker cables, so I could put to rest for myself whether I can really hear a difference or not (and likely find out I have been fooling myself).

With XLR cables in a balanced system (just because a system uses XLR cables does not mean it is actually balance), any potential difference in cables should be absolutely minimized because of their inherent noise rejection. Ralph, the owner of Atma-Sphere, who is a regular poster on another forum has regularly stated that one of the advantages of balanced connections is that is removes any sound differences between cables when properly manufactured cables are used (and this is a manufacture that believes that cables can have a significant effect).

I remember a time when a Benchmark engineer was stating that Asynchronous USB receivers in DACs were unnecessary and did not improve jitter performance. That is until they started using them. Not sure if that engineer still works for them?

I think Bruno's opinion about removing cable differences is more accurate.

 

[Julf]

There seems to be a lot of confusion in this thread between XLR connectors and balanced/differential connections. Yes, balanced connections tend to use XLR connectors, but not always, and unbalanced connections can use XLR connectors. Can we try to be clear about which situation we are talking about.

XLR connectors tend to be mechanically better than RCAs or jacks, but the difference is rarely audible. Balanced/differential connections bring the benefits of rejection of common mode noise and avoidance of ground loops.

 

[ajawamnet]

Here's what we do for proaudio installs - this was Canare multipair with EDAC's for the console, the rest is West Penn or Belden. I've designed quite a few large installs - both broadcast and pro-audio - and this is typically how it's done; everything terminates in a wiring closet. This way, any gear that is having grounding issues can be easily isolated from shield and troubleshot. The XLR's were Neutrik and Switchcraft.

Interesting story concerning that installation of how a brand new state of the art Studer 827 had hideous hum - and how it was related to a VFW across the street:
http://www.ajawamnet.com/ajawamnet/studiohum.html

The looms for that particular install were done by a major connector termination company. Most of the mil I now do is the same way - for instance the Daniels crimper for the Amphenol stuff is stoopid expensive - https://dmctools.com/oscar/catalogue/dmc216_1270/
Even molex gets picky about stuff like that - https://www.molex.com/images/products/apptool/qual_crimp.pdf

When you get into super hi rel stuff like Glenair sells, they won't even let buy crimp tools - has to be done in house. There were some D-subs to XLR that we had to fab that they had to do

As to soldering - I've done thousands of XLR's - I noticed that some are gold plated in the solder cups. Bad mojo - see this:

Embrittlement is an issue that arises when the percentage of gold in the solder joint causes a non-metallic to migrate and eventually fail. As that IPC thing rec's it needs to be removed.

A lot of companies love ENIG and ENEPIG on their bare boards. The recent J-STD in section 4.5.1 indicates that even these electroless plating methods might cause issues. So for hte Hi Rel stuff I do we do HASL on our PCB's... even if it has BGA.

The PCB mount connectors from places like Amphenol need to be tinned prior to soldering... the gold is a surface protectant that will dissolve into the solder.

https://www.eptac.com/j-std-001-rev...ng-the-requirements-for-gold-plating-removal/

 

[Labjr]

It's worth noting that you don't have to solder XLR connections. Amphion and Neutrik used to sell "IDC" XLR connectors that had v-shaped knives to tap into the cable leads. I don't remember what Amphion's were called but the Neutrik like was called "ezCON." You don't need any tools to install them in theory, not even a wire stripper. In practice, Amphion works fine toeless but ezCON comes together much easier with a pliers wrench than finger pressure. Unfortunately both Amphion and Neutrik discontinued the IDC connectors. I asked Neutrik about it, and they told me they were great connectors but weren't purchased in sustainable volumes. They are probably better for hobbyists than pro users, and hobbyists don't buy a lot of connectors. Both might still be available from electronics surplus places, or as eBay sellers.

Neutrik has current production XLR connectors with crimp pins instead of solder cups available from Markertek and other sources.
https://www.neutrik.us/en-us/neutrik/products/xlr-connectors/xlr-cable-connectors/xx-crimp-series

Neutrik's crimp tooling is rebranded standard Rennsteig parts. However, they only work with shield twisted pair cable cable. The open barrel crimp pins are designed to capture both the insulation and the wire, and with two conductors that does not work.

The other thing about star quad is I don't know of any snakes made with star quad cores. A snake is what you want in a multichannel system, unless all the speakers are active with plate amps. All the wires stay together and the cores are individually labeled on their jackets. Even inside a media cabinet with a 7 channel system, it's easier to run an 8-core snake than 7 individual signal cables and a 12V trigger. If you hate soldering, as I do, you can even buy 1/4" TS or TRS connectors with terminal blocks on the other end for the trigger.

Neutrik Speakon connectors use a little screw clamp that works pretty well. But it uses a small pozidriv screw, which is like a Philips head screw. I almost wish they'd use a screw with an allen head or squarebit head, so I could apply a bit more tightening force. Who knows, maybe they don't want you to overtighten.

 

[jhaider]

Neutrik Speakon connectors use a little screw clamp that works pretty well. But it uses a small pozidriv screw, which is like a Philips head screw. I almost wish they'd use a screw with an allen head or squarebit head, so I could apply a bit more tightening force. Who knows, maybe they don't want you to overtighten.

Yes, but Speakon are speaker level, not line level. I've never had cam-out issues when using the correct size screwdriver or bit (PZ1).

I like the Speakon design though. The set screw acts on a metal clamp instead of directly on the wire. I still prefer to use an end sleeve to eliminate stray strands, but it is probably not necessary.

 

[Labjr]

Yes, but Speakon are speaker level, not line level. I've never had cam-out issues when using the correct size screwdriver or bit (PZ1).

I like the Speakon design though. The set screw acts on a metal clamp instead of directly on the wire. I still prefer to use an end sleeve to eliminate stray strands, but it is probably not necessary.

I think the Speakon design is great too. And when working with higher current, speaker connections it becomes more important to use genuine Neutrik and not the cheap knock-offs. At first, I didn't believe 10-12 gauge standed wire would fit in the terminals. But it does! I wonder if similar clamp terminals would work for XLR plugs?

 

[Cbdb2]

After reading this, and getting off the floor, l would stay clear of fanboy. " Sound characteristics: high frequency smooth; medium frequency density strong, vocal sweet; low frequency full of elastic feet."

 

[Cbdb2]

Now to add to the confusion. As already mentioned, an xlr can be ballanced or not, a 1/4 inch can be ballanced ( tip, ring, sleeve) or not, but an rca is always unballanced. A pro interface is almost always xlr. But there is also a ( 10db?) level difference between a pro interface and a commercial one so connecting a pro out to a comercial in ( all xlr) will overdrive it. Some pro gear even has a +4dbu/-10dbv switch on the outs. So check your manuals for your equipment levels.

 

[watchnerd]

"low frequency full of elastic feet."

what?

 

[CDMC]

After reading this, and getting off the floor, l would stay clear of fanboy. " Sound characteristics: high frequency smooth; medium frequency density strong, vocal sweet; low frequency full of elastic feet."

Yea, that elastic feet comment is really stretching the truth.

 

[watchnerd]

Yea, that elastic feet comment is really stretching the truth.

bad pun!

 

[MediumRare]

bad pun!

Nice bounce back. You’re on your toes.

 

[CDMC]

Nice bounce back. You’re on your toes.

Quit dancing around and get back on topic.

 

[Jukka]

Great stuff, thanks. Makes me feel a little better about choosing a DAC with (hopefully) properly implemented balanced output. The DX7 Pro reports balanced output impedance of 200 ohms. Is that "low"? And how high is "high" input impedance? My amp has 22k ohms input impedance.

The balanced input impedance of Brunos NC400 is 100k Ohm, which I think is quite normal for balanced. For single ended it's usually half of that, 50k Ohms. These values I most often see in specs.

 

[jeffbook]

I have a Nord 8 channel nCore amp.

The Hypex NC252 and NC122 modules have an input impedance of 47K ohms for both balanced and unbalanced inputs. The NC100HF modules has an input impedance of 10K ohms for either input. My minidsp 2x4HD has an output impedance of 560 ohms. The adjustable gain in the DSP section of the miniDSP takes cares of any relative signal loss due to the various impedance matches between the miniDSp and the Nord.

 

[Jaysz]

How about using xlr splitters from dac to monitors and sub
Would there be any signal loss or any other changes

 

[RayDunzl]

How about using xlr splitters from dac to monitors and sub
Would there be any signal loss or any other changes

Yes, but likely a minimal (possibly measureable, probably not audible) degradation.

I'd expect the improvement of getting the sub in there would outweigh any negatives.

 

[LTig]

I read this is another forum. I've also read about it in several other places over the years. Thus I would tend to choose two conductor Mogami over the Star Quad:

"In listening tests we found Mogami Quad to be inferior to W2549, particularly as the length increases. This is an expected side effect and was a bit more dramatic than we expected to hear.

Expectation bias literally raises its ugly head again.

 

[RayDunzl]

Is expecting "no difference" an acceptable expectation?

 

[LTig]

Is expecting "no difference" an acceptable expectation?

No.