Long interconnects or speaker cable?

[amt]

Hi there,

Im thinking of doing an active speaker thing, really just putting my amps and Marchand XOs next to my speakers, and want to know what to do about a 7-8 foot interconnect. Id like to do it DIY, like maybe a Kimber style Cat5 or something. What are my options? Is it sonically better or worse than keeping the electronics together and running longer speaker wires? BTW, I will be using an Adcom GFP-750 to drive them.

thanks amt

 

[OliverD]

From my experience, it is definitely better to have long interconnects than having long speaker cables. I strongly recommend balanced interconnects. They don't need to be expensive. BTW, my system is an active two-way setup (using AMT-1 for mids and upwards)

 

[Da5id4Vz]

9451

Definitely balanced interconnects. I've always been happy with Belden 9451 with Nuetrik XLR, but I'm sure in the world of audiophilia this is a rather pedestrian solution.

 

[piertr]

Re: Long interconnects or loudspeaker cable?

This may or may not be helpful:

http://www.empiricalaudio.com/frAudio_FAQ's.html

 

[amt]

Good information link. So it appears that balanced interconnects are the best bet for longer runs and that what everyone suggests. My Adcom has balanced outputs but the XO boards are not balanced as far as I can see so am I out of luck or is there a way to modify/adapt them?

amt

 

[purplepeople]

If they would make all drivers dual voice coil, then speaker cable could also be balanced. Unless... does someone know a way to do that?

🙂ensen.

 

[B.VDBOS]

If they would make all drivers dual voice coil, then speaker cable could also be balanced. Unless... does someone know a way to do that?

The output of a bridged amp is balanced ,but at output power levels it wouldn't matter much

Billy

 

[runebivrin]

So far, all speakers I've seen lack a ground connection, so I'd say they're automatically balanced. At least from a CMRR point of view.

Rune

 

[OliverD]

But then, what kind of EMI would you need to produce audible artifacts by radiating into the speaker cable (terminated by 8 Ohms on one side and almost zero on the other)?

 

[runebivrin]

Beats me. Maybe if you split a zip cord and run the halves on either side of the room, you could pick up interference from your mains wiring or a badly shielded power transformer. But then we're quickly leaving the realm of balanced connections anyway.

Rune

 

[Jennice]

Is EMI a real concern to these cables?

Hi everyone,

Interesting thread you got going, even though it's on the verge of falling off the deep end. 😀

I have a lot to say about this (my experiences and oppinions), but instead of repeating myself, why don't you get my notes on the subject (amongst others?)
You can find them at http://www.sylvest.dyndns.dk:8080
and enter the user name "diyaudio" and password "booklet".
Select the folder called DIY audio, and get the PDF file.

Comments (of all sorts) welcome - annoyingly few (2 to be precise ) have ever followed my request and given me feedback on it. I know that a LOT more than 2 have downloaded it, so please do take your time one of these days...

Jennice

 

[runebivrin]

I've just had a chance to scan it ever so quickly, but I'd say it's refreshingly heretic. But then, I've got wooden ears and an engineers heart.🙂

Rune

 

[mrfeedback]

Balancing Mods.......

Good information link. So it appears that balanced interconnects are the best bet for longer runs and that what everyone suggests. My Adcom has balanced outputs but the XO boards are not balanced as far as I can see so am I out of luck or is there a way to modify/adapt them?

amt

Which Marchand model do you have ?.
The ones I have seen have provision for balanced output.
Diagrams are available on the Marchand site.

Eric.

 

[phase_accurate]

So far, all speakers I've seen lack a ground connection, so I'd say they're automatically balanced. At least from a CMRR point of view.

Just as an info: Some Tannoy models can be grounded (I. e. the drivers' baskets and other metal parts that don't carry any signal). But I would still regard them as symmetrical.


Regards

Charles

 

[purplepeople]

Balanced interconnects are not prone to the interference that can intrude on a normal cable run. While the signal to "noise" ratio is much higher for speakers cable, I'm suggesting that running a 3-wire balanced cable will also create the same resistance to external signals. The push-pull amps would then only need a common and speakers with DVCs would be instant conversions. Now if there is a way to turn your normal single ended output into a balanced output and then reconvert it back for a typical speaker, then we can also enjoy the benefits of a balanced run to the speakers.

🙂ensen.

 

[Jocko Homo]

Go balanced......but......

Most gear is not properly designed for good CMRR. My guess is that AdBomb would rank up there as one that is incorrect.

The guys at Jensen Transformer have some good info on how to do it right. Shouldn't be hard to find on their site. Recommended reading.

Jocko

 

[Da5id4Vz]

A great site:

http://www.jensen-transformers.com/apps_sc.html

Also good:

http://www.benchmarkmedia.com/
- Select application notes from the tool bar.

and both work well with my workhorse 9451 cable

and I've only skimmed the article, and have no hands on expereice with it, but here is Uwe Beis/Rod Elliot's aproach:

http://sound.westhost.com/balance.htm

 

[Steve Eddy]

Balanced interconnects are not prone to the interference that can intrude on a normal cable run. While the signal to "noise" ratio is much higher for speakers cable, I'm suggesting that running a 3-wire balanced cable will also create the same resistance to external signals. The push-pull amps would then only need a common and speakers with DVCs would be instant conversions. Now if there is a way to turn your normal single ended output into a balanced output and then reconvert it back for a typical speaker, then we can also enjoy the benefits of a balanced run to the speakers.

What's a "balanced" interconnect? A simple twisted pair is "balanced." And its susceptibility to interference is the same whether it's used in a balanced interface or an unbalanced interface. What matters in terms of common-mode rejection is what's attached at each end.

And why would you need a dual voice coil driver? A driver's voice coil is fundamentally no different than the primary winding of a transformer and will float depending on the source. So as long as your source is balanced, you'll get the best common-mode noise rejection.

What you would need however, unless you're using a single driver with no intervening crossover network, is a symmetrical crossover. Most crossovers are asymmetrical and will unbalance the line and degrade common-mode rejection. Though I suspect that like a transformer, the voice coil will have a considerably high common-mode input impedance and should still give very good common-mode rejection even if driven from a wholly unbalanced source.

se

 

[amt]

Im lost...

Im not sure what the answer is. Heres the setup Im going with:

Adcom Gfp750 (btw I bought it specifically because its a Pass design so I hope its engineered properly) will be feeding two Marchand XM1s, one per channel, which in turn will be integrated with a stereo GC amp and power a twoway speaker. The XM1 boards only have +/- input and hi/low/ground output.

I assume that if the cable between the pre and xo is balanced, that between the xo and amp (in one chassis) shouldnt matter, unless it has to be balanced on both input and output and the amp.

I really only have a basic grasp of this so dont worry about a remedial approach.

amt

 

[Da5id4Vz]

I'm not sure that I understand the rational of trying to achieve mili-volts common mode noise reduction on a speaker level interconnect that is swinging a hundred volts or more.

On line level interconnects they make a lot of sense, the noise floor is very close to where induced noise will start to appear. This is why balanced interconnects are typically referenced 14 dB hotter than unbalanced signals. Cleanly raise the peak output levels, and you can ignore the bottom 14 dB of hash while maintaining or extending the SNR. Getting less than 3 dB of attenuation over 1000 feet of 9451 is also a nice plus.