[dminches]

I know this is an frequent discussion for the online forums and although there seems to be a consensus of long interconnects, I was wondering what those who have Modwright preamps and amps are doing.  And, if you don't, feel free to discuss anyway.

And, Dan, any thoughts from you?

[Daedalus Audio]

I prefer 3 meter speaker cables with 1 meter IC's to the reverse....  if that is the choice.
lou

[ted_b]

I prefer 3 meter speaker cables with 1 meter IC's to the reverse....  if that is the choice.
lou

+1

Although Dan's preamps have good output buffers to power longer IC's (balanced) they MUCH prefer shorter IC's in my opinion.  But the answers would be more relevant if you told us your setup..i.e how far from pre to amp, amp to speakers. 

[dminches]

Ted, in your room it looks like the speakers are closer to the amps than the pre-amp.  Is that correct?

[dminches]

+1

Although Dan's preamps have good output buffers to power longer IC's (balanced) they MUCH prefer shorter IC's in my opinion.  But the answers would be more relevant if you told us your setup..i.e how far from pre to amp, amp to speakers.

I am redoing my room.  Currently, my amp is on the side which requires 10-12' speakers runs and a 5 foot IC run to my LS-100.  I was thinking of putting the KWA-150 in between the speakers and then running ICs to the pre which would be on the side of the room, similar to your room.

[avahifi]

In the overall scheme of things in the audiophile universe, it is pretty well understood that power amplifiers are designed to be load driving devices.  Preamplifier and other "small signal" devices are in general not designed to drive loads as a first priority.

A long interconnect cable with its distributed capacitance along with the input capacitance of the amplifier it is driving, can be a pretty difficult load for a preamplifier.  A speaker wire is a pretty simple load for a power amplifier, especially in comparison to the really complex load the speaker crossover may represent.

Thus, my thoughts would be that if the preamp is not designed to drive hard loads, one should keep the load represented by the interconnect cable as low and simple as reasonable possible - - short, well shielded, and low capacitance interconnects.

Since the speaker wire probably is not the worse case problem for a well designed power amplifier, I doubt if it will be in trouble driving 10 to 30 feet of decent speaker wires.  Of course if the cables are highly capacitive, then all bets are off.

This idea is probably pretty east to test.  Simply have two lengths of the same kind of interconnects and speaker wires available, one short pair and one long pair of each, and do some listening tests (preferably double blind) with both combinations, and see if we can really hear any difference. Actually, if you did not mind doing the tests in mono, you could hook up one channel with short interconnects and long speaker wires, and the other with long interconnect and short speaker wires, switch the system to mono, place the speakers side by side, and do instant AB testing with your balance control.

Just an idea and suggestion.

Best regards,

Frank Van Alstine

[dminches]

Frank, you make good points and I have no basis to not agree. What confuses me is that every time I see a high end system the amps seem to be right behind the speakers, indicating the just the opposite of what you are saying.

[ted_b]

Ted, in your room it looks like the speakers are closer to the amps than the pre-amp.  Is that correct?

Yes, I have redone my room and reduced the "clutter" between my speakers, so i have elected to go longer IC's and shorter speaker cables.  But the preamp I use for this (not Dan's currently, but my 36.5 DM would work) has huge output buffers and is designed, as Frank said, to drive long IC's...as long as they are very low capacitance and balanced.  I use 6.5M of Furutech custom made Reference III IC's to Dan's wonderful KWA 150 monoblocks, then run the equally wonderful ASI Liveline speaker cables to the (wonderful ) Revelations (actually to their outboard crossover boxes).

I agree with Frank that it's "easy" to do a test.  HF rolloff would be the first culprit, as well as noisefloor.

[dminches]

FWIW, my current preamp is the LS-100 which is not a balanced set up.  Does this change the answer?  Although I could use balanced cables between and amp and pre, does it matter if the pre isn't a true balanced design?

[audioengr]

I did a few simulations to answer this a few years back when I sold only cables and here are the results.  The conclusions may be different if your cables have much different characteristics:

http://www.empiricalaudio.com/computer-audio/audio-faqs/short-versus-long-cables

Steve N.
Empirical Audio

[avahifi]

Thanks Steve for the good engineering answer to this question.

However, I suggest you try the numbers again with about 100 pF per foot of capacitance for both the speaker wires and the interconnect cables and see what results you get then.

Regards,

Frank Van Alstine

[arthurs]

What about a 32' pair of IC's and 6' pair of SC's with the following specs?

IC's
Capacitance:  170 pF/m ±15% (pin-to-shield)
Resistance: 13.2 mΩ/m (conductor)
Gauge (effective):  22 AWG

SC's
Resistance: 13.3 mΩ/m (conductor)
Gauge (effective):  9 AWG 

Preamp specs
Output impedance: < 150 ohms, 20 Hz - 20 kHz, static (i.e., not dependent upon feedback)

[avahifi]

What about your 32' long 170 pF per meter interconnects?  That comes out that you are tying a .002 microfarad capacitor across each ouptut terminal on your preamp!   

Just for fun, try this with a load bank and a square wave generator.  See what a 10KHz square wave looks like with, and without that relatively big capacitor tied across the output of your preamp (from hot to ground).

Methinks you will not like the result at all, namely overshoot and ring city.

In any event, this will give you a result, not an opinion.

Regards,

Frank Van Alstine

[audioengr]

Thanks Steve for the good engineering answer to this question.

However, I suggest you try the numbers again with about 100 pF per foot of capacitance for both the speaker wires and the interconnect cables and see what results you get then.

Regards,

Frank Van Alstine

Frank - my advice would be to get different interconnects.  Particularly for unterminated, unbalanced, this is really high.  30pf/foot is even high.  Low-pass filter into 50kohms.

Steve N.

[dmatt]

Hmmm,

My 35 foot interconnect capacitance isn't quite that large ... 20 AWG wire with 14 pF/ft (conductor to conductor) and 0.0095 ohms/ft (conductor) in balanced configuration.  Would that come out to .00049uF and 0.33 ohms per conductor? 

My speaker wires are only about 2 ft. a piece since each mono amp is behind its speaker.

I have no engineering basis for why I use long interconnects and short speaker wires.  Only that my speaker manufacturer suggested keeping the speaker wires as short as possible and my amp manufacturer suggested short/stout speaker cables and balanced interconnects for long runs, so that's where I started.

I am reading the comments from Frank and Steve with great interest.

DM

[modwright]

I believe that if you are going to run LONG IC's, then simply go balanced.

As to which is better, assuming worst case = RCA IC's, then it really depends on the cables.  As Mr. Van Alstine points out, high capacitance IC's will not produce the best sonic results.  Our balanced preamp has a Zou = 100 ohms.  Our SE preamp has a Zout = 300 ohm.  Both with well-buffered output stages.  Assuming your cables are good quality, I would expect the difference sonically to be relatively small.

I believe the reason that you often see long IC's and short speaker cables in high-end systems, is because the amps are LARGE, often monoblocks, and often on amp stands, by the speakers.  It is more convenient to have the rack on a sidewall or closer to the listening position.  In this case the system is also often balanced.  If balanced, then the length of the IC's really isn't an issue.

In the end, we can do the math, based on measured capacitance of cables, etc., but it really comes down to your layout, physical constraints, system and how it sounds best to you!

If you use a well-buffered preamp with low output impedance (Zout), the effects of cable length will be negligible.

Thanks,

Dan

[mikel51]

I haven't acted on this, but in his book Get Better Sound, Jim Smith suggests that one of the common mistakes that audiophiles make in their sound room setup is to put all of their equipment between the speakers.  He says that the best thing you can do is to move all of your gear off to the side and stop it from interfering with imaging.  He goes on to say that you would be better off with cheap long cables and moving your gear, than expensive short cables.   To implement this, you would need some long cables, either speaker or interconnects.  I find this an intrigueing suggestion, but am not sure when I will have the energy to move it all (TV and audio gear) to check it out. 

[arthurs]

So a 32' run of balanced (XLR) presents less or different issues than a 32' run of unbalanced (rca)?  This would run from my pre to amp which both have fully balanced option.

[ted_b]

So a 32' run of balanced (XLR) presents less or different issues than a 32' run of unbalanced (rca)?  This would run from my pre to amp which both have fully balanced option.

Art, I'd make darn sure the cables are first and foremost very low capacitance (i.e. 15 pf/ft) and then having them balanced will insure low or no noise.

[arthurs]

Thanks Ted, any idea who manufactures cables with capacitance that low?

Hope you're well!