Long Speaker Cables vs. Long Interconnects
Hi MIT,
I posted this question in another group. A couple of members of the site gave me their opinions (thanks guys!), but I'd really like a cable company to tell me what they think.
I'm just going to copy and paste my original post from over in the other group. Can you tell me what you suggest?
"If I have to choose between long interconnects and long speaker cables, which should I choose? For example, I currently have my system rack on the same side of the room as my speakers. I have a small stereo amp that is on the bottom of the rack. In this configuration, I have one meter interconnects throughout the whole rig and 12 foot speaker cables. But what if I wanted to put my rack in a more convenient location? Say I wanted to put it close to the listening position so I don't have to get up to change CD's. What would be the better option? Should I keep the amp in my rack and get 30 foot speaker cables? Or should I keep my 12 footers, put the amp between my speakers (in which case I really would only need 8 foot speaker cables) and get a long pair of interconnects?"
Thanks in advance.
~Rick
Discussion started by Rick Trace , on 1142 days ago
Rick,
There is some debate on this issue throughout the industry. It seems that there is a balancing act between performance and convenience that one must consider. For our part, we always side with performance. Therefore, the answer to your question is that you should always use interconnects that are as short as possible. When faced with the choice of whether to use long speaker cables or long interconnects, always choose long speaker cables.
Why? Well, it is actually because of the nature of the components that the cables are interfacing. Audiophile amplifiers are always “wide open.” There is a lot of energy flowing through them, so they are less likely to be affected by field coupled noise. Interconnects however, because the components they interface are only sending minuscule amounts of energy, tend to act like antennas: the longer the length, the bigger the antenna. Noise that they pick up is then passed through your components and amplified by your amplifier. This “muddies up” the signal and the end result is less contrast between voices and instruments, which affects the dimensionality of the soundstage and overall quality of the playback.
There are ways to counteract the problems associated with long interconnects. MIT’s networked interfaces include a small network that is designed to capture and dissipate noise on the circuit. So, we have seen systems with long MIT interconnects that sound quite nice. But we also know that those systems would perform better if they shortened the interconnects and lengthened the speaker cables. :)
MIT Cables
There is some debate on this issue throughout the industry. It seems that there is a balancing act between performance and convenience that one must consider. For our part, we always side with performance. Therefore, the answer to your question is that you should always use interconnects that are as short as possible. When faced with the choice of whether to use long speaker cables or long interconnects, always choose long speaker cables.
Why? Well, it is actually because of the nature of the components that the cables are interfacing. Audiophile amplifiers are always “wide open.” There is a lot of energy flowing through them, so they are less likely to be affected by field coupled noise. Interconnects however, because the components they interface are only sending minuscule amounts of energy, tend to act like antennas: the longer the length, the bigger the antenna. Noise that they pick up is then passed through your components and amplified by your amplifier. This “muddies up” the signal and the end result is less contrast between voices and instruments, which affects the dimensionality of the soundstage and overall quality of the playback.
There are ways to counteract the problems associated with long interconnects. MIT’s networked interfaces include a small network that is designed to capture and dissipate noise on the circuit. So, we have seen systems with long MIT interconnects that sound quite nice. But we also know that those systems would perform better if they shortened the interconnects and lengthened the speaker cables. :)
MIT Cables
Good points Bruce, but would running balanced cables (in a true balanced system - that being the amplifier and preamp are actually balanced designs) make any difference? Have heard mixed opinions on this... like anything in audio, ask 10 people and get 10 different answers.
Dear Dave,
It is true that a balanced cable is a better option than a single-ended cable for long interconnect runs because a balanced cable is less susceptible to external noise. But it is not altogether unsusceptible. So now we should ask ourselves, “Is a speaker cable less susceptible to external noise than a balanced interconnect?” The answer is yes.
Again, we at MIT Cables always side with performance over convenience (though we understand that each audiophile must do their own balancing act). So, we stand by our original answer: It is always better to use long speaker cables instead of long interconnects.
MIT Cables
It is true that a balanced cable is a better option than a single-ended cable for long interconnect runs because a balanced cable is less susceptible to external noise. But it is not altogether unsusceptible. So now we should ask ourselves, “Is a speaker cable less susceptible to external noise than a balanced interconnect?” The answer is yes.
Again, we at MIT Cables always side with performance over convenience (though we understand that each audiophile must do their own balancing act). So, we stand by our original answer: It is always better to use long speaker cables instead of long interconnects.
MIT Cables
Room architecture and layout also play a part we sometimes cannot avoid. My own room is a good case in point - because of its architecture and layout, the said dilemma doesn't exist for me, as my audio is stacked up 2 feet to the left of where I sit, and the speakers are some10 feet away. Hence, I have to use short interconnects and long speaker cables, in my case, 2 x 6 m (app. 20 ft) van den Hul 352 Hybrid (a proprietary mix of carbon fibre and silver plated OFC copper).
What is often forgotten in such instances is that as the distance between the amp and the speakers increases, you need thicker and thicker cabling - if you want all those Amperes of current properly transferred. Mine has a thickness of 5.5 mm (no idea of what gauge that is) and uses 2x256 strands. Rough calculations show that is sufficient for impulse currents of more than 150 Amperes, which should be enough even for me. My wife's and son's systems are good with half that length, so they have van den Hul CS 122 cables, which are exactly the same as mine, but are only 3.5 mm thick and use 2 x 128 strands.
Also, since speaker cables appear as series resistance, if they are flimsy, they will play funnies with your damping factor. This is not too much of an issue with transistor amps, but can produce VERY unpredictable results with tube amps.
While there is no definitive answer, think of it this way: interconnects carry signals in the milivolt range, with currents well below 1 mA, while speaker cables carry signals well into the volt range (1 Watt into 8 Ohms is 2.83V rms) and with currents into the Ampere range in transients. Knowing this, which would appear to be more sensitive?
What is often forgotten in such instances is that as the distance between the amp and the speakers increases, you need thicker and thicker cabling - if you want all those Amperes of current properly transferred. Mine has a thickness of 5.5 mm (no idea of what gauge that is) and uses 2x256 strands. Rough calculations show that is sufficient for impulse currents of more than 150 Amperes, which should be enough even for me. My wife's and son's systems are good with half that length, so they have van den Hul CS 122 cables, which are exactly the same as mine, but are only 3.5 mm thick and use 2 x 128 strands.
Also, since speaker cables appear as series resistance, if they are flimsy, they will play funnies with your damping factor. This is not too much of an issue with transistor amps, but can produce VERY unpredictable results with tube amps.
While there is no definitive answer, think of it this way: interconnects carry signals in the milivolt range, with currents well below 1 mA, while speaker cables carry signals well into the volt range (1 Watt into 8 Ohms is 2.83V rms) and with currents into the Ampere range in transients. Knowing this, which would appear to be more sensitive?
Of course I guess the question could be at what length does it really make a difference? I am running 3-meter balanced to my amps and 4 foot speaker cables from the amps to the speakers. Only because I want to place the amps to the outside of the speakers as opposed to the inside where I would probably only need 2-meters. I do get noise/hum on occasion, but have found that it is more an issue of the cable being used (make) as opposed to the length. Something is in the wall behind my system that causes an issue on occasion and in looking as to what I am perplexed as there appears to be nothing there electrically.
Let me turn this around since this is an experiment I've always wanted to conduct. Does a 1 meter IC from manufacturer A sound the same as a 3- 5 meter pair from the same manufacturer? Obviously this IC is being used at a line level. And obviously this may have something to do with the input impedance of the say the preamp or the amp, drive of the preamp, etc, etc.
@Myles
Myles, the best cable is the one not there at all. The prime postulate of electronics has always been to make any and all connections as short as possible. Hence, the shorter the IC the better. I have a pair I made myself using van den Hul wire that is just 0.5 m long, no need for longer as the DAC sits on top of the preamp.
To answer your question, with everything else the same, the shorter cable run should always sound better than a longer one. However, will this difference be actually audible, even if it could possibly be measured, is a moot point, because there are many other variables to consider. Which signal is it carrying, the one from a turntable, possibly in the microvolt range, or the one from a DAC or CD, well into full Volt range (Red Book 0 dBVU is specified as 2.0 V)? The smaller the signal voltage, the more it is susceptible to cable properties corruption. Next come the source device's output impedance and the target device's input impedance, and so fort. Which means that there could be a difference, ranging from inaudible to very much audible, depending on actual conditions. I have heard such differences, but admittedly, they were not great, more in the nuance range, unless the signal differences were also great.
But by and large, the shorter the cable, no matter how good it may be on its own, the better. Cable manufacturers are in the rut here, because they have to cater for a very wide audience, with wildly differing requirements, so they have to be very careful. Some go for universal values, typically 0.7 ... 1 m cable lengths, others offer varying lengths, etc.
Myles, the best cable is the one not there at all. The prime postulate of electronics has always been to make any and all connections as short as possible. Hence, the shorter the IC the better. I have a pair I made myself using van den Hul wire that is just 0.5 m long, no need for longer as the DAC sits on top of the preamp.
To answer your question, with everything else the same, the shorter cable run should always sound better than a longer one. However, will this difference be actually audible, even if it could possibly be measured, is a moot point, because there are many other variables to consider. Which signal is it carrying, the one from a turntable, possibly in the microvolt range, or the one from a DAC or CD, well into full Volt range (Red Book 0 dBVU is specified as 2.0 V)? The smaller the signal voltage, the more it is susceptible to cable properties corruption. Next come the source device's output impedance and the target device's input impedance, and so fort. Which means that there could be a difference, ranging from inaudible to very much audible, depending on actual conditions. I have heard such differences, but admittedly, they were not great, more in the nuance range, unless the signal differences were also great.
But by and large, the shorter the cable, no matter how good it may be on its own, the better. Cable manufacturers are in the rut here, because they have to cater for a very wide audience, with wildly differing requirements, so they have to be very careful. Some go for universal values, typically 0.7 ... 1 m cable lengths, others offer varying lengths, etc.
@Dave Clark
Exactly so! But you have to factor in the output impedance of the source and the input impedance of the traget device as well. It's hardly the same thing if the output impedance is say 600 Ohms and in some Japanese devices, 120 Ohms as specified by IEC recommendations, or say 47 Ohms or less as found in many High End line devices. The target's input impedance is worse yet, as it varies from relatively benign 20kOhms or so to 120kOhms as found in Electrocompaniet products. The capacitance component of the cables will act as a low pass filter in conjuction with the input impedance, so you will naturally have a 6:1 ratio in the above example. Assuming merely 3 pF of total cable capacitance, the cutoff point of -3 dB will vary from about 2.5 MHz with 20k to about 420 kHz with 120k Ohms.
Now, here's the paradox: by 400 kHz, any power amplifier still linear causes a phase shift at 20 kHz of around 2 degrees, WAY below the threshold of hearing (taken as 5 degrees by convention, but it does vary individually), so this may at first glance appear to be even beneficial. However, if the 3 pF is related to the cable itself, you also have to factor in the stray capacitance on both sides, the inevitable contact point capacitance. When added, this may well be quite a bit more, which is why it is highly advisable to use cotton buds and medicinal alcohol to regularly clean up your contact points on both sides, male and female.
Lastly, just to spite us, cables need to be viewed as a particular set of RLC variables, because all three act together one way or another. Each cable, each material, has its pros and cons, and it's their joint action that defines a cable. Hence the differences, some of which can be heard, and some which cannot, but all in a single system, i.e. in conjuction with system parameters.
As I understand it, the whole point of MIT's work was to reduce the effects of such interactions. But it would be better for MIT to explain that.
Exactly so! But you have to factor in the output impedance of the source and the input impedance of the traget device as well. It's hardly the same thing if the output impedance is say 600 Ohms and in some Japanese devices, 120 Ohms as specified by IEC recommendations, or say 47 Ohms or less as found in many High End line devices. The target's input impedance is worse yet, as it varies from relatively benign 20kOhms or so to 120kOhms as found in Electrocompaniet products. The capacitance component of the cables will act as a low pass filter in conjuction with the input impedance, so you will naturally have a 6:1 ratio in the above example. Assuming merely 3 pF of total cable capacitance, the cutoff point of -3 dB will vary from about 2.5 MHz with 20k to about 420 kHz with 120k Ohms.
Now, here's the paradox: by 400 kHz, any power amplifier still linear causes a phase shift at 20 kHz of around 2 degrees, WAY below the threshold of hearing (taken as 5 degrees by convention, but it does vary individually), so this may at first glance appear to be even beneficial. However, if the 3 pF is related to the cable itself, you also have to factor in the stray capacitance on both sides, the inevitable contact point capacitance. When added, this may well be quite a bit more, which is why it is highly advisable to use cotton buds and medicinal alcohol to regularly clean up your contact points on both sides, male and female.
Lastly, just to spite us, cables need to be viewed as a particular set of RLC variables, because all three act together one way or another. Each cable, each material, has its pros and cons, and it's their joint action that defines a cable. Hence the differences, some of which can be heard, and some which cannot, but all in a single system, i.e. in conjuction with system parameters.
As I understand it, the whole point of MIT's work was to reduce the effects of such interactions. But it would be better for MIT to explain that.
The input impedance on my amplifiers is 100k Ohms (Clayton Audio M200 which are a true balanced design) and the output on the Cary SLP-05 is 400 Ohms (also said to be a true balanced design). Cables in house are Kubala-Sosna Emotions, Purist Audio Proteus Provectus, and Jade Audio. So all this would mean what to me - sorry but I am not an electrical engineer!
FYI, the input impedance on the Cary is 100k and the output on the Playback is 22 ohms.
FYI, the input impedance on the Cary is 100k and the output on the Playback is 22 ohms.
The effect of being able to account for impedance matching (MIT and Transparent) are in my system improved dynamics, lowered noise floor and improved sense of space.
@Dave Clark
What that means to you Dave is quite simple that you have absolutely no problems with signal transfer functions. A postulate on that says that if the difference between the source output impedance and target input impedance is 10:1, you have no problems with transfer functions. As you described it, you have (120,000:400) 300:1, WAY over the minimum.
As for the cables you named, I'm sorry, but I must admit I have never even heard of them. The Great Cable Game originated from the US, and has thankfully remained mostly confined to the US. Frankly, I am irritated by that game of packaging and repackaging, OEM cables are resold at typicallt 5...7 times their OEM price, and far too many claim far too much. To me, the only worthwhile attempts. beaside the actual cable materials themselves, are what Transparent and MIT have done.
I have no ties to MIT whatsoever, in any form or fashon, and I don't use their cables - but my son does. But I've been around for a long time, since 1970, and I've seen them come and go, and at 57, I still believe I can tell a rip-off from a sincere attempt at something. I've always preferred MIT to Transparent Audio for two reasons: 1) TA was always the glitzy one, never saying outwardly but always insinuationg that they were The One, and this never fails to irritate me, and 2) I've not been able to confirm any sonic advantage of TA's cables over similar MIT cables, so I see no reason why I should pay so much more for the same sound. Hence MIT for the son. I am a silver cable man myself.
And trust me on this, I AM a true freak, I'm the kind of guy who designs and manufactures headphone amplifiers which will fill a room with sound via 8 Ohm speakers with a low, low efficiency of just 82 dB/2.83V/1m. Their output impedance will yield a damping factor (the very difference between input and output impedances) of better than 200 into headphones of just 20 Ohms, 20 ... 20,000 Hz, at levels higher than those required to split your eardrums. All in pure, true class A, your choice between a bipolar, FET input MOSFET output, or all tube version. So, as an active member of the industry, it's not easy to pull wool over my eyes.
BTW, the knob on my avatar is the headphone amp. :-)))))))
What that means to you Dave is quite simple that you have absolutely no problems with signal transfer functions. A postulate on that says that if the difference between the source output impedance and target input impedance is 10:1, you have no problems with transfer functions. As you described it, you have (120,000:400) 300:1, WAY over the minimum.
As for the cables you named, I'm sorry, but I must admit I have never even heard of them. The Great Cable Game originated from the US, and has thankfully remained mostly confined to the US. Frankly, I am irritated by that game of packaging and repackaging, OEM cables are resold at typicallt 5...7 times their OEM price, and far too many claim far too much. To me, the only worthwhile attempts. beaside the actual cable materials themselves, are what Transparent and MIT have done.
I have no ties to MIT whatsoever, in any form or fashon, and I don't use their cables - but my son does. But I've been around for a long time, since 1970, and I've seen them come and go, and at 57, I still believe I can tell a rip-off from a sincere attempt at something. I've always preferred MIT to Transparent Audio for two reasons: 1) TA was always the glitzy one, never saying outwardly but always insinuationg that they were The One, and this never fails to irritate me, and 2) I've not been able to confirm any sonic advantage of TA's cables over similar MIT cables, so I see no reason why I should pay so much more for the same sound. Hence MIT for the son. I am a silver cable man myself.
And trust me on this, I AM a true freak, I'm the kind of guy who designs and manufactures headphone amplifiers which will fill a room with sound via 8 Ohm speakers with a low, low efficiency of just 82 dB/2.83V/1m. Their output impedance will yield a damping factor (the very difference between input and output impedances) of better than 200 into headphones of just 20 Ohms, 20 ... 20,000 Hz, at levels higher than those required to split your eardrums. All in pure, true class A, your choice between a bipolar, FET input MOSFET output, or all tube version. So, as an active member of the industry, it's not easy to pull wool over my eyes.
BTW, the knob on my avatar is the headphone amp. :-)))))))
@ Dave - Give Steve Holt, our National Sales Manager, a call. He can talk to you about lending our some cables for review.
You need to be a member of this group before you can participate in this discussion
