Which is the optimal length of speaker wires?

We use the Tune Method to evaluate performance

Moderator: Staff

User avatar
mrco99
Very active member
Very active member
Posts: 723
Joined: 2009-12-10 17:14
Location: The Netherlands
Contact:

Post by mrco99 »

Charlie1 wrote:If I post mine back, will you please solder back on 30cm. Tx ;)

Seriously, that's great information. Must have taken you ages.
And some buckets of soldering silver maybe?
I imagine you´d start with a certain ´excessive´ cable length and then cut down and resolder until you get to point where the signal quality starts to decrease instead of increasing?

Alternatively the whole process of optimalisation will require some serious reels of cable until finding the sweetspot in length.

Respect Fredrik, great work.
Charlie1
Very active member
Very active member
Posts: 4831
Joined: 2007-12-11 00:30
Location: UK

Post by Charlie1 »

rowlandhills wrote:Unusual for you to order 2.18m cables originally, Charlie!
There's nothing wrong with my maths skills, - well, not that my 5-year old can detect
User avatar
lejonklou
Administrator
Administrator
Posts: 6524
Joined: 2007-01-30 10:38
Location: Sweden
Contact:

Post by lejonklou »

Ok, here comes a long and thorough summary. If you don't want to read it, the important stuff is this:

* K200/400/600 sounds fantastic at a length of 2.5 m.
* Do not go shorter than 2.4 m, as your system will sound restless.
* Longer than 2.5 m is ok, but you will gradually loose performance.

This is what I did when determining the 2.48 m length:

The speaker cable was always Linn K400, often but not always cut in half (so that it becomes two pairs of "K200"). If you pick any other cable, the optimal length is likely to be different. I used both newly purchased rolls of K400 and some older runs (3-5 years?) I had lying around. They all sounded the same. The length of the cable was always measured from beginning of copper to end of copper, before any termination took place.

I mostly terminated them with Linn Knekt bananas, but on some occasions with Deltron 550. Linn Knekt sound better, but the optimal length of cable didn't seem to be affected (I say "seem to", because I am not 100% certain of this). In all of the final comparisons between 2.4 and 2.8 m, I used only Linn Knekt. They were soldered with my own 60/40 solder wire at 353°C with my Weller WSD80 soldering station and WSP80 pen. The temperature is valid only for this particular station, pen and exact soldering technique. For example, my other Weller WD1 station differs by 7°C.

Worth knowing: Linn Knekt bananas have changed, perhaps a year ago. The bend of the four aluminium "flaps" that determines the contact pressure is different. The old version had a mild bend right in the middle of the flaps. The new one has a sharper bend, closer to the end of the flaps. The reason for the change could be an attempt to make it last longer - they have been known to loose their springiness pretty quickly, and for that reason many shops don't use them.

The new version of Knekt fits considerably harder than the old! As contact pressure is an important parameter for sound quality, this change affects performance. All connectors sound best when they fit "just right", which is tight but not so tight that it becomes difficult removing it, or the materials making loud cracking noises when their surfaces move against each other. If the contact pressure is too high, the music becomes hard and distorted. If the pressure is too low, the sound becomes sloppy and uncontrolled. While Knekts appear to have very little variation in size, banana holes can vary quite a bit (and on older Linn speakers such as Kan, Sara, Isobarik etc, Knekts don't fit at all as the holes are smaller). So when you connect your Knekt, notice how hard it grips into the hole. If it feels very tight, unplug and replug it many times, until it looses some of it's resistance and fits more easily.

When evaluating, I always did at least two pairs of cables at the same time, in the exact same way, but with different lengths. It's not possible to compare cables which are of different age, unless they are all more than 4 months old and have stabilised in performance. I started at 3.9 m and went as short as 0.5 m. Pretty quickly I concluded that the optimal length was somewhere in between 2 and 3 m.

I used Linn Majik, Linn Akurate, Tundra and Tundra Mono power amps. I used Linn Majik 109, Akurate 212, 242 and Klångedang T1 loudspeakers, both single and multi-wired. On a few occasions, I also used aktiv 242's (no passive filter involved). All of these units appeared to perform best with the same length of loudspeaker cable, which leads me to believe that the length I found optimal is a property of K400 itself.

In addition to any component's "quality in itself" there is however also a degree of interaction with the environment in which it operates. This became evident with the Klångedang T1's, which have their passive filters in separate boxes outside the loudspeaker cabinets. Now, the optimal length of the single wire speaker cable required between the power amplifier and the T1 filter boxes is as short as possible - not 2.48 m. I use 0.4 m, but recommend 0.5 m as it's more practical. But between the filter boxes and the T1 loudspeakers, where a bi wire cable is required, the optimal length was 2.48 m.

Passive loudspeakers do perform better single wired than bi- or multi wired. The closer I came to the optimal speaker wire length, the more obvious this became. Almost always the tweeter + and - inputs should be used.

Why exactly 248 cm? Ok, I confess: The very last cm is my guess. I had arrived at 2.40 m versus 2.50 m. Now, just like when using a precision torque tool on a fastener somewhere in your HiFi system, I find that the closer one gets to the optimal cable length, the more pronounced the differences get. They also become easier to understand and appreciate, as if one is approaching a peak where things just seem to snap into place. Therefore I actually find it easier to compare 2.40 m with 2.50 m than to compare 1 with 4 m.

2.40 is very upfront, agile, willing to attack and almost a bit tiring to listen to. Anything shorter than 2.40 gets tiring and restless, with an emphasis on the attack of each note. Avoid! Too long is much less of a problem - what happens then is that the music gradually looses its dynamics, in particular the attack of each note, and becomes gentle and cozy. Not bad, but less precise and thrilling.

2.50 takes one half step back from 2.40, finds a good groove and pulls you in. It's less in your face and slightly more forgiving. Yes, it's better than 2.40! But perhaps a tiny bit too laid back and easy going? I made more than ten comparisons back and forth between 2.40 and 2.50 with different types of music before I decided that 2.50 sounds like it might be a little too long. Half way in between the two - 2.45 m - didn't seem right to me. I made a pair at 2.48 m and it is the best I have heard so far.

I have not yet tried 2.46, 2.47 or 2.49. The problem is that once you are on that peak of performance, small random variations begin to affect the comparisons. So I'm not even dead sure that 2.48 is better than 2.50, it could have been those particular soldered joints that turned out a tad better. To get from dm to cm in precision, we need to compare more pairs than two.

EDIT: Added Akurate 212 to the list of loudspeakers
Last edited by lejonklou on 2013-11-08 23:13, edited 1 time in total.
User avatar
lejonklou
Administrator
Administrator
Posts: 6524
Joined: 2007-01-30 10:38
Location: Sweden
Contact:

Post by lejonklou »

I made this post a sticky. That does not mean that my K400 length of 2.48 m is the final answer. It means that I want my findings questioned, challenged and proven wrong.

I would also be happy to learn about the optimal length of other cables than K400. And wouldn't it be fantastic if there was an even better cable around the corner? Lord knows I've tried and will keep trying.

Objections to making this subject a sticky? Think other threads should become sticky? Please pm me and I'll bring it up with the moderator!
ChrBea
Active member
Active member
Posts: 113
Joined: 2012-02-22 15:15
Location: Luxembourg

Post by ChrBea »

Many thanks for that detailed report. Well, I'm glad that I have the chance to listen to those 248 cm cables at home. I will let you know how they compare to my current K200 cables. But as I made some huge changes to my system, I will switch items one by one to be able to make some comparison.
Charlie1
Very active member
Very active member
Posts: 4831
Joined: 2007-12-11 00:30
Location: UK

Post by Charlie1 »

Yes, thanks Fredrik!

Mine are 2.2m as mentioned previously. I certainly recognise the fast attacking sound compared to my previous 6m long cables, although you didn't solder those and they used bananas instead of knekt. I have a feeling that the edgier presentation is less of an issue and perhaps helpful at very low volume, which is how I mostly listen. However, I'd rather have more musicality and optimal length, but it was just an observation that sprung to mind.

One question. If you had a non-AV, music only room, do you think you would prefer your stereo sat between your speakers and benefit from optimal speaker length, or would you prefer the system off to one side and benefit from better imaging? I think I know the answer to this one, but worth checking...
User avatar
ThomasOK
Very active member
Very active member
Posts: 4358
Joined: 2007-02-02 18:41
Location: United States
Contact:

Post by ThomasOK »

Fredrik, thanks for the detailed reply (although from the length of it you might have thought it was one of mine) and even more for all the work that went into it. I just wish I had known it before I cut the 2.4m lengths to make a new set of speaker cables a couple of weeks ago - so it goes! At the time I cut them I noticed that they were very close to an 8 foot length. Just calculated 8 feet and it equals 243.84 cm. Maybe it just wants to be cut in Imperial! (Although, in actuality it would have to be 1.5 inches m longer than 8 feet.)

By the way, a while back I rewired the external crossover boards for my rear Isobariks with K400 and Knekt bananas. While they were a bit tight they do work OK with the last generation Isobariks as well as with the penultimate version.
fatjulio
Active member
Active member
Posts: 159
Joined: 2012-01-13 23:22

Post by fatjulio »

Very interesting information. A couple thoughts came to mind.

What about built in active amp speakers? (350a, Akubarik) Their cable lengths will be less than 1 m. Do you think the their sound has been tuned with the on board amp setup, or passive?

I have an AV setup, so the gear is off to the side. Consequently I have 5m cables for the front L/C/R. Would a better compromise be to shift the power amps to within 2.5m, and have longer interconnects from the preamp? This would be an unbalanced (non XLR) connection length of 2-4m.
User avatar
lejonklou
Administrator
Administrator
Posts: 6524
Joined: 2007-01-30 10:38
Location: Sweden
Contact:

Post by lejonklou »

ChrBea wrote:I'm glad that I have the chance to listen to those 248 cm cables at home.
Yes... The best sounding length was sent to ChrBea. And here I sit with a pile of other lengths. If anyone wants to ignore my advice and get some shorter-than-2.4 m K400, I have a pile of mixed lengths that will be sold cheap. :)
Charlie1 wrote:Mine are 2.2m as mentioned previously.
Charlie1 wrote:I have a feeling that the edgier presentation is less of an issue and perhaps helpful at very low volume, which is how I mostly listen.
Sorry, but no, it isn't helpful. I did some of the comparisons late at night on low volume and the results were the same: Too short sounds stiff and less skilled. Closer to optimal length is more groovy, the musicians appear more skilled, it sounds more impressively dynamic in a "correct" way etc.

Should I post you a pair of 2.4 m just for trial? I think I have three K200 pairs around that length.
Charlie1 wrote:One question. If you had a non-AV, music only room, do you think you would prefer your stereo sat between your speakers and benefit from optimal speaker length, or would you prefer the system off to one side and benefit from better imaging? I think I know the answer to this one, but worth checking...
It might depend on how much stuff there is between the speakers in the optimal speaker length case, and some other factors too. One example: At home I have my system on the left side in the room and have compared
1. Having two Monos in the same Harmoni rack as the rest of the system and >4 m of speaker cable, versus
2. Having the Monos on a low and wide IKEA table in between the speakers, with 4.5 m black interconnect going to the Monos and optimal length of speaker wires going out. Option 2 is better! But in another room and system, other factors might be at play and the result different.
anthony
Very active member
Very active member
Posts: 786
Joined: 2007-02-04 22:39
Location: UK

k400

Post by anthony »

My cables are 5 metres, and had tried 1.5m lengths. I much preferred the 5 m, consequently I dismissed the short speaker cable length being better. Will try some 2.48m although cat5/6/7 is the way forwards, a whole new can of worms.
User avatar
lejonklou
Administrator
Administrator
Posts: 6524
Joined: 2007-01-30 10:38
Location: Sweden
Contact:

Post by lejonklou »

ThomasOK wrote:Just calculated 8 feet and it equals 243.84 cm
Doesn't sound too bad! 2.4 m was my best length for most of this year. Beats 2.2 and 2.8 hands down.
ThomasOK wrote:By the way, a while back I rewired the external crossover boards for my rear Isobariks with K400 and Knekt bananas. While they were a bit tight they do work OK with the last generation Isobariks as well as with the penultimate version.
Really? This surprises me, because I tried Knekts on an early pair of Kans - those that had two metal holes straight into the cabinet, no colours, no nuts, just metal banana jacks - and they completely flattened the Knekts that were forced in. Switched to Deltron 550's, which Linn used for everything back then, and they fitted much better.
In any case, pay attention to contact pressure, as it has an important effect on sound quality.
fatjulio wrote:What about built in active amp speakers? (350a, Akubarik) Their cable lengths will be less than 1 m. Do you think the their sound has been tuned with the on board amp setup, or passive?
I think that in aktiv speakers with the amps inside the speakers, they simply ignore the optimal loudspeaker cable length. There are many other things to optimise and problems to solve.
fatjulio wrote:I have an AV setup, so the gear is off to the side. Consequently I have 5m cables for the front L/C/R. Would a better compromise be to shift the power amps to within 2.5m, and have longer interconnects from the preamp? This would be an unbalanced (non XLR) connection length of 2-4m.
In the example of my living room that I mentioned above, the answer is yes, clearly better with longer interconnects and 2.5 m cables! There might be something with your room or system that could affect the outcome, but I would certainly recommend you try it.
hcl
Very active member
Very active member
Posts: 360
Joined: 2008-01-13 11:03
Location: Göteborg
Contact:

Post by hcl »

lejonklou wrote:
fatjulio wrote:What about built in active amp speakers? (350a, Akubarik) Their cable lengths will be less than 1 m. Do you think the their sound has been tuned with the on board amp setup, or passive?
I think that in aktiv speakers with the amps inside the speakers, they simply ignore the optimal loudspeaker cable length.
Irrespective of the other difficulties of integrating filters and power-amps in active speakers, there surely are other (even better e.g. more controlled) ways of optimising the connection between the power amps and drivers in an integrated active speaker.
Charlie1
Very active member
Very active member
Posts: 4831
Joined: 2007-12-11 00:30
Location: UK

Post by Charlie1 »

lejonklou wrote:Should I post you a pair of 2.4 m just for trial? I think I have three K200 pairs around that length.
Maybe, any 2.5m left over :)
User avatar
lejonklou
Administrator
Administrator
Posts: 6524
Joined: 2007-01-30 10:38
Location: Sweden
Contact:

Post by lejonklou »

hcl wrote:
lejonklou wrote:
fatjulio wrote:What about built in active amp speakers? (350a, Akubarik) Their cable lengths will be less than 1 m. Do you think the their sound has been tuned with the on board amp setup, or passive?
I think that in aktiv speakers with the amps inside the speakers, they simply ignore the optimal loudspeaker cable length.
Irrespective of the other difficulties of integrating filters and power-amps in active speakers, there surely are other (even better e.g. more controlled) ways of optimising the connection between the power amps and drivers in an integrated active speaker.
Can you please explain what you mean, hcl? I don't understand.
User avatar
lejonklou
Administrator
Administrator
Posts: 6524
Joined: 2007-01-30 10:38
Location: Sweden
Contact:

Post by lejonklou »

Charlie1 wrote:
lejonklou wrote:Should I post you a pair of 2.4 m just for trial? I think I have three K200 pairs around that length.
Maybe, any 2.5m left over :)
Sorry, but no. For some reason they're all gone.

Seriously, I can trade your 2.2 for a 2.4 m pair. It would be interesting to hear what you think of the difference between them.
hcl
Very active member
Very active member
Posts: 360
Joined: 2008-01-13 11:03
Location: Göteborg
Contact:

Post by hcl »

lejonklou wrote:
hcl wrote:
lejonklou wrote: I think that in aktiv speakers with the amps inside the speakers, they simply ignore the optimal loudspeaker cable length.
Irrespective of the other difficulties of integrating filters and power-amps in active speakers, there surely are other (even better e.g. more controlled) ways of optimising the connection between the power amps and drivers in an integrated active speaker.
Can you please explain what you mean, hcl? I don't understand.
For starters, it is not likely that the optimum connection between the amp and driver is the same for all pairs of amps and drivers. Second, it is perfectly possible to achieve the same (if not identical) electrical properties from a completely different and much more compact circuit. I would guess that You could use a different cable length and optimise the output of your amps for that specific cable length.

In addition, Your research regarding speaker cable lengths could well provide a valuable input for Linn (and others) if they (for example) sets out to optimise the connections in the active speakers.
Charlie1
Very active member
Very active member
Posts: 4831
Joined: 2007-12-11 00:30
Location: UK

Post by Charlie1 »

lejonklou wrote:
Charlie1 wrote:
lejonklou wrote:Should I post you a pair of 2.4 m just for trial? I think I have three K200 pairs around that length.
Maybe, any 2.5m left over :)
Sorry, but no. For some reason they're all gone.

Seriously, I can trade your 2.2 for a 2.4 m pair. It would be interesting to hear what you think of the difference between them.
Great idea, I'll contact you privately
User avatar
lejonklou
Administrator
Administrator
Posts: 6524
Joined: 2007-01-30 10:38
Location: Sweden
Contact:

Post by lejonklou »

hcl wrote:For starters, it is not likely that the optimum connection between the amp and driver is the same for all pairs of amps and drivers.
Why is that unlikely? Could you give an example of when a different connection was more optimal?
hcl wrote:Second, it is perfectly possible to achieve the same (if not identical) electrical properties from a completely different and much more compact circuit. I would guess that You could use a different cable length and optimise the output of your amps for that specific cable length.
You are assuming that it's the amplifier that wants the particular load that the optimal cable length provides. I don't think that is the case, as it doesn't matter which amplifier I have used in the comparisons.

I have tried optimising amplifier outputs for different cables and drive units. It didn't work. The best solution was the best regardless of what the amp was driving. I am thinking that the amplifier's quality in itself has been more important than the relatively small changes in it's environment that the cable and drive unit provides.

In addition, I have a bunch of cables in my lab with identical electrical properties to K200. This is easy to achieve, just make an almost identical cable. Do they sound the same? No.
User avatar
Music Lover
Very active member
Very active member
Posts: 1673
Joined: 2007-01-31 20:35
Location: In front of Lejonklou/JBL/Ofil

Post by Music Lover »

Sorry to partly interrupt this interesting discussion with a question; did you ever test with other termination than bananas?
Did you end up with same cable length?

As you know I have Speakon and like to get your thoughts regarding performance and cable lenght (same optimum length if I would guess).
Also do you think it's possible to later when I change amps, replace the Speakon with bananas without ruin the cable performance?


Back to topic, Linn has clearly stated they are not interested in optimizing small items as cables. (length and direction). Pity
It's all about musical understanding!
anthony
Very active member
Very active member
Posts: 786
Joined: 2007-02-04 22:39
Location: UK

Post by anthony »

Music Lover wrote:Sorry to partly interrupt this interesting discussion with a question; did you ever test with other termination than bananas?
Did you end up with same cable length?

As you know I have Speakon and like to get your thoughts regarding performance and cable lenght (same optimum length if I would guess).
Also do you think it's possible to later when I change amps, replace the Speakon with bananas without ruin the cable performance?


Back to topic, Linn has clearly stated they are not interested in optimizing small items as cables. (length and direction). Pity
I have changed to neutrik to bananas twice, obviously it was a different twin but banana always sounded better.
hcl
Very active member
Very active member
Posts: 360
Joined: 2008-01-13 11:03
Location: Göteborg
Contact:

Post by hcl »

lejonklou wrote:
hcl wrote:For starters, it is not likely that the optimum connection between the amp and driver is the same for all pairs of amps and drivers.
Why is that unlikely? Could you give an example of when a different connection was more optimal?
hcl wrote:Second, it is perfectly possible to achieve the same (if not identical) electrical properties from a completely different and much more compact circuit. I would guess that You could use a different cable length and optimise the output of your amps for that specific cable length.
You are assuming that it's the amplifier that wants the particular load that the optimal cable length provides. I don't think that is the case, as it doesn't matter which amplifier I have used in the comparisons.

I have tried optimising amplifier outputs for different cables and drive units. It didn't work. The best solution was the best regardless of what the amp was driving. I am thinking that the amplifier's quality in itself has been more important than the relatively small changes in it's environment that the cable and drive unit provides.

In addition, I have a bunch of cables in my lab with identical electrical properties to K200. This is easy to achieve, just make an almost identical cable. Do they sound the same? No.
What I was trying to convey is that the validity of such, empirically verified, findings may well seem generally applicable. On the basis of electro magnetic field theory (although I am no expert in this field) I would think such results only are valid in a limited set of cases (a limited set of amplifier designs, canle types and maybe for a limited type pf speaker types). As such, it would be possible to change both cables and amp-design, ending up with another, equally valid optimum cable length.

If considering the wide band electro-magnetic field properties of the complete design I would think it to be possible to make a different kind of optimisation. This approach could be used to take full advantage of an integrated design as the new integrated active Linn speakers.
User avatar
lejonklou
Administrator
Administrator
Posts: 6524
Joined: 2007-01-30 10:38
Location: Sweden
Contact:

Post by lejonklou »

Music Lover wrote:did you ever test with other termination than bananas?
Did you end up with same cable length?
I have only used bananas in this test. But the bananas have been two types and their female counterparts in amplifier and speaker end three types. In addition, I have played with contact pressure. I don't exactly KNOW that the optimal cable length was unaffected by these change of parts and pressure, but I have a feeling they are separate qualities. With "have a feeling" I mean that the changes in sound are of different characters and it doesn't sound like they would affect each other.

If we assume that the choice of connector DOES affect the optimal cable length, my guess is that it's very little. Perhaps the last centimetres or so? Should be easy to check by making two K200 with Speakon. At 2.60 and 2.50 m perhaps? Then if 2.50 m wins cut both down to 2.55 and 2.45? Etc.

Regarding changing connectors: The way I solder K200, they need to be cut down a minimum of 2 cm when changing connectors in one end. So if I make a 2.50 m pair with Speakon, you can later change them into a 2.48 m with bananas.
usjaz
Member
Member
Posts: 19
Joined: 2012-05-11 07:13

Post by usjaz »

lejonklou wrote: I used Linn Majik, Linn Akurate, Tundra and Tundra Mono power amps. I used Linn Majik 109, Akurate 212, 242 and Klångedang T1 loudspeakers, both single and multi-wired. On a few occasions, I also used aktiv 242's (no passive filter involved). All of these units appeared to perform best with the same length of loudspeaker cable, which leads me to believe that the length I found optimal is a property of K400 itself.

In addition to any component's "quality in itself" there is however also a degree of interaction with the environment in which it operates.
Linn has always suggested multi-wire & shortest speaker cables to a passive speake, so amps have more controls of the drivers units. I noticed the same thing described here. Shorter the cable, more unrest the music appears.

I have each solo/d located next to 350P, quad-wired (w/ 1.1m K200s) at the moment. Too bad, Lejonklou, you haven't tested on solos. Akurate/majik/Tundras are not nearly as powerful as solo/d. Also solos are Class AB design right?

Wonder if the result will be different in the case of a more powerful class AB solos/d.
Klimax RDS -> Kisto -> Klimax Solo/D -> 350P & REL B1
Oppo 103 -> M6100/D -> M140, M112, 2xC102
User avatar
lejonklou
Administrator
Administrator
Posts: 6524
Joined: 2007-01-30 10:38
Location: Sweden
Contact:

Post by lejonklou »

usjaz wrote:Too bad, Lejonklou, you haven't tested on solos. Akurate/majik/Tundras are not nearly as powerful as solo/d.
I am 99% certain that the result will be the same on Solo. So far every test has indicated that the optimal cable length has nothing to do with power or type of design.
User avatar
Music Lover
Very active member
Very active member
Posts: 1673
Joined: 2007-01-31 20:35
Location: In front of Lejonklou/JBL/Ofil

Post by Music Lover »

hcl wrote:
If considering the wide band electro-magnetic field properties of the complete design I would think it to be possible to make a different kind of optimisation. This approach could be used to take full advantage of an integrated design as the new integrated active Linn speakers.
We all know the limitations having cables, PSU's, source components, Xovers, power amps and speakers close together. And to make matters worse, in a small vibrating enclosure.
Basically the main optimization would be reducing the number of short cables/sockets by using an integrated design of the mother board as in Akubarik.
BUT, this going to affect the performance in other ways!

It's going to be rather difficult to manage the various electromagnetic fields and it's impact on other objects.
If you have any ideas, I'm eager hearing them.
It's all about musical understanding!
Post Reply