As I mentioned in my first post here (#15), I hear the effect of bi-wiring as changes in the phase relationship of the drivers, affecting instrument images and overall stage. I'm not sure if looking at response in a sim will offer much insight. B&W's explanation (reducing crossover interaction) makes perfect sense to me.
Not everyone hears the difference, and among those who do, not everyone welcomes it. But as with many things, once you hear it it's hard to unhear.
I'd agree. Of course these three things are actually a single issue. Source resistance creates crossover interaction which changes the response and the phase. However fixing the response will also fix the phase, or to put it another way if you return to the original response you'll return to the original phase, regardless of your wiring scheme. This may involve you breaking in to the crossover to compensate. It also raises the question of which way B&W intended it to be, and was it previously incorrect?
I had a friend who was obsessive about making his speaker wires the same length. I told him there is no perceptible difference by human beings or even dogs, and that he wouldn't hear a difference unless he had relativistic hearing. Also the cable coiled up underneath the bookshelf had a much larger effect on the sound.
My biggest problem is my wife moving the speakers around all the time.
From a schematic viewpoint the circuit is identical whether there are two cables or one. It's just that they either meet at the speaker or they meet at the amplifier. The only difference is a halving or doubling of cable effects. Some effects can be good some bad, but for me they are inaudible especially in my house of cooks and animals. Would it matter if the crossover was outside of the loudspeaker?
There has to be an element of 'me too' in this. If your competitors are doing it and you're not, and the golden ears are insisting they can hear a difference, you'd be mad not to include extra binding posts and a justification in your blurb. Especially on your 'audiophile' product. Even top end speaker manufacturers have marketing departments.
If someone can show me the double blind test results that clearly demonstrate a difference I'll be interested in measurements that may explain the observed effect.
It takes courage to go against the consensus and tell the king he has no clothes. And, yes, I'm aware that cuts both ways 🙂
Here's a test with and without a bi-wired speaker. Measuring mic at 1m distance, no DSP.
Absolutely nothing changed in the room or the setup, only the wiring.
First wire was connected to the bass speakers, second to mids and tweeter.
There is quite some difference in the 250Hz to 1.5kHz range.
I leave it to the seasoned speaker builders to draw a conclusion. 🙂
Speaker specs:
Absolutely nothing changed in the room or the setup, only the wiring.
First wire was connected to the bass speakers, second to mids and tweeter.
There is quite some difference in the 250Hz to 1.5kHz range.
I leave it to the seasoned speaker builders to draw a conclusion. 🙂
Speaker specs:
There's no doubt one would hear a difference with that change. I would guess that voices will be further back, meaning depth of soundstage is changed. Why response changes is more interesting, amp not stable? All I can say, I never heard a difference with bi-wiring, NAD, Rotel and Yamaha amps involved...
Btw, measuring can be tricky. I have seen this kind of differences, back-back measuring just seconds apart...
Btw, measuring can be tricky. I have seen this kind of differences, back-back measuring just seconds apart...
Thank you for actually performing an experiment! Interesting that the bi-wired speaker has up to 2db less output in the midrange.
To narrow down what is going on here can you run two cables to the speaker but not bi-wired? I'd like to know if the difference is caused by adding more copper or if there is something going on in the amplifier.
Edit: Can you measure your cables? That might help us too.
To narrow down what is going on here can you run two cables to the speaker but not bi-wired? I'd like to know if the difference is caused by adding more copper or if there is something going on in the amplifier.
Edit: Can you measure your cables? That might help us too.
I should add that the amp is a JLH and maximum measured RMS output was 124mV, most likely at the 42Hz peak.
Measurement was done with a true RMS Fluke 87 IV in record mode (Min, Max, AVG)
I will perform these measurements tomorrow, Arthur.
Measurement was done with a true RMS Fluke 87 IV in record mode (Min, Max, AVG)
I will perform these measurements tomorrow, Arthur.
Interesting. It's 2-3db between 450-700hz. I guess some might be able to pick it out in ideal conditions, but reliably? The shift in the peaks is interesting too.
Thanks @Netlist for taking the time to do the tests.
I look forward to seeing the additional results.
@picowallspeaker BD249C 🙂
I tried measuring the cables after nulling the Ohm meter but it reads 0.00 Ohm.
In the kOhm range I get 0.0001 Ohm
I tried measuring the cables after nulling the Ohm meter but it reads 0.00 Ohm.
In the kOhm range I get 0.0001 Ohm
The cables that connect to the bass speakers are about 3m and the ones to the mids and tweeters about 5m I had lying around.
Both are from the same no-name brand, 4mm² with very thin stranded wires.
I can not measure the resistance.
Both are from the same no-name brand, 4mm² with very thin stranded wires.
I can not measure the resistance.
With lengths like that, wouldn't be surprised if you could get some substantial improvement with star-quad. Even 1m of cable is enough that it starts to have some effect.
Questions:
1) were all the speaker cables equal length? He must compare only equal lenght, otherwise each cable is picking up different rf based on length.
2) were speaker cables open at the end or terminated with 4-8 ohm resistance? Because speaker wires from amplifier are terminated by speakers, typically 4-8 ohms. Not open. Open is irrelevant. No one listens to open wires.
With real speakers sometimes it can be beneficial to put a 100R non-inductive resistor across the speaker terminals. Why? Because the speaker impedance may be very high at RF frequencies. 100R is about the right termination for twisted pair or parallel conductor transmission line. What the resistor will do is absorb RF energy picked up by the cables when they act as antennas. Otherwise its possible for RF energy to build up on on the cables and affect the amplifier output. Sometimes the load resistors can help and other times not. Just one of those things easy to determine by a quick experiment. IIRC, the original non-inductive resistors used for this were made by Caddock. Maybe like this: https://www.digikey.com/en/products/detail/caddock-electronics-inc./MP930-100-1%/1284421
Last edited by a moderator:
Measurements taken with cables swapped.
First graph is the shortest length to the mids and tweeter, on the second one I reversed the cables, the shortest length to the woofers.
You can see that the different cable lengths have a very insignificant effect.
Here are the graphs with one and two cables. 'More copper' as Arthur suggested.
There's a minor difference, however, very likely inaudible.
- Status
- Not open for further replies.