Another "Wiring" Demonstration

[D-MAN]

Had a buddy of mine who is a professional musician and recording fanatic come over and give my system a listen.

He agreed that the BMS midrange driver is a little hot, and since I can't attenuate them with any more granularity than the ESN xformer is capable of, I mentioned that I intended to use wiring to make the fine-tuning adjustments, which he questioned the validity of (kind of like some of you)...

Being that the crossovers are exposed and are easy to get to, I then proceeded to use a single piece of cheasy 12 AWG wire from Lowes about 3 inches long and jumped the wire braid that Al K. uses for bi-wiring between the woofer and midrange/tweeter "section", which in the circuit, simply shortens the "run" of flat braid cable by about 4 inches. Played a tune, and then took the jumper wires off.

He imediately and definitely noticed a change in the response, jumped up and went to look at the xover and asked whether I had talked with the xover dude about changing wires, etc. My response was that the xover dude was an engineer and would not listen to me (ahem!) about wires making any difference! Manufacturing something and going whole-hog are two different things, besides, so I can't hold it against Al for that.

So, as some of you in-the-know already know, wires (even very short lengths) has a profound effect on a sensitive speaker, and can be used to "fine-tune" the response to a degree.

By the way, as a further experiment, I swapped out all of the internal wiring between the xover and the drivers with some leftover (GASP!) Monster 14 ga. speaker wire that I had sitting around, and although I REALLY hate Monster, it improved the response compared to the cheap stuff that I was previously using (bulk 12 ga. speaker wire from Lowes). The 12 gauge copper (one lead)and silver-colored (the other lead) cheap bulk wire from Lowes produces an "edgy" high-end and attenuates the lows. The Monster 14 ga. is a more well-balanced effect complete with the lows that I had previously been missing somewhat. So if you want to accentuate a treble bandpass, or tilt it toward the high-end, the cheap Lowes stuff is a good bet. It presents a clean and clear soundstage without alot of grain, but like I said, it accentuates the highs and attenuates the lows.

Changing internal wires alters the response to the degree that one can explore the attenuation settings that previously were deemed unsuitable - that is a ramification of careful wire selection. The prospects for achieving the desired presentation is more attainable for relatively small expense. The alternative (which I personally find distasteful) is an active EQ or perhaps (and I am thinking about this one) - an active crossover with attenutators, perhaps a bit more expensive for a good one than AL K.'s ESNs...

I have noticed that ALL of the internal wires should be the same type, size and brand of wire all the way around, that is, between the crossover and woofer, crossover and midrange driver, and tweeter, etc. Otherwise, there seems to be a noticable difference in the "timing" and overall coherence of the soundstage which really throws off the "believability" of the presentation, if you know what I mean. Those of you who have tried this know what I mean.

If you really need convincing, swap out the midrange driver wires on one side only and listen to some familiar music. It will sound "strange" - that is an indication of the unbalancing effect that unequal wires have on the soundstage - it's not quite right, its a perturbation, and it's NOTICABLE!

Just ordered some bulk pure silver wire to experiment further with the interal wiring, maybe the crossover, too, as I know the effect that it made to the response when I went to silver interconnects.

DM

[tofu]

i just went from 10 year old 14 guage to brand new 12 guage last night.

with all that effort getting it to stay in the terminals, you'd think it'd be nice enough to atleast sound different []

[johnyholiday]

A hand wired point to point gee-tar amp will cost ya up the kazoo,cause

it just sounds better ,Johny'Holiday knows that ,a Vox AC 30 TBX hand

wired P2P will cost ~$2k more.TBX=top boost

[steamer]

I have heard similar results in wire or differences I guess

To me silver gets toooo edgy with SS but works great with tube power(takes a little warmth away)resulting in a clearer presentation.

Copper is good with SS but adds to much warmth with tube power

Tonight I changed my Bluejeans speaker wire(6 in total per side)with some older Audioquest Indigo(same length/litz design/better shielding)and there was a noticable difference for the better IMO,the bottom line was there was a difference!

I have no idea what is going on,maybe interaction between components,but yes I have always found wire can make a difference.

Greg

[Erik Mandaville]

"He agreed that the BMS midrange driver is a little hot, and since I can't attenuate them with any more..."

This is a good example of where an variable L-pad would be convenient.

Erik

[JJkizak]

In my opinion you should be using the proper nomenclature involved here which would be the "characteristic impedance" of the wire. Since the impedance of the speaker drivers is all over the place depending on frequency and the crossover performance the wire is the third variable. In VHF/UHF applications this is reduced almost to "0" by using abnormally high input signals reduced down to operating parameters with a 20 db resistive pad. The pad in effect matching the three different impedances---antenna, wire, tv set. Now if the impedance of the speakers remained constant and the crossovers then it would be a piece of cake. Do you think that PWK had the formulas for this problem? They did publish the impedance variations verses frequency of the K-horns as I recall. And it was real squiggly. So what I am saying is that if the crossover itself had a pure 20 db resistance between the three drivers, itself, the wire, and the amplifier and the amplifier power was increased 20 db your wire changing would be a non-changing medium as the impedance would be constant for all things.

JJK

[tigerwoodKhorns]

In an old Edgar interview he stated that swapping Sidreal Caps (now Auricaps) and solid core copper wire made a big difference in his horns.

Then again, in an article based on some testing done by Mcintosh and other, they said that there is no real difference as long as the wire is of sufficient guage.

I would not have a problem buying more expensive wire (to a point) if it made an audible and predictable improvement. I just don't want to spend a fortune on wire and guessing. Between tube rolling and wire you can spenda fortune and not get an improvement. It seems so much easier to use heavy guage copper wire (cheap stuff) and a Behringer DEQ2496 (yea, a cheap piece of crap, I know...) to make attenuation adjustments instead of swapping wire.

Dana,

I am using an active crossover and I can make huge adjustemtns on the fly for the different drivers. This seems like an ideal setup for someone who is always building speakers and swapping drivers and such out. For about $1,000 or so you can get a digital crossover, a beefy SS pro amp for woofers, and a few teac's for the mids and highs. You can dial in the speakers, then using the settings that you attain, calculate a passive crossover that you need for the setup.

Just a thought,

Chris

[homemade]

What about upgrading the small gauge wire used in the coil? I'm guessing there is something like 50-100ft of 18-20 gauge wiring in there?

[PrestonTom]

What about upgrading the small gauge wire used in the coil? I'm guessing there is something like 50-100ft of 18-20 gauge wiring in there?

Homemade: you make an excellent point. Would it need to be cyrogenically treated also?

Good Luck,

-Tom

[Def Leper]

Sure am glad I don't have that disease. Sounds contagious. And expensive.

[DrWho]

Let me get this straight - "timing differences" between different wires

sounded bad until "corrected"? Different diameter drivers are going to

have different group delays for the same frequency so by making

everything perfectly aligned up stream, you are effectively forcing the

drivers to be misaligned - even if the two drivers are equidistant from

the listener. Secondly, you have huge path length differences between

the different drivers which will even compounding this issue. Also, the

acoustic origin for the same driver (especially in horns) changes with

frequency. Heck, even passive crossovers have their own group delays.

These are all factors that are working on the order of milliseconds. I

don't think it's even measurable the time differences that different

wire size and compositions can make - especially at those distances.

It sounds to me like something wasn't connected properly before you

made the switch. I would love to see you verify your conclusions with a

double blind test...in fact, if you can show that you can hear a

difference then there is a guy waiting to pay you a million dollars.

[formica]

I am using an active crossover and I can

make huge adjustemtns on the fly for the different drivers. This

seems like an ideal setup for someone who is always building speakers

and swapping drivers and such out.

Very good point Chris... I was wondering myself why DMan was still

playing around with passive x-overs. Going active also allows for

time alignement... which can be factor when working with folded

horns. Afterall, mixing and matching DIY horns requires some

experimentation....

ROb

[D-MAN]

My original idea is that passives would be more appealing at the consumer level, i.e., look what comes in your Khorns.

I AM leaning in the direction put forth by Chris, an active, for testing and design at least.

I would go with silver even if I was using an active, too! I've wanted to for a long time, and it's not that expensive. Cheaper than buying a new set of midrange horns!

DM

[efzauner]

JJKIZAK

I presume you are joking here right? Maybe I am naive and think that this is a serious post. Just in case it is: characteristic impedance is only usefull at high frequencies where the cabling is a about a wavelength or longer and we care about distributed impedances. At audio frequencies it does not matter, everthing can be considered a lumped element. Your statement "In VHF/UHF applications this is reduced almost to "0" by using abnormally high input signals reduced down to operating parameters with a 20 db resistive pad" makes no sense. What is reduced to zero? In RF applications it is important to have things impedance matched, that is why there are converters to convert 300 ohm twinlead to 75 ohm coax etc. Pads attenuate the signal, but also can have impedance converters or matching networks included. Since pads do attenuate the signals, they will also attenuate any reflections that are caused by impedance missmatches. Again, all irrelevant at audio frequencies.

[DrWho]

The physics and signal theory for VHF/UHF is the same for audio signals

too - just that they're operating at different frequencies which puts

the emphasis of the circuit into different variables. But impedance

"matching" is every bit as important with audio as it is with any other

electrical circuit - it's just a ton harder with audio because a

speaker presents a time and frequency variant impedance (so most amplifiers try to "iron-fist" their way through it).

[efzauner]

DrWho. Yes, I do agree that designing with impedance in mind is important, but in Audio, you do not care about the characteristic impedance of the wire that connects the source to the load. You care about its resistance, but not characteristic impedance. They are not the same. In RF, the cables are a good portion of a wavelength or more and therefore the voltage along the line changes with the phase of the wave. But at audio, the wavelegth is so long that the time difference and therefore voltage difference along a speaker cable is so small to be essentially zero. Thus we can consider a speaker wire a lumped LRC network. Tube amps typically have out put transformers to match a high impedance tube output stage to a low impedance speaker. This is not needed with transistors because their internal impedance is so low already. Also at RF, generating power and amplifing is very expensive thus there is a huge effort to impedance match power devices to loads so as to have maximum power transfer, that is, to have a conjugate match. Also power reflected back into an RF transistor because of a badly matched load can blow output transistors etc. At audio we can think of things in the same terms: A short causes all the power available from the output transistors to be reflected back with a 180deg phase shift so that the positive outgoing wave is cancelled by the reflected, inverted incomming wave to give a zero volt sum. A good writup on RF load matching and other stuff that is relevant at RF but not at audio can be seen at http://www.microwaves101.com/encyclopedia/maxpower.cfm There are lots of other good writeups on impedance matching in interconnects such as preamp to amp, 600 ohm balanced vs unbalanced etc. I am sure there have been many such links before on this forum, and on audioholics etc.

[DrWho]

Well you are obviously more knowledgeable on the subject, but that is

what I was getting at....the emphasis of the circuit shifts to

different areas.

[efzauner]

DrWho. One of the problems with all the snake oil is that they often quote or refer to perfectly valid electromagnetic theory, but avoid saying that at low frequencies these phenomenon are insignificant. Things like skin effect come to mind. Its like saying: Einstein's theory of relativity states that time slows down as your velocity increases up to the speed of light, therefore, you should only record CDs at the slower record speeds because recording at higher speeds like 48x slows down the waves and results in audible frequency shifts. True on the Einstein thing, garbage on the rest. A big part of an engineer's job is to shift thru the theory and decide what is relevant, what do I need to take into account. So, Electromagnetic theory defines transmission line theory, but is not important at low frequencies and short distances. However it is very significant for power companies because 60cycle power lines are very long transmission lines.

The big problem is that unless one has a degree in electrical technology or engineering one is not equiped to fend of the snake oil salemen. Few consumers are so equipped, thus lots of $1000 HT systems with $500 monstercable sold at big box stores. Its the little guy I feel sorry about, not the guy who buys $100K of gear with the $1000 power cord. If you have that sort of spare cash, you can spend it however you like! Have a good evening. I am on a conference call with China...Learning about Wimax

[PrestonTom]

DrWho. One of the problems with all the snake oil is that they often quote or refer to perfectly valid electromagnetic theory, but avoid saying that at low frequencies these phenomenon are insignificant. Things like skin effect come to mind. Its like saying: Einstein's theory of relativity states that time slows down as your velocity increases up to the speed of light, therefore, you should only record CDs at the slower record speeds because recording at higher speeds like 48x slows down the waves and results in audible frequency shifts. True on the Einstein thing, garbage on the rest. A big part of an engineer's job is to shift thru the theory and decide what is relevant, what do I need to take into account. So, Electromagnetic theory defines transmission line theory, but is not important at low frequencies and short distances. However it is very significant for power companies because 60cycle power lines are very long transmission lines.

The big problem is that unless one has a degree in electrical technology or engineering one is not equiped to fend of the snake oil salemen. Few consumers are so equipped, thus lots of $1000 HT systems with $500 monstercable sold at big box stores. Its the little guy I feel sorry about, not the guy who buys $100K of gear with the $1000 power cord. If you have that sort of spare cash, you can spend it however you like! Have a good evening. I am on a conference call with China...Learning about Wimax

Very well put!

I also feel sorry for the little guy.

We are in an age where even the "average Joe" can have a decent stereo system for not all that much money. In some cases, such a system uses ideas/technolgy from half a century ago (my K-Horns come to mind....). But look what happens, the snake oil salesmen come in and muddy the waters.

Good Luck,

-Tom

[JJkizak]

I guess Bell Labs/Western Electric was all wrong then by utilizing 600 ohm patchcords (all of there mainframe interconnect hookup wire had 600 ohm characteristic impedance at audio frequencies)

They also used 135 ohm patch cords for land lines (audio frequencies) and 75 ohm patch cords for Multiplex patching. (12khz to 12mghz). But I guess impedance doesn't matter at audio frequencies.

One thing that was not mentioned when a 20 db pad is inserted is the noise level is also reduced 20 db. Every industrial electronic setup on receivers that I have seen has had a 20 db pad in the front end to eliminate VSWR and cut the noise by 20 db.

I guess PWK was all wrong when he published his papers on how hard it was to maintain a constant impedance throughout the frequency range of the K-horn. He actually then pretty much said the impedance of the K-horn was a compromised figure.

JJK