Another "Wiring" Demonstration

[D-MAN]

Shawn's point about the problems still existing is not lost on me.

So now I'm thinking about getting some nice wood tractrix horns, ala Martinelli - after all, why try to weasel around cheap horns by throwing more money at them?!

As I noted before in prior posts, the driver is only as good as the horn its in.

Here is the BMS driver response as published by the manufacturer.

DM

[DrWho]

" And you can get a rather good measuring setup for under $100 too....$50 behringer mic and $40 mixing console (for the mic preamp and phantom power). Get yourself some free RTA and waterfall software on the net and you're home free..." And if you measure FR electrically (what is feeding the driver basically) instead of acoustically it can be even cheaper. Literally could be free if he has a decent sound card in his computer. You tap into the feed to the driver and put it into the computer. Then using the freeware out there you could measure it. If you are working with higher voltages (about about 2v) then you would need a couple of bucks worth of resistors to make a circuit for measuring. Shawn

I've always wondered how well such a measuring method corresponds to the anechoic response....

How hard is this to do? Would you be willing to walk me through it? Does such measuring take into account the acoustic impedance the driver is seeing? I can imagine this would be a real powerful way to see how the horn influences the output of the driver, and likewise the effect of the room too. You could even plot all 3 at once with some rather inexpensive ETF software and get some really cool information too....I definetly need to play around with this.

[D-MAN]

You need a full-duplex soundcard.

DM

[DrWho]

Ya, I've got an Audigy2 ZS...

[sfogg]

"I've always wondered how well such a measuring method corresponds to the anechoic response...."

Not at all. You are measuring the FR of the amp/crossover if you do it

electrically. You won't see the acoustic FR of the driver/horn in this

method. If an interaction between the drivers impedance/crossover/amp

is electrically altering the FR you *will* see that.

"How hard is this to do? Would you be willing to walk me through it? "

Very easy really if the signal level voltage is low. You just basically

wire to the sound cards line level input in parallel with the driver.

The input impedance of the sound card is orders of magnitudes higher

then the driver so very very little actual power will flow to it and it

will have a negligable loading on the circuit. Most sound cards can

only take a volt or two into them so you can only do low powered

testing this way. If you need to test high voltage levels you can build

a buffer/voltage divider (some resistors) to drop the level down for

the sound card. Do a search on a 'Wallin Jig' for what you need in that

case.

Dana,

"You need a full-duplex soundcard."

Not necessarily. If you are just doing a simple FR measurement you can

just use pink noise on a CD as the source and use a half duplex

soundcard for the measurements. ETF has the option of using a test CD

for full MLS measurements too.

Shawn

[formica]

You are measuring the FR of the amp/crossover if you do it

electrically. You won't see the acoustic FR of the driver/horn in this

method. If an interaction between the drivers impedance/crossover/amp

is electrically altering the FR you *will* see that.

Mike, in other words... we aren't really concerned with the actual FR

improvement of this tweek, but rather if there actually is any difference. If there was one, Dana's EQ'ing wire should alter the electrical FR.

"You need a full-duplex soundcard."

Not necessarily.

True it's not necessary, but very few soundcards are not "full

duplex". The only one's I've come across in a while were in

ultra-light notebook computers (which were duplex, but mono in and

stereo out).

D-Man's just looking for a way out of doing the measurements, IMHO... []

ROb

[Dan1045]

Please tell which model of the behringer microphone you are talking about. I am thinking of doing this.

thanks.

Dan

[D-MAN]

D-Man's just looking for a way out of doing the measurements, IMHO... []

ROb

Great! Now Rob's busting my b****, too. Thanks, guys.

DM

[DrWho]

http://www.partsexpress.com/pe/pshowdetl.cfm?&DID=7&Partnumber=248-625&scqty=1

[efzauner]

JJ

Sorry to go on, but you may be mixing up characteristic impedance with simple inductance-capacitance-resistance. For short distances it is only the L-C-R that is important.

Characteristic impedance is Zo=squareroot of(L/C) were L and C are the inductance and capacitance per unit lenght. If you know L and C per unit lenght you can calculate Zo but for any electrical purpose that is a fraction of a wavelength, all you have to do is use the actual L and C value, depending on the length of wire you have. The Characteristic impedance just becomes a bulk L and C attached to the amp/speaker connection.

for more details please see: http://sound.westhost.com/cable-z.htm

[JJkizak]

My main problem is understanding how much of a change in resistance, capacitance, inductance of the wires causes a change in the response of the system. Is it minute quantities or more substantial. I would like to see this in charts or graphs as to are we splitting hairs or logs. This would be a kind of benchmark for all speaker systems. In other words how long the wire is, diameter, insulation type and thickness and even the metal composition of the wires, connectors and screws, what effect on the amp and crosover circuits. Does it make a difference? Why doesn't Klipsch have anything to say about these subtle anamolies? Or does it affect the sound by .00000000000001%? If I throw an audio 20 - 20khz sweep on the amp and measure the ouput of the speakers with a corrected mic and throw that on a scope what do I get? What happens when the wires are changed? You should see it on the scope if it's a good one.

theoretically it should be flat oooops, not so with K-horns, well then what do we really want, to match the factory spec or make are own that is perfectly flat? Many variables crop up out of the woodwork. Even the test setup which would I assume would be different for each man who sets it up. And then at different levels. Way too many parameters here. Pure science bumping heads with ears.

JJK

[efzauner]

JJ

the material is out there. You just need the time to find it and read it. My previous link to westcost sound has excellent theory and measurements on cables and how they affect sound. There is also a good section on electronic crossovers and how, by eliminating the passive crossover, you dramatically improve damping factor and save amp headroom. I really like the West coast sound site. Audioholics also has a really good section on cabling.

I am thinking of delving deeper into this. I have an HP impedance analyazer, and other test gear. I just recently simulating a Hafler/Rochford Fosgate power amp in software and am going to build on that to add cabling and cross over effects. Right now I have it simulating clipping and I can see the amount of harmonics generated. I am also simulating intermodulation distortion etc. My next step will be to simulate TIM: Transient intermodulation distortion. This is a long project as my job and 3 kids drains most of my time. The plus side is that playing with the simulator as a hobby helps my day job and vice versa.

Alk: I can simulate x overs with C, L, ESR vs frequency easily with table date, no need to create polynomials. You might have difficulty with that in spice because frequency dependant resistor are non causal in time domain (for non tech types: non-causal basically means that the signal comes out before the signal goes in!) Let me know if you would like me to try something. However, I think we agree that it is diminishing returns.

PS: for those into car audio, the Rockford Fosgate amps are all nearly identical to their Hafler pro stable mates. Very impressive!. I got a schematic from Rockford Fosgate and compared it to schematics of pro gear that is included in the pro owners manuals. All basically the same. I am simulating 2 amps: a regular classAB amp and a Class G subwoofer version of it.

One last note on Characteristic Impedance Zo: The formula Zo=sqrt(L/C) is derived with calculus solving for the current and voltage pulse traveling down a transmission line with L and C chopped up into infinitesimally small sections. The result, Zo is propagating wave version of ohms law for an infinitely long transmission. It defines what the amplitude of a current pulse traveling down the line is as a result of a voltage pulse traveling down the line. If the transmission line is a small fraction of a wavelegth then all you need is the L and C (and R of course) times the short length of cable. At audio frequencies and the distances involved in home installations, you do not have to consider traveling waves. If you measure the voltage along a speaker cable, it is the same voltage everywere.

You seem to have a background in communications. You may recall college level courses in introductory high frequency where you measured the VSWR in a slotted line by moving a probe along its length. At RF/microwave frequencies the voltage along a coax cable is indeed different depending on where you measure!

Boy, I like this thread, I is forcing me to open my textbooks and study. Trying to explain/teach something is the best way of learning it yourself!

Enjoy the music,

[JJkizak]

efzauner:

I was just hoping that someone smarter and more motivated than I would experimemt with this stuff so I could just buy it and listen to the results. You know, add a little magic device like a crossover network to enhance to the max everything in the path. When I installed Dean's crossovers the affect was substantial. They cleaned everything up, balanced everything better, and improved the Cornwall better than new. I assume a lot of this was because of new components and newer design methods. I could also drive them a bit louder than before. So if silver wires or coat hangers make it work better, why not? The electronic history of the coat hanger is prolific, especially with Muntz tv's.

JJK