buzzy said:
On this page:
Shouldn't this read higher instead of lower? Both the rankings and the tests (R,L,C) seem to suggest that higher is better. For example, compare the Monster cable to the top line.
The main reason that the factors for resistance, Inductance, and capacitance read from a high of ten down to a low of one is that I originally used an exponential curve that I set up for another project and wanted to use it on this project. The use of an exponential curve sort of created a greater seperation at the high (good) end and bunched things up at the low end. This method was developed in excel and was based on weighing various engineering design principles that I worry about when designing a ships propulsion plant (you know, gas turbines, diesel engines, gears, propellers, bearing, and shafts etc....). Anyway Gene, wanted to use a simpler linear method of generating each parameters rank factor. Gene won. But I kept the 10=high=good, 1=low=bad concept.
The reason the reason the "Technical Rank" and "Final Rank" are the opposite; 1=high=good, 10=low=bad, is human nature. People equate #1 with being #1. Plus is was a simple ranking procedure performed in excel. Plus it was easier (for me at least) to see that one particular cable was #1 out of 20 and another cable was #17 out of 20.
j_garcia said:
So do the measured variances correlate to actual performance differences? Am I just missing this aspect of the comparison?
That is something I am still working on.
Mathmaticaly, you can compare two cables and come up with a number. But to do this you need to know your speakers impedance at all frequencies, you need to know your amps output impedance at all frequencies. When connected, you would have a circuit that looks like Figure 1. Mathmatically, you can treat the cables capacitance and speaker impedance as a seperate parallel impedance and the amps output resistance along with the cables resistance and inductance as a seperate series circuit as shown in Figure 3. This makes it easier to solve.
My next phase is to actually compare and measure the entire circuit, amp/cable/speaker to see how they actually correlate to a mathmatically determined dB delta. To do this you need to have your speaker impedance plot and know the amps output resistance at specific frequencies (I'll be using an Alesis RA150 amp for everything).
I have the calibrated measurement mic (Earthworks M30), I have the sillyscopes (Instek GDS-840C 2 channel and Tektronix 2246 4 channel), I have the other equipment necessary (I hope) including a Metex 9170 multimeter/waveform generator, Instek LCR821 LCR meter, Instek GOM-802 milliohm meter, Fluke 189 handheld digital multimeter. Now I just have to measure several cables that have distinctly different resistances, inductances, anc capacitances at the octave center frequencies (i.e. 16, 20, 100, 1000, 2500, 5K, 7.5K, 10K, & 20K Hz) to see if I can detect any voltage deltas significant enough to be heard.
The test set up will be as follows:
1) CD player to amp using generic interconnects playing test tones of above mentioned frequencies.
2) Aamp gain/volume knob set to output 2 volts at the amps output terminals measured with a Metex 9170 multimeter.
3) Voltage signal measured at the speaker input terminals using an Instek GDS-840C Oscilloscope.
4) Mic positioned one meter from reference speaker (In my case, a Sapphire SB because I have the impedance curves for it). The mics output goes into Channel two of the sillyscope.
5) Subtrack one signal from the other using the oscope
6) Convert voltage difference to dB and that would be the drop across that particular cable.
7) Repeat using a Tektronix 2246 Oscillscope to see if there are anomalies.
8) Repeat again using a Metex 9170 Frequency generator instead of a CD of test tones to compare more complex wave forms (more complex than a sine wave, but not as complex as music)
If I find that I need a mic preamp I'll try and run everything through a TASCAM mixer I have.
I hope I've answered all of you questions. But Just wait, I've got something that'll blow your socks off. But It will take a month or so until I'm ready to make it public.