Technical and scientific discussion of amps, cables and other topics.
Return to Propeller Head Plaza
Message Sort: Post Order or Asylum Reverse Threaded
Cable Test Method
202.67.64.154 |
||
Posted on November 17, 2002 at 12:42:30 | ||
Posts: 90
Joined: August 11, 2002 |
Hello All, Differences in peoples opinions about they way a cable sounds (or doesn't) has always made me want to try to develop a reproducible test method to quantify any differences in cables. I belive I have a setup that can measure cable differences down to about -100 dB (into the tens of microvolts anyway-see below) and I would like to sound out your views on the technique and develop it further if possible. I would like to drive an interconnect with a 'suitable' signal form a 'suitable' source and measure the differences between input and output of the cable whilst connected to a 'suitable' load, using the differential amplifier plug-in (Tektronix 7A22) of my oscilloscope (Tektronix 7834). The TEK 7a22 amplifier has a bandwidth of 1 MHz and at its most sensitive is able to measure a differential signal down to a gain of 10 microvolts per division, with some residual noise (16uV or 0.1 divisions on screen), whilst differentially rejecting the signal in the output down to below 100 dB. (It has a common mode rejection of over 100 dB in the 10Hz to 100KHz range when DC coupled.) OK then, say I use a one volt p-p signal, then this signal would be removed from the output by the amplifier down to a level of atleast 10uV, which coupled with the residiual noise means that 26 uV (10+16) would be the lowest measurable difference. The noise part of this figure can be improved on considerably by reducing the bandwidth of the amplifier (choices are 1MHz, 300kHz, 100kHz, 30kHz, 10kHz, 3kHz, 1kHz). As an example at 100Khz, (my preferred choice), the 16uV noise can be reduced to about 4uV, making 14uV the lowest measurable difference, not attributable to the equipment. To get real world results I have always wanted to connect the cable to a real world source (say a CD player) and drive a real world load (a preamp or amp) and thereby cause the cable to be subject to real world component and source impedances. BUT, this does not make for a test method which is easily reproducible by others. So then I am faced with selection of a test source, say a signal generator, appropriate signal, appropriate artificial load etc.. My thoughts so far have been to try a triangular wave (as suggested in a similar Cardas test method), or a square wave or pulse signal and use a load impedance which simulates a 100kOhm source (some resistor/inductor/capacitor combination). Any further thoughts or comments, particularly regarding an appropriate load, source and test signal? Thanks in advance, HAVE FUN, JOEY.
|
Don't be. There will always be someone like that., posted on November 19, 2002 at 05:12:00 | |
Posts: 4425
Location: Long Island Joined: October 3, 2002 |
But your question is still very valid. TTFN, John |
Re: mud measurements., posted on November 19, 2002 at 14:52:28 | |
Posts: 4708
Joined: May 16, 2000 |
No, I just have my ears. |
Re: Should I just yell "PULL"??, posted on November 19, 2002 at 15:17:32 | |
Posts: 4708
Joined: May 16, 2000 |
That is why I sent you to him. |
Thanks guys, that was interesting. [nt], posted on November 19, 2002 at 17:16:55 | |
nt |
Re: Ooooooh, I forgot........, posted on November 19, 2002 at 17:32:33 | |
Posts: 4708
Joined: May 16, 2000 |
Of course he does, we talked this week. We are writing an LTE together. Get real, please. |