Quote:
Originally posted by eric343
Doug- that's an excellent point. To be perfectly honest, I don't have anything capable of generating a spike in the audio band... the closest to that is the 500nS setting on the TDR, and it shows little to nothing if I remember correctly (though I don't think I've tried it with the Nites, and I don't want to interrupt the burn-in process at the moment to try). The problem is that the cables are much too short - any reflection would be swamped by the outgoing signal. |
The same problem occurs in a radar system. The minimum radar range detection is the inverse of the output pulse width. IOW, it's impossible to detect something too close to the transmitting unit since the unit is still ouputting its signal and can't transmit and receive at the same time. In this case, a broadcast signal will travel at about 300M meters per second, which is just a few pico seconds per inch IIRC. At that rate of travel the shortest test path you could measure would be somewhere in the meter range, counting for recovery time from transmit to receive.
Quote:
| The other problem is that much of what we can hear so well, our test instruments can't detect so well unless they're running in the range of signals that *should* be waaay out of the band that we're supposed to be limited to hearing in. I suppose I'm learning that I can hear things that strictly speaking there's no way I should be able to hear - as far as I can tell, music is a much more complex and difficult to reproduce signal than most people think. |
That was a very succinct and honest answer. I guess that those who have tried to slam Rick at VD for not knowing WHY his cables work really are saying that they don't know enough about the subject to understand that they CAN work, or that they also don't know why or how they do work.
Quote:
| What's also possible is that all I'm hearing is the placebo effect, and it really DOESN'T matter what cables look like on a TDR screen. Barring that possibility, though, as far as I can tell the TDR is the best way of analyzing cables because it gives them a fairly tough to handle signal and sees how they react, since it's through stress that you can see faults you otherwise wouldn't. |
At the frequencies you are testing at, I would say the test data is about 110% irrelevant to put in rather bluntly. To test an AF cable at very high RF frequencies will yield nothing relevant nor revealing as to how they would sound and perform at any frequencies other than as an antenna to broadcast those tested frequencies.
Your conclusions could be related to the placebo effect, but it's too soon to tell. Did you hear this smearing of the sound before or after testing on the TDR? Did you verify what you heard, or what you saw?
And to correct myself, a 100nS signal is, I believe, 10GHz, not the orignal 1GHz I originally thought. I only have the cheap calculator that came with Windows, but if anyone has a real calculator which you could punch in 100x10 ^-9 and then invert it I would appreciate it. Even if it is 10GHz and not 1GHz, it is still so high a frequency that the skin effect is most likely coming into play here. Even the 500nS pulse would only yield results at 2GHz if I am correct, and this is still quite a bit too high a frequency.
Eric, how low will an FDR go frequency wise? And, why not just use a nice, cheap, signal generator and inject an audio frequency into the cables and skip the TDR all together? You could always read the results on an o'scope instead of using the TDR for this purpose.
