Hi guys. I'm taking some online lessons in circuits and have a question about interconnects and speaker cables. In my current lesson, it says that capacitors pass voltage when it changes and acts like a short with dc. So, if it will pass changing voltage, which music is, changes all the time, why is capacitance bad in a cable?? Seems like it wouldn't matter, as the message(music) would get through just fine?? On the other hand, I want to make some interconnects and all the sites i have been to mention capacitance as an evil that must be avoided. How come?? Don't flame me too badly guys, I'm only trying to learn something. Thanks and regards, J.D.
Interconnects... two cables running together (the inner core and outer shield) form a capacitor. In effect they are two plates side by side. Longer lead length = larger value of capacitance. So far so good 🙂
The output of a source component is not zero ohms but some other (usually low but not always) value. So if the output impedance was highish then the capacitance of the lead would form a simple R/C filter and roll off or attenuate the high frequencies. In practice it's not such a problem with normal length interconnects.
Speaker leads... same applies in theory... in practice the output impedance of an amplifier is always low so the attenuation thing isn't an issue.
Also small levels of capacitance can cause instability if say an interconnect were connected directly to the output of a high speed opamp.
The output of a source component is not zero ohms but some other (usually low but not always) value. So if the output impedance was highish then the capacitance of the lead would form a simple R/C filter and roll off or attenuate the high frequencies. In practice it's not such a problem with normal length interconnects.
Speaker leads... same applies in theory... in practice the output impedance of an amplifier is always low so the attenuation thing isn't an issue.
Also small levels of capacitance can cause instability if say an interconnect were connected directly to the output of a high speed opamp.
The two conductors bound together act like the two plates of a capacitor, so that's where the capacitance comes from. But it's not capacitance from one end of the conductor to the other (which would pass AC, like you're talking about*), it's from one conductor to another. So at (very) high frequency the capacitive cable acts like a short. Well, that's the extreme case, more likely it just throws something else off.
* Your cables would make a very small capacitance, most likely, and wouldn't be useful for passing music even if it was in this alignment.
My understanding, anyway.
* Your cables would make a very small capacitance, most likely, and wouldn't be useful for passing music even if it was in this alignment.
My understanding, anyway.
Last edited:
A coupling capacitor (passes AC, blocks DC) is in series with the signal; together with the input resistance of the next stage it forms a high-pass CR filter so the capacitance value must be high enough to pass the lowest frequencies. Have a look at CR filters.
Cable capacitance is in parallel with the signal, so together with the output resistance of the previous stage it forms a low-pass CR filter so the capacitance value must be low enough to avoid reducing the highest frequencies.
You need to understand these basics before worrying about other effects which may or may not happen. Bear in mind that cables are a topic that people argue about.
Cable capacitance is in parallel with the signal, so together with the output resistance of the previous stage it forms a low-pass CR filter so the capacitance value must be low enough to avoid reducing the highest frequencies.
You need to understand these basics before worrying about other effects which may or may not happen. Bear in mind that cables are a topic that people argue about.
Thanks guys. Knew this was the place to ask. Thank you for keeping it simple. I understood what you fellas were saying!!!!! Must be learning something. Much appreciated guys. Best regards, J.D.
One more thing, and I hope it will not confuse the issue too much: capacitance in a cable is not the same as pure capacitance in a component f.e.: it is partially or completely balanced by the inductance of a cable.
This means that you can carry broadcast quality signals on hundreds of feets of cable.
There are second order effects that limit the bandwidth, but anyway, capacitance is not the primary limiting factor: in a speaker cable, inductance is the dominant one, because the charateristic impedance of the cable (something like 80 ohm, for ordinary cable) is much higher than the impedance of your speakers.
This means that you can carry broadcast quality signals on hundreds of feets of cable.
There are second order effects that limit the bandwidth, but anyway, capacitance is not the primary limiting factor: in a speaker cable, inductance is the dominant one, because the charateristic impedance of the cable (something like 80 ohm, for ordinary cable) is much higher than the impedance of your speakers.
Thank you all. Well spoken fellas in language i could understand!!! Never occurred to me that it was in parallel instead of series. Duh. Well, like i said, i'm learning. Thanks again fellas, really appreciate it. Best regards, J.D.
True for RF, less true for audio. In most audio cables the inductance is swamped by resistance, so the characteristic impedance becomes complex and frequency dependent.Elvee said:
(OP please ignore this post if it confuses you - Elvee and I are talking about more complicated effects than simple CR filters).
It confuses most people!
But more correctly it's also the leakage resistance of the insulator (conductance) that makes the formula very messy at audio frequencies. Once we get to radio frequencies we can use the simpler formula.
Another way they describe the parallel capacitance of a cable (say an RCA jack cable) is "parasitic" capacitance, that is it takes current to drive it, that doesn't get to the end and do anything useful. Current costs money. Right now, LM4560 op amps that will drive a lot of current very fast, are over $2, whereas 4558's that won't put out much current fast at all, are about $.25. The power supply that will drive a 24 channel mixer board with LM4560's is a lot more expensive than the power supply that only has to support 24 channels of 4558 op amps. That 100' snake cable from mixer to power amp, has a whole lot of parasitic capacitance. Before "electret" mikes with a little internal amp, you couldn't put the mikes 100' from the mixer, because the mikes wouldn't drive the capacitance of the cable without a mike amp up there on stage. "Dynamic" mikes that would drive a lot of cable, didn't pass many high frequencies so they weren't much use for symphonies, only voice and pop bands. High frequencies from a medium source impedance like a dynamic mike, get lost in the capacitance of the cable.
Another example- a popular phono preamp from 1961 was the tube PAS2. It only put out enough current to drive the parasitic capacitance of a 6' RCA cable, plus the input of the power amp. There were other preamps that would drive a lot more cable, but they didn't sell as well because they cost a lot more, using more capable, lower volume tubes and a bigger power supply. As well, the preamp had to be less than 6' from the magnetic phono cartridge on the turntable, because the un-amped cartridge can't drive the parasitic capacitance of more RCA cable.
Another example- a popular phono preamp from 1961 was the tube PAS2. It only put out enough current to drive the parasitic capacitance of a 6' RCA cable, plus the input of the power amp. There were other preamps that would drive a lot more cable, but they didn't sell as well because they cost a lot more, using more capable, lower volume tubes and a bigger power supply. As well, the preamp had to be less than 6' from the magnetic phono cartridge on the turntable, because the un-amped cartridge can't drive the parasitic capacitance of more RCA cable.
Last edited:
Yes, but in practice I think it turns out that at audio frequencies insulator conductance is negligible when compared to capacitive reactance, but conductor resistance dominates inductance.Speedskater said:
Jim Brown has a good paper on Transmission Lines at Audio Frequencies:
http://www.audiosystemsgroup.com/TransLines-LowFreq.pdf
Dynaudio Ocos Loudspeaker Cable has very high insulator conductance:
http://www.dynaudio.com/eng/pdf/DYN_Ocos_TechnGrundl_INT.pdf
A solution in search of a problem.
http://www.audiosystemsgroup.com/TransLines-LowFreq.pdf
Dynaudio Ocos Loudspeaker Cable has very high insulator conductance:
http://www.dynaudio.com/eng/pdf/DYN_Ocos_TechnGrundl_INT.pdf
A solution in search of a problem.
That's debatable: a typical figure-of-eight speaker cable has a lineic resistance of ~15milliohm/m, and a lineic inductance of ~1µH/m.
At a frequency of 10KHz, this is equivalent to 62milliohm/m, 4 times the resistive part.
At 1kHz, assuming your figures, it would be 0.4 times the resistive part so in the LF-mid region where audio signals are strongest the resistance dominates. OK, I will qualify my statement: for most of the audio band resistance dominates. If you consider the audio band to have ten octaves, then only the top three(ish) have inductance dominating. My main point was that you have to consider inductance, resistance and capacitance but you can usually ignore insulator conductance.
I completely agree: all current polymers, including the humble PVC can be considered as perfect in this respect, and for this application.
- Status
- Not open for further replies.