Thanks for the explanation. As it happens, my non-Circlotron preamp also has no output coupling cap, but at the cost of a negative supply...
All else isn't equal. You're familiar with the data on frequency response variations from speaker cables that Peter Aczel presented. What happens when you double the length of that cable? It doubles those variations (assuming you haven't gone over to the Dark Side and are now denying basic physics 😀). There's nothing even vaguely like that in preamp-to-power-amp cabling.
Agreed. Excellent for DC coupled SS power amps. As I recall italian Graaf 6C33C OTL using Cyclotron schematic for poweramp, however not sure about my preamp graaf 13.5B II (love it).
Mu-followers have become very popular as well.
Regarding the pre amp/power amp interfaces controlling cables/speaker cables:
The pre amp which I will be using shortly has an output imp of 100 ohms - at what frequencies, I don't know.
The manufacturer states:
Up to 10m (33') each output (@30pF / ft).
This seems a good candidate to move the mono-block power amps to the speaker locations, driven from the pre amp at a distance of about 7 feet per channel.
I'll purchase some Canare RCA interconnects LV-77S at 21 pF/ft.
Does this sound like a good plan?
Any recommendations for some other cables?
What about speaker cable recommendations?
The mono block amps don't list the output imp, but output 125 watts@8/250@4
Regarding power amp sensitivity for full output, here is my theory and plan:
My DAC has user adjustable outputs via ten-turn potentiometers - max output with 0dBFS, is +18dB ref. to 0 dB =.775 volts RMS.
If theoretically, 0 dBFS from the CD data represents max out (there are cases where more output would be produced with inter sample overs), and the pre amp set to unity gain, and the power amp's sensitivity is @ 1.25 volts rms, then would it make sense to play the CBS CD Test Disk 0 dBFS track, measure the pre amp out with a true rms volts meter and set the pre amp out level to just a little under 1.25 volts by adjusting the DAC output trim?
This seems logical, such that max level from CD data, passing through pre amp at unity will never exceed pushing the amp past max output?
Note that the pre amp does have an attenuator moving from unity downward, so one wouldn't be stuck with max data from CD to power amps.
The pre amp which I will be using shortly has an output imp of 100 ohms - at what frequencies, I don't know.
The manufacturer states:
Up to 10m (33') each output (@30pF / ft).
This seems a good candidate to move the mono-block power amps to the speaker locations, driven from the pre amp at a distance of about 7 feet per channel.
I'll purchase some Canare RCA interconnects LV-77S at 21 pF/ft.
Does this sound like a good plan?
Any recommendations for some other cables?
What about speaker cable recommendations?
The mono block amps don't list the output imp, but output 125 watts@8/250@4
Regarding power amp sensitivity for full output, here is my theory and plan:
My DAC has user adjustable outputs via ten-turn potentiometers - max output with 0dBFS, is +18dB ref. to 0 dB =.775 volts RMS.
If theoretically, 0 dBFS from the CD data represents max out (there are cases where more output would be produced with inter sample overs), and the pre amp set to unity gain, and the power amp's sensitivity is @ 1.25 volts rms, then would it make sense to play the CBS CD Test Disk 0 dBFS track, measure the pre amp out with a true rms volts meter and set the pre amp out level to just a little under 1.25 volts by adjusting the DAC output trim?
This seems logical, such that max level from CD data, passing through pre amp at unity will never exceed pushing the amp past max output?
Note that the pre amp does have an attenuator moving from unity downward, so one wouldn't be stuck with max data from CD to power amps.
Last edited:
As a result power amps would drive interconnects quite well 🙂
But in driving speakers, the longer the cable the greater loss in resolution. I think I suggested this earlier: drive 100 feet of speaker cable and see how well you do. Compare that to 3 feet. The difference is night and day, obvious to anyone that can chew gum. It is more subtle when the speaker cables are shorter, 10 feet as opposed to 3, but you can still hear it. The lower the impedance of the speaker, the more obvious the difference. Also the higher the output impedance of the amp, the more obvious the difference.
OTOH if a preamp is doing its job, it is controlling the interconnect cable, and 100 feet or 3 feet there will be no difference.
This is assuming for your inquiry an integrated amp of no compromise, an animal that does not exist in the real world.
Last edited:
Pre-amps generally drive high impedance loads and don't have to deliver significant power. My objection to very long speaker cable runs is two fold, IR losses over 50' or so can be appreciable, and most cables have significant capacitance which can create stability issues with some power amplifiers - particularly solid state. (All of these issues can obviously be addressed)
Driving a very long 600 ohm balanced pair properly terminated is relatively easy by comparison.. (Note "properly terminated" in this case that means the cable inductive and capacitance reactances actually appear purely resistive to the driving source. This is otherwise not the case.)
I'm pretty agnostic on the issue overall. My listening room has relatively short runs of 12ft from amp to speaker and the amp is located within 2 ft of the line stage which can actually also drive relatively long lines without a problem.
Very low efficiency/low impedance speaker systems would argue for the amp being placed very near the speaker to reduce power losses in the speaker cables. I have seen some ridiculously large speaker cables to drive speakers like some of the larger Apogees which have impedances of just 1 - 2 ohms depending on model.
Last edited:
What's the difference? Hell, in my system I use the same cable for both.
I'll give that a try next time I need 100 feet of cable to drive my loudspeakers. Which will be... never.
I'm also told you can hear the difference if you put photographs of yourself in your freezer. Or if the copper's 99.999999% pure versus 99.99% pure.
The output impedance of the amplifier can have effects that are wholly irrespective of the cable.
And I see absolutely no fundamental difference between an amp and preamp when it comes to "controlling" the cable. What exactly is it about the cable that needs to be "controlled"? And what is it about a preamp that it can "control" the cable better than a power amp all else being equal?
100 feet of cable isn't the "real world" for any home systems to speak of.
se
V = IR.
edit: Fascinating software glitch. I capitalized all the variables but it keeps showing up as lower case.
No. But I don't see where the amount of power delivered to the load is of any particular relevance.
Yes.
The only point at which the cable's inductive and capacitive reactances would appear purely resistive to the driving source in the audio band would be at the cable's resonant point. Which typically would be well above the audio band.
There is no "characteristic impedance" in the audio band for cables of lengths you'll find in an audio system.
se
Interesting crusade, Steve.
Everybody 'knows' it, but nobody can explain it.
I wonder how this will end ? 😉
jd
Everybody 'knows' it, but nobody can explain it.
I wonder how this will end ? 😉
jd
Inductance introduced in a long run to speakers from amplifiers have a greater tonal balance effect than that of interconnects driven by the pre -amp.
Hence long speakers cables are less acceptable than long interconnects of course this is very system and topology dependent as to the degree of how much is noticeably ...
You can tell the sound of any cable by measuring them , magnitude / phase vs frequency tell the story !
Hence long speakers cables are less acceptable than long interconnects of course this is very system and topology dependent as to the degree of how much is noticeably ...
You can tell the sound of any cable by measuring them , magnitude / phase vs frequency tell the story !
Last edited:
Yes. In interconnects, it's capacitance that's of more concern.
Yes. And how long one defines "long."
Sure, if you have a rather large room and you have your loudspeakers on one end and your sources on the other and are faced with running 30 feet of interconnects versus 30 feet of speaker cables, then yeah, that situation might favor 30 feet of interconnects.
But that doesn't make an integrated amp a compromise. The compromise was putting your source components on the other side of the room from the loudspeakers.
se
Well I'm sure you could put the worlds best amp, Pre-amp, phono stage and there associated power supplies all in one and call it an integrated amplifier , most would call it a condominium !
Now why would that make SS better than tubes .......... 😀
Now why would that make SS better than tubes .......... 😀
Years ago a couple bored hours waiting for downtown accident traffic to clear so I could head home was wiled away with an Audio Precision examining the effect of Zout driving 200 feet of cabling between studios. All generic twisted-pair CAT 3 stuff. The details are hazy but I recall an optimum Zout of 75 or 150 ohms into a bridging load, the main deviation being some minor peaking around 75 kHz. Afterward I stopped fretting about the magnitude of interconnect capacitance.
Everything. Let's say you have a power amp input that varies 2:1 in impedance over the audio range (perhaps it's got a high input capacitance). Let's say 50k-100k. Voltage is of the scale 1 V. Current is then 1V/50k = 0.02mA to 1V/100k = 0.01mA. Now, the long interconnect, with perhaps 5 ohms of resistance, forms a voltage divider with the amplifier input impedance. At one extreme, the drop across the interconnects is 0.1mV, at the other it's 0.05mV, for a difference of 0.05mV. That's 0.005% of the signal. And that's for an impedance deviation much greater than one normally sees.
Now, what about the speaker? Speaker load is on the order of 10R. Let's say that also varies 2:1 over the frequency range (it will probably be worse, but apples to apples). Power amp output is on the order of 20V. Let's pick a speaker cable of 0R5 resistance. Current is then, at one extreme 2A, on the other extreme 1A. What's the frequency response deviation because of variations in the IR drop through the interconnect? At one extreme, the drop across the cable is 1V, at the other, it's 0.5V. That's a 0.5V difference, which is 0.5% of the signal. One hundred times worse than the deviation from the interconnect. And that's for an impedance deviation much smaller than one normally sees.
But the deviation of the signal is a function of the source vs. load impedance. Not how much current you may have flowing at the time.
Drop your power amplifier output voltage from 20 volts to 1 volt. You have a lot less current flowing, but the signal deviation will be the same.
Similarly, increase your preamp output voltage from 1 volt to 20 volts. You have a lot more current flowing, but the signal deviation will be the same.
What distinguishes the interconnect example from the speaker cable example isn't "current." It's the difference in source/load impedance ratios.
se
- Status
- Not open for further replies.