[Rclark]

Seems to be the crux of a good system and I'm assuming is the true meaning behind the phrase "system synergy".

Can someone please break it all down.. Also also explain how ne would measure it on their components so they know for sure what they're dealing with?

[*Scotty*]

This is not what is meant by the phrase "system synergy". This ONE of the variables that determines whether a source component is compatible with a pre-amp and whether or not the pre-amp can drive the input impedance of a particular poweramp.
 The phrase "system synergy" refers to a subjective judgment made by a listener upon hearing a collection of audio components and speakers assembled into system. The listener decides that they like or dislike the sound of system or the sound of the system after a new component has been added.
 Obviously when the judgment call is in the favor of the system it is said of the components that they have positive "system synergy" together.
There is no objective basis for this value judgment on the listeners part," system synergy" exists entirely in the ear of the beholder.
Scotty

[Rclark]

Thank you Scotty. How do you measure it, and what are good ranges to have between parts?

[*Scotty*]

Measuring input and output impedances frequently cannot be done with a simple DMM. Many times it will require a signal generator and a more sophisticated measurement technique.
 It would nice if the sources and preamp had as low an output impedance as possible. Power amp input impedances should be what ever sounds best as dictated by the circuit design and the devices used.
As a rule input and output impedances are speced by the factory and don't need to be measured.
Scotty

[Rclark]

well then, if we don't need to measure it, how do you know when your pieces have a good fit?

 What are the acceptable ranges per device?

[munosmario]

well then, if we don't need to measure it, how do you know when your pieces have a good fit?

 What are the acceptable ranges per device?

Rclark, hope the link below helps you to get a better idea.http://www.soundonsound.com/sos/jan03/articles/impedanceworkshop.asp

munosmario

[Speedskater]

Stereophile magazine has been very good at measuring details like this.  For the most part you should be OK.  Only if your system has units with vacuum tube outputs or pro-audio inputs or passive attenuators might you be concerned.

[Davey]

http://www.sengpielaudio.com/calculator-InputOutputImpedance.htm

[opnly bafld]

There is no objective basis for this value judgment on the listeners part," system synergy" exists entirely in the ear of the beholder.

IMO there is more often than not an objective (measurements that will correlate) reason for someone calling a pairing "system synergy", the listener is just unaware.
So, instead of having 0 + 0 = 0* some end up with +1 + -1 = 0
    *0 represents a particular sound or result a listener is looking (listening?) for


Lin

[Ethan Winer]

Seems to be the crux of a good system and I'm assuming is the true meaning behind the phrase "system synergy".

It's vastly simpler than that. The SOS article is mostly right on, but I can simplify further:

The ideal with line level signals is for the output impedance to be as low as possible, and the following input stage to have an impedance at least ten times higher. A low output impedance helps drive long cables with less HF loss due to capacitance, and it also reduces crosstalk between channels and radio interference because the output acts like a short circuit to anything that arrives at the wires.

The only thing that can happen as an output impedance rises is for the audio to become less accurately transferred. Either HF loss, or an interaction with other reactive properties (subsequent input transformer), or reduced damping (power amps), and so forth.

--Ethan

[munosmario]

Gentlemen--Ethan, Davey--the SOS article, in the box labeled "Impedance and Frequency Response" the author says:

"The output impedance of a device and the capacitance of its connecting cable form a simple first-order low-pass filter, producing a 6dB/octave attenuation above a certain frequency. However, you need either quite a low output impedance or quite a long high-capacitance cable to bring the turnover of this filter into the audio band. Even so, it's best to select cables which have as low a capacitance as possible, and to keep cable runs as short as practicable."

If the low pass formula is F= 1/(2x3.1416xRxC), where F is  the -3 dB turnover point of the low pass filter, and R is the output impedance of the source (plus the interconnect's basic resistance, if I remember correctly), and C is the interconnect capacitance, then, that statement in the box is incorrect. Both R and C work in the same not in opposite directions. A low value of R (source's low output impedance) or C (low interconnect capacitance) will make the denominator smaller, making F larger; that is, moving the low pass filter's turnover point higher and higher in the frequency spectrum (pushing beyond the 20kHz treshold), letting all the relevant audio information pass through. Am I right? I believe Ethan's post agrees with what I am saying but I wanted to re-confirm.

munosmario

[Steve]

Hi R,

There are two criteria for low and high frequency response. I shall use capacitive output coupling as that is the vast majority of what systems employ. As mentioned above we are dealing with two simple 6db filters with both filters.

The output coupling capacitor of the source/preamplifier, and the input impedance (Z) of the amp constitute a high pass filter. As the frequency lowers from midband, the source output lowers, approximately 6db/octave.

The second filter comprises the output Z of the source/preamplifier and both the IC capacitance and the Amp input capacitance. Let us call the total capacitance "Ct". We have a low pass filter. As the frequency rises the source output lowers from midband.

First rule, according to the RCA Radiotron Designers Handbook, 1960, 26 RCA engineers, is that the input impedance (Z) of the amp (midband) be minimum 5 times that of the output Z of the source. I recommend 7-10 times to be conservative. But there is more.

For accurate bass response, the amplifier input Z should be very high for a given output capacitor. Even then the output capacitor should be large for relatively constant amplifier input Z vs frequency.

The high frequency response, as mentioned above, is almost totally determined by the output Z of the source/preamplifier and Ct.

As an example, let's say we have Ct of say 200pf. With source output Z of 3000 ohms (3k ohms) the frequency response (FR) drops only 0.6db at 100khz.

With 100 ohms output Z, the response is down less than .0007db at 100khz. As one can see, total overkill? Worse is the sonic degradation due to another stage added, with its parts, power supply etc just to lower the output Z.

Even with 400pf Ct, 100 ohms output Z, the frequency response is still excellent, better than .002db down at 100khz. The only time one needs very low output Z and high current drive is for very high capacitances due to very long runs of ICs.

Of course there are other criteria such as distortions of any kind (distortion is any variance from input).

Cheers.

[Davey]

Yeah, the statement in that blue box on the SOS page is incorrect.  It should say:

"However, you need either quite a HIGH output impedance or quite a long high-capacitance cable to bring the turnover of this filter into the audio band."

Probably just an oversight.  It should be intuitively obvious to the author since if output impedance reduces to zero you don't have a filter anymore. 

Cheers,

Dave.

[munosmario]

Thanks, Dave

[Rclark]

Always above and beyond. Thanks guys.

[Steve]

Correction. This:

Even then the output capacitor should be large for relatively constant amplifier input Z vs frequency.

Should state this:

Even then the output capacitor should be large for relatively constant preamplifier output Z vs frequency.

Sorry about the confusion.

Cheers.

[Ethan Winer]

if I remember correctly ... that statement in the box is incorrect.

Sorry, late to the game.

Yes, the box is wrong on that one small point, and that's probably a brain fart. Hugh Robjohns definitely knows his stuff. And as Steve explained, the input impedance and capacitance of the subsequent input stage are also involved. So is the available output driving current, which is independent from the output impedance itself. You could have an output impedance of 0.00001 Ohms, but with too little available current to charge the capacitance of even three feet of wire. Not that this usually happens in practice (though it does inside a condenser microphone).

--Ethan