A rule of thumb is to have the input impedance 10 times the value of the output impedance of the driving equipment. If the input impedance of the amp is 10K ohms (most are at least that) then the output impedance could be as high as 1000 ohms.
An analogy to a battery might help to understand the situation. A battery can be modeled as a perfect voltage source (maintains its voltage even if the current is infinite) in series with a small value resistor. As you pull more current out of the battery, the voltage drops a little because of the low value resistor that is "inside" the battery. As the resistance of the load gets smaller it will try to draw more current from the battery, but the "resistor" limits that current somewhat. One can see that the "resistor" inside of 12 volt lantern battery is much larger than the one inside a automotive battery. If a low resistance load (like a automotive headlight) is connected to an automotive battery it is going to be brighter than when hooked up to the little flash light battery.
The idea is the same in a passive volume control. (several CDP's use this on the output while others put active stages following the volume control and then there are the ones that use digital attenuation in the output stage) A passive's maxium output impedance is TYPICALLY at mid position (electrically) and is approximately half the value of the variable resistor. This means that a 10K pot will have about a 5K output impedance and might be suitable for use with a tube input amp as long as the IC cables were low in capacitance. If the volume adjust is followed by a electronic stage then the design of the stage determines the output impedance. (typically a 200 ohms or below)
I do agree that a well designed CDP could be better than a preamp just because it will have less connections in the signal path. The big question is what entails a "well designed" CDP. I believe that a good output will sound good and will look forward to comments from others regarding the sound of CDP with variable outputs. Chris
An analogy to a battery might help to understand the situation. A battery can be modeled as a perfect voltage source (maintains its voltage even if the current is infinite) in series with a small value resistor. As you pull more current out of the battery, the voltage drops a little because of the low value resistor that is "inside" the battery. As the resistance of the load gets smaller it will try to draw more current from the battery, but the "resistor" limits that current somewhat. One can see that the "resistor" inside of 12 volt lantern battery is much larger than the one inside a automotive battery. If a low resistance load (like a automotive headlight) is connected to an automotive battery it is going to be brighter than when hooked up to the little flash light battery.
The idea is the same in a passive volume control. (several CDP's use this on the output while others put active stages following the volume control and then there are the ones that use digital attenuation in the output stage) A passive's maxium output impedance is TYPICALLY at mid position (electrically) and is approximately half the value of the variable resistor. This means that a 10K pot will have about a 5K output impedance and might be suitable for use with a tube input amp as long as the IC cables were low in capacitance. If the volume adjust is followed by a electronic stage then the design of the stage determines the output impedance. (typically a 200 ohms or below)
I do agree that a well designed CDP could be better than a preamp just because it will have less connections in the signal path. The big question is what entails a "well designed" CDP. I believe that a good output will sound good and will look forward to comments from others regarding the sound of CDP with variable outputs. Chris