author=Response Audio link=topic=5079.msg65278#msg65278 date=1408814160]
All this is a perfect example of is meant when people mention "system synergy". Too many people don't look at impedance differences when combining components which sometimes leads to bad or incorrect consumer reviews.
From what I have read in forums, a LOT of Class D based amps are getting wrongful bad feedback due to improper matching of a preamp. I saw one consumer bad mouth a Class D amp (6k ohm input) and after checking further, his tube preamp had an output impedance of 2,500 ohms and a gain of 26dB. Simply not a fair evaluation. That preamp is meant more for a tube amp with high input Z and very low gain.
The impedance matching itself is not a problem as is the coupling method (see below). IF capacitance output coupled, the bass response will definitely suffer. I have seen some components use a poor size .67uf output coupling capacitor (see below).
If transformer coupled, one then has to worry about ringing, +/- frequency response besides tilting of the frequency response.
In most cases, a preamp (any output stage) with an output impedance of under 1k ohms does not pose a problem with [most] tube amplifiers and many solid state amps.
Preamplifier output Z is the main determining factor of high frequency response in most amplifiers. 20k, 50k, 100k amp input Zs. However, most preamps already have tremendous FR as per my quote above.
With about 2k preamplifier output Z and 100pf, the FR is somewhere around 300khz or higher, minus 3db. As one can see, less than 2k means little except another stage, unless if one has very high interconnect cable capacitance.
My whole article is at:
http://www.sasaudiolabs.com/sasaudio.htmThe vast loss in FR in a preamplifier is due to volume control to input device, whether tube or SS.
So going by the 10:1 ratio, this requires a minimum 700 ohm output from a preamp. The lower input of many of today's amps, especially Class D/Ice, is what lead to the design of our single ended preamps having a very low output impedance. One of our full size LeAnna models and out Segue tube buffer has an output of <90 ohms and the other LeAnna models no higher than 630 ohms.
It is too bad that class D amp designers do not raise the input impedance because with 6k, 7k, a huge, gigantic output capacitor is necessary to not heavily skew the overall response, the lower the more deviation from accuracy.
A 1uf capacitor has 8k reactance at 20hz, 2uf has 4k of reactance at 20hz. Either small value will negatively influence the response into midband. 8k will be at least 6db down at 20hz. That means approximately minus .25db at 320hz, and some even at 640hz. (Of the top of my head.)
Either a huge output cap is needed, or transformer, or dc coupled. But they also have their problems.
I agree that with 26db of gain, a low gain amplifier would work nicely as one has more volume control rotation to work with.
Cheers
Steve