|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
76.90.38.219
In Reply to: RE: Output impedance of a DAC to input impedance of Power amp?????? posted by georgehifi on August 11, 2014 at 02:49:35
To add to what Nelson said,
Maybe the output impedance of the passive preamp is too high and the highs are being sucked out of the music by the low pass filter created by the output impedance vs. the shunt capacitance of the interconnecting cable plus the Miller capacitance of the first stage of the amplifier?
At 10k ohms the (worse case) output impedance will be 2500 ohms at the -6db setting.
That should only cause HF roll off with very high shunt capacitance but not all passive preamps are that low of an impedance.
With a 100k passive resistive attenuator the worst case output impedance is 25k ohms and it would only take a little shunt capacitance to roll off the highs and shift the phase well down into the audio band.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Follow Ups:
Hi Tre', just to further add to what you also said for the interested readers.That is why a 10K external passive pre is the way to go, never 100kohm. As you said worst a 10kohm will be is 2.5kohm output impedance.
Good interconnects these days are always less than <100pf (picofard) per foot. I always say keep them under 1.5mts long when I suggest use with my Lightspeed Attenuator which mimics a 10k passive pot, but say you have 2mts at 100pf per foot. That's approx 600pf of capacitance per channel.This combined with the 2.5kohm ouput impedance of the 10k passive pot equates to a HF roll of -3db at 106khz!!!!!so there is far from any phase shift down into the audio band. Most amps can't reach this figure, let alone digital sources.
Cheers George
Edits: 08/11/14 08/11/14 08/11/14
While a -3db point of 106kHz will yield a (more or less) flat frequency response to about 53kHz it takes a decade for the phase shift to go away.That means there will be some phase shift all the way down to 10.3kHz
This is why all the tech books talk about bypassing cathodes to 2Hz if you want 20Hz flat and phase shift free (it's the same thing but on the high pass side of the equation) and if we want 20kHz flat and phase shift free, all of the low pass filters (and they are seemly everywhere in a system) need to have their -3db points no lower than 200kHz.
(It's impossible to reach that goal when an audio transformer is involved but that's no reason to introduce more phase shift when you can help it)
You can find good cables with far less than 100pF per foot.
I'm using 1 meter cables that measure 70pF. Not 70pF per foot but 70pF total! Measured on a Sencore LC53.
BTW Don't over look the input capacitance of the power amp, that would be added to the cable capacitance.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 08/12/14
Of course you can find cables less than 100pf per foot, and I used this as a maximum, that is why I said < less than 100pf.
As for frequency response no tube poweramp's output transformer I know of, can attain 100khz without masses of feedback around it, and then they sound rubbish.
Cheers George
"As for frequency response no tube poweramp's output transformer I know of, can attain 100khz......"
OK, but is that a good reason/excuse to add more phase shift in the audio band when it could otherwise be prevented?
Have a listen to 2.5k driving 600pF then drop that to 150pF. I would bet you will hear the difference even through an output transformer that can't make 50kHz flat.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Hey Tre" I'm all for speed, my amps are -3db at 180khz. Why do you think I called my product Lightspeed Attenuator it goes from dc - ghz in speed.
The simple fact is all digital sources are limited to 20khz and DSD to 50khz. And vinyl is a joke trying to reproduce a 20khz sine let alone square wave.
When I give my specs for interconnects to be used with the Lightspeed Attenuator I say <100pf per foot and 1.5mt long, this with the Lightspeed's output impedance works out to be -3db at 141khz.
Cheers George
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: