Here's another example - I don't know if Andreas Koch calls himself an audiophile, but you might be forgiven for thinking he knew better than to draw the noise floor line for 24bit PCM where he's put it in this graphic : http://www.playbackdesigns.com/wp-content/uploads/2012/11/Picture4.png
<edit> by the way PMA, the factor is 10 in your log expression as we're dealing with powers, not voltages 😉 You must've used 10 yourself to get the result to be 45dB. Ooops, no sorry you're right - missed out the sqrt - easier to write it and calculate it as 10 log (24000/0.73).
Last edited:
Oh no, for it is AMPLITUDE SPECTRUM (see the image properly)
For amplitude spectrum, we have 20 log V1/V2
For power spectrum, we have 10 log P1/P2 and get the same dB. There are only one dB's in this world 😀
Please do not argue, I have the FFT measurements verified.
My dad gave me a totally disassembled 1964 VW bug engine in a box...I was 12. Best erector set ever!!
He did some really amazing things.
Ground loop. Total lack of control over current paths and Faraday's law of induction.
I thought I was living on the edge with purpleheart..😕
I'm far more interested in how mutual coupling affects input to output of each filter cap.
jn
Well then measure it. We are shipping a fourteen rack system tomorrow. A bit busy till then. After that a day or too off.
Of course if you have an idea about the technique.
Pavel
Very nicely stated.
ES
MY estimate of an IC's 'audio quality' is related to the steepness of the distortion waveform. If it looks like 'Switzerland' you are in trouble. If it looks like 'rolling hills' then it is probably good sounding.
Personally, I don't care if this non-linearity has to be measured with 'heroic' methods. In my personal opinion, you don't have to measure significant distortion that is probably buried in noise or is very small, because the intrinsic non-linearity of the transfer function appears to be the key, as this non-linearity may appear to the ear as a different (non-measured) distortion such as PIM, but that is only one distortion that we know can be generated from the intrinsic non-linearity of the device (without feedback covering it up).
Personally, I don't care if this non-linearity has to be measured with 'heroic' methods. In my personal opinion, you don't have to measure significant distortion that is probably buried in noise or is very small, because the intrinsic non-linearity of the transfer function appears to be the key, as this non-linearity may appear to the ear as a different (non-measured) distortion such as PIM, but that is only one distortion that we know can be generated from the intrinsic non-linearity of the device (without feedback covering it up).
As I said, run the capacitor return current over the cap body using a braid. That will do a lot towards decoupling input current and output circuit.
jn
Like this? Return path is a cylinder, removed for photo.
Attachments
Looks like a flux shunt.
( ~dozen years later, Shocking with/without Blue : ✎ M?n Journaaltje: Jerney Kaagman Naakt )
( ~dozen years later, Shocking with/without Blue : ✎ M?n Journaaltje: Jerney Kaagman Naakt )
I though Ron measured PIM, most modern IC's had none while revered audiophile components had lots. So I guess PMA is right these distortions are sought out by the GEB.
How about that IC resistor non-linearity, Scott? You know the ones you use in most IC's. Please say something useful.
arent you listening John?
do you deliberately ignore answers given you, so that you can more readily repeat the question 100 pages later?
do you deliberately ignore answers given you, so that you can more readily repeat the question 100 pages later?
- Status
- Not open for further replies.