Sorry to inform you the voltage and current in a capacitor are 90 degrees out of phase in SPICE. I mean really, come on.
Hmmm. If you noticed the real part is NOT 90 degrees at all freq and THAT is not modelled in SPICE, as you just stated.
THx-RNMarsh
The point I made was the difference was 0.1dB or more below 150hz -- audible.
It is audible on an amplitute comparison basis.... with and without.
The design without (the +/> .1dB rise) is the more accurate one. Although some comments here seem to think some arbitrary amount of inaccuracy is OK.
The 'slight' change in thd may be true or not depends on freq from 90 degrees to 0. Again, the comparision is not that it is 'slight' or not but is it detectable (with and without) and which is more accurate.
Anything Added that should not be there can be removed, then do so. This is what the High-End of JC and others think about and eliminate -- at every place and detail so the total sum is greater accuracy.
Not for everyone, maybe, but perfection is a goal, too. And, you cannot do your best by ignoring the many details of parasitics and their direct or indirect affect.
On the more practical side is that electrolytic are like the LP... they degrade with use/time. I didnt test junk parts so this is the best results and they Can be detectable. In time of use the performance and difference between an old or used electro and nothing will be More noticeable.
Detectable -- with and without the cap is the critieria i am using and concerned with. Note too that in many circuits today we have at least 6 such detectables in the path... maybe 8. Remove all the minor detectables and it is a rather more dramatic change. In that case, 'easily' heard by listeners as they have been saying for decades now.
THx-Richard
Last edited:
Interestingly enough, many of those self-acclaimed High-End products (and I am not speaking specifically about JC's designs, please bear this in mind) are not accurate at all. Opposite is very often truth and very often these products belong to those with very poor measurements. I think you do not want for me to bring the names, but there are at least two that are often highly regarded here that both design products that are very inaccurate in technical terms, though supposed to be very good sounding by crowd of supporters and reviewers. The problem of these debates is that almost no one comes with real data, real measurements, facts. Instead, general statements and stories are provided, without any quantification, without any possibility of qualified discussion. Then it looks like a spectacle for admiring unqualified crowds.
Last edited:
That is your choice. Not the one percenters choice - such as JC. Some listeners do actually have large rooms and even acoustically treated rooms. And, those are the one's that want better and get catered to by some designers. I remember an Arab when asked which product -- brand x or Y. he said, 'which is the best? Thats the one I want.' What do you tell him... Oh this other one is good enough and cheaper?
But, you dont have to be wealthy beyond all reason to get more accurate sound. Sometimes, its just a collection of many 'small' things.
THx-RNMarsh
Last edited:
However in your case you have a slew of DSP to flatten the response of your M2s. correcting that 0.1dB at 150Hz (which I actually doubt is audible that low) is just another biquad for you. Therefore in YOUR case a total non-issue.
And it's not my choice to not have a listening room the size of a small auditorium. It's not feasible in UK unless you are in the 0.01%
And it's not my choice to not have a listening room the size of a small auditorium. It's not feasible in UK unless you are in the 0.01%
's
What are you trying to say, that SPICE does not use complex variables? You add ESR or parasitic inductance that's what's known as a model. But you said perfect capacitor so the voltage and current are always in quadrature.
What are you trying to say, that SPICE does not use complex variables? You add ESR or parasitic inductance that's what's known as a model. But you said perfect capacitor so the voltage and current are always in quadrature.
Last edited:
Where the rise starts and how much just depends on the component choices and values. That example isnt fixed in stone.
I know, most people have too small of room volume to get the best out of their gear. Headphones, now come to mind.
There is always a way or work-around.AND, i showed a way to EQ the headphone to be flat at the ear (Linear Audio). In case the HP isnt perfect enough.
-Richard
It is not. At that frequency, the room effects swamps everything to multi-dB amplitude comb filter effects.
I believe it is not useful to reply to your other posts. Against reasoned arguments you seem to just repeat your opinion. I am not prone to monologues.
Jan
Last edited:
Tell it to Harmon then. they did tests in a room also. They are detecting if they heard any difference, also. You do Not need a response flat to better than .1dB to hear a .1dB change against no change.
You would be hard pressed to have a room response to better than 1dB variation anywhere but could detect an added 1dB... etc etc.
As I also said in #251, first sentence.... so, it can be at higher freq as well, depends on values selected. You can get higher than .1dB with choices depending on bad assumptions and worse parts.
I am illustrating potential issues with large capacitor parasitics as found in electrolytic capacitors. it is for the reader/designeer to figure out if any of this polar cap parasitic stuff is relevant to a given design.... including degradation with time. No need to reply.
What is your gripe with dc-servo... any tests... audible or otherwise? Give specific topology. or are you saying all dc-servo implementations give audible errors? Which is how your generalization appears.
THx-RNMarsh
Last edited:
This is what a good film does to phase.
It is my patented design built by Rel-Cap. PPFX
REL-Invested $50K USD to have a custom cap making machine to do it.
It is only company who can make it. Co-patented by Bas Lim - owner of REl-Cap. I figured out what to do and how and Bas figured out how to make the machine work to make it to my design.
Its self resonance is much higher than typical PP caps. The esl and esr are lower. Up to 1/10th lower.
It is the esr + esl which causes the change in phase.
90 degrees at 150Khz
The cap you use in fb to ground, should be like this; Using PP film. But, its going to be huge for large values.
THx-RNMarsh
It is my patented design built by Rel-Cap. PPFX
REL-Invested $50K USD to have a custom cap making machine to do it.
It is only company who can make it. Co-patented by Bas Lim - owner of REl-Cap. I figured out what to do and how and Bas figured out how to make the machine work to make it to my design.
Its self resonance is much higher than typical PP caps. The esl and esr are lower. Up to 1/10th lower.
It is the esr + esl which causes the change in phase.
90 degrees at 150Khz
The cap you use in fb to ground, should be like this; Using PP film. But, its going to be huge for large values.
THx-RNMarsh
Last edited:
Hi Richard, what is the value of your cap, just for reference? I was going to ask you to try your amazing impedance analyzer on one of your caps, but you did it first. Good!
By the way, Walt Jung just called me on a different subject, and I tried to update him about what is going on on this thread at the moment. I told Walt about your amazing HP impedance analyzer. I wish that I had one!
By the way, Walt Jung just called me on a different subject, and I tried to update him about what is going on on this thread at the moment. I told Walt about your amazing HP impedance analyzer. I wish that I had one!
Call me and I'll tell you. To 20Khz, though, most any film cap is close to 90 degrees. 80 degrees worse case.
You can also use a network analyzer for phase info. If you can find one for low frequencies. Thats what I had been using but they have limited Z input range -- 50/75 Ohms without a probe.
So this is easier and shows all terms used in components like caps and inductors and goes much higher in freq than typical LCR meters. Which makes it also good for designing HF filters.
THx-RNMarsh
Last edited:
There is that
Of course the Crown iTech sidestep all these worries and could be seen as the REAL step forwards. Digital input, full EQ and FIR for room correction and the ADCs drive the amplifiers directly so the design team have full control of the signal path. Makes worrying over input coupling seem so last century!
- Status
- Not open for further replies.