The permanent magnet serves to cancel the field caused by the average DC current flowing through the inductance. Obviously, such a scheme is only usable when there is such a current, as in the filter chokes of SMPS. For a class D amplifier, the scheme is unusable (unless you intend to create a single ended class D).classd4sure said:
Regarding the drum cores, I've already dismantled a number of cheap chokes including them, and one thing that always strikes me is the disproportion between the diameter of the central rod and the general size of the core. With such a small Aeff in the ferrite part of the path, saturation begins at modest currents.
In addition, the ferrite material is always of the Ni/Zn variety; this is necessary because the leads are embedded in the material, requiring a high resistivity material (and by the way, resistivity of ferrites is complex and I wouldn't be very optimistic about the linearity of the leakage impedance at the switching frequency).
The problem is that these ferrites have a significantly lower induction capability than Mn/Zn types, compounding the saturation problem.
I understand the central pillar is made as thin as possible to leave the maximum of space free for the copper, and maybe this is OK for lots of applications, but for class D, you want a magnetic path as homogenous as possible to avoid partial saturations.
LV
Lars Clausen said:
Saturation of the coil, see below photo. 10 Ampere per div.
That oscilloscope capture speaks for itself. Nice linear low-hysteresis high-saturation inductor indeed. BTW: I use that method to measure parameters and test linearity of all my inductors, as you have probably figured out from some similar captures that I posted some time ago.
Eva said:
Now you have made me curious, what is the easiest method of acquiring this graph of a given inductor? Construct some monstrous square voltage waveform source that is preceded with a current sense resistor to allow measurement of the inductor current and plot this on the scope triggered from the rising edge of the square voltage step?
Eva said:
But if I want to measure Isat up to maybe 40 Amps, how do I accomplish those monstrous currents in any practical manner? My lab PSU only manages 10A, I guess I could charge capacitors to high voltage and use them for short bursts of current, but then I also need to add feedback to the inductor voltage..
Caps are all you need... in an extreme case, you may wish to measure the voltage across the inductor differentially (2 probes, A minus B) or some other means. In this way you can compensate (correct on paper) the current slope measurement for losses in voltage as the caps discharge and ESR consumes voltage.
Classd4sure:
Any audio designer knows that there are two aspects of a design, the measured difference, and the audible difference. What you do with any design is to try and find a relationship between what is heard and what is measured in order the fix the problem the right way. If all the experts knew the mathematical solutions, you could design anything just by a perfect simulation, which anyone in the audio world knows it's impossible.
If someone cannot pin point specifically what is audibly different so that others cannot duplicate the experience, then lots of dicussion cannot be correlated.
Any audio designer knows that there are two aspects of a design, the measured difference, and the audible difference. What you do with any design is to try and find a relationship between what is heard and what is measured in order the fix the problem the right way. If all the experts knew the mathematical solutions, you could design anything just by a perfect simulation, which anyone in the audio world knows it's impossible.
If someone cannot pin point specifically what is audibly different so that others cannot duplicate the experience, then lots of dicussion cannot be correlated.
As you may tell from nature of the posts, measurements in this case are easily indicative of linearity /distortion. There's no need to resort to a "how does it sound" emperical method, which would leave you as in the dark as before, only able to swap parts blindly having no real clue why one might almost work and the other not. That's more guesswork, luck of the draw, and may very well never get you there.
It takes a firm grasp of the principles at hand in order to do any worthwhile empiracle work as well. So let's first establish a firm level of understanding, and there'll be no need whatsoever to resort to "how's it sound", because you'll be able to do the job right everytime, and so it will automatically sound just fine.
The alternative will just not get us anywhere, or anything, but more subjective babble, marketing or otherwise, it's just not something that'll help you in the "design" aspect.
It takes a firm grasp of the principles at hand in order to do any worthwhile empiracle work as well. So let's first establish a firm level of understanding, and there'll be no need whatsoever to resort to "how's it sound", because you'll be able to do the job right everytime, and so it will automatically sound just fine.
The alternative will just not get us anywhere, or anything, but more subjective babble, marketing or otherwise, it's just not something that'll help you in the "design" aspect.
soongsc said:
"how does it sound" will NEVER make for a designed inductor, period!
We're discussing one component out of the entire amp, it happens to be probably one the most critical ones to have as optimal as possible. It alone will never be fully responsible for how it sounds, but it will play a very large roll in THD vs Frequency /Power, which is only optimised by careful design, after which, how it sounds, is simply how it sounds.
Asking a question like that in a design based thread just makes you a troll in it, you won't convince me otherwise. Why not sit back and try to soak in what others are discussing instead? Use their cues to do your own research.... oooh you're not interested in design.... you just want to know how it sounds
The question remains...
Distortion in this environment would be about core and copper losses... and in particular the non-linearities.
Feedback rears it ugly head... once again.
It is less about the material and technique that it is about how generously an inductor is sized for the application. No decent audio app should be at 1/2 of Bsat.
Want something small? Pay for exotic materials. Want something cheap... make it oversized. Want quality? Do both.
Distortion in this environment would be about core and copper losses... and in particular the non-linearities.
Feedback rears it ugly head... once again.
It is less about the material and technique that it is about how generously an inductor is sized for the application. No decent audio app should be at 1/2 of Bsat.
Want something small? Pay for exotic materials. Want something cheap... make it oversized. Want quality? Do both.
classd4sure said:
"how does it sound" will NEVER make for a designed inductor, period!
We're discussing one component out of the entire amp, it happens to be probably one the most critical ones to have as optimal as possible. It alone will never be fully responsible for how it sounds, but it will play a very large roll in THD vs Frequency /Power, which is only optimised by careful design, after which, how it sounds, is simply how it sounds.
Asking a question like that in a design based thread just makes you a troll in it, you won't convince me otherwise. Why not sit back and try to soak in what others are discussing instead? Use their cues to do your own research.... oooh you're not interested in design.... you just want to know how it sounds
The goal for discussion is to make an amp sound better. So if something seems good in theory, the next logical step is to ask how it sounds. Why would you even spend time in technical discussion if it's not going to sound better?
Now to think of it, it might just also be a social thing?Personally I think to know where to go in improving the output inductor, is to conduct measurements at variouse input levels and signal type, and look at the output voltage/current relationship. If someone had that kind of data and cannot interpret it, then that's the beginning of real engineering discussion. A whole bunch of theory is really irrelevent to any specific situation, but is a good chat if that's what people like. People in politics do the same thing before deciding on something.
We've been waiting on the RMAA measurement you've said you would post, hope you didn't burn your sound card doing it.
"The goal for discussion is to make an amp sound better. So if something seems good in theory, the next logical step is to ask how it sounds."
Once again it's not a tweak thread, its' a design thread.
"Why would you even spend time in technical discussion if it's not going to sound better? " To learn to better design
"Now to think of it, it might just also be a social thing?" troll
"Personally I think to know where to go in improving the output inductor, is to conduct measurements at variouse input levels and signal type, and look at the output voltage/current relationship. If someone had that kind of data and cannot interpret it, then that's the beginning of real engineering discussion."
Which was the topic of discussion..... but, "how does it sound
"A whole bunch of theory is really irrelevent to any specific situation"
If you believe that I feel sorry for you.
"We've been waiting on the RMAA measurement you've said you would post, hope you didn't burn your sound card doing it. "
Wrong thread for that topic, but when I do post them, I expect you'll either ask how it sounds, or if my measurements respected absolute polarity.
Once again it's not a tweak thread, its' a design thread.
"Why would you even spend time in technical discussion if it's not going to sound better? " To learn to better design
"Now to think of it, it might just also be a social thing?" troll
"Personally I think to know where to go in improving the output inductor, is to conduct measurements at variouse input levels and signal type, and look at the output voltage/current relationship. If someone had that kind of data and cannot interpret it, then that's the beginning of real engineering discussion."
Which was the topic of discussion..... but, "how does it sound
"A whole bunch of theory is really irrelevent to any specific situation"
If you believe that I feel sorry for you.
"We've been waiting on the RMAA measurement you've said you would post, hope you didn't burn your sound card doing it.
" Wrong thread for that topic, but when I do post them, I expect you'll either ask how it sounds, or if my measurements respected absolute polarity.
soongsc said:
Definitively, NO:
The goal is to make the amplifier as transparent, linear and faithful as possible from an objective (i.e. measurable) point of view.
For all sorts of reasons, this might not be the "best sounding" option.
The reasons could be psychoacoustic, compensation for another distorsion elsewhere in the signal chain, etc, but this does not matter.
You wouldn't want to compensate for an oxidised mike connector in a recording studio, would you?
You have to make your bit as well as possible and hope other people upwards in the chain have done the same. Any other approach makes no sense.
Once you've reached the technical perfection, as set by the measurement floor of your instruments, you're free to try to improve further by other means. If you can't measure any difference between a polysulfone and a PTFE cap, you can try A/B comparisons, but first, you've to do your homework.
LV
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.