jlsem said:
So what is the mechanism? can you point to a reference available online?
The remanence is not affected by permanent magnetization (supposedly induced by the applied dc offset). Low remanence alloys/grain size modifications would have a beneficial effect, but I fail to see how dc offset would reduce distortion.
In fact, an offset would seem to negatively impact the half cycle by moving the core closer to saturation on that side of the loop, inducing nonlinear response. maybe this non-linear increased distortion is mistaken for a "smoother" sound?
Is the thinking here that this is somehow analogous to crossover distortion in a p-n junction. If so, that would be incorrect.
rdf said:
If anything, from what I've read (and recollect), dc offset increases the coercivity (moves the small signal loops further out the loop where it is flatter), requiring greater field strength to reverse the domains. This would seemingly be the wrong direction to take, as the system would seem to exhibit greater losses (and most likely distortion atrifacts) than it would if the loops were closer to the origin and area of high reversibiltiy.
"Another important factor with soft magnetic materials is incremental permeability with DC Magnetizing Force ‘H’. The effective permeability of the material will decrease with the introduction of DC current in the coil winding. The lower permeability materials (14 to 60ìi) have higher DC Bias levels where the inductance remains more stable with high surges of current."
http://www.elnamagnetics.com/librar...t/Introduction to Soft Magnetic Materials.pdf
John L.
With all the above qualifiers magnified 10x, my understanding of the reasoning is the domains don't flip in SE. The lines of flux are always in one direction, the sum of DC + AC current never results in a net flow - using the normal convention for current - from ground to B+. Problem is, I'm not visualizing how this isn't true of push-pull too from a core magnetization perspective.
Not sure I understand how an ac signal can pass through a magnetic material without "flipping the domains". This would seem to be a fundamental necessity for a non steady state signal; else, what other mechanism transfers the signal through the core?
and if so, why not permanently magnetize the core and use high remanence and coercivity materials and then not need the dc bias at all?
i agree as to why this wouldn't be equally applicable to pp vs. se
and if so, why not permanently magnetize the core and use high remanence and coercivity materials and then not need the dc bias at all?
i agree as to why this wouldn't be equally applicable to pp vs. se
With single-ended amps the magnetic field never collapses. The magnetization curve is not the same as a hysteresis loop. It has been years since I have read the article and the book below. I was able to work out the problem to my own satisfaction a long time ago.
http://www.amazon.com/Magnetische-W...ef=sr_1_1/105-2327131-8375647?ie=UTF8&s=books
http://www.amazon.com/Magnetische-W...ef=sr_1_1/105-2327131-8375647?ie=UTF8&s=books
Picture is a snapshop of a Figure from the SE vrs PP Part 2 article by Eddie Vaughn (both a good read about the pros & cons of each topology)
http://www.vaughnaudio.com/tech-papers.html
What i was referring to was that an SE amp operates around the yellow spot, a PP (or parafeed SE) operates around the red spot... in the latter case, as the AC signal changes polarity energy has to be expended to flip the field (one of the reasons a little cobalt isn't a bad thing) and it has to work thru that wiggly bit of the curve. Looking just at this and nothing else, the advantage will go to an SE amp when very low power levels are required.
dave
http://www.vaughnaudio.com/tech-papers.html
What i was referring to was that an SE amp operates around the yellow spot, a PP (or parafeed SE) operates around the red spot... in the latter case, as the AC signal changes polarity energy has to be expended to flip the field (one of the reasons a little cobalt isn't a bad thing) and it has to work thru that wiggly bit of the curve. Looking just at this and nothing else, the advantage will go to an SE amp when very low power levels are required.
dave
Attachments
planet10 said:
That all seems rather misleading to me.
First is the curve itself. The curve shown assumes no DC magetization. With DC magnetization, the curve's not going to be the same as the one shown. So how can it be said that with DC magnetization that the SE amp will operate around the yellow spot when the DC magnetization changes the curve from what's shown?
Second, the dashed line is the curve of initial magnetization, i.e. it begins with both B and H at zero. Because of remanence, I don't see how that situation would ever exist under AC conditions.
Third, I don't see how he figures that DC magnetization will somehow cause the transformer to operate hysteresis-free. The core will still exhibit hysteresis. If it didn't, then when the magnetization force drops back down from the saturation point, it would follow the same curve as the initial magnetization curve.
se
rdf said:
Basically, yes.
His argument is based on his claim that the DC offset eliminates hysteresis (which would mean there is no remanence) which is just complete nonsense.
The offset DC creates a static magnetic field that holds the transformer in a highly linear region of it's magnetization curve, even at very low signal levels. It avoids ever seeing a zero flux condition, whereas PP does not. In a PP OPT there is no offset DC, so with no signal present there is no mmf, and no magnetic flux. The PP OPT depends solely on the AC signal to provide the magnetizing or "excitation" current, while a SE OPT stays magnetized 100% of the time. It's very important to understand this, because it's a critical reason why SE amplifiers are more coherent and detailed at low volumes than PP! There is no hysteresis or core loss...
So thus far, I haven't seen any evidence that a DC bias on the transformer improves any objective performance parameter.
se
A cursory check of Allegheny Ludlum's "Electrical Materials Handbook" shows a significant decrease in the area of the 60 hertz hysteresis loop of Permalloy with a small amount of superimposed DC. A DC "bias" of only .198 Oersteds reduces the hysteresis loss by approximately 90%. So while not completely eliminating core loss from hysteresis, a significant enough reduction lends some truth to that assertion.
Obviously it would take more superimposed DC to have the same effect on GOSS and I'm not sure what the effect of the gap is yet, but it is easy to see why bass notes saturate a single-ended OPT core faster than HF tones ( the position of the hysteresis loop moves up the induction axis and to the positive end of the magnetizing force axis).
John
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.