In the mid-80's, I was hired to do some acoustical research in Brittany, France. I drove from Geneva and overnighted in Paris; I parked my car in a very expensive guarded parking lot, and, of course, they broke into my car and stole everything, including a near pristine copy of the Acoustical Engineering.
I somehow managed to find a copy locally and xeroxed all of the 719 pages! I know it's not legal, but it was an emergency, and I was the owner of the original copy. I still have the xeroxes, but don't call the cops. 🙂
Assuming a resistive ring, the induced EMF and current in it are in 90 deg phase shift relative to the VC current and field (Faraday's law). Thus the ring field is also in phase quadrature with the VC field and hence doesn't oppose it. The VC inductance itself is also actually not reduced in this, similarly as the inductance of a transformer in not reduced by the load attached to it. The apparent VC impedance is, however, reduced because the image of the ring (load) resistance now appears electrically across the primary inductance (VC).jneutron said:
Same with pole piece currents.
Last edited:
BTW, my professor of electroacoustics at the University was Josef Merhaut. His book Theory of Electroacoustics, Mc Graw Hill 1981, is still sold at high price. I have a Czech version of the book.
Theory of Electroacoustics: Josef Merhaut: 9780070414785: Amazon.com: Books
1 New from $735.00 5 Used from $150.00
Theory of Electroacoustics: Josef Merhaut: 9780070414785: Amazon.com: Books
1 New from $735.00 5 Used from $150.00
The fluid filled suspension parts is strange but interesting.
The coaxial is odd in that he attaches the hf cone to the woofer cone so that woofer excursion move the tweeter vc in it's own gap. And, the hf coil seems to have a spider at the end of the vc as well.
I didn't find anything that had to do with magnetic theory or application.
And yes, lots of water. It's nice to appreciate how much he drove the boat, but considerable advances have been made in the intervening time.
jn
Sigh, and I was so talking you up.😀
The rate of change of the vc flux field is 90 behind the vc current.
The ring current is 90 degrees behind that.
The magnetic field energy map in 3d space sums up to a total less than the vc alone. Inductance is the relationship between the total energy of the field and the current in the system. Less stored field energy, less inductance.
I'll look for IEEE 1050.. It was about two computers ago, but I think it made the transition..may take a while to find however.
Currents induced in a conductive surface try to oppose the rate of change of the flux through that surface. The hole doesn't change that effect, it just adds circulating currents around the hole.
jn
JN, not for one thing or another, because we agree, as I knew we would, but I fail to get your explanation under my hood. The copper pole piece does not exclude magfield. It is too thin for that. So, I also can't grasp the relevance of the measurements you have shown to the working of a shortening ring, interesting as they are in their own right.
I also fail to grasp how the dissipation in the shortening ring is velocity dependant. Its major constituent is the flux in the pole piece being modulated by the varying magfield generated by the VC, which induces a secondary current in the shortening ring. This makes it situationally dependant, as you correctly describe. In some overhung speaker, behind the rest position, more of the VC encapsulates the pole piece, and the higher the inductance is. Which equites to higher flux modulation. This is dependant on situation, not velocity.
For speaker builders, the relevance of your measurements is to not put aircoils in a passive xover inside a conducting box, even not a non-magnetic conducting box. But it does not describe a mechanisme I can see going on in a shortening ring in any significant way.
Yes, it is Olson, just a mis-spelling. The real spelling challenge is the other guy:Beranek, Leo L. who is almost as interesting. Back in 1972, we made a number of complete Zerox copies of Olson's book (it was out of print) and used it as sort of a bible for years.
I think that Richard's idea of adding a relatively small current sense resistor and putting it in the return path of the voltage feedback network is elegant, difficult to understand exactly how it works, and what value is best, and a wonderful candidate for discussion here.
I think that Richard's idea of adding a relatively small current sense resistor and putting it in the return path of the voltage feedback network is elegant, difficult to understand exactly how it works, and what value is best, and a wonderful candidate for discussion here.
From what I've seen here, the problem is some of the debaters refuse to learn, so an expert would not make any difference.
The debate is not even at "expert level." It's at "AC circuit analysis" level, and is covered in second-year (perhaps first-year at better schools) classes at any EE college.
It doesn't differ essentially from ordinary transformer operation. Adding a load to the secondary doesn't change the original flux unless the primary voltage is allowed to change.jneutron said:
It is more complex than one inductance. Inductance is always the number of turns squared divided by the overall magnetic reluctance. Neither of these changes when a secondary coil is added to the system.
'Less stored energy' is the result of the shorting of the original inductance by the transformer loading effect.
Behind VC current. Yes, that is what I said.
Well, Scottjoplin, why don't you explain it to me? Is it positive or negative current feedback that is added? How does it change the output of the amp? Give equations please. What value current sensing is 'best'? What about more resistance? Less?
You will find the answers here Variable Amplifier Impedance and elsewhere including the references quoted, one of which both I and Richard have provided links to, have fun.
This is a rough draft of a mixed voltage/current feedback amp I made on the basis of an LM3886 to do some testing a looong while ago. Easy to put together and nice to play with.
R6 and R7 are two halves of a potmeter and when r6 is zero, it is voltage feedback, when r6 is zero it is voltage feedback. Allows for seemless adjustment of the Zout of the amp.
And yes, of course there is always some bleed through, so it is never fully current or voltage feedback.
View attachment mixxed.asc
R6 and R7 are two halves of a potmeter and when r6 is zero, it is voltage feedback, when r6 is zero it is voltage feedback. Allows for seemless adjustment of the Zout of the amp.
And yes, of course there is always some bleed through, so it is never fully current or voltage feedback.
View attachment mixxed.asc
You could make the feedback across the sense resistor variable and look for the distortion null point.
But, we still have no clear explanation as to why this works ie why it causes speaker distortion to reduce.
Looks like we cross posted vacuphile
There are some answers here https://pdfs.semanticscholar.org/2e6b/5b6ec83e229c1ab06108f731582de7368298.pdf
For those who are serious in speaker distortion reduction
https://pdfs.semanticscholar.org/c7b0/dfac1f520bb8027ad063b24e25992da88577.pdf
https://pdfs.semanticscholar.org/c7b0/dfac1f520bb8027ad063b24e25992da88577.pdf
Don't think of the copper in terms you would a magnetic material..think in terms of conductivity.
My graphs I linked to were copper on a pc board. What would that be, 2 or 3 oz material..pretty thin.
jn
The vc has a magnetic field associated with it. The motion is trying to bring those flux lines through the conductive ring.
jn
A coil energized by an ac current....the magnetic field is proportional to the current.
The rate of change of that magnetic field is the derivative, so is 90 degrees lagging behind the vc current.
That time varying magnetic field induces a current in a second ring. That current is proportional to the time rate of change of the magnetic field through the ring, so is 90 degrees behind the magnetic field.
The current in the ring produces a magnetic field directly proportional to the ring current, that is 90 degrees behind the magnetic field of the vc, and 180 degrees behind the vc current.
jn
Thanks Scottjoplin for the link. This a very clear explanation of what Richard recommends and even gives me a chance to express myself more clearly about what I have used over the last 50 years.
I first found the circuit for the 'improved Howland pump'(fig.6 in Rod Elliot's article) in an engineering article in the 1960's, but I had to modify it to be useful for bridge operation, but I did it. A decade later, I used the single sided version of the 'improved Howland pump' as a voltage to current driver for an analog tape recorder tape record drive. It is very effective for that.
It would appear that what Richard has put forth is an artificial resistance drive for driving his loudspeaker. That makes sense to me. Back in the early 60's, I added a series resistance (3.3ohms) to a 4 ohm speaker to get best subjective performance, but that approach wastes power, bigtime. Richard's approach drops the series resistor to 0.15 ohms or so, a great improvement.
Somehow I thought that Richard was using negative impedance drive, but that is not so, according to RE. There would have to be a positive input connection point, like RE shows. The 'improved Howland current pump' uses both positive and negative inputs for the feedback connection and is most elegant. When I showed this circuit to Dr. R.G. Meyer at UCB in 1973, he didn't think much of it. Now it is in the textbooks. See how history changes as new ideas get 'invented' then 'debated' then 'accepted'.
I first found the circuit for the 'improved Howland pump'(fig.6 in Rod Elliot's article) in an engineering article in the 1960's, but I had to modify it to be useful for bridge operation, but I did it. A decade later, I used the single sided version of the 'improved Howland pump' as a voltage to current driver for an analog tape recorder tape record drive. It is very effective for that.
It would appear that what Richard has put forth is an artificial resistance drive for driving his loudspeaker. That makes sense to me. Back in the early 60's, I added a series resistance (3.3ohms) to a 4 ohm speaker to get best subjective performance, but that approach wastes power, bigtime. Richard's approach drops the series resistor to 0.15 ohms or so, a great improvement.
Somehow I thought that Richard was using negative impedance drive, but that is not so, according to RE. There would have to be a positive input connection point, like RE shows. The 'improved Howland current pump' uses both positive and negative inputs for the feedback connection and is most elegant. When I showed this circuit to Dr. R.G. Meyer at UCB in 1973, he didn't think much of it. Now it is in the textbooks. See how history changes as new ideas get 'invented' then 'debated' then 'accepted'.
You're welcome
@JN There is a section on motional feedback at the end here (section 3.2) which might interest you using a sensing coil https://pdfs.semanticscholar.org/2e6b/5b6ec83e229c1ab06108f731582de7368298.pdf
@JN There is a section on motional feedback at the end here (section 3.2) which might interest you using a sensing coil https://pdfs.semanticscholar.org/2e6b/5b6ec83e229c1ab06108f731582de7368298.pdf
- Status
- Not open for further replies.