I agree completely with Dr. Geddes. "Tweaking" in the usual audio sense is a huge waste of time - in my experience, you just go round in circles, going from one fad to another. It takes measurements to make any substantive progress - finding the right measurement, that correlates with the subjective world of experience, is what takes serious effort and R&D time.
In other words, subjective and objective must work hand in hand to make any forward progress - but I would be the first to admit the necessary measurements can be very difficult and a long way from the "industry standard" set of graphs you see in the magazines. One simple example - the inter-driver phase angle between LF and HF drivers has a major subjective effect on crossover integration, but this is never seen in magazines, because it is hard to measure without cracking open the crossover and making a series of adjustments for yourself.
In other words, subjective and objective must work hand in hand to make any forward progress - but I would be the first to admit the necessary measurements can be very difficult and a long way from the "industry standard" set of graphs you see in the magazines. One simple example - the inter-driver phase angle between LF and HF drivers has a major subjective effect on crossover integration, but this is never seen in magazines, because it is hard to measure without cracking open the crossover and making a series of adjustments for yourself.
I was wondering, because I have never heard a system that I was satisfied with for more than a month. Yet I see people defend their decisions to death like they have the ultimate solution. I find new things to explore that gets on my list every week, and find things in measured data that are very interesting.gedlee said:
soongsc said:
Working hypothesis: although a copper-plated pole piece provides a degree of isolation between the magnet and the induced flux coming from the voice coil, this is not 100%. No form of shielding (magnetic or electrostatic) is ever 100% - you have to be satisfied with whatever attenuation is practical.
Given there is a degree of interaction between the magnetic system and the voice coil (this is where almost all of the VC inductance comes from), the question arises, how nonlinear is this inductor? We can ask our transformer-building friends this question, and we'll find that even small changes to core material are not only measurable, but quite audible. Transformer guys go on and on about the differences between M6, nickel, and amorphous cores, with ferrite only considered suitable for RF use.
So my guess is that the magnetic structure is, to some degree, audible - but you'd have to have several speakers side by side, each identical except for magnet structure, and then do extensive IM and harmonic distortion measurements. I'd look for high-order terms (5th on up) appearing at 500 Hz and higher frequencies, and do a series of tests from very low-level to 10 or more watts. Measuring the distortion on the current waveform (looking at the back-EMF) might be a useful shortcut, and would avoid S/N issues with acoustic measurements.
soongsc said:
I too find this poignant. Although it seems to take forever to actually finish a design, I generally live with them for many years. I was quite happy with the LO-2's from 1979 to 1994, and the Ariels from 1994 to the present. I used the Audionics CC-2 from 1979 to 1995, built the Amity in 1995, and am still fiddling around with the Karna after having it for several years. I get it to where I want it, and then stop.
I believe the hysteresis issues inherent in the classic steel-cored voice-coil motor represent one of the biggest compromises audiophiles have had to live with. I think the future of dynamic loudspeaker motor design lies in minimizing steel's role in the circuit. There are ways to do this, but it means a total re-think of the classic design.
I always smile to myself as I read posts by audiophiles discussing the finer points of high-zoot air-core inductors in the crossovers of speakers featuring great asymmetrical lumps of mild steel at the ends of their speaker leads.
I always smile to myself as I read posts by audiophiles discussing the finer points of high-zoot air-core inductors in the crossovers of speakers featuring great asymmetrical lumps of mild steel at the ends of their speaker leads.
Lynn Olson said:
A copper "plated" pole is basically ineffective. Copper plating can get to maybe .003" thickness or so, which has little effect. You need to have a solid rigid sleeve of a much greater thickness to be very effective. Ours use a sleeve that is .025" thick and covers the full length of the pole. There was an AES paper that I don't have access to anymore that explained the correlation between thickness of the sleeve and how effective it was. Basically the thicker the sleeve, the lower in frequency it was effective, and the greater effect it had going up more in frequency.
The core material can be a contributing factor. You want the highest magnetic permeability with the lowest electrical conductivity. A powdered iron core for example is basically electrically non-conductive and totally kills the inductance of the driver. The issue is that it is also less magnetically permeable, taking away as much as 20% of the flux in the gap. Hiperco 50 (permendur) is a good material choice, but is ridiculously expensive. To get bar stock in 2" diameter is roughly $200 per INCH! Laminated cores with layers of steel/copper/steel/copper much like a transformer core are also an option but are quite expensive to make. They also have a loss of BL as well.
A simple way to see the effectiveness of a copper sleeve or other shorting ring is by playing a low frequency tone at high excursion, then another higher frequency tone. For example 20hz with a 1KHz tone. You can measure and hear a huge difference in the 1KHz tone between the drivers. The one with copper sleeve will show much less modulation of the 1KHz tone.
Some other notes on this issue. The magnetic material alone doesn't necessarily make a huge difference. Majority of the modulation of flux is not in the magnetic material but where there is the least amount of flux. This means first in the air gap, and then secondly in the top plate/pole. The wider the air gap, the more area of low flux you have and more easily the field there modulates. The key is to get gap tight minimizing the airgap. Then to get the steel near the gap to be as close to saturated as possible. In the case of an Alnico slug in the pole piece with a small washer on top, you have a very minimal area of steel directly on top of the magnet and in most cases it is completely saturated. I like the idea of both alnico in the pole and a ring on the OD. Consider though an alnico ring around the OD with a traditional top plate and modulation looks closely similar to a motor with equivalent flux from a ceramic magnet.
John
Bill F. said:
Hi Bill, good to hear from you on here. We need to get in touch again.
One thing that was done recently was a motor where they fixed the coil to the pole permanently, then moved attached a steel tube to the cone to move it in and out of the gap. As the coil is in a fixed position on the pole you have some clear benefits as the core is always staying the same. Also being physically anchored to the pole it has great heat transfer.
John
Hello Earl,
That's funny, here in France a majority of audiophiles considers that I am on the objectivist side, probably because I don't dismiss theory (see my "work" on horns and on crossovers) nor measurements.
I cannot accept when someone argues that I have chosen the "easy road" ...
But I also rely a lot on practice and experience and this leads me by times to be sceptical about some scientific results. As an example I am sceptical about results on the "inaudibility of non linear distortion in compression drivers" ...
Old Scientific reviews are full of results that were considered reliable and true in their time.
Best regards from Paris, France
Jean-Michel Le Cléac'h
That's funny, here in France a majority of audiophiles considers that I am on the objectivist side, probably because I don't dismiss theory (see my "work" on horns and on crossovers) nor measurements.
I cannot accept when someone argues that I have chosen the "easy road" ...
But I also rely a lot on practice and experience and this leads me by times to be sceptical about some scientific results. As an example I am sceptical about results on the "inaudibility of non linear distortion in compression drivers" ...
Old Scientific reviews are full of results that were considered reliable and true in their time.
Best regards from Paris, France
Jean-Michel Le Cléac'h
gedlee said:
Well, after listening for some time, I always can find places where improvement it necessary, but that does not mean I don't enjoy the system. As a matter of fact, I still have a set of active speakers of about 20 years. It always reminds me of what my thoughts were then when I put the speaker together.gedlee said:
John_E_Janowitz said:
Interestingly this is a byproduct in the Lambda TD motor that happened by accident. I designed the original prototypes around an off the shelf edgewound VC designed for some JBL speaker. So it was a thin coil. Once we went into production the 4 layer flatwires were designed to fit inside this same thin gap. I was also forcing the gap tighter for heat conduction more than anything else at the time. We went to really tight tolerances and clearances there as a result.
John_E_Janowitz said:
John
Why do you say the lowest electrical conductivity? I don't see the rational there, especially when you surround it with high conductivity material. Am I missing something?
This requirement would make an ALNICO slug magnet a poor choice as it has very high electrical conductivity and I had always thought of that as an asset.
<snip>
.003" in 1 hour @ std. conditions. No limit with proper high speed setup >1" thick easily attained. Rotogravure cylinders, electrotype, microwave waveguides come to mind. Where did you get this (mis)information?
Perhaps one of the high permeability/low conductivity nickel/cobalt/phosphorous alloys as a core used in the ubiquitous hard disc platters could be of some value with heavy copper electroplating. Much work has been done on these materials (for obvious reasons).
John L.
John_E_Janowitz said:
.003" in 1 hour @ std. conditions. No limit with proper high speed setup >1" thick easily attained. Rotogravure cylinders, electrotype, microwave waveguides come to mind. Where did you get this (mis)information?
John_E_Janowitz said:
Perhaps one of the high permeability/low conductivity nickel/cobalt/phosphorous alloys as a core used in the ubiquitous hard disc platters could be of some value with heavy copper electroplating. Much work has been done on these materials (for obvious reasons).
John L.
auplater said:
The problem is that once you start to get very thick with any kind or plating it is very hard to keep the same thickness tolerances all around the round part. Also the process gets extremely expensive due to the time involved. Calculate the cost to do a 2" diameter round pole that is 3.5" long to .025" thick with electroplating. We're talking quite a lot of surface area. The copper tube we have drawn costs significantly less. I don't doubt that you can technically deposit this kind of copper on a part, but it isn't practical to do so.
In theory it is a good option, but the costs really puts it out of any kind of a reasonable price range. Look at any of these materials and figure what a 2" diameter x 4" long round bar would cost. Now look at the cost of 2" round 12L14 or 1010 steel. Then factor in the machinability. If you are aware of something that can be bought reasonably in 2" round bar that we can work with I would be all for trying it. Remember we still have to be able to sell these at prices people can afford to buy.
John
Whatever the core material, the ideal goal is a linear BH curve. Anything with a "knee" in the curve is going to display hysteresis.
The BH curve of steel alloys tend to be more linear when operated in saturation, above the knee. The trouble is that even a nicely saturated magnetic circuit doesn't stay that way when VC signal is introduced. Half the waveform will push the magnetic circuit out of saturation--across the knee--and the other half will drive it deeper into saturation. The net effect is a dynamic transfer function.
This is especially true at low frequencies where eddy currents drop off.
John_E_Janowitz said:
Well...maybe a commercial plater would charge an "arm and a leg" for a 1-up or 100-up order, but trust me, the actual costs to do such under high speed conditions are pretty low, at least if you're not a hack.
John_E_Janowitz said:
Again, this is true with any R&D endeavor, nothing unique about EF... I was mainly offering what could be theoretically better performance. Cost/value analysis comes down the pike after the technical issues are resolved, in my experience. Mostly, though, folks don't want to "pony up" for the necessary R&D.
John L.
soongsc said:
Live unamplified performance in Austria.
Is it Vienna?
Couldn't find out for sure - but that place where the musicians play reminds me on bronze figures (in the background) I haven't seen again since my first visit more than three decades ago.
Do you remember the name of the place, soongsc?
some thousand pix that might give an ("slightly" polished
) impression of Vienna http://www.foto-julius.at/images_wien49.html
at Julius Silver's page
Michael
Does this place look familiar to anyone? Old movie?mige0 said:
Is it Vienna?
Couldn't find out for sure - but that place where the musicians play reminds me on bronze figures (in the background) I haven't seen again since my first visit more than three decades ago.
Do you remember the name of the place, soongsc?
some thousand pix that might give an ("slightly" polished
http://www.foto-julius.at/images_wien49.html
at Julius Silver's page
Michael
Hint: Located close to a border.
Beautiful place.
I really enjoyed the Art Museum in Vienna. Walking around, it seemed you can walk into many places that do various live performances.
