Separate names with a comma.
Discussion in 'Meadowlark Audio Forum' started by Prime Minister, Apr 30, 2018.
Wouldn't the answer to this problem be a bigger driver?
I suppose that could be AN answer. But larger driver cones mean all the driver parts will be larger and heavier, requiring far larger magnets and amplifier power to control than that required for just an increase in cone mass. Also, the larger the cone, the harder it is to keep rigid. If I understand correctly, a loudspeaker cone that flexes under load is undesirable.
This is a fascinating thread! Thank you for starting it!
Your're right, Guy. Bigger isn't better.
The whole idea is that cone motion precisely track the rapidly changing input signal. So you want plenty of force because the amp should have perfect say over where the coil is at any moment. It shouldn't be flapping around on its own.
It's basic Newtonian stuff: F=MA. You want maximum A. So big F and small M.
With current technology, there is a practical limit on the intensity of available magnetic flux. So you really can't scale F to keep up with an increasing M.
Keep in mind that larger cones are both heavier and face a higher load meeting the air. So M is actually it's physical mass plus the air load.
Happily, stroke and diameter are perfectly interchangeable. A driver's 'displacement' determines how much work it can do. You figure it just like automotive engine displacement: working area times stroke times number of cylinders. It's a volume specified in liters. For speakers the notation is: Emissive Area (Sd) times peak-to-peak linear excursion (Xmax p-p).
So you easily see why two motors driving two smaller cones is nicer than one equivalent motor driving a larger one when the total displacements are equal. There's twice as much F being applied to about the same load.
As usual, all of these things lead us down the road to spending more money ;-)
Just out of curiosity, is reverting to an older technology something that could be beneficial at times? I'm wondering if it's possible to create a stronger magnetic field with a field coil than with a permanent magnet structure.
What a fun idea! But I'm too ignorant of magnetic circuit engineering to figure it out.
The power requirement calculations should be pretty easy, though.
Assuming equivalent efficiency, the static field will need at least as much power as the moving coil. And it will have to be held at the coil's peak power - otherwise the system will clip the peaks. Our amp power ratings are for continuous power; a 200W amp can peak at 2 or 3 times that. So we'd want probably 600W for the static field.
So: 600W of noise free DC. Hmmm. In most amps the power supply is the largest and most expensive system. This is heading in a rough direction.
Oh, but wait: the whole idea is to generate more force at the motor. I'd say this baby is a nonstarter for less than 6dB. That would put us at 2400W. Ouch! That draws 20A from thru your house wiring, times 2 for stereo. Call an electrician!
Maybe the thing works by using two amplifiers, with the second one to sending an inverse but otherwise identical signal to a static coil?
I'll bet that's been tried. Or maybe we just invented something. Who wants to do a patent search?
Ok, I get the picture. It's likely possible, but quite impractical. Same as it was 60 (70?) some years ago when field coils fell out of favor.
I wish I knew more about magnetic fields. My dad was a TV engineer starting back in the vacuum tube days who worked with the high voltage circuits that modulated the electron beam deflection yokes in cathode ray tubes. The problem that tormented those guys was called hysteresis.
That problem pretty much limited how fast they could scan the beam, limiting both resolution and frame rate. I'm just guessing, but I suspect that may be a problem with field coils.
Fortunately for us, the magnet guys are doing great things. Have you noticed how tiny a car's starter is now? They're getting big flux from smaller and smaller hunks.
Driven by the impending need to transduce HD audio with 140dB of DR, the best driver guys are catching on. Here's a pic of the motor on the latest Scan Speak:
In this driver they're taking advantage of the stronger flux to both lengthen the coil and improve sensitivity - two opposing objectives - an amazing feat.
It looks as though they could continue to stack those slugs several times over. So I think maybe we're may be on the right track using magnets.
I just wanted to tie a bow on this tech workout by touching on two additional problems in typical speakers that hurt dynamics and the perception of dynamics.
First, many crossovers are designed from an F/A perspective with scant attention paid to other issues like timing perturbations. They tend to be complex with numerous caps and coils. The problem is that both capacitors and inductors work by absorbing power then releasing it later. While this doesn't interfere with dynamics per se, it does interfere with the perception of "smack" by snubbing the leading edge of transients, turning a nice sharp "whack' into a softer "drub". The factor that I call "bark and growl" is impaired. The effect is to make the music seem lifeless even if the amplitudes are scaling well. For this reason I've always held filter design to a bare bones minimum, have never put a capacitive shunt in a woofer circuit, though that is far-and-away the norm. It does take much more time and effort to design a simple circuit so I can understand why guys take the easier path and go with complex filters because they're easy, quick and safe. Sometimes your choice of drivers may not work with simple filters, but maybe you don't get to find that out until you're well into prototyping. So you can see why the corporate fellows stay on the low risk path to avoid do-overs that can look exactly like incompetence to their higher ups. The bosses are only gonna look at the F/A graphs anyway, so why fool around with tricky filter work? That's just one of the diseases that infect "design by committee" projects; just one of the reasons that they seem to produce pond water. I design alone and happily toss my boo-boos into the dumpster while my wife laughs at me.
Second, the best dome tweeters run out of poop pretty early on. Owing to the wavelengths involved, 20KHz being less than an inch, the diaphragm needs to be small to avoid becoming narrowly directional. And, F=MA, the mass must be kept tiny to wiggle so fast. So the coils tend to be dainty, so they tend to get hot. The usual fix is to add a heat conducting magnetic oil called ferrofluid between the coil and the gap which has the side effect of snubbing dynamics by damping motion. Take your pick: it's either that or fry your coils. Those factors pretty much hem in the dynamic envelope for dome tweeters. That didn't really matter much for LP, reared its ugly head for CD and is now a real problem for HD.
Now, it makes no sense to build an LF system that outruns your HF system; the whole thing's just going to sound thick when pressed. So some guys have resorted to using multiple tweeters to grab more headroom but, with domes, that can easily result in comb filtering in the top octave, an unpleasantly spooky effect.
Our best solution so far is Oskar Heil's clever invention, the Air Motion Transformer. A large diaphragm is compressed by folding it, pleating it like window drapes, so a large working area is squeezed down to speak thru a small aperture, killing two birds with one stone. Large working area, small emissive area. Three birds, actually, since the conductive trace is spread out over the large area that pumps air like a bellows, handily whipping the thermal issue at the same time.
The rub is that AMTs are difficult to fabricate properly, read: expensive. The cheap ones really suck. They're easily as annoying as cheap domes. So please don't be dissuaded if you hear a crappy AMT.
The top performers are pricey but worth it. We've found that Mundorfs can sail right on up to the human amplitude limit with superior linearity and with ease and grace. Beautiful detail and naturalness. Plus they can handle bandwidth. In our new Nightingale we hit the limit in our 24'x32'x10' starting at 1.5KHz with a medium sized Mundorf. Just for giggles, here are two pics of 40" AMT arrays that worked down to one kilocycle at 10dB above the limit in a very large room. On that build we ended up limiting the whole system by 13dB to prevent the owner's grandchildren from hurting themselves ;-) At normal listening levels the drivers were barely working at all so the dynamics sounded....ummm.....realistic.