If neodymium is all that great . . .

[thoriated]

Dan was refering to permanent demagnitization, not AC flux modulation.

Try not to forget that part of what I was doing was to make an effort to clearly differentiate between the two phenomena, something that nobody else appears to be overly interested in pointing out. Sure, alnico's coercive force is lower than the other common types, but it's kinda silly to dwell on that to obscure its superior characteristics for high quality speaker applications. Proper alnico magnetic circuit design can even stymie the worst efforts of warranty-voiding audio abusers who are liable to hook your speakers to an arc welder amp and then clip it into them indefinitely.

[MBK]

Pan,

I always try polite and I seem to fail.

Resolution and distortion is related don´t you think?

Err. Yes. But what distortion? and do you, as you claim, find the relevant info easily on a manufacturer's website? Vifa, Peerless etc mainly publish FR graphs. Seas has THD graphs in addition to that. Some manufacturers may have waterfall plots with their inherent interpretation problems due to windowing etc. Typically the measurements used different setups each time, even within a same manufacturer. Take simple FR: some have full space anechoic measurements (say, Seas) in a small box. The FR shows a clear baffle step. Others, say Peerless, use half space measurements. No baffle step.

"Can anyone deduct resolution performance (and dynamic compression etc) from frequency response graphs and manufacturer's spec sheets alone?"

I don´t thihnk anyone ever claimed that but to some degree yes, if the measuremetns are good and detailed enough. Resolution needs low HD/ID distortion, flat FR and fast decay.

You went to two manufacturer's websites and deduced relative performance from the data published there (Aura 10" vs Excel 10" woofers). So, *you* claimed that.

I generally commented on what I saw as your naive view of taking a few published and not well standardized measurements as basis of comparison. For instance:

you´ll see that the performance are not superior on the Aura woofer. Even the W22 has better upper bass than the Aura woofers

This sounds suspiciously as if you took the naturally faster rolloff of a low Q woofer as an indication for "inferior bass response".

Back to resolution. THD ad IMD will impact resolution to some degree (by producing masking effects if severe enough and if in a vulnerable point of the Fletcher Munson curve, i.e. in lower and mid bass where the curve still slopes down heavily), but as steady state measurements actually only tell you what tones the driver will produce which were *not* in the original signal. Resolution means, how well does the driver render tones which *were* in the original signal. Specifically we talk about time domain here, not frequency domain. As such, for instance, the driver with higher resolution may have the same THD and IMD, and a *less* even FR, than another driver with less resolution. Why? because an even FR may indicate a smoothening out through woolly (inexact) rendering, due to high energy storage. Besides, FR deals with steady state response in the frequency domain.

And I really don't see much info on manufacturer's specs regarding readily comparable data on time domain performance decay (energy storage). The best I have seen are Siegfried Linkwitz's measurements (and comments!) on www.linkwitzlab.com, for instance

http://www.linkwitzlab.com/frontiers-2-M2.htm

and

http://www.linkwitzlab.com/mid_dist.htm

If you take the time to peruse that material, you may come to a more nuanced view on judging drivers on the basis of a few data.

[Bill F.]

In addition to muddying the discussion by being another Bill F., I'd like to caution those who are eager to compare different drivers for the sake of determining the relative merits of their magnet types.

Unless the two drivers being compared are otherwise absolutely identical, I believe the sonic signature of the magnet type will be so buried beneath other differences that comparison would be silly.

Even where T/S parameters are identical, subtly different diaphragm/suspension systems can have very different sonic signatures.

Setting that aside and looking just at the motor, a lot of sonically significant things happen at the gap/VC interface. The topology (over/underhung), the ratio of VC height to gap height, inductance modulation, pole and top-plate shape and dimensions, the presence and position of shorting rings, the gap flux-density gradient from top to bottom, etc., etc. All probably have quite a bit more to do with the final sound than magnet type.

Even comparing drivers from the same manufacturer would probably be pointless since motor components likely aren't the same.

[pinkmouse]

The 10" mid driver, (PD sourced, I think, but it could have been Eminence) used in older Turbosound PA systems was replaced with a new model with a neodymium magnet several years ago. It was an exact replacement, and sounded no different to the older alnico model, but was just smaller and lighter.

[thoriated]

I have nothing 'against' neodymium, and if I were a speaker designer, I would probably use it in preference to alnico for some designs where maximum flux/magnet volume and/or light weight were the overriding considerations. But it simply is not as temperature insensitive as alnico. And if the overwhelming indictment against alnico is supposed to be its low coercivity, why not go to samarium cobalt which doesn't have that problem or the particular weaknesses of neo?

[pinkmouse]

Originally posted by thoriated
But it simply is not as temperature insensitive as alnico.

So all those PA companies that use neo magnet drivers, usually running at near 100% duty rating, night after night, end up with demagnetised drivers?

[thoriated]

So all those PA companies that use neo magnet drivers, usually running at near 100% duty rating, night after night, end up with demagnetised drivers?

They sure do, until they cool down. An additional nonlinearity (even in smaller amounts) you would be better off without when high quality is your prime objective, IMO. None of which is meant to imply that said nonlinearity will in any way make neo speakers sound 'bad'. Probably the effect would be mild soft euphony and a degradation in image stability and loss of small detail in dynamic material. Again, if you insist on the argument that one must need massively overpower his speakers, therefore he mustn't use alnico drivers, I strongly disagree.

[Bill F.]

Perhaps alnico's demagnetization threshold makes it a good candidate for a HyperDynamics dual-gap topology. The counter-wound VCs cancel each other's inductive "push," so (unless I'm having a Friday moment) I believe the magnet's field should remain undisturbed at any VC power level. I find this a facinating topology.

[Pan]

MBK,

please read my first couple of post cerfully once again and I think you get what I tried to say. I´ll get back later, dinner is served.

/Peter

[DanWiggins]

Originally posted by thoriated


They sure do, until they cool down. An additional nonlinearity (even in smaller amounts) you would be better off without when high quality is your prime objective, IMO. None of which is meant to imply that said nonlinearity will in any way make neo speakers sound 'bad'. Probably the effect would be mild soft euphony and a degradation in image stability and loss of small detail in dynamic material. Again, if you insist on the argument that one must need massively overpower his speakers, therefore he mustn't use alnico drivers, I strongly disagree.

I'd love to see some data where the magnet structure reached 70 or 80 deg C, yet the voice coil still holds together, and the increase in DCR from thermal heating doesn't completely swamp the effect from magnet heating.

By the time a voice coil can generate enough heat to raise the magnet to 150+ deg C, it's well above 200 deg C itself, and possibly above 240 deg C, thereby setting the cone on fire (yes, I have seen that happen as well! Amazing what happens when one gets a little crazy with a QSC PLX3402 on a speaker - 3400W will melt down and literally catch a speaker on fire in a hurry).

The voice coil will exhibit a LOT more sensitivity loss than the magnet, at a quicker rate as well. Thermal mass of a typical pro-sound driver is extremely high; take a 6 pound magnet, add 8 pounds of steel. That's a LOT of mass (and a LOT of radiating area) to heat up. The little 40 gram voice coil will heat a lot quicker and to a much higher degree.

I've thermal-stressed drivers by the hundreds in our test lab. I've gotten motors to the temperature where they are too hot to keep your hand on them for more than 4-5 seconds (around 60 deg C), but never been able to push one past that without the voice coil dying. Even neo units.

Neo is making a BIG push into PA use because it offers low weight, small size (easier to pack multiple drivers into a given box size), and more than enough reliability - even SH grade. Ferrite is still choice number 2 because of cost, ease of assembly (everyone's lines are set up to build with it) and reliability. AlNiCo really comes up short in price and coercivity. And SmCo? That's REALLY pricey, and nearly unobtainable - it's virtually unused in the loudspeaker industry. It's not as strong as neo, and its big advantage in thermal issues aren't that big of a deal in reality. That's why it's pretty much limited to mil/aero uses where operation over extreme temperature ranges is needed.

IMHO, if you have a budget and aren't concerned about weight, go ferrite. There's really not much better in terms of price/performance. If weight or physical size is a concern, then neo. AlNiCo should be reserved for nostalgic use or where you need extremely high temperature operation (such as in industrial motors - nothing you'd get in a speaker).

Dan Wiggins
Adire Audio