Woofer noise, VC rubbing or something else?

[TLS Guy]

A majority of VC are overhung in most consumer, and lots of pro gear. You seem to make a lot of assumptions on the woofers based off of a crappy phone video that doesn’t provide any objective data.

If the drivers had poor linear excursion, there’s no way the 5.25” driver in the smaller speakers could reproduce 100dB at 50hz with 3.6% distortion at a distance of 11’.

Yes, Overhung voice coils are used a lot but should not be. Again it comes to quality and price. The fact is that overhung voice coils, especially the one shown have problems with heat and thermal compression.

The VC out of the gap, does not have any heat sinking. This makes it more prone to burnout and thermal compression.

I believe that the complete VC winding should stay in the gap during the whole of linear travel.
This alone does not make for a uniform magnetic field with travel. However modern techniques use methods like copper rings above and below the coil to keep the magnetic field linear.

This is how it is accomplished in the drivers I use: -

By using two heavy copper rings fitted above and below the magnet gap defined by a T-shaped pole piece, which was press-fit into a bumped back plate. To further enhance the heat transfer capability, a solid copper phase plug was fitted to the top of the upper ring. The stationary phase plug replaces a conventional dust cap and thereby eliminates the acoustic resonator behind the dust cap. At the same time, the excellent thermal conductivity of the phase plug aids tremendously in heat dissipation, while the air movement from the cone over the phase plug also serves to cool the motor.

My Dynaudio drivers are also underslung. When working on the JWs many years ago, we were all convinced they needed to be underslung. There is a lot more to building a good driver then meets the eye, and getting the motor system optimized is crucial.

It seems to me that there is a lot of coil out of the gap on those drivers. One certainly wonders in that event if ding your testing the former did not leave the gap and incur some damage.

There really is a world of difference in the sound quality of run of the mill drivers, and those designed using best practice.

 

[shadyJ]

Yes, Overhung voice coils are used a lot but should not be. Again it comes to quality and price. The fact is that overhung voice coils, especially the one shown have problems with heat and thermal compression.

The VC out of the gap, does not have any heat sinking. This makes it more prone to burnout and thermal compression.

I believe that the complete VC winding should stay in the gap during the whole of linear travel.
This alone does not make for a uniform magnetic field with travel. However modern techniques use methods like copper rings above and below the coil to keep the magnetic field linear.

This is how it is accomplished in the drivers I use: -

By using two heavy copper rings fitted above and below the magnet gap defined by a T-shaped pole piece, which was press-fit into a bumped back plate. To further enhance the heat transfer capability, a solid copper phase plug was fitted to the top of the upper ring. The stationary phase plug replaces a conventional dust cap and thereby eliminates the acoustic resonator behind the dust cap. At the same time, the excellent thermal conductivity of the phase plug aids tremendously in heat dissipation, while the air movement from the cone over the phase plug also serves to cool the motor.

My Dynaudio drivers are also underslung. When working on the JWs many years ago, we were all convinced they needed to be underslung. There is a lot more to building a good driver then meets the eye, and getting the motor system optimized is crucial.

It seems to me that there is a lot of coil out of the gap on those drivers. One certainly wonders in that event if ding your testing the former did not leave the gap and incur some damage.

There really is a world of difference in the sound quality of run of the mill drivers, and those designed using best practice.

Underhung drivers are a great solution if money is of no consequence. Sadly, underhung needs a very powerful magnetic field to have a good amount of flux throughout the entire gap. That means a very large ferrite magnet or neodymium. That means lots of weight and lots of $$. It is nice in theory, but there are plenty of linear excursion graphs for overhung drivers that have excellent results as well. I would not get 'hung' up on coil/gap topologies (BOOM! pun INTENDED!)

You mention overhung drivers having problems with thermal dissipation, but underhung is even worse in this respect, since they typically have a smaller, lighter coil.

What matters most is not design preferences but measurable results. Design topologies are only a means to an end, the end being high-fidelity audio.

 

[TLS Guy]

Underhung drivers are a great solution if money is of no consequence. Sadly, underhung needs a very powerful magnetic field to have a good amount of flux throughout the entire gap. That means a very large ferrite magnet or neodymium. That means lots of weight and lots of $$. It is nice in theory, but there are plenty of linear excursion graphs for overhung drivers that have excellent results as well. I would not get 'hung' up on coil/gap topologies (BOOM! pun INTENDED!)

You mention overhung drivers having problems with thermal dissipation, but underhung is even worse in this respect, since they typically have a smaller, lighter coil.

What matters most is not design preferences but measurable results. Design topologies are only a means to an end, the end being high-fidelity audio.

True you need a higher flux density. However the good transducer designers I have talked to have a strong preference for underhung drivers. The VC does not have to be short, and the general wisdom is that heat transfer is better with underhung drivers. The SEAS EXCEL drivers in particular are really well engineered and well worth the money.

 

[yepimonfire]

Yes, Overhung voice coils are used a lot but should not be. Again it comes to quality and price. The fact is that overhung voice coils, especially the one shown have problems with heat and thermal compression.

The VC out of the gap, does not have any heat sinking. This makes it more prone to burnout and thermal compression.

I believe that the complete VC winding should stay in the gap during the whole of linear travel.
This alone does not make for a uniform magnetic field with travel. However modern techniques use methods like copper rings above and below the coil to keep the magnetic field linear.

This is how it is accomplished in the drivers I use: -

By using two heavy copper rings fitted above and below the magnet gap defined by a T-shaped pole piece, which was press-fit into a bumped back plate. To further enhance the heat transfer capability, a solid copper phase plug was fitted to the top of the upper ring. The stationary phase plug replaces a conventional dust cap and thereby eliminates the acoustic resonator behind the dust cap. At the same time, the excellent thermal conductivity of the phase plug aids tremendously in heat dissipation, while the air movement from the cone over the phase plug also serves to cool the motor.

My Dynaudio drivers are also underslung. When working on the JWs many years ago, we were all convinced they needed to be underslung. There is a lot more to building a good driver then meets the eye, and getting the motor system optimized is crucial.

It seems to me that there is a lot of coil out of the gap on those drivers. One certainly wonders in that event if ding your testing the former did not leave the gap and incur some damage.

There really is a world of difference in the sound quality of run of the mill drivers, and those designed using best practice.

An overhung voice coil should remain linear so long as there is an equal length of windings in the gap as the depth of the gap. Xmax for an overhung VC is given by VC length - magnetic gap / 2, while an underhung VC is given by VC length/2. The issue with underhung voice coils is they’re not practical for drivers designed with high xmax (like subwoofers or bigger woofers with lower fs). You lose some sensitivity and the cost to performance benefit (due to the fact you need a substantially larger magnet), so an underhung VC doesn’t make sense for a pair of $600 speakers designed with a 45hz f3 and efficiency in mind.

Based on my own testing, those drivers remain linear at low frequencies (~45hz-100hz) until xmax is exceeded, at which point distortion jumps from around 3% to 10% or higher. With the RP-150m woofers, that 100dB measurement was at the edge of xmax. An increase in 2dB takes the distortion measurements from 3.6% to 12% or worse, and it’s definitely audible at that level.

Off the top of my head, I don’t remember if the woofers use shorting rings, I do know the new drivers in the Reference Premier series replaced the aluminum former with titanium. This gives the benefit of the rigidity and thermal conductivity of aluminum, without the issues introduced by eddy currents in al VC formers, since titanium is a poor conductor. The thermal expansion of titanium is also half that of aluminum. There is no indication that these drivers were designed poorly. Designed to achieve the best performance (low distortion, high efficiency, and decent bass extension) within the price range of the speakers? Sure, but that doesn’t mean it’s a crappy, cheap off the shelf Chinese woofer.

Like everything in the world of audio reproduction, everything involves a trade off.

Btw the speaker that bottomed out was the 150m, these are brand new.

One thing I’m curious about is the visible VC assembly, this is the first driver I’ve seen that has the VC/pole piece assembly visible, and I’m wondering if it was designed that way to increase cooling. None of these speakers have an impressive power handling rating (75w for the 150m and 100w for the 160m), but it could also just be that the power rating given is more honest. I believe I have read from some of the employees over on the Klipsch board that their power handling figures are derived using pink noise with a 6dB crest factor, hence the 100w continuous/400w peak rating.

 

[yepimonfire]

+1 this. Sensitivity measurements can vary wildly depending on methodology. There isn't any widely used standardized way of taking this measurement, so it can go all over the map, even for the same speaker. I look at it as a relative measurement; I would only really use one tester's data against that tester's other results.

True. I believe S&V measures sensitivity from 500hz-1khz. The owm3 has less of an impedance dip in the mid range vs the Klipsch, which display their lowest impedance from about 100hz to 1khz. Measuring the 150m from 500-1khz does nothing but give you the average sensitivity of the woofer, which is obviously going to be lower in the ~4ohm region of the woofer vs the horn loaded tweeter or higher impedance (100hz-60hz) mid bass region of the woofer.

I don’t remember who it was that said it now, but there was one guy who frequently tests speakers who advocated that a 2.83v sensitivity rating actually made more sense than a 1w rating, since solid state amplifiers are a voltage source, not current source. Of course, many cheaper SS amps are limited in their current output, but a 2.83v rating tells you that if you apply 2.83v to a speaker, it will play this loudly. Impedance varies so wildly among speakers that you can’t accurately give a 1w rating. If you tested it from 50hz-100hz, or 2khz-20khz for example, you’d get something like 6dB+ due to the impedance spike. I suppose one could use pink noise, but that never gives a constant spl reading.

I’ve also wondered if it wouldn’t make more sense to rate amplifiers by rms voltage and continuous/peak current output vs watts. If an amp was rated for 28v rms with a minimum continuous current of say, 14A instead of 100wpc @ 8 ohms or 200w @ 4 ohms, one could easily figure out how much power an amp could provide into any load across any number of channels. A 28v rms amplifier with a minimum continuous current output of 14A could provide 58w into 5 8ohm speakers, or 100w into two.

 

[shadyJ]

True. I believe S&V measures sensitivity from 500hz-1khz. The owm3 has less of an impedance dip in the mid range vs the Klipsch, which display their lowest impedance from about 100hz to 1khz. Measuring the 150m from 500-1khz does nothing but give you the average sensitivity of the woofer, which is obviously going to be lower in the ~4ohm region of the woofer vs the horn loaded tweeter or higher impedance (100hz-60hz) mid bass region of the woofer.

I don’t remember who it was that said it now, but there was one guy who frequently tests speakers who advocated that a 2.83v sensitivity rating actually made more sense than a 1w rating, since solid state amplifiers are a voltage source, not current source. Of course, many cheaper SS amps are limited in their current output, but a 2.83v rating tells you that if you apply 2.83v to a speaker, it will play this loudly. Impedance varies so wildly among speakers that you can’t accurately give a 1w rating. If you tested it from 50hz-100hz, or 2khz-20khz for example, you’d get something like 6dB+ due to the impedance spike. I suppose one could use pink noise, but that never gives a constant spl reading.

I’ve also wondered if it wouldn’t make more sense to rate amplifiers by rms voltage and continuous/peak current output vs watts. If an amp was rated for 28v rms with a minimum continuous current of say, 14A instead of 100wpc @ 8 ohms or 200w @ 4 ohms, one could easily figure out how much power an amp could provide into any load across any number of channels. A 28v rms amplifier with a minimum continuous current output of 14A could provide 58w into 5 8ohm speakers, or 100w into two.

Almost everyone who really tests speaker sensitivity does so at 2.83v. Only manufacturer specs will state the 1w rating, and of course, that means very little. Harman has the best way to state sensitivity that I know of, 2.83v at 1 m in an anechoic and averaged over a bunch of different frequencies. That is the no nonsense way to specify sensitivity.

 

[yepimonfire]

Almost everyone who really tests speaker sensitivity does so at 2.83v. Only manufacturer specs will state the 1w rating, and of course, that means very little. Harman has the best way to state sensitivity that I know of, 2.83v at 1 m in an anechoic and averaged over a bunch of different frequencies. That is the no nonsense way to specify sensitivity.

Be nice if the FTC stepped in and updated their requirements for claimed specs. I’d at least wish manufacturers were required to disclose how they got their number, ie half space of full space/ anechoic. Some do, most don’t. All of Klipsch’s speakers are half space or (in room) interpolations, for example, they measure the speakers in an anechoic chamber and then add 4dB to that to account for the room. They don’t disclose that, which is why their sensitivity ratings look unrealistic. Yes, most of their speakers are anywhere from 3-6dB more efficient than comparable direct radiating speakers, but a 96dB sensitivity rating for a 6.5” bookshelf just seems impossible. Even the fusion 10 with a 10” driver and a compression tweeter only gets up to 94 dB. Subtracting 4dB to yield 92dB gives a much more realistic number, but if you don’t regularly research this stuff, the average consumer wouldn’t know that.

 

[highfigh]

Be nice if the FTC stepped in and updated their requirements for claimed specs. I’d at least wish manufacturers were required to disclose how they got their number, ie half space of full space/ anechoic. Some do, most don’t. All of Klipsch’s speakers are half space or (in room) interpolations, for example, they measure the speakers in an anechoic chamber and then add 4dB to that to account for the room. They don’t disclose that, which is why their sensitivity ratings look unrealistic. Yes, most of their speakers are anywhere from 3-6dB more efficient than comparable direct radiating speakers, but a 96dB sensitivity rating for a 6.5” bookshelf just seems impossible. Even the fusion 10 with a 10” driver and a compression tweeter only gets up to 94 dB. Subtracting 4dB to yield 92dB gives a much more realistic number, but if you don’t regularly research this stuff, the average consumer wouldn’t know that.

With the lack of funding for some agencies, don't hold your breath. This isn't going to be seen as 'important' by Congress OR the FTC.

If consumers were told about the fudge factor, what would they do with that knowledge? I think you're over-estimating the number of people who actually consider this when buying equipment.

 

[highfigh]

Almost everyone who really tests speaker sensitivity does so at 2.83v. Only manufacturer specs will state the 1w rating, and of course, that means very little. Harman has the best way to state sensitivity that I know of, 2.83v at 1 m in an anechoic and averaged over a bunch of different frequencies. That is the no nonsense way to specify sensitivity.

Except for power testing, which usually DOES use pink noise.

 

[everettT]

Waiting for more numbers vs real world experience .

 

[yepimonfire]

With the lack of funding for some agencies, don't hold your breath. This isn't going to be seen as 'important' by Congress OR the FTC.

If consumers were told about the fudge factor, what would they do with that knowledge? I think you're over-estimating the number of people who actually consider this when buying equipment.

Personally, I’m a numbers and data person. While the real world translation obviously introduces variables that slightly change the theoretical predictions, it still gets you close enough.

Honestly, I think it’s even more important that your average poorly informed “Best Buy” shopper be given legitimate figures with which to make a purchasing decision. I still have the box for a Sony STR-DH550, and it clearly states in big numbers on the front that it’s 140w x5. Of course this figure is at 6 ohms on the brink of clipping. Not surprisingly, Best Buy takes the 140w figure and multiplies it times 5 and lists it on their website as a “700w receiver”, even though there is no way in hell its capable of delivering that. Real world values are 90w 2ch at 6 ohms and about 67w at 8ohms.

 

[yepimonfire]

Except for power testing, which usually DOES use pink noise.

We hope

In pro audio, there are certain standards for testing things like power handling, sensitivity, max spl etc. in consumer audio, for all we know, the manufacturer could have just read the power handling specs included in the driver specs and added them together to get that number, ie, a 70w rated woofer and 15w tweeter equating to 85w.

 

[highfigh]

Personally, I’m a numbers and data person. While the real world translation obviously introduces variables that slightly change the theoretical predictions, it still gets you close enough.

Honestly, I think it’s even more important that your average poorly informed “Best Buy” shopper be given legitimate figures with which to make a purchasing decision. I still have the box for a Sony STR-DH550, and it clearly states in big numbers on the front that it’s 140w x5. Of course this figure is at 6 ohms on the brink of clipping. Not surprisingly, Best Buy takes the 140w figure and multiplies it times 5 and lists it on their website as a “700w receiver”, even though there is no way in hell its capable of delivering that. Real world values are 90w 2ch at 6 ohms and about 67w at 8ohms.

I am, too- that's why I'm so annoyed that the power ratings are stated in the way you posted WRT Beast Buy- it impresses the ignorant.

As far as pro- the reason they show a lot more than most people will ever need to know is due to the fact that if someone bases their choice on the specs and it doesn't work out, someone's job is on the line. Manufacturers can't afford to have equipment returned because it goes t&ts up in the field, from the standpoint of financial cost for repairs or the damage it would do to their reputation.

I have had Sony integrated amps that were the opposite of your experience, but they were made in the '70s and '80s. The first was rated at 30W/ch and when the magazines tested them (not the same piece, but the same model), they always output 56W/ch at rated distortion, bandwidth, etc. These were rated for far better than 20Hz-20KHz, the next model up (in the link) is rated for 2Hz-100KHz, -2dB.

http://www.thevintageknob.org/sony-TA-5650.html

I had their TA-F6B, too- the specs shown in the link are wrong and very incomplete but when I had mine tested on a BPI distortion analyzer into 8 Ohm load, it output 176W/ch, wide band. IIRC, it's bandwidth was .1Hz-200KHz/+0, -1dB.

http://www.thevintageknob.org/sony-TA-F6B.html

I had the tuner in the next link- that thing was amazing! 34 pounds, FM only. Both of the last two models had been on display at the Consumer Electronics Show in Chicago.

http://www.thevintageknob.org/sony-ST-A7B.html

 

[Johnny2Bad]

If it is heard on Sine Waves, it's probably frequency doubling, more commonly referred to as second harmonic distortion.