The significance of high Qms..? - diyAudio
Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 31st March 2008, 05:44 PM   #1
Elbert is offline Elbert  Norway
diyAudio Member
 
Join Date: Mar 2008
Default The significance of high Qms..?

I started another thread regarding midranges, and from some of the answers I got and quite a bit of frantic reading up and investigation, I made some observations which might be interesting to bounce of the forum...

it really started with an interview with Joachim Gerhard of Audio Physics, posted on Troels Gravesens web-page: http://www.speakerbuilding.com/content/1039/


Here Mr Gerhard comments that many new drivers are very heavily damped in order to give flat frequency response, something which is characterized by low Qms. He then goes on to say that many of the older drivers with much less mechanical damping and high Qms were much more dynamic.

In my previous midrange thread, a surprising many people recommended using pro-drivers for good dynamics and response.
Looking at the data for many of these drivers, I found that most of them had very high Qms, typically 4-8, whilst msot Hi-fi drivers seem to be in the 1-1,5 range.

What I also found was that most of the pro-drivers seemed to have rather ragged frequency response, perhaps a result of poor damping?

It is within this context I find two drivers very interesting as candidates for 200-3000Hz midranges for my active speaker project;

The Vifa WN 17 255 suggested to me by Calvin and the Focal 6W 4311B, which I found on the Zalytron web page.

Looking at the efficiency, suspension compliance and the Qms of these drivers, they seem to have relatively low mechanical damping, yet the frequency plot is very flat, at least compared to a lot of the aforementioned pro-drivers.

The Focal driver has a Qms of over 8, a number many times higher than for any other hi-fi unit I have seen. The frequency response is perhaps not as smooth as that of the Vifa unit, but still looks OK.

So, what I'm thinking is that these drivers represents examples of drivers which are actually reasonably linear without relying to heavily on mechanical damping, which to the best of my understanding is conductive to good dynamics and transient response.. (?)

Or am I making totally stupid conclusions here??


Of course, drivers with low mechanical damping will still need to be controlled somehow if they are not to resonate and flap about uncontrolled so to speak, but an amplifier with good damping factor should contribute towards this i suppose?

Anyway, some viewpoints,and by all means corrections, on this topic would be very interesting!
  Reply With Quote
Old 1st April 2008, 02:01 AM   #2
diyAudio Member
 
Join Date: Feb 2002
Location: Western Sydney
Usually, I just look at Qts, beacuse it is what is used for simulators & and is a combination of Qms & Qes,
Qts = Qms * Qes / (Qms + Qes)
and I think this is what most people look at when deciding the suitability of a driver for a particular task.
However, I think a discussion of how the individual Qs affect the performance would be interesting and enlightening.

From http://web.media.mit.edu/~meyers/mcg...udspeakers.pdf
(looks like a not-to-well-written undergrad paper, but it’s a starting point)
“Electrical damping relies on magnets to act as the restoring force in the speaker. Cheap, small magnets result in multiple resonance’s in the diaphragm because they are too slow in restoring the vibrating mechanism to its equilibrium point. This causes a poor transient response”

http://www.speakerplans.com/index.php?id=faq1 says “low values of Qts give a tight and punchy sound but with little weight or low bass and high Qts values give a slow and heavy sound that will give you lots of low frequency output”, yet ” Drivers with a very high mechanical Q can sound more open, cleaner and have a better dynamic range. This is because they have less loss. The surround is more flexible, the spider is better constructed, they have better air flow and usually have higher sensitivity. So a high mechanical Q is a very good indicator of energy storage behaviour”

I wonder how real these generalizations are in practise, and what they’re leaving out

more discussion here: http://www.diyaudio.com/forums/showthread/t-44570.html
as Ron E says in that discussion, there is an apparent contradiction.

I’m thinking that High Qms drivers will resonate more because the damping is not well controlled, i.e. the output does not as accurately follow the input as lower Qms drivers because of the suspension. (Of course, some may prefer that...). I suppose there must be a point of compromise between lack of damping and over damping, and the box needs to be taken into account...???
__________________
Impedance varies with frequency, use impedance plots of your drivers and make crossover calculations using the actual impedance of the driver at the crossover frequency
  Reply With Quote
Old 1st April 2008, 06:27 PM   #3
Elbert is offline Elbert  Norway
diyAudio Member
 
Join Date: Mar 2008
Pete,

Some very interesting links you posted, enjoyed reading those and it certainly inspires further investigation.

As I don't really understand the Thiele Small parameters properly (yet), this is all fairly speculative yet on my account, but it is interesting to try and make some deductions.

Hopefully, other more knowledgeable forum members will chime in..

The fundamental concept of damping and harmonics is fairly understandable concepts.

If there is no or little damping, mechanical swinging and uncontrolled oscillation is free to occur, and in the context of loudspeakers, this will result in cone motion not called for by the input signal.

This can of course never be good, so some degree of damping is obviously a must.

But can we get to much damping, or damping of the wrong kind?

Lets think of a car suspension. There is a spring for allowing motion of the wheels independent of the car body. If the dampers are bad or missing, the car will bounce uncontrolled after the suspension has absorbed energy after hitting a bump.
So we need dampers to absorb and control this excessive energy so the suspension can be brought back to its normal state.

But what will happen if we apply to much damping? The dampers will then prevent the suspension from responding quickly enough, thereby allowing the energy of a sudden bump to pass straight to the chassis of the car. That's why a sportier and more firm/ stable suspension often gives a harsh ride.

So what would be ideal then? That would be a damper that was firm enough to give stability over undulating road surface, yet compliant enough to respond quickly to sharp bumps and minor road imperfections.

In real life, there is often a compromise involved. The ideal would of course be a suspension system that was adaptive or active. (and these do exist, the damping is actively and continuously adjusted to suit the real-time requirements)

It is also known that cars, and in particular motorbikes, with relatively heavy wheels and suspension components (unsprung mass) place higher demands on the suspension, it is more difficult to damp the motion of a heavy wheel than a light one.

So we need damping, I'm not gonna argue against that.

But what I find interesting is that we have two components in the damping equation, Qms and Qes.

As I have understood it Qms is mechanical damping in the form of suspension stiffness and inherent energy absorbing qualities of the cone and its suspension parts. and the higher the Qms value, the less these factors contribute to dampen the cone and bring it back to its neutral state.

then there is Qes. Now, this one is a bit harder for me to grasp at the moment. As far as I understand, this describes the degree of control the motor can exert on the cone. This degree of control is a function of the strength of the motor BL, and the mass it has to control, lower mass and higher BL equals greater motor control.

The total Q, Qts, is a resultant of Qms and Qes. This means that two drivers with similar Qts, can have any variation of ratio between Qes and Qms. The Qts is important for selecting the correct enclosure, as the system damping is the sum of the acoustical damping of the box and the total damping built in to the driver it self.

but what about the ratio between Qes and Qms?

What would the ideal loudspeaker look like compared to the car suspension?

From what I have understood so far, I'm inclined to compare the mechanical (or passive) damping of a speaker with the shock absorber of a car. a degree of passive damping would reduce oscillations and resonant modes, but at the same time resist sudden, i.e. transient responses. The fact that many linear and heavily damped (to damp out resonances and achieve linear frequency response) are often described as dull and engaging is worth considering.
I have also listened to small heavily damped studio monitors with +/-1 dB frequency response and they DID sound terribly dull although the tonal balance was perfect.

Too little mechanical damping, well, there's the church bell...

But what about the electrical damping then?

A weak motor might be able to push the cone in to motion, but it might be hard pressed to exert enough force to overcome the momentum of the cone and reverse it as the polarity of the signal waveform reverses. this will become especially noticeable if there is relatively little mechanical damping to brake the "runaway" cone.

But what if the motor is strong?
Ideally, the motor should be powerful enough to exert absolute control of the cone, an electromagnetic fist slaving it to the signal waveform, be it a "slow" sinuous low frequency or a sudden transient detail.

In my mind this sounds like the active damping I outlined for the car suspension.

Now, I've so far assumed the cone being a "perfect" piston without any internal resonances. All cones and diaphragms have internal resonances and these are obviously beyond the control of the damping a strong motor can provide. Hence any cone resonances must be taken care of by mechanical damping inherent in the cone material, or the driver must be operated outside the frequency range where these resonances occur. (like with some metal cones)

But back to Qes....

A strong electric damping not relying on efficiency sapping and transient flattening mechanical damping must be very good then?

Surely, this is to simple to be true, or is it?

In my somewhat simplistic view of this issue, I can only find one culprit.

Ultimately, the control (and damping) provided by the motor system, depends on the amplifier and its damping factor.

What is the damping factor of an amplifier? To my limited knowledge, it is the amplifiers ability to exert control over the driver.

As current is fed in to the voice coil of a driver, it sets up an electromagnetic force in the coil which reacts with the permanent magnetic force of the motor magnet, thereby causing the coil to move. Fine.

But as the voice-coil moves in the magnet gap of the motor, the opposite also takes place, the permanent magnetic force induces current in the voice coil! Just like a dynamo. The amplifier will have to counter this "back-current", and its ability to do so determines how well the amplifier controls the motion of the voice coil.

Try this experiment.
With your amplifier switched of, gently tap the cone of the woofer with your fingertip and listen carefully.
Now switch the amplifier on and try again. You are likely to notice that the tapping sound is now much more damped. this is your amp trying to keep the voice coil in position against your "input" like it would also try to counter the unwanted "input" originating from resonances and uncontrolled cone-motion.

I was quite fascinated the first time i tried this!









This could perhaps go towards explaining why so many manufacturers build drivers with relatively high Qes and low Qms.
A driver with a high Qms would rely highly on good amplifier control and damping in order to work well, i.e. rely on high quality amplifiers, beefy low resistance speaker cables and crossovers which would not compromise the damping factor of the amp.
This means that if you and I, the consumer, buy a mediocre amp and spindly speaker wires, the loudspeaker manufacturters otherwise excellent product could end up sounding s**t.

The other way around, in a active speaker where a high quality amp module (which is of course what we build anyway), hooked straight up to the driver with a short length of beefy wire and no x-over network with nasty inductors and resistors, a high Qms/ low Qes driver might provide the best compromise! (?)

From my rambling deductions and other snippets of info found on the web so far, this is the direction I seem to be heading at the moment.

So unless someone comes along and puts me right, I am for the time being very much inclined to go out and source a midrange driver with very low moving mass and high Qms for my 3-way active speaker project!
  Reply With Quote
Old 1st April 2008, 10:09 PM   #4
diyAudio Member
 
Join Date: Feb 2002
Location: Western Sydney
Elbert,
While I think you’re correct, most modern SS amplifiers have damping factors low enough that it doesn’t matter. Active xovers are probably better because the effects of back EMF are negated, but active xovers still cause phase variation. And low mass cones will respond faster, but at the expense of mechanical strength, i.e. less rigidity > higher distortion.

Reading around the net, (and this is getting away from Q), for best transient response the driver needs to be able to reproduce a square wave as close as possible – for this, flat frequency response and flat phase response are necessary. This means the effects of the box Q, crossover and time alignment need to be taken into account, although there is some debate about the audibility of time alignment: http://sound.westhost.com/ptd.htm

Ignoring the external factors, perhaps the best we can hope for is to buy the most linear motor we can afford, with a frequency response as linear as possible outside the passband of interest? In the end it probably comes down to personal preferences and the compromises one is prepared to make e.g. metal cone vs. paper...
__________________
Impedance varies with frequency, use impedance plots of your drivers and make crossover calculations using the actual impedance of the driver at the crossover frequency
  Reply With Quote
Old 1st April 2008, 10:33 PM   #5
Elbert is offline Elbert  Norway
diyAudio Member
 
Join Date: Mar 2008
.. and reasoning one self out of a position one has reasoned oneself in to might be even harder!

You probably have a good point about most amplifiers having a good enough damping factor, so that might not be such a significant point after all..

However, I had one very interesting experience once...

In a studio, we had a pair of (i believe it was acoustic energy) 2-way monitors, vented boxes with an 8 Seas driver and a seas aluminum dome.

Very nice speakers, in fact better sounding than a pair of active Meyersounds we borrowed. (the meyers were about 10 times as expensive)

We ran the monitors of a fairly standard Sony amp, and were then we were given a Crown 300 studio amp on trial.

In terms of thd, etc., the Crown was probably not that much better than the Sony. The difference in these parameters are usually much smaller than the distortion generated in most loudspeakers anyway.

But what an amazing difference! the sound became so much tighter and detailed, not to mention dynamic! I actually found the difference quite astounding at the time.

Could it be that the hefty crown amp (you could probably use it for TIG welding) somehow just exerted better control of the drivers? Just speculating, these things are hard to quantify...

Again, interesting link, I'm probably gonna be late in bed again!

I absolutely agree on the cone weight issue, something too thin and flimsy will inevitably start to flex and play its own tune, I guess this is where frequency response plots come in handy! I would expect a driver with diaphragm of insufficient rigidity will be revealed by a ragged frequency response?

You probably have a good point about active crossovers and phase shifts and I should look closer in to this as it is something I'm not to familiar with. My assumption up until now was that you get phase shifts in both passive and active filters so that it wouldn't really make any difference at the end of the day? looking at the cost often involved with good passive filters, i figured I could just as well go for an active solution with all the other benefits that has!

Anyway, trying to get a grasp on how to select a good midrange... it's just like red wine, you just can't taste them all!
  Reply With Quote
Old 2nd April 2008, 12:59 AM   #6
diyAudio Member
 
Join Date: Feb 2002
Location: Western Sydney
"it's just like red wine, you just can't taste them all!"
- I'm giving it my best shot
(actually with speakers too - I've never tried a metal cone mid-woofer, so my next speaker will use them, when I can afford it...)
cheers,
Pete McK
__________________
Impedance varies with frequency, use impedance plots of your drivers and make crossover calculations using the actual impedance of the driver at the crossover frequency
  Reply With Quote
Old 3rd April 2008, 08:01 PM   #7
Elbert is offline Elbert  Norway
diyAudio Member
 
Join Date: Mar 2008
Taste them all or die (from liver failure) trying!

Read the last link you posted...
Although in many respects a theoretical excersise, some food for thought regarding time and phase.

My belief is that interference, due to phase shifts and reflections is inevitable, when one considers the nature of sound, it is actually unavoidable and unless excessive, a perfectly natural imperfection?

Having said that, I hope to keep the problems to a minimum by keeping the tweeter and the midrange reasonably close together (less than the wavelength of the crossover frequency as recommended in some literature), and by using electronic filters with relatively steep slopes.

But regarding the time aspect, this is something I agree on the importance of! In the 200-2000Hz area, spatial information is mainly based on the arrival time differential between the ears, and this time domain is the same where early reflections are most pronounced.

One way I intend to minimize this effect, is to use absorbing material on the baffles. I'm also going to use the new diffraction/ waveguide controlled DXT tweeter from SEAS which should minimize baffle effect on the high frequencies.

Anyway, Ordered a pair of Access 6A's (focal) from Zalytron.. with high Qms, so I guess I'll soon find out if it has anything going for it! .)
  Reply With Quote
Old 4th April 2008, 01:01 PM   #8
diyAudio Member
 
Join Date: Mar 2008
Hi Elbert,

i tend to agree, that speakers with high mechanical resistance
(mostly due to the surround material chosen) tend
to have poor dynamic behaviour, especially at low signal
levels.

20 Years ago i measured some woofers with a very resistive
surround, i think it was made from PVC .

By the way: The stiffness of the surround was very dependent
on temperature ...

The woofer had very different parameters at different signal
levels. As far as i remember fs was shifted about 25% .
At low signal levels fs was higher than at high levels near Xmax.
(How to properly design a BR cabinet for such a speaker ??? )


Such a speaker tends to sound dull at low levels and wakes
up when played loud. Speaker with high Qm are to be preferred, because of better performance at low levels. But this is just a tendency, it depends on the materials used to achieve mechanical damping.

If you build surrounds with low resistance it is more difficult to control resonant modes of the cone. For woofers of small size however this is not such a big problem, for fullrange speakers it is ...
__________________
Oliver, RFZ believer (?)
www.dipol-audio.de
  Reply With Quote
Old 4th April 2008, 03:13 PM   #9
Elbert is offline Elbert  Norway
diyAudio Member
 
Join Date: Mar 2008
Allthough a 20 year old woofer with a PVC (!) surround is perhaps a bad example of a bad example.. (uh..).. it is interesting to hear your general views and observations.

Considering the reduced ability of high Qms driver suspensions for controlling resonances, this would just mean that cone and motor design becomes even more critical, something which is perhaps more demanding and costly than just adding mechanical damping?

Really looking forward to getting my drivers now, better start building a test box so I can determine the optimum enclosure size and port dimensions (if i go for a vented enclosure).
  Reply With Quote
Old 4th April 2008, 04:15 PM   #10
diyAudio Member
 
Join Date: Mar 2008
Quote:
Originally posted by Elbert
Allthough a 20 year old woofer with a PVC (!) surround is perhaps a bad example of a bad example..
That was a misunderstanding, the woofers were new
at that point in time. But i did that measurement around
20 years ago ...

But nevertheless it IS a bad - but instructive - example.
__________________
Oliver, RFZ believer (?)
www.dipol-audio.de
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Wanted - High Power, High Voltage, Audio Frequency Transformer Manual. kimbal Tubes / Valves 4 11th May 2009 08:05 PM
Significance of CEO? Colescuttle Parts 1 24th March 2007 07:10 PM
Power Response::What is it, and what significance does it have? sardonx Multi-Way 3 6th December 2005 07:10 AM
ps cap significance Mark Hathaway Pass Labs 2 14th October 2002 10:39 AM


New To Site? Need Help?

All times are GMT. The time now is 05:47 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2