Low Q drivers are very useful in that they can be put into horns, vented boxes, APs, etc. Semi-high Q (0.4-0.8) speakers can be put into sealed boxes, TLs, TQWTs, open baffles, etc. What I don't get are High Q speakers (0.9-1+). You can't really build a proper enclosure for these, and whatever you make the system Q will never be lower than that. Considering the optimal Q is 0.707, why would anyone make a driver with Q higher than that? I feel like I'm missing something here...
Short & sweet answer: because people buy them. That literally is, it. Forget performance. Forget sound quality. Forget all that. It's irrelevant. It is purely a question of commercial manufacturing reality and profit. If people, whether they be DIYers like here, or large brands purchasing OEM units, didn't buy them, they wouldn't be made.
In the case of home and pro audio, there is something of a corner case "legitimate" application of high-Q drivers. A Butterworth third order alignment uses a high Q (= 1 IIRC) 2nd order section in series with a first order section. Many higher order alignments are similarly constructed (i.e., an under-damped 2nd order section in series with additional sections). Thus, high-Q drivers can be of use in constructing hybrid acoustic/electronic alignments.
In the case of guitar amplification, a high-Q driver can add some of the body that the guitar, er, body doesn't.
In the case of guitar amplification, a high-Q driver can add some of the body that the guitar, er, body doesn't.
It is true that any driver with a high or low Qts can be equalised to provide any arbitrary final target Q when filtered via a high pass. The B&W FST midrange driver is a prime example of this, with at least the old generation with the Kevlar cone having a Qts of around 1 and these aren't inexpensive. It isn't just midrange drivers mind you many tweeters have high Qts figures too. Most of the time it's the small neo tweeters that have very little in the way of air space. The high Qts is of very little consequence when the tweeter is correctly filtered.
If we're talking about drivers that aren't going to have a high pass fitted then it's as POOH says, they are cheap. To lower a drivers Qts you usually need to increase its magnet strength and doing so requires either more magnet material or a higher grade of material. Both of these cost money.
There is most likely an element of creating a sound with more fullness too, but if you're using the loudspeakers in an open backed enclosure, such as some TVs etc you're going to get dipole cancellation occurring at some point and theoretically the rising response could counter that but that'd require a lot of correct design elements being put together, which is unlikely when a loudspeaker of such low cost is being used.
If we're talking about drivers that aren't going to have a high pass fitted then it's as POOH says, they are cheap. To lower a drivers Qts you usually need to increase its magnet strength and doing so requires either more magnet material or a higher grade of material. Both of these cost money.
There is most likely an element of creating a sound with more fullness too, but if you're using the loudspeakers in an open backed enclosure, such as some TVs etc you're going to get dipole cancellation occurring at some point and theoretically the rising response could counter that but that'd require a lot of correct design elements being put together, which is unlikely when a loudspeaker of such low cost is being used.
Open Baffle and high Q drivers
When trying to adjust for baffle step loss, high Q drivers are useful, and often times more efficient, than lower Q drivers, when used as open baffle speakers.
I myself have not done a critical sound review of low vs high Q drivers. Thorsten Loesch apparently did such a review and found that, despite the use of smaller magnets, the sound quality of high Q drivers did not suffer, compared to low Q drivers, until the Q of the driver exceeded 1.3.
Retsel
When trying to adjust for baffle step loss, high Q drivers are useful, and often times more efficient, than lower Q drivers, when used as open baffle speakers.
I myself have not done a critical sound review of low vs high Q drivers. Thorsten Loesch apparently did such a review and found that, despite the use of smaller magnets, the sound quality of high Q drivers did not suffer, compared to low Q drivers, until the Q of the driver exceeded 1.3.
Retsel
Hi,
High Q drivers are well suited to open backed guitar speaker cabinets.
In HiFi they are old school, giving a Bass bump to speakers
designed with no BSC (baffle step compensation). They
still proliferate in the budget markets of the world.
Many budget drivers are not built for the "rich" markets.
rgds, sreten.
High Q drivers are well suited to open backed guitar speaker cabinets.
In HiFi they are old school, giving a Bass bump to speakers
designed with no BSC (baffle step compensation). They
still proliferate in the budget markets of the world.
Many budget drivers are not built for the "rich" markets.
rgds, sreten.
Flow resistance damping and clang
Hi,
a high ratio between air loading and motor strength reduces harmonic distortions ("clang"), because air load usually is more linear than electromagnetic motor. Therefore, high-Q drivers have less clang than low-Q drivers.
Low-Q drivers can become current-driven, raising Q and lowering clang, see also current-feedback amplifiers.
High-Q drivers can often be damped by flow resistance damping, in explanation providing for a low mechanical Q by letting back sound press thru wool. In order for this to work, enclosure must be a compression chamber, say smaller than lambda/2Pi.
Flow damping raises efficiency: While electric damping transforms all power loss at resonant frequency to power gain above it, mechanical damping transforms at least and most half of that.
Backdraws: For one, vegetable and to a lesser degree animal and synthetic wool changes characteristics with temperature, humidity and smoke. For another thing, if flow resistance becomes too great, it saturates. This is, why, if full level is transmitted at resonance frequency, port should be big but may be smaller, if an electric highpass runs over it.
Cut short, high-Q need not be a problem but may be a challenge or even blessing.
Uli
Hi,
a high ratio between air loading and motor strength reduces harmonic distortions ("clang"), because air load usually is more linear than electromagnetic motor. Therefore, high-Q drivers have less clang than low-Q drivers.
Low-Q drivers can become current-driven, raising Q and lowering clang, see also current-feedback amplifiers.
High-Q drivers can often be damped by flow resistance damping, in explanation providing for a low mechanical Q by letting back sound press thru wool. In order for this to work, enclosure must be a compression chamber, say smaller than lambda/2Pi.
Flow damping raises efficiency: While electric damping transforms all power loss at resonant frequency to power gain above it, mechanical damping transforms at least and most half of that.
Backdraws: For one, vegetable and to a lesser degree animal and synthetic wool changes characteristics with temperature, humidity and smoke. For another thing, if flow resistance becomes too great, it saturates. This is, why, if full level is transmitted at resonance frequency, port should be big but may be smaller, if an electric highpass runs over it.
Cut short, high-Q need not be a problem but may be a challenge or even blessing.
Uli
pretty obvious no one here has tested the old speakers of the tube era.
Almost every one I have tested (pushing 100) have just stupid high Q.
I've seen them in open back and 2 foot square cubes.
One thing nice about them is they sound better then what software claims.
Makes me wonder about speaker modeling software.
Almost every one I have tested (pushing 100) have just stupid high Q.
I've seen them in open back and 2 foot square cubes.
One thing nice about them is they sound better then what software claims.
Makes me wonder about speaker modeling software.
I am curious about your experiences testing and using these drivers. High(er) Q drivers are often characterized as underdamped, while low(er) Q drivers are characterized as overdamped (all, of course, depends on the box they are used in). Thus, SS amps are claimed to increase the damping of high Q drivers, while tube amps (which often have a higher output impedance) are claimed to balance or lessen the damping of low Q drivers.
Have you found a preference of what style of amp to use with low and high Q drivers?
Retsel
The first claim is untrue, because an amplifier with zero output impedance does not reduce Quality. In vacuum valve times electromagnetoacoustic motors rarely had high linear excursion but often depended on some source impedance to lessen clang.
An amplifier, which feeds back a voltage according to its output current to its input, has an output impedance inequal zero, that is positive impedance for negative feedback and vice versa. Such amps need a shunt resistor, which measures current.
It's not a matter of 'often characterising' -it's just engineering reality. Q = 0.5 is critically damped. By definition, a driver (note caveat) with a Q > 0.5 is underdamped, while one with a Q < 0.5 is overdamped. Simple as that.
As far as system Q is concerned, that inherently varies depending on the output Z of the amplifier and any series R present in the circuit. A low output impedance solid state amplifier cannot increase the damping of a high Q driver (or any other Q for that matter) -system Q is entirely a function of the driver's innate Q, any changes from box loading, and the series R or output impedance of the amplifier. It just isn't going to go any higher. A high output impedance amplifier or some series R can however raise the effective system Q. Which may, or may not be useful depending on the specifics of the implementation. The old Fostex FExx6E series (and presumably their En replacements) for example was specifically designed on the assumption that they would be used on the end of high output impedance amplifiers -roughly in the 2ohm - 4ohm category, or with appropriate changes to provide similar results. The same applies to Lowthers. Vintage drivers with a high Q were largely either cheap, and / or intended for use in open-back boxes where said was necessary if you wanted to have any LF output. A significant number though had massive magnetic power & a low Q because they were specifically intended to be used on the end of the variable output impedance amplifiers that were popular at the time. Sadly that has resulted in a large amount of myth about xyz being 'required' because that was how they did it in the days of yore -forgetting the rather significant point of what they were intended to be driven by.
As far as system Q is concerned, that inherently varies depending on the output Z of the amplifier and any series R present in the circuit. A low output impedance solid state amplifier cannot increase the damping of a high Q driver (or any other Q for that matter) -system Q is entirely a function of the driver's innate Q, any changes from box loading, and the series R or output impedance of the amplifier. It just isn't going to go any higher. A high output impedance amplifier or some series R can however raise the effective system Q. Which may, or may not be useful depending on the specifics of the implementation. The old Fostex FExx6E series (and presumably their En replacements) for example was specifically designed on the assumption that they would be used on the end of high output impedance amplifiers -roughly in the 2ohm - 4ohm category, or with appropriate changes to provide similar results. The same applies to Lowthers. Vintage drivers with a high Q were largely either cheap, and / or intended for use in open-back boxes where said was necessary if you wanted to have any LF output. A significant number though had massive magnetic power & a low Q because they were specifically intended to be used on the end of the variable output impedance amplifiers that were popular at the time. Sadly that has resulted in a large amount of myth about xyz being 'required' because that was how they did it in the days of yore -forgetting the rather significant point of what they were intended to be driven by.
Last edited:
You seem to feel that critical damping is an ideal value for the damping design parameter.
Why?
Ben
Any serious sports car (or motorcycle) enthusiasts out there? Think about "unsprung weight" and this whole discussion will become clear.
Critical damping (Q = 0.5) offers the maximum extension with no time-domain peaking, ringing, what-have-you. Some feel this is a useful design goal. Which goes to show that for every Q, there is an admirer.
I think you may need to read my post again. What I said is that a Q of 0.5 is critically damped, a Q < 0.5 is over-damped and a Q > 0.5 is underdamped. That is engineering fact, there is no opinion about it. What provides the optimum varies according to the specifics of implementation & what you are doing.
I've read that statement too. It's twaddle. I note with interest that the site in question (I assume it's the one I'm thinking about, as I haven't run across it anywhere else thank goodness) neglects to provide any supporting evidence of any kind whatsoever. 
I suspect they're using the peaking to try to prop the midbass up a bit.
Re your success with a 0.88 Q unit, I don't doubt it -but how well something works depends on how it is implemented & what the design objectives are. Obviously yours matched your design requirements well. That's what good design is about.
I suspect they're using the peaking to try to prop the midbass up a bit. Re your success with a 0.88 Q unit, I don't doubt it -but how well something works depends on how it is implemented & what the design objectives are. Obviously yours matched your design requirements well. That's what good design is about.
Last edited:
Some of this is age. I have some TC Sounds fifteens that over a decade old, and their Q has gone up noticeably over the years.
Another thing to consider is the actual bass response of the speaker when it's in a room. We fuss about a dB here or there, bit of a peak, a little bit of overshoot. But put it in a room, measure the actual response, look at it in that light. It will not even vaguely resemble the pretty modeled curves.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.