|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
||Thread Tools||Search this Thread|
|20th February 2004, 02:10 PM||#11|
Wizard of Kelts
diyAudio Moderator Emeritus
Qts, Fs, Vas predict the driver's bass response curve in a box of various types-or even on a dipole. That is it.
Even then, sensitivity, linear excursion, power handling, all come into it.
Generally, these numbers are less critical in drivers that are to be used in midbass or midrange applications. And has been pointed out, these numbers tell you nothing of smoothness through the midbass and midrange regions, which is what most consider important.
"A friend will help you move. A really good friend will help you move a body."
|20th February 2004, 02:31 PM||#12|
Join Date: Feb 2004
Now, let's assume for a moment (and I am not very happy about it) that these mechanical losses also have an effect on other resonances in the mechanical system, for example a cone breakup. Wouldn't a *high* loss be desirable then? A high loss would bring down the Q of the breakup and the "peakedness" would be smaller. Assuming that the Qms resistance has anything to do with cone breakup at all...
|20th February 2004, 02:32 PM||#13|
Join Date: Dec 2003
a driver cannot be defined by ,
to consider its t/s parameters , never actually suggests how it will play
or the materials it may be constructed with , ofcourse if it was to be used by the police force - kevlar is very important
but when it is to ultimately convert source to analougus sound , what are on line is its realism , accuracy , depth , tone / timbre , nature etc in hi fi
in pro audio - directivity , effective spl , accuracy , decay etc
it cannot be defined by a mechanical formula
what has all this got to do with diy ?
when one has to choose a drive unit or units for ones amp / speaker enclosure in mind etc
what are the factors worth considering ?
considering a purchase of a loudspeaker unit
ts parameters defines this unit mechanically , through these mechanical parameters one can proceed to use / utilise the unit to its best , including efficiency , and is of utmost importance to its enclosure design enclosure
but one has also paid for the materials used in the construction of the drive unit ,
this will decide its longetivity , compression at peak power , overall accuracy - not to mention the tonal quality or the nature of sound produced
where am i heading , - the internal volume of a cone may be same as that made of ordinary hard pressed paper or for eg a kevlar woofer
1 set of compliance figures can be easily duplicated between
a rubber surround cone / and a foam surround cone
all the manufacturer needs to do is control the hardness factor of the rubber surround material (ie the rubber)
the density of the foam material _ of the foam surround
with this surround material differance the sound quality of both the drive units will be entirely different
- along with the materials used the construction process is of importance
so is the x- max the voice coil inductance , the cooling efficiency
distortion - can be reduced by a symmetrical field magnet structure , shorting rings for eddys , shorting rings built into the coil to limit the coil from leaving the magnetic gap during extreme excursions
these important issues relating the drive unit is not reflected very much in the t/s matters
where am i going ?
i can copy a standard 100 $s subwoofer unit' s t/s parameter using
ordinary components for eg a pressed paper instead of a non press one , same coil dia , wire winding etc but with a poor insulation
and it will never play , last , sound as the original 100 $ s one
this is of utmost importance to diy constructors , along with t/s parameters
Reverting to important issues governing diy and speaker design
1 . The speaker matching with the amplifier ,
I can provide a double magnet (155 to 230 mm) subwoofer in a lambada extended pole underhung sfg topology magnet structure with a kapton voice coil with imported aluminum wire voice coil lengths in excess of 60 mm with a xmax of 25 mm at 4 ohms , a imported non-press paper rubber surround cone (which totally absorbs all mid + hi frequency due to the cone design)
if the user constructs the most powerful IC amp it will not match
infact it wont match a 150 w rms transistorized design using latest Toshiba ets transistors
the reason the magnet structure is over damped
now if it was the crescendo mosfet amp then we are going somewhere ,
the type of the transducer , its power handling , has the voice coil size been increased blindly to increase power handling without magnetic efficiency and x max in mind ?
the type of cone used , pressed paper , semi press paper , non press paper , felted composite , reinforced ets ( this decides how the speaker will sound )
the motor structure , sfg (symmetrical field geometry magnet structure) , for full range and pro drivers if there are eddy current shorting rings , other shorting rings employed to extend frequency response , linearise impedance , improve transience
the vc material and wire , the vc inductance 2 layer , 4 layer , dual vc type etc
the suspension system and compliance
surround material , eg rubber then type natural or synthetic ets , the extent of the roll of the surround and its match with the spiders
the construction at the end all parts of a speaker is held together by adhesives , the quality of adhesives used / and the process of manufacturing , for eg if I repair a toy with a epoxy hardener type adhesive and do not allow for the correct drying time
at the end all manufacturers are increasing production speeds
transducer design engineer
|20th February 2004, 02:46 PM||#14|
Join Date: Dec 2003
then a good driver ?
this statement can be said as correct that :
a power rating given to a driver , means that at all input levels of the driver rating within its specified limit , the driver :
1 ) should perform within the rating in long term use , robustness
2) should perform within very acceptable limits of distortion
3) and should show a certain limit of linearity in relation with the , actual accoustic power output increase to the input power supplied from the amps
the distortion is now adays mostly the deciding factor of the power rating including long term use
Here it needs to be devided in two sections
in a broad context
distortion / you will find some speakers distorting _ much before the voice coil actually burns out
for eg - considering a driver in which this takes place :
a driver or if they are pro audio or drivers with simple pole vent cooling with the coils set in a tight magnetic gap - the coil can be on a paper former , oridinary wires used in the vc s
the vc s temperature will not rise easily normally 60 deg max , at this time it will distort
then this distortion is not from the vc temperature
(though in most books vc temp is one of the main source of distortion , why ? read on ....)
there are 3 basic types of non linearity in a transducer , these generate distortion , due to this distortion the voice coil heats up / at this moment no amount of cooling can help ,
the distortion effect from driver non linearities are such that it may lead to vc temp rise exponentially leading to vc burn out
the non linearities are
1. non linearity in the suspension mechanism
2. non linearity in the motor sturucture
3. non linearity caused when a coil moves in a permanent magnet motor structure (this is bound to happen even if the above 2 non linearities are addressed well )
to check for these non linearities while purchasing drivers
1. suspension non linearity
for the x - max of the driver you would like to purchase (or the one specified by the manufacturer) meaning you either know the x-max or have one in mind , then :
push the driver cone downwards by that amount , it should go down linearly (the stiffness of the suspension increasing linearly) to the amount of the x-max
pull the cone upwards towards you from beneath and it should come up to the specified x-max , again stiffness increasing linearly
if this does not take place the driver will distort -
unless the transducer manufacturer identifies specifically that such a non - linearity has been addressed , by making suitable compensations in the vc itself
2. motorstructure non linearity
fringe fields exists at the magnetic gap - the motor structure should be - sfg a symmetrical field geometry _ topology
therefore while purchasing a driver look if the motorstructure is - sfg or not
since expenses goes up in a sfg structure it will be specifically mentioned mentioned by the driver manufacturer , to inform customers
3. Non linearity caused when a coil moves in a permanent magnet motor structure
explaining this nature of non linearity is complex , anyone interested can see the sections on shorting rings in my thread
what is important
is that : -
has this been addressed by the transducer manufacturer
it is addressed by shorting rings placed in the motorstructure
since this too involves expence , and a considerable slow down in the manufacturing process this too will be specified by the driver manufacturer , to inform customers
the above are the main sources of distortion that determines the safe operating power of a driver , and are not the only sources of driver distortion (other sources of distortion are eddys , bottoming to name a few)
due to such complex stuff good driver manufacturers all over the world can be counted on finger tips
in good drivers context
good drivers are in which all the above criterias have been addressed
then the rise in vc temperature over long time use / the cooling efficiency comes in , and mostly all speaker failures of reputated manufacturers result from vc burn out (as mentioned by Navin)
mostly these limits are tested thoroughly and drivers tested to distruction and safe regions specified
i have found driver malfunction due to negligence of owners to go in for a service contract when pro audio drivers are used eg - movie halls , discs
in home audio , sources of reputated driver burn outs are :
amp malfunction and dc component generation in the drive signal or signal clipping
speacially when high power speakers are used by under powered amps , many people do not even consider that a 35 w amp can damage a 100 w speaker system _ plain simple signal clipping can distroy well made drivers
if you have an under powered amp please ask a sound engineer to determine the clipping levels of your amp and restrict from running at clipping levels (in high end audio amps from manufacturers like Nad and onwards this is built in as standard , in the circuitry , therefore the prices)
or the suspension parts have outlived _ may be the owner !
thanks navin for the reccmn.
Suranjan Das Gupta
Transducer design engineer
|20th February 2004, 10:01 PM||#15|
diyAudio Member RIP
Join Date: Nov 2003
Location: Brighton UK
the simple answer here is Qe is a predictable parameter
and a consequence of the magnetic and electrical parameters.
Qms is not. It cannot be treated as a lumped parameter.
It is the mechanical damping of the system and as such
reflects the way the drive unit is made and its subsequent
Consider a driver with low Qms. Is this due to the cone surround
or the voice coil spider ? It should be obvious that the point of
application affects unit behaviour.
Another point is is Qms entirely resistive ? What is certain is that
cheap drivers with low Qms (allowing savings on magnet induced
Qe damping) are certainly not, the type of damping applied
exhibits hysteresis, that is it is non-linear.
Of course simply measuring Qms will not indicate this hysteresis.
and where is this hysteresis applied ? at the centre or edge ?
By implication a driver with high Qms will be less likely to suffer
from hysteresis type effects.
I believe it is generally accepted a driver with a low Qms due
to the surround (for optimum midrange termination) does not
make a good bass driver.
And when used as a bass/mid there is a bass quality versus
midrange quality trade off, you could argue that a high Qms
driver must have compromised midrange.
Note that hysteresis effects of the surround are negligible at
midrange frequencies / excursions, it acts more as a damped
cone edge terminator.
|20th February 2004, 11:04 PM||#16|
Join Date: Feb 2004
So, sreten, you are basically saying that a high Qms is desirable, because this means that the mechanical losses are low, and that low mechanical losses is good since they mostly are nonlinear and hysterical (or whatever the adjective should be )
Did I get you right?
|21st February 2004, 07:14 AM||#17|
Join Date: Aug 2002
let's pretent a car with a certain weight.
It has a certain spring/dampening optimum related to it's weight.
You can choose stiffer springs or harder dampers, each combination will give you another 'character'.
For racing you choose stif springs/ relatively hard dampers, for off road racing weak springs/ soft dampers. Why? I recon you see why.
Fact is that in off road racing dampers have to dissipate a lot of energy because they have to absorb the long excursions of the suspension. But they enhance driveability.
On the other hand off road suspension don't give that much information to the driver's steering wheel. There's a loss of information because suspension in absorbing alot of that info.
In circuit racing it's more important that the driver feels the exact moment when tires are loosing grip, so suspension has to be more 'direct'.
Note that an off-road suspension is usefull even on circuit, not optimal, but useable, but circuit-suspension is unuseable off-road.
|21st February 2004, 08:16 AM||#18|
Join Date: Feb 2004
Re: car comparision
OK, completing the analogy with loudspeakers, this tells me that Cms and Fs is important (which is set by the suspension stiffness and the mass of the car) but also that the *Qts* is important (which is set by the shock absorbers (is that the english word?))
There is no Qes in the car analogy, so Qts and Qms are the same, so we cannot draw this parallel too far.
|21st February 2004, 08:54 AM||#19|
Join Date: Aug 2002
not physically comparable with the Qes, because Qm and Qes work in paralel (Qts= (Qm*Qe)/(Qm+Qe)), but in cars you have also the undamped weight namely the tyres (including brakes) wich are also a system on its own with different tyres characteristics and inflation pressure. This system is in series with the suspension so for cars it is Qsusp+Qtyres=Qsystem.
The problem is the same, more specific tuning cause less overall useability.
The solution is also the same: dsp-processing wich alters the Q - characteristics along the way.
|21st February 2004, 09:01 AM||#20|
Join Date: Feb 2004
Well, the tyres also have a mass, so the analogy does not hold in detail, really. My point was that in speakers damping is damping, regardless if it comes from Qes or Qms. sreten had a point about the mechanical damping being more non-linear, which may be true. Otherwise I see little difference between the two.
|Thread Tools||Search this Thread|
|Thread||Thread Starter||Forum||Replies||Last Post|
|Please explain what I did||amo||Solid State||22||16th July 2008 01:56 PM|
|can please someone explain this to me?||krusty75||Tubes / Valves||6||18th February 2008 10:38 AM|
|Please explain OS vs. NOS to me...||Lyra||Digital Source||18||5th January 2004 04:45 AM|
|Can someone explain to me?||MaXiZ||Solid State||3||12th November 2003 10:36 PM|
|Can someone explain to me this?||MaXiZ||Solid State||17||1st October 2003 07:43 AM|
|New To Site?||Need Help?|