taking a W3-871S to 80 hertz

[Douwe]

Hello everyone,

I want to build my own computerspeakers and was thinking of a vented box with 1 or 2 tangband W3-871S in them. I think I want to put the -3dB point at around 80 hz but don't know if this can destroy those tangbands because its FS is 110hz.

 

[:tom:germany:]

Hello
In Cyburgs "Brick" or "Needle" the bass goes down to about 50hz with the W3-871S

greetings Tom

 

[leadbelly]

I think I want to put the -3dB point at around 80 hz but don't know if this can destroy those tangbands

You've not stated the most important consideration: How loud will you play them? Are they PC speakers or for a full room?

 

[Douwe]

Thanks Tom but do you also know how these speakers sound? Are the lower frequencies not almost fully produced by the bass pipe? And if this is so how does it sound ???

 

[Douwe]

I'm going to use them as PC speakers and with 12 watt they won't go harder then 97dB if I should use one speaker.

 

[leadbelly]

I'm going to use them as PC speakers and with 12 watt they won't go harder then 97dB if I should use one speaker.

I think you'll run out of excursion long before you get 94dB @ 80 Hz out of the TB's. Look at their efficiency and Xmax numbers. But are you being realistic? Do you listen to your PC speakers at 97dB?

 

[Bob Brines]

Something that is apparently not well understood is that drivers mounted in BR's and TL's run out of excursion one octave above cut-off. That means you are going to get all of your distortion starting at 160 Hz if you are going for 80 Hz F3. Those distortion products are going to appear at 300 Hz and above. This is the problem with using a 3" full-range driver as anything but a mid-tweeter. I think that you should be happy with 80 dB for a pair of 871's, and this means the speakers 2-3 feet from your head.

Bob

 

[soongsc]

Something that is apparently not well understood is that drivers mounted in BR's and TL's run out of excursion one octave above cut-off. That means you are going to get all of your distortion starting at 160 Hz if you are going for 80 Hz F3. Those distortion products are going to appear at 300 Hz and above. This is the problem with using a 3" full-range driver as anything but a mid-tweeter. I think that you should be happy with 80 dB for a pair of 871's, and this means the speakers 2-3 feet from your head.

Bob

This that based on listening or measurements?

 

[Douwe]

I have looked at a design called the brick(with one W3-871S driver) and this one is cut of at around 60hz and there is a peak at 120hz...... http://www.diyaudio.com/forums/showthread.php?s=&threadid=47558&highlight=But are these harmonics easy to hear? My objective is to make two cheap pc speakers. Ofcourse I want them to sound as good as is possible for the price i'm willing to pay.

 

[Bob Brines]

This that based on listening or measurements?

This is based on observation and theory. ALL resonant enclosures have a peak in the excursion plot one octave above cut-off. That's the way it is. This peak excursion is what limits the maximum undistorted SPL you will get from your speaker. The maximum undistorted SPL is far lower than implied by simplistic programs like WinISP. You need to look at the excursion plot when designing speakers.

Bob

 

[soongsc]

This is based on observation and theory. ALL resonant enclosures have a peak in the excursion plot one octave above cut-off. That's the way it is. This peak excursion is what limits the maximum undistorted SPL you will get from your speaker. The maximum undistorted SPL is far lower than implied by simplistic programs like WinISP. You need to look at the excursion plot when designing speakers.

Bob

So if one is able to lower the resonance then a this means the maximum undistorted SPL is greater? Isn't Xmax also supposed to be defined at 10% distortion? Is the excursion plot the ones Martin's Mathcad worksheets show?

 

[Bob Brines]

So if one is able to lower the resonance then a this means the maximum undistorted SPL is greater?

Lowereing the system resonance generally means coming closer to the infinite baffle responce. This is good or bad depending on what you intend to do. After you reach an optimal resonant point, further lowering of the resonance will DECREASE the bottom end SPL.

Isn't Xmax also supposed to be defined at 10% distortion?

You can use that definition. Do you want 10% distortion in the midde of the voice range?

Is the excursion plot the ones Martin's Mathcad worksheets show?

Yes it is.

Bob

 

[:tom:germany:]

Thanks Tom but do you also know how these speakers sound? Are the lower frequencies not almost fully produced by the bass pipe? And if this is so how does it sound ???

I've build the Needles with the Visaton FRS8. The Tang Band sounds a little bit better, but the bass is the same (opinion from people who heard both). For normal hearing in our living room (about 25m2) the volume is enough (so that you can still talk with someone normaly). The sound itsself is very surprising for those little speakers, I think. But I must say that I never heard real high end speakers for reference.
For PC speaker use the volume should be quite enough, I think but don't know if it will go up to 97dB.
Maybe this speakers are also a alternative for you:
http://www.blueplanetacoustic.de/pdf/169/W3-315 Longtall-2.pdf

nice greetings Tom

 

[Douwe]

If the BETA version of winISD is right then I can't get below 224 hz because the cone excursion would get larger then 0,5mm. If I look at the W3-871S I almost can't belief that its Xmax is only 0,5mm and not larger. Can someone tell me where my thinking goes wrong because if my thinking is right shouldn't there be a lot of dead W3-871S speakers?

 

[thalis]

The real problem remains under Fb for most applications. Excursion above the cut-off frequency depends on system frequency tuning. The deeper you tune the greater excursion you get a little above the Fb (not an octave above – where exactly it depends on Q). But excursion remains bigger in almost every circumstance in the lower than Fb frequencies (except if you tune at 20Hz for instance with a really big driver and volume and you don’t care about what’s happening bellow 20 Hz).
Typical excursion of a ‘normal’ BR 8 inch driver 33Hz tuning, attached.
Regards,
Thalis

 

[thalis]

Douwe, drivers don’t easily pass away if they reach their limits
This tiny driver has about 5mm of maximum (not linear) excursion capabilities and as it progresses to its max excursion it begins to ‘soft clip’ the waveform. In practice you will be able to get decent levels without a lot of audible distortion (say 92-95dB in room) - you don’t expect to hear really loud I presume.

 

[soongsc]

Lowereing the system resonance generally means coming closer to the infinite baffle responce. This is good or bad depending on what you intend to do. After you reach an optimal resonant point, further lowering of the resonance will DECREASE the bottom end SPL.



You can use that definition. Do you want 10% distortion in the midde of the voice range?



Yes it is.

Bob

I think there are lots more considerations since some Xmax definitions are different, near resonance areas are also where the amp cannot have control of the driver to some degree, total air displacement, etc. Mathcad worksheets seem to give you a good idea on parts of the aspects, and Martin seems to have made them quite well for the DIY community.

 

[soongsc]

I've build the Needles with the Visaton FRS8. The Tang Band sounds a little bit better, but the bass is the same (opinion from people who heard both). For normal hearing in our living room (about 25m2) the volume is enough (so that you can still talk with someone normaly). The sound itsself is very surprising for those little speakers, I think. But I must say that I never heard real high end speakers for reference.
For PC speaker use the volume should be quite enough, I think but don't know if it will go up to 97dB.
Maybe this speakers are also a alternative for you:
http://www.blueplanetacoustic.de/pdf/169/W3-315 Longtall-2.pdf

nice greetings Tom

Not having tried the 871's before, but having tried the 1284SB which has about 0.4mm Xmax, it sounded quite well for a small full range during initial listening, but you would want more Xmax if 80Hz is realy desired.

If the BETA version of winISD is right then I can't get below 224 hz because the cone excursion would get larger then 0,5mm. If I look at the W3-871S I almost can't belief that its Xmax is only 0,5mm and not larger. Can someone tell me where my thinking goes wrong because if my thinking is right shouldn't there be a lot of dead W3-871S speakers?

TB speakers seem to take lots of abuse. 1231SH has 1.2mm Xmax or somewthing like that.

 

[seanzozo]

The real problem remains under Fb for most applications. Excursion above the cut-off frequency depends on system frequency tuning. The deeper you tune the greater excursion you get a little above the Fb (not an octave above – where exactly it depends on Q). But excursion remains bigger in almost every circumstance in the lower than Fb frequencies (except if you tune at 20Hz for instance with a really big driver and volume and you don’t care about what’s happening bellow 20 Hz).
Typical excursion of a ‘normal’ BR 8 inch driver 33Hz tuning, attached.
Regards,
Thalis

I think what bob meant is that excursion above Fb is often GREATER than at Fb for the entire octave above Fb (which your graph demonstrates perfectly).

Above Fb in a ~80hz tune is where lots of energy is in non synth recording (kickdrum etc). The generally increased levels of information above Fb relative to below Fb can lead to noticible distortion if one does not pay attention to the excursion plots above Fb. That distortion then climbs into the vocal registers where the ear is sensitive. 2nd harmonics from 160-320, thirds from 240-540.

Sean

 

[Bob Brines]

I think what bob meant is that excursion above Fb is often GREATER than at Fb for the entire octave above Fb (which your graph demonstrates perfectly).

Above Fb in a ~80hz tune is where lots of energy is in non synth recording (kickdrum etc). The generally increased levels of information above Fb relative to below Fb can lead to noticible distortion if one does not pay attention to the excursion plots above Fb. That distortion then climbs into the vocal registers where the ear is sensitive. 2nd harmonics from 160-320, thirds from 240-540.

Sean

This is exactly what I ment. Most resonant full-rangers runout of gas first in the 80 Hz range because of the excursion problem and the program overload in this range. This leads to a thick, muddy sound, and the reason full-rangers don't normally do well with heavy metal, regge, etc.

Bob