I have 2 Visaton W200S-8 units here. I would like to use them as woofers in a sealed three way arrangement. For the mid I want to use a 3" wide range unit, that can be filtered low (say 250-300Hz) at the low end and high (say 4-5Khz at the high end). Thinking of a 3/4" soft dome tweeter for this.
I have worked out the sealed air volumes for the W200S-8, and think to have decided on a volume somewhere between 28.7L (Bessel) and 16.2 L(Butterworth). and erring on the bessel side. Maybe even going below it with 30L total.
Is that a good idea? Any input here is welcome, as long as it remains sealed (I want no smear on the low end).
Another big question is this one: How to choose a sealed air volume for a full range mid mid driver? For instance if I where to use a Tang band w3-881SJF from 300Hz upwards.. what sealed alignment or volume would be appropriate and why?
Suggestions about alternative mid drivers are welcome, as are tweeters or other advice! I do however want to keep crossoverdesign simple as I will be doing it by ear (more ears here though..).
Kind regards!
I have worked out the sealed air volumes for the W200S-8, and think to have decided on a volume somewhere between 28.7L (Bessel) and 16.2 L(Butterworth). and erring on the bessel side. Maybe even going below it with 30L total.
Is that a good idea? Any input here is welcome, as long as it remains sealed (I want no smear on the low end).
Another big question is this one: How to choose a sealed air volume for a full range mid mid driver? For instance if I where to use a Tang band w3-881SJF from 300Hz upwards.. what sealed alignment or volume would be appropriate and why?
Suggestions about alternative mid drivers are welcome, as are tweeters or other advice! I do however want to keep crossoverdesign simple as I will be doing it by ear (more ears here though..).
Kind regards!
I think that woofer is better suited for BR duties, its Qts is too low for sealed alignment to produce the bass that you want.
I simulated it in Basta in 50L total volume (for the two drivers) crossed over at 300 Hz with 1.5 - 1.8 mH coil.
In that alignment Qtc is 0.59 and it appears that the bass will be too anemic, hardly reaching 51 Hz @ -6 dB.
Not a great result.
Otherwise its Le (2.2 mH!) will help with the crossover.
Maybe look for a woofer with Qts of 0.6 - 0.7, which will help with the bass and also give you the option to go Aperiodic if you wish.
I simulated it in Basta in 50L total volume (for the two drivers) crossed over at 300 Hz with 1.5 - 1.8 mH coil.
In that alignment Qtc is 0.59 and it appears that the bass will be too anemic, hardly reaching 51 Hz @ -6 dB.
Not a great result.
Otherwise its Le (2.2 mH!) will help with the crossover.
Maybe look for a woofer with Qts of 0.6 - 0.7, which will help with the bass and also give you the option to go Aperiodic if you wish.
I have no guidance on your woofer selection other than if a tower design, a .5 way inductor/coil is essential between the two woofers to prevent steering the forward lobe of the mid driver towards the floor…..otherwise just build a 2 way with a sub as you’ll degrade all the advantages of your efforts.
3” driver to 300hz?……..it can do ‘ok’ but it will be the SPL limiting driver in your design. As to the sealed midrange enclosure?…….whatever volume yields a .7 QTS. If after you’ve measured your prototype and there’s an impedance peak from the mid driver in its passband?………add or create an aperoidic vent….or in other words a little leak. This will smooth the impedance dramatically but also reduce power handling…..pick your poison…….there’s no free lunch.
If it were me….and I like the idea of a wide range/full range mid driver…….you only need one eight inch here. If you were using a 5-5.25” mid I’d say go for 2.
Keep the 3” mid anyways?…………and still wanna go double 8’s?…….Scanspeak Discovery 10f has the power handling but better cross it higher…..say 400hz. Let that top 8” woofer be the mid-woofer it was always meant to be in a smallish sealed chamber.
I personally favour a large Qtc 0.5 enclosure for open backed drivers used for midrange. I find I can get more airspace and more stuffing and damping material behind the driver to help absorb back reflections. I get the smoothest midrange impedance curves with this approach when doing a box design.
For the woofers, it all depends on what low end you are looking for and whether you will augment any bass loss with a subwoofer or boundary reinforcement in room.
For the woofers, it all depends on what low end you are looking for and whether you will augment any bass loss with a subwoofer or boundary reinforcement in room.
This particular woofer has Qts = 0.33. Unfortunately, this makes it much more suitable for a vented-box enclosure than for a closed-box enclosure. Using it in a closed-box enclosure will result in not very much low-frequency output.
With two drivers in series in a closed-box of 62 litres with a Qtc = 0.577 we obtain the following response. It has f3 = 69.2 Hz, f6 = 48.3 Hz, and f10 = 35.1 Hz.
With a Qts = 0.33, this driver is well suited to a QB3 vented-box alignment, which has quite a good transient response. This comes about due to its 3rd-order quasi-Butterworth initial roll-off.
For a vented-box enclosure of 63.9 litres in volume with a vent tuning frequency of 36.1 Hz, we obtain f3 = 43.0 Hz, f6 = 35.6 Hz, and f10 = 30.0 Hz. The dash-dot light-blue line is the previously computed closed-box response. At 40Hz and 50Hz, the vented-box enclosure has about 4dB more bass output than the closed-box enclosure, which is a substantial difference.
It is interesting to note that the vented-box and closed-box enclosures are almost identical in volume. Hence, you could build a vented-box system and, by simply blocking the vent, you could compare it directly to the closed-box system.
If using a 3" diameter midrange driver, it would probably be better to cross it over no lower than about 400–500Hz or so. This is still well within the smooth operating range of the woofer.
Last edited:
You could look into the series cap GHP trick https://www.diyaudio.com/community/threads/passively-assisted-sealed-alignments.360786/ - might find some suitable drivers that could do the job.
Brief explanation in one of Tony Gee's designs https://www.humblehomemadehifi.com/download/Humble Homemade Hifi_Black Box.pdf
Brief explanation in one of Tony Gee's designs https://www.humblehomemadehifi.com/download/Humble Homemade Hifi_Black Box.pdf
I like an 8” woofer to a 3” fullrange (WAW) type loudspeaker. They work great in average sized rooms. No tweeter is required. You may want to raise the crossover to the mid a bit (300hz+) for lower distortion and greater power handling. The Visatons are for vented enclosures. A Peerless SLS 830667, 8" woofer would work better in a sealed enclosure, which should get an f3 down around 45 hz in a 1-1.5cu ft sealed box. Then put some Fountek FE85 3" full ranges in as the mid-tweets. This should give you a pretty nice system for not too many Euros. If you have ears better than my dog’s then add a tweeter. If you want better sound then try the Markaudio CHP 90 as the midrange. If you want surround then add three more CHP 90s in their own boxes for center and rear channels. You find the volume of the midrange box the same way as for the woofer box, it’s the same math.
Another thought, you really shouldn’t press a woofer into an alignment that it wasn’t designed for. Putting a woofer designed for vented system in a closed box will only result in extremely limited bass, and forcing a woofer designed for a sealed alignment into a vented system often results in a self destructive boom box. If you like the Visatons then put them in a properly designed vented enclosure.
I was curious about the Thiele–Small parameters of the Peerless SLS 830667 8-inch woofer. With a Qts = 0.66, it is definitely suited to a sealed-box low-frequency alignment.
However, if a large enclosure is used, a vented-box enclosure produces the following response when using a 2000µF capacitor in series with the woofer. The −3dB point is a very low 22.1 Hz, while the response is quite flat. Apart from helping achieve this alignment's overall flatness, the capacitor reduces woofer cone excursions below the tuning frequency.
Thanks for all the replies! I think I made a mistake in my first post: I only want to use 1 8" woofer per speaker. (I said I had two, but I meant for the stereo pair). I will deliberate on changing the woofer type, or the bass alignment. Probably the latter, as I have these woofers lying about.
I think it's more likely a misunderstanding on my end. To convert my twin-woofer simulations to a single-woofer case, simply halve the enclosure volumes that I used.
Last edited:
The reason I recently did favour a sealed alignment was because my main concern here was the better impulse response of a sealed driver. I do have the W200s-8 on a shelf, and thought I could use these sealed for the better impulse response. I am not so sure a different driver with a series capacitor is going to bring me this fabulous impulse response.. But I remain interested: Does this series capacitor provide better impulse response?
I very much doubt it, as it also decouples the damping factor of the amplifier. But this is my two cents.. I will do something different with these visaton W200S-8 units, but that is not this thread.
I have become interested in other 8" (or 10") drivers that might fit the bill. Its not really about the low end extention and low end efficiency I care about here, but rather about extending a good impulse response low down, and yes I would gladly trade in efficiency here, as my living room size and placement will compensate. These speakers will be relatively close to the back wall in a non ideal position. (not in the length of the livingroom, but accross it on one side). At the moment there are 2 way bassreflex units in this location: (https://www.jkacoustics.com/jk-producten/jk-reference-junior) and they are very good. But placing Chartwel ls3/5 there blew them away. But I do worry about the low end.. so making a sealed 3 way with a bigger woofer seemed a good idea.. But remaining sealed is very important here.
I very much doubt it, as it also decouples the damping factor of the amplifier. But this is my two cents.. I will do something different with these visaton W200S-8 units, but that is not this thread.
I have become interested in other 8" (or 10") drivers that might fit the bill. Its not really about the low end extention and low end efficiency I care about here, but rather about extending a good impulse response low down, and yes I would gladly trade in efficiency here, as my living room size and placement will compensate. These speakers will be relatively close to the back wall in a non ideal position. (not in the length of the livingroom, but accross it on one side). At the moment there are 2 way bassreflex units in this location: (https://www.jkacoustics.com/jk-producten/jk-reference-junior) and they are very good. But placing Chartwel ls3/5 there blew them away. But I do worry about the low end.. so making a sealed 3 way with a bigger woofer seemed a good idea.. But remaining sealed is very important here.
I understand your concerns regarding impulse response. Consider a classical closed-box and vented-box maximally flat alignment. If they both have the same −3dB cut-off frequency, then the closed box has the better transient response. That's because of the dynamic response behaviour arising from its second-order roll-off rate (12dB/octave), while the vented box has a 4th-order roll-off rate (24dB/octave). The addition of a series capacitor to any system will add an extra order of roll-off at low frequencies, which will make the transient response a little bit worse. The series capacitor changes the system dynamics of the closed box from a second-order system to that of a third-order system, which is still of lower order than a vented-box loudspeaker.
Below is the transient response of a second-order Butterworth high-pass filter with a −3dB point of 50Hz, which simulates a closed-box enclosure, to a 50Hz single pulse transient signal. It's quite well-behaved, but there are now three peaks rather than just one. The attenuation of the amplitude of the filtered waveform is a result of the filter's −3dB point being 50Hz.
Below is the transient response of a fourth-order Butterworth high-pass filter with a −3dB point of 50Hz, which simulates a vented-box enclosure, to a 50Hz single pulse transient signal. This isn't as good as the result for the second-order filter, as there are now five peaks present in the response instead of three. The build-up in the response of this filter is also a bit slower.
The effect of the series capacitor results from it changing the dynamic behaviour of the loudspeaker system with which it is interacting, not because of its placing itself in series with the output resistance of the amplifier. The damping factor of a typical solid-state amplifier does not serve to dampen the woofer's response.
Keep in mind that, if we lower the −3dB cutoff frequency of the enclosure and extend its low-frequency response, we do get some benefit. If we now apply the 50Hz pulse in a region above the cut-off frequency, then the filtered response of the pulse will look more like the original waveform. The oscillatory transient response behaviour shown in the two examples above has been shifted lower down in frequency, and a simple pulse will have purer, more accurate starting–stopping behaviour.
Below is the transient response of a second-order Butterworth high-pass filter with a −3dB point of 50Hz, which simulates a closed-box enclosure, to a 50Hz single pulse transient signal. It's quite well-behaved, but there are now three peaks rather than just one. The attenuation of the amplitude of the filtered waveform is a result of the filter's −3dB point being 50Hz.
Below is the transient response of a fourth-order Butterworth high-pass filter with a −3dB point of 50Hz, which simulates a vented-box enclosure, to a 50Hz single pulse transient signal. This isn't as good as the result for the second-order filter, as there are now five peaks present in the response instead of three. The build-up in the response of this filter is also a bit slower.
The effect of the series capacitor results from it changing the dynamic behaviour of the loudspeaker system with which it is interacting, not because of its placing itself in series with the output resistance of the amplifier. The damping factor of a typical solid-state amplifier does not serve to dampen the woofer's response.
Keep in mind that, if we lower the −3dB cutoff frequency of the enclosure and extend its low-frequency response, we do get some benefit. If we now apply the 50Hz pulse in a region above the cut-off frequency, then the filtered response of the pulse will look more like the original waveform. The oscillatory transient response behaviour shown in the two examples above has been shifted lower down in frequency, and a simple pulse will have purer, more accurate starting–stopping behaviour.
Last edited:
That's true @R. Urbach. As stated by V. Dickason in his well known Loudspeaker Cook Book : the best Bass-Reflex will always offer inferior impulse and step response than the best Sealed enclosure. This is also true for the group delay.
IMHO, you should have a look to speakers offering a QTS at circa 0.4 to 0.6, and a low FS. That said, it may be difficult to find 8" speakers in that category, I don't know... Maybe the Monacor SPH-200 ?
If you keep the Visaton W200S8 that you have in stock :
Here are quick online simulations with it :
The Sealed box (blue) is for a QTC at 0.7 (the better compromise between volume / linearity / step response) and there's two BR propositions : the Ideal BR (black) and the Custom BR, with same volume as Sealed box, with low FB tuning (red), offering a better step response but less bass than the Ideal (black).
Here's the comparison of the group delays ( which represents the speaker's ability to transmit various simultaneous low frequencies at the same time, the smallest value being the best).
Indeed, the best one is offered by the Sealed enclosure (blue), the worse is for the Custom BR (red), but it is located at 10Hz instead of 30Hz, so in a rather inaudible range. It's a compromise favoring the impulse response with more bass than the Sealed enclosure...
Another view :
The Ideal BR remains the same (black), the Sealed box at QTC=0.7 is also the same (blue), and the red is a Sealed box with QTC=0.5, that is to say larger volume and even better step response... But the bass response will be different : deeper, but less present... Again, it's compromise...
T
The simulations provided in Post #15 are very interesting and helpful. However, I'm not entirely sure as to the broad usefulness of the step-response comparisons. The step input is a very unusual, wide-band input signal. Although it is a stalwart of transient response analysis, I am not convinced as to how well it applies to typical music program material. Although the closed-box in the examples above has a "better" step response, that is achieved largely through the much greater amount of low-frequency attenuation that is present in the closed-box system relative to the vented-box system.
My thinking is that when evaluating low-frequency transient response, we also need to take into account the amount of low-frequency attenuation that is present in the system. For example, the first track on Pink Floyd's album Dark Side Of The Moon has a heartbeat-sounding tone burst whose main frequency content comes in at 30Hz. An excerpt of that music signal is shown below.
If a closed-box loudspeaker has a −3dB cut-off frequency of 50Hz, it is incapable of reproducing that signal with a reasonable degree of fidelity. On the other hand, a vented-box loudspeaker with a −3dB cut-off frequency of 30Hz can attempt it to some degree, albeit with some inherent attenuation. So, in this instance, does the closed-box system have a "better" transient response? I'd tend to say no because when using the closed-box loudspeaker I can't hear the music as the artists had intended.
It would be nice if it were possible to use the simulation setup reported in Post #15 to allow a single 1/2 cycle of a raised sinewave pulse to be fed into the models. This pulse signal would need to be variable in frequency. It could be started at the −3dB cut-off frequency of the system with the lowest cut-off frequency, and then progressively increased in say 1/6-octave steps. That way we could better see what the transient response is doing, rather than being somewhat misled by using the step response input and its subsequent output, as a loudspeaker dynamical system cannot reproduce DC signals.
My thinking is that when evaluating low-frequency transient response, we also need to take into account the amount of low-frequency attenuation that is present in the system. For example, the first track on Pink Floyd's album Dark Side Of The Moon has a heartbeat-sounding tone burst whose main frequency content comes in at 30Hz. An excerpt of that music signal is shown below.
If a closed-box loudspeaker has a −3dB cut-off frequency of 50Hz, it is incapable of reproducing that signal with a reasonable degree of fidelity. On the other hand, a vented-box loudspeaker with a −3dB cut-off frequency of 30Hz can attempt it to some degree, albeit with some inherent attenuation. So, in this instance, does the closed-box system have a "better" transient response? I'd tend to say no because when using the closed-box loudspeaker I can't hear the music as the artists had intended.
It would be nice if it were possible to use the simulation setup reported in Post #15 to allow a single 1/2 cycle of a raised sinewave pulse to be fed into the models. This pulse signal would need to be variable in frequency. It could be started at the −3dB cut-off frequency of the system with the lowest cut-off frequency, and then progressively increased in say 1/6-octave steps. That way we could better see what the transient response is doing, rather than being somewhat misled by using the step response input and its subsequent output, as a loudspeaker dynamical system cannot reproduce DC signals.