Sunday, March 16, 2008

Designing Aperiodic Enclosures

Parts Express Tech Talk

Curious: how does one design an aperiodic enclosur

Posted By: zobsky <zobsky@yahoo.com>
Date:
Friday, 14 March 2008, at 2:53 p.m.

I have an old pair of vintage 12" co-ax(Fs around 30 Hz, Qts around 0.5 , .. need to re-measure though) and was wondering if aperiodic would work .

I'll probably try open baffle first,but if that doesn't provide the desired low end , I'd like to investigate a simple aperiodic design. From what I understand, the vent resistance is best determined experimentally by attempting to minimize the impedance peak , but how does one size the enclosure in the first place?

Thanks

Re: Curious: how does one design an aperiodic encl

Posted By: dthomas <dthomas@turinnetworks.com>
Date:
Friday, 14 March 2008, at 3:07 p.m.

In Response To: Curious: how does one design an aperiodic enclosur (zobsky)

> I have an old pair of vintage 12"
> co-ax(Fs around 30 Hz, Qts around 0.5 , ..
> need to re-measure though) and was wondering
> if aperiodic would work .

> I'll probably try open baffle first,but if
> that doesn't provide the desired low end ,
> I'd like to investigate a simple aperiodic
> design. From what I understand, the vent
> resistance is best determined experimentally
> by attempting to minimize the impedance peak
> , but how does one size the enclosure in the
> first place?

> Thanks

Not a lot of info on this type of enclosure. I have built one and worked very well. First you want a high Q driver. Or a driver designed to work well in a sealed box. Look for drivers with Qes of .6 to 1.0.

The approach I took was to build a box equal to 1/3 Vas. Then I built my own aperiodic vent. I used a 7" woofer with a Vas of 30L. I put it in a 10L box with a 2.5" homemade vent. All I did was drill a hole 2.5" hole in the rear of the cabinet I installed fiberglass in the hole and varied the amount of fiberglass until I got the smallest impedance peak. I covered each side of the hole with black canvas fabric to hold the fiberglass in the hole. I cut a circular wooden cover for the rear to dress the vent up.

All in all it worked out very well. I ended up with an f3 of about 48Hz and very good woofer control. If you can't measure the impedance peak you vary the amount fiberglass until you have the best combination of cone control and bass for your application.

Dave Re: Wow, that was excellent, thank you

Posted By: Undefinition <Undefinition@aol.com>
Date:
Friday, 14 March 2008, at 4:47 p.m.

In Response To: Re: Curious: how does one design an aperiodic encl (dthomas)

Not to mention timely (I have a strange project up my sleeve currently where I was considering aperiodic)

Re: Curious: how does one design an aperiodic encl

Posted By: zobsky <zobsky@yahoo.com>
Date:
Friday, 14 March 2008, at 4:59 p.m.

In Response To: Re: Curious: how does one design an aperiodic encl (dthomas)

Thanks, ... I'll measure the Vas of the driver and go from there. Vintage drivers were notorious for having large Vas values. I'm pretty sure this driver was used in a sealed box, .. or a box with a passive radiator. Doesn't seem to be a very powerful magnet, in any case (further measurements needed to verify this).

Re: Curious: how does one design an aperiodic encl

Posted By: aduncan <aduncan5000@yahoo.com>
Date:
Friday, 14 March 2008, at 5:00 p.m.

In Response To: Re: Curious: how does one design an aperiodic encl (dthomas)

Great answer. This is something I've been curious about.

> Not a lot of info on this type of enclosure.
> I have built one and worked very well. First
> you want a high Q driver. Or a driver
> designed to work well in a sealed box. Look
> for drivers with Qes of .6 to 1.0.

> The approach I took was to build a box equal
> to 1/3 Vas. Then I built my own aperiodic
> vent. I used a 7" woofer with a Vas of
> 30L. I put it in a 10L box with a 2.5"
> homemade vent. All I did was drill a hole
> 2.5" hole in the rear of the cabinet I
> installed fiberglass in the hole and varied
> the amount of fiberglass until I got the
> smallest impedance peak. I covered each side
> of the hole with black canvas fabric to hold
> the fiberglass in the hole. I cut a circular
> wooden cover for the rear to dress the vent
> up.

> All in all it worked out very well. I ended
> up with an f3 of about 48Hz and very good
> woofer control. If you can't measure the
> impedance peak you vary the amount
> fiberglass until you have the best
> combination of cone control and bass for
> your application.

Nice post Dave!

Posted By: Curt C <cc00541@neb.rr.com>
Date:
Friday, 14 March 2008, at 6:00 p.m.

In Response To: Re: Curious: how does one design an aperiodic encl (dthomas)

Certainly I'm no expert on aperiodically damped boxes, but I will suggest that The enclosure volume can be any realistic volume, with smaller than the optimum sealed size gaining the most benefit. One possibility: choose the volume where the target Q is met when the box loss provides the lowest impedance maxima.

Done properly, the trade off is less low end extension for a smaller enclosure. But an additional, and perhaps more significant advantage is the benefit of better bass resolution around system resonsnce.

Re: Curious: how does one design an aperiodic encl

Posted By: Æ
Date:
Friday, 14 March 2008, at 6:44 p.m.

In Response To: Re: Curious: how does one design an aperiodic encl (dthomas)

> Not a lot of info on this type of enclosure.
> I have built one and worked very well. First
> you want a high Q driver. Or a driver
> designed to work well in a sealed box. Look
> for drivers with Qes of .6 to 1.0.

I'm sure you can use drivers with a lower Qts.
I have the original DYNAUDIO spec sheet for their VARIOVENT, they make no mention of woofer or woofer Qts, only enclosure volume and stuffing.

Here is what Professor Bullock wrote to me many years ago on the subject. Back when I was a subscriber to Speaker Builder magazine.

Mathemetics and Statistics Department
Oxford, Ohio 45056

December 3, 1986

Dear Mr. (Æ),

The aperiodically damped enclosure is just a vented-box with the vent resonance frequency set too low to provide any useful low frequency reinforcement. To make doubly sure that the vent provides no output, it is resistively loaded as well. It is claimed by some that this type of an
enclosure produces a flatter impedance curve than a closed or vented-box, but modeling does not support this claim. It shows that impedance peaks
can be substantially reduced by using a filling material, but not by introducing a lossy vent.

Robert M. Bullock, III
Contributing Editor, Speaker Builder

Re: Curious: how does one design an aperiodic encl

Posted By: Andy_G <gradds55@optusnet.com.au>
Date: Friday, 14 March 2008, at 7:30 p.m.

In Response To: Re: Curious: how does one design an aperiodic encl (Æ)

my vague understanding is that using aperiodic loading allows you to reduce the volume from what might generally be used for some of the older style drivers, that t/s specs would have shown needed a HUGE vented enclosure (if they had had t/s specs back then).

From what I remember from when Dad was building his (like 40 years ago) the aperiodic was some 1.5 times what a sealed box might need, and with the use of a reasonable amount of stuffing and a large resistive vent, were able to produce very smooth clean bass, particularly suited to the classical style of music.

Top of Form

Re: Hmmmmm....Sorry but,....

Posted By: Jeff B. <jeffb1836@yahoo.com>
Date:
Friday, 14 March 2008, at 8:00 p.m.

In Response To: Re: Curious: how does one design an aperiodic encl (Æ)

> I'm sure you can use drivers with a lower
> Qts.
> I have the original DYNAUDIO spec sheet for
> their VARIOVENT, they make no mention of
> woofer or woofer Qts, only enclosure volume
> and stuffing.

> Here is what Professor Bullock wrote to me
> many years ago on the subject. Back when I
> was a subscriber to Speaker Builder
> magazine.

> Mathemetics and Statistics Department
>
Oxford, Ohio 45056

> December 3, 1986

> Dear Mr. (Æ),

> The aperiodically damped enclosure is just a
> vented-box with the vent resonance frequency
> set too low to provide any useful low
> frequency reinforcement. To make doubly sure
> that the vent provides no output, it is
> resistively loaded as well. It is claimed by
> some that this type of an
> enclosure produces a flatter impedance curve
> than a closed or vented-box, but modeling
> does not support this claim. It shows that
> impedance peaks
> can be substantially reduced by using a
> filling material, but not by introducing a
> lossy vent.

> Robert M. Bullock, III
> Contributing Editor, Speaker Builder

I hate to disagree with Mr. Bullock, but I don't believe his explanation is correct. If you judge an Aperiodic enclosure by the characteristic impedance, it is nothing at all like a vented box that is tuned too low to have appreciable vent output. Although, if he is talking about 0-2 Hz, then maybe it would be, but that simply behaves as a normal sealed box. It really models best as a sealed alignment with a very leaky box. This results in both an impedance plot and a response plot that matches an actual Aperiodic enclosure pretty closely. George Short at North Creek did his Thesis on these and I beleive he has some info posted on them as well. For modeling in software select a sealed enclosure and then use a Ql value of around 2 or 3 and it will be very close.

Jeff B.

Re: Hmmmmm....Sorry but,....

Posted By: dthomas <dthomas@turinnetworks.com>
Date:
Friday, 14 March 2008, at 8:59 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (Jeff B.)

> I hate to disagree with Mr. Bullock, but I
> don't believe his explanation is correct. If
> you judge an Aperiodic enclosure by the
> characteristic impedance, it is nothing at
> all like a vented box that is tuned too low
> to have appreciable vent output. Although,
> if he is talking about 0-2 Hz, then maybe it
> would be, but that simply behaves as a
> normal sealed box. It really models best as
> a sealed alignment with a very leaky box.
> This results in both an impedance plot and a
> response plot that matches an actual
> Aperiodic enclosure pretty closely. George
> Short at North Creek did his Thesis on these
> and I beleive he has some info posted on
> them as well. For modeling in software
> select a sealed enclosure and then use a Ql
> value of around 2 or 3 and it will be very
> close.

> Jeff B.

Actually he has a nice article on a dual chamber aperiodic enclosure. The woofers are put in a very small sealed box with an aperiodic vent that goes into another much larger sealed box.

I was able to measure the change in impedance and could vary it by changing the amount of fiberglass I used in the vent.

It turned out nice with excellent bass output for a 7" woofer in a 10L box. IIRC it ends up approximating a 3rd order roll off.

Dave

Re: Hmmmmm....Sorry but,....

Posted By: Jeff B. <jeffb1836@yahoo.com>
Date:
Friday, 14 March 2008, at 9:10 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (dthomas)

> Actually he has a nice article on a dual
> chamber aperiodic enclosure. The woofers are
> put in a very small sealed box with an
> aperiodic vent that goes into another much
> larger sealed box.

> I was able to measure the change in
> impedance and could vary it by changing the
> amount of fiberglass I used in the vent.

> It turned out nice with excellent bass
> output for a 7" woofer in a 10L box.
> IIRC it ends up approximating a 3rd order
> roll off.

> Dave

Ok. I've debated this before too, but mathematically it simply can't approach a third order roll-off. A driver only rolls off second order below resonance. A vent, if it has output, is also a second order bandpass device. They combine for a fourth order roll-off acoustically due to the changing relative phase relationship between them. An aperiodic box has no mechanical mechanism that would introduce an additional first order high pass response to the woofer cone's response. It can be a damped second order response, but unless that port has appreciable output (and it doesn't) the roll-off has to asymptote to second order. I've done the math on this model, and I just can't see any first order element in the model.

Re: Hmmmmm....Sorry but,....

Posted By: dthomas <dthomas@turinnetworks.com>
Date:
Friday, 14 March 2008, at 9:27 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (Jeff B.)

> Ok. I've debated this before too, but
> mathematically it simply can't approach a
> third order roll-off. A driver only rolls
> off second order below resonance. A vent, if
> it has output, is also a second order
> bandpass device. They combine for a fourth
> order roll-off acoustically due to the
> changing relative phase relationship between
> them. An aperiodic box has no mechanical
> mechanism that would introduce an additional
> first order high pass response to the woofer
> cone's response. It can be a damped second
> order response, but unless that port has
> appreciable output (and it doesn't) the
> roll-off has to asymptote to second order.
> I've done the math on this model, and I just
> can't see any first order element in the
> model.

You may well be right I was just repeating what George Short of NorthCreek stated in his paper. He states that a aperiodic systems has a 3rd order roll-off below resonance.

Dave

Re: Hmmmmm....Sorry but,....

Posted By: Jeff B. <jeffb1836@yahoo.com>
Date:
Friday, 14 March 2008, at 10:00 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (dthomas)

> You may well be right I was just repeating
> what George Short of NorthCreek stated in
> his paper. He states that a aperiodic
> systems has a 3rd order roll-off below
> resonance.

> Dave

Yes, I have seen that he has said that. I know I am really sticking my neck out there, disagreeing with both Bullock and Short on the topic, but I do know a little bit, and on this topic I really believe they are wrong and haven't thought it through well enough. If I am wrong someone is welcome to correct me, and I will be fine with that, but I expect them to explain the mechanical-acoustical mechanism at work that I am not seeing. Otherwise, this seems quite intuitive given how cone and vent response model mathematically (and accurately) in the standard box model.

Re: Hmmmmm....Sorry but,....

Posted By: Curt C <cc00541@neb.rr.com>
Date:
Friday, 14 March 2008, at 11:48 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (Jeff B.)

> Yes, I have seen that he has said that. I
> know I am really sticking my neck out there,
> disagreeing with both Bullock and Short on
> the topic, but I do know a little bit, and
> on this topic I really believe they are
> wrong and haven't thought it through well
> enough.

I'll have to take your side too, Jeff. As I understand it, the vent is a purely resistive acoustic impedance, so there is no extra pole.

Re: Hmmmmm....Sorry but,....

Posted By: Feyz
Date:
Saturday, 15 March 2008, at 1:12 a.m.

In Response To: Re: Hmmmmm....Sorry but,.... (Curt C)

> I'll have to take your side too, Jeff. As I
> understand it, the vent is a purely
> resistive acoustic impedance, so there is no
> extra pole.

> C
Vent is usually modeled as an acoustic mass, as an inductor. When stuffed, it becomes an inductor with a high DCR. With a variovent, the length of the vent is small, which makes the modeled acoustic mass a small inductor. The stuffing has high resistance value, so especially at low frequencies, vent's inductance becomes insignificant wrt its DCR coming from the stuffing in the vent.

Re: Hmmmmm....Sorry but,....

Posted By: John Nail <SpkrNuts@aol.com>
Date:
Friday, 14 March 2008, at 9:07 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (Jeff B.)

> I hate to disagree with Mr. Bullock, but I
> don't believe his explanation is correct. If
> you judge an Aperiodic enclosure by the
> characteristic impedance, it is nothing at
> all like a vented box that is tuned too low
> to have appreciable vent output. Although,
> if he is talking about 0-2 Hz, then maybe it
> would be, but that simply behaves as a
> normal sealed box. It really models best as
> a sealed alignment with a very leaky box.
> This results in both an impedance plot and a
> response plot that matches an actual
> Aperiodic enclosure pretty closely. George
> Short at North Creek did his Thesis on these
> and I beleive he has some info posted on
> them as well. For modeling in software
> select a sealed enclosure and then use a Ql
> value of around 2 or 3 and it will be very
> close.

> Jeff B.
I'm with you on this one, Jeff. I model aperiodic usually by using a Ql of about 3....and "typical" amount of stuffing. The trick to the "variovent" or aperiodic vent is to NOT block the airflow between rear of driver and vent... So far, every one that I have modeled as a sealed box with high loss (Ql of 3) have turned out to be very very close to the actual end result.

John

Re: Hmmmmm....Sorry but,....

Posted By: Andy _G <gradds55@optusnet.com.au>
Date:
Friday, 14 March 2008, at 9:59 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (John Nail)

> So far, every one that I
> have modeled as a sealed box with high loss
> (Ql of 3) have turned out to be very very
> close to the actual end result.
How does the volume compare to what one would use if going with normal sealed cabinet ??

Posted By: Jeff B. <jeffb1836@yahoo.com>
Date:
Friday, 14 March 2008, at 10:04 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (Andy _G)

> How does the volume compare to what one
> would use if going with normal sealed
> cabinet ??

The enclosure losses simply reduce the air spring within the box, thereby reducing the system Q, so Qtc drops. Unfortunately, Fc increases, and so then does F3. So, you can get a flatter response in a smaller box with no peaking, but you will get a higher F3 as well. There's no free lunch, you always give up something. In this case it is a little bass extension that is lost as well.

Jeff B.

Re: Hmmmmm....Sorry but,....

Posted By: Andy _G <gradds55@optusnet.com.au>
Date:
Friday, 14 March 2008, at 10:08 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (Jeff B.)

> The enclosure losses simply reduce the air
> spring within the box, thereby reducing the
> system Q, so Qtc drops. Unfortunately, Fc
> increases, and so then does F3. So, you can
> get a flatter response in a smaller box with
> no peaking, but you will get a higher F3 as
> well. There's no free lunch, you always give
> up something. In this case it is a little
> bass extension that is lost as well.

Pity those old Goodmans drivers were before T/S, might be interesting to see how they modelled compared to what Dad's cabs are.

> Jeff B.

Re: Hmmmmm....Sorry but,....

Posted By: John Nail <SpkrNuts@aol.com>
Date:
Saturday, 15 March 2008, at 3:38 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (Andy _G)

> How does the volume compare to what one
> would use if going with normal sealed
> cabinet ??
Hi Andy, The last woofer I modeled for an aperiodic alignment was a 11" AR woofer.....it called for a 2.28 cu.ft. sealed box with 100% fill with Ql of 17 to get a final Qtc of .707.
Modeling that same woofer in an aperiodic, I set the Ql to 3 and reduced the fill amount to about 50% and the required box volume to get final Qtc of .707 was about 1.49 cu.ft. So, as you can see, it was a significant reduction in bo size. As cab be expected, there was also a bit of a loss of extension using the smaller box....but not as much as you might think.
The F3 comparison was 41.5hz sealed VS. 48hz for the aperiodic. However, the F9 was less of a difference......27hz sealed, vs 30.5 aperiodic.

john

Re: Hmmmmm....Sorry but,....

Posted By: Andy_G <gradds55@optusnet.com.au>
Date:
Saturday, 15 March 2008, at 4:36 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (John Nail)

> The F3 comparison was 41.5hz sealed VS. 48hz
> for the aperiodic. However, the F9 was less
> of a difference......27hz sealed, vs 30.5
> aperiodic.

any idea what the box volume would have been on the aperiodic if the F3's were matched to 41.5Hz?

Re: Hmmmmm....Sorry but,....

Posted By: Andy_G <gradds55@optusnet.com.au>
Date:
Saturday, 15 March 2008, at 8:10 p.m.

In Response To: Re: Hmmmmm....Sorry but,.... (John Nail)

I talked to Dad about this. It seems I had it the wrong way around. The normal thing was that the sealed box would have been 1.5-2 times BIGGER.
(Its sometimes hard to remember stuff from 30 odd years ago ;-)

This all makes sense with what is being modelled.

Re: Curious: how does one design an aperiodic encl

Posted By: Andy_G <gradds55@optusnet.com.au>
Date:
Friday, 14 March 2008, at 5:23 p.m.

In Response To: Curious: how does one design an aperiodic enclosur (zobsky)

I have an old 1963 "Goodmans" catalog with several aperiodic designs in it.

I could scan and post if interested.

Dad still has his Goodmans Axiom 12" aperiodic cabs in use. They do really smooth bass. I know he used soft 1/4" felt sandwiched between two plates of expanded metal for his "Acoustic Resistance Unit"

Posted By: BrianP <caspian@peak.org>
Date:
Friday, 14 March 2008, at 11:56 p.m.

In Response To: Re: Curious: how does one design an aperiodic encl (Andy_G)

if you could put that up on the Argos website. I bet a lot of folks over at AA Vintage Asylum would appreciate it too.

Back in the "dark ages" before Theile & Small, Linkwitz & Riley, et. al. codified the hard science of speaker design, there were some talented designers who, by guess and by golly (and a lot of careful listening), came up with some great sounding stuff.

Can aperiodic be modeled in WinISD

Posted By: JeffKnob <audionut21@gmail.com>
Date:
Friday, 14 March 2008, at 6:32 p.m.

I have two Dayton RSS390HF subwoofers each in a PE 3.0 cu.ft. enclosure. I know this is a little small for this subwoofer. Is there a way to model this enclosure and subwoofer combination with maybe 3-4 of the Scanspeak vents? As expensive as the enclosures are I would hate to cut 3-4 holes in the enclosures without having some idea of what it will do to the sound.

Aperiodic is a waste of time.

Posted By: Daryl <quasar@frontiernet.net>
Date:
Friday, 14 March 2008, at 7:45 p.m.

In Response To: Can aperiodic be modeled in WinISD (JeffKnob)

Just use a sealed enclosure resulting in a Qtc less than 1.0 and you will be able to use a series capacitor to knock the peak flat.

It will extend bass response lower while knocking the peak flat.

Just enter the capacitor into the crossover modeling section of your software along with the model of your driver and enclosure and you will be able to determing the capacitor value.

Re: Aperiodic is a waste of time.

Posted By: JeffKnob <audionut21@gmail.com>
Date:
Friday, 14 March 2008, at 7:58 p.m.

In Response To: Aperiodic is a waste of time. (Daryl)

> Just use a sealed enclosure resulting in a
> Qtc less than 1.0 and you will be able to
> use a series capacitor to knock the peak
> flat.

> It will extend bass response lower while
> knocking the peak flat.

> Just enter the capacitor into the crossover
> modeling section of your software along with
> the model of your driver and enclosure and
> you will be able to determing the capacitor
> value.

Thanks, All I know how to use is WinISD. Where can I do that in there?

Posted By: Rudy Jakubin <rjakubin@yahoo.com>
Date:
Friday, 14 March 2008, at 8:11 p.m.

In Response To: Re: Aperiodic is a waste of time. (JeffKnob)

> Thanks, All I know how to use is WinISD.
> Where can I do that in there?

For an Aperiodic box you can lower Qa in the Box window and click on Adavanced than click on Qa an lower that number.
This advice came from Kristian Ougaard of UniBox fame.
http://www.pvconsultants.com/audio/boxmodel/unibox.htm
As for Daryl's filter I've played with it in WinIsd and he replyed that what I did was wrong.
Changing Re in the Signal box and adding a 1st order cap under the Eq/Filter box and changing the Hz for a high pass filter.
Although Jeff Bagby replied it can be done in his PCD spreadsheet almost the same way.

Re: Aperiodic is a waste of time. link to filter

Posted By: philip ARCARIO <philipARCARIO@YAHOO.COM>
Date:
Friday, 14 March 2008, at 8:50 p.m.

In Response To: Re: Aperiodic is a waste of time. (Rudy Jakubin)

> For an Aperiodic box you can lower Qa in the
> Box window and click on Adavanced than click
> on Qa an lower that number.
> This advice came from Kristian Ougaard of
> UniBox fame.
>
> http://www.pvconsultants.com/audio/boxmodel/unibox.htm
> As for Daryl's filter I've played with it
> in WinIsd and he replyed that what I did was
> wrong.
> Changing Re in the Signal box and adding a
> 1st order cap under the Eq/Filter box and
> changing the Hz for a high pass filter.
> Although Jeff Bagby replied it can be done
> in his PCD spreadsheet almost the same way.
I built daryl's filter it works. I did overkill on all components and it cost under 60 bucks. look at first picture in this link. the caps cost me 40 and the resistors cost 12

http://s237.photobucket.com/albums/ff73/philipma1957/Dayton%20rss390hf-4%20%20build/

I disagree with Darryl.

Posted By: NepaEric <nepamuskiehunter@yahoo.com>
Date:
Saturday, 15 March 2008, at 7:39 p.m.

In Response To: Re: Aperiodic is a waste of time. link to filter (philip ARCARIO)

While his filter may, in fact, work to solve problems, an aperiodic "Variovent" can also obtain very good results in a smaller than optimum box. It can also relieve pressure buildup on one side of the cone, causing "oil-canning," and other problems that electrically, you cannot solve. More often than not, aperiodic enclosures come to be as a result of using a driver in a box volume which is less than suitable, or the builder chooses to in order to obtain a similiar result that would require a larger box.

Regards,
Eric

Re: Can aperiodic be modeled in WinISD

Posted By: Jeff B. <jeffb1836@yahoo.com>
Date:
Friday, 14 March 2008, at 9:19 p.m.

In Response To: Can aperiodic be modeled in WinISD (JeffKnob)

> I have two Dayton RSS390HF subwoofers each
> in a PE 3.0 cu.ft. enclosure. I know this is
> a little small for this subwoofer. Is there
> a way to model this enclosure and subwoofer
> combination with maybe 3-4 of the Scanspeak
> vents? As expensive as the enclosures are I
> would hate to cut 3-4 holes in the
> enclosures without having some idea of what
> it will do to the sound.

Yes, in WinIsd, model a sealed enclosure with a Ql set to 3 and you will have an Aperiodic box model. It's nothing tricky or magical, and it's only use is to flatten a peak in an otherwise high Qtc sealed alignment.

You can't really model an aperiodic box

Posted By: OldMongrel <mongrelaudio@comcast.net>
Date:
Saturday, 15 March 2008, at 12:03 a.m.

In Response To: Can aperiodic be modeled in WinISD (JeffKnob)

> I have two Dayton RSS390HF subwoofers each
> in a PE 3.0 cu.ft. enclosure. I know this is
> a little small for this subwoofer. Is there
> a way to model this enclosure and subwoofer
> combination with maybe 3-4 of the Scanspeak
> vents? As expensive as the enclosures are I
> would hate to cut 3-4 holes in the
> enclosures without having some idea of what
> it will do to the sound.

There is no "aperiodic box theory". If you look up just "aperiodic" on the AES site...

http://www.aes.org/e-lib/institutional.cfm (scroll down to see the search section)

...you will find three matches, and none are close to what we would need. (This is from about 11,000 articles spanning 50 years).

I think one reason for this is the difficulty of measuring or predicting the characteristics of the damping material in the vent. There's probably a whole list of parameters, and you would need to know all of them. If you did, and could plug it all into a formula, it would be great. Of course, even then, if you stuffed the vent too tightly or loosely (compared to the measured or known properties of the stuffing) or whatever, everything would change. The stuffing would probably not meet spec anyway. Look at all the drivers out there now that vary widely from sample to sample.

WinISD might indeed be useful for determining the box size if you can devise a damped vent to give you the enclosure Q changes you want. You would have to do the vent by trial and error. There are also sites out there that will give you some interesting rules-of-thumb. (They might present them as established science, so be careful).

This is not to say that aperiodic is no fun. I'm building a pair right now. They will be of the sealed type (damped vent mounted in an internal partition, similar to the Dynaco A-50.) I don't like the simpler external damped vent because there should be no vent radiation with aperiodic, if I understand the basic idea. If there is, you just have a quasi-randomly (and poorly) designed vented system.

I hope I'm wrong and that someone points me to an aperiodic design program or set of formulas that takes all the necessary variables into account. I bet the stuffing-measurement process that would need to go with it would be about as involved as driver measurements. Or, maybe that would be part of the program. You punch in the driver parms and box size and it spits out "126 grams of Acme #2 polyester evenly distributed in a 4" ID x 8" ABS tube" to get the Ql you want.

Let us know if you go with the Scanspeak vents. I'm curious about what you would get.

Re: You can't really model an aperiodic box

Posted By: Æ
Date:
Saturday, 15 March 2008, at 2:15 p.m.

In Response To: You can't really model an aperiodic box (OldMongrel)

I would have to agree with you substantially.
But you could always reverse engineer an aperiodic alignment.

I did an "Aperiodic" once. I used some PEERLESS TP165 polypropylene woofers with rubber surrounds, in some pre-made ready to go oak veneered enclosures I bought from SPEAKER CITY. The enclosures weren't very big, satellite sized. Maybe only about .4 cu.ft. each.
Anyway these enclosures had precut holes for installing woofer, tweeter, terminal and port tube. Instead of installing a port tube though, I installed a piece of perforated metal "screen" in the hole. I then added poly pillow stuffing to the enclosures, screwed in the drivers and ran impedance curves the old fashioned way with a constant current source and VTVM. I continued to add or subtract stuffing until I was able to get the flattest impedance possible. It was very time consuming to have to keep removing and installing the woofers in order to change the amount of stuffing in the enclosures. If I remember correctly I was able to get the impedance at resonance down to something like 12 ohms!

The way I did it was very empirical. And the results turned out good.

Re: You can't really model an aperiodic box

Posted By: OldMongrel <mongrelaudio@comcast.net>
Date:
Saturday, 15 March 2008, at 11:49 p.m.

In Response To: Re: You can't really model an aperiodic box (Æ)

I'll be using a similar method. I like to tune vented enclosures empirically, as well, although the theory gets you close enough to do the job *if* you don't make a mistake (I do that a lot).

John

Re: You can't really model an aperiodic box

Posted By: Æ
Date:
Sunday, 16 March 2008, at 1:25 a.m.

In Response To: Re: You can't really model an aperiodic box (OldMongrel)

> I'll be using a similar method. I like to
> tune vented enclosures empirically, as well,
> although the theory gets you close enough to
> do the job *if* you don't make a mistake (I
> do that a lot).

Mistakes aren't that bad, as long as they are correctable. As long as you learn from them.
Wishing you success. Maybe when you are finished you could post the results and/or some pictures.
Thanks
Æ

Re: Can aperiodic be modeled in WinISD

Posted By: Feyz
Date:
Saturday, 15 March 2008, at 1:39 a.m.

In Response To: Can aperiodic be modeled in WinISD (JeffKnob)

My understanding is variovents were for drivers that had very high Qts, like some old Dynaudio woofers. The variovent was providing a way to lower the sealed box Q for these drivers within a reasonable box.

RSS390HF has a low Qts, it doesn't need this treatment. And the flattening of the impedance peak which variovents provide is actually not a good thing unless you are using passive equalization. Because it means more current will be drawn from the amplifier by the driver. This can lead to more distortion.

I used an aperiodic vent in the same application..

Posted By: NepaEric <nepamuskiehunter@yahoo.com>
Date:
Sunday, 16 March 2008, at 12:49 a.m.

In Response To: Can aperiodic be modeled in WinISD (JeffKnob)

...and obtained a much more satisfactory result using a simple (but effective, and inexpensive) method of concocting my own "Variovent," in the form of compressed fiberglass in a roughly 2-1/2" hole in a 2.9 cubic ft enclosure for my RSS390HF driver. I also used 3 lbs of fiberglass in the box, which I believe is superior to Dacron polyester fiberfil or Acoustastuf. This will also help increase your "effective" box volume and increasing internal damping.

In retrospect, I believe I would have obtained a better result for that driver in a 4.25 cu ft box, as it modeled well for this, (or even a vented alignment in a much larger box, yet.)but the 3.0 ft3 box you have ought to obtain a similiar result as mine if you do as I have with it. You need to allow that sucker to breathe! That was my problem, and it sounded like crap before I did the following:

3# of glass lining all 6 interior walls, and cut a hole in the box, obtain some window screen mesh, staple on the inside and outside with a "sandwich" of fiberglass in a slight compression. The difference, for me, was like night and day. I thank and credit Pete Schumacher for my success and advising me on this. I really blamed alot of the problems on the driver, and it is really a pretty nice unit when utilized correctly. Hope this helps.

Best Regards,
Eric

1 comment:

ras said...

has any tried an AP MAT