[rick57]

Hi

I’ve tried to research the effect:  driver Qt + baffle size =  Qts?
 . . but haven’t found anything yet
(I only seek to cover the ‘midbass’ range, from my Phoenix woofers, up to 350-450 Hz - I have ideas I’m comfortable with to cover the rest).

I want to use a spare pair of high efficiency midbass drivers I have, with Qt between 0.24 – 0.33.
Baffle size will be what WAF compromise allows, aiming/ hoping for a height of about 900-1000 mm (36-40”), with a fixed baffle + wings flattened out to about 1200 mm ie 48 inches wide.
Active EQ to flatten humps/ rises. I have no idea if Qt will go up or down (would be nice if there were a formula)??

With driver Qt of eg 0.3, it probably would be good if Qts went up to 0.5 if that optimizes impulse response, if other things are equal (are they?).

Does anyone know the relationship, does MJK’s spreadsheet model this??

Thanks

[maxro]

I think the formula for calculating Qt on a flat baffle would be Qt/1. 

In other words, driver Q remains virtually unchanged.

[tubamark]

Hi

I’ve tried to research the effect:  driver Qt + baffle size =  Qts?
 . . . With driver Qt of eg 0.3, it probably would be good if Qts went up to 0.5 if that optimizes impulse response, if other things are equal (are they?).

Rick,
Maxro's right. Flat (or nearly flat) baffles leave Q 'bout the same. Maybe a smidge higher (.0__ more) if mounted low to the floor.
However, before getting too attached to the .5 target, I would suggest a careful reading of John K's Q-versus-room articles:
http://www.geocities.com/kreskovs/Box-Q.html

The ol' .5 target is really most useful for a monopole working below the lowest room resonance, where speaker rolloff is compensated for by room pressurization ("cabin gain").  With a Dipole, cabin gain doesn't happen.  If your lowest room mode is above your driver resonance, you'll never even hear the Fs in your room with a dipole.

I don't want to ruffle any purists feathers, but I would wager that double-blind testing would prove that most folks cannot hear a 1 or 2 db bump at an Fs below, say, 45 Hz.  I'd also wager that under double-blind tests most would choose clean dipole bass with a Qts as high as 1.0 over a .5 monopole producing either lean, or flabby bass (depending on room dimensions).

Hey, lots of purists also love tube amplification, which ironically "warms up" bass response via underdamping. 

In the end, there are no wrong answers, just preferences.

[rick57]

Thanks guys,

I've heard this question addressed before and not really grasped it, but now it's clear.   

I’m open to either tube or SS amps. As I doubt there’s yet a formula linking a tube amp’s output impedance or the damping factor (incorporating the driver’s Zmin) to Q, other than in general ‘direction’ I guess it’s one to find what you like by experimenting.

Cheers

[mcgsxr]

Not sure I qualify as a purist (!) but I do prefer my b200 OB with the Magnavox EL84 amp - great synergy with the driver, and the best bass I have heard.

I also believe that it does relate to the nature of the b200, the connection with the damping factor of the old amp, the room size and shape, and the idiot with the ears!

[hurdy_gurdyman]

I also believe that it does relate to the nature of the b200, the connection with the damping factor of the old amp, the room size and shape, and the idiot with the ears!

"and the idiot with the ears!"
This may be the single most important factor of any choice of audio components.

Dave

[dewar]

I've wondered about Q for some time now and still dont understand it completely.

If one can end up with flat FR through the use of EQ, is Qt still an issue?

I'm looking at subwoofer driver to make H-baffle dipole subs, with a Behringer BFD to flatten response. I'm thinking of using some Peerless XLS's with Qt=0.2. Seems low but they are what Linkwitz uses in his W-baffle dipole subs so I figure I cant go too wrong.

Cheers,

B.

[ttan98]

Use Martin King MathCad spreadsheet, it will tell you the spl vs freq. response curve, if you input the necesssary parameters like, baffle width, x-over, no of drivers, etc. You don't have to guess or worry whether Qts increased, decreased, or remains unchanged.

By adjusting the baffle size and other parameters the output will give you the frequency response and directivity display.

Go to his website, the s/w is only $25 for DIYers. Look for project 7.

http://www.quarter-wave.com/

cheers...

[johnk...]

Qts isn't effected much by a baffle. Peerless used to piblish driver specs with both baffled and unbaffles Q. I don't seem to be able to locate any of the old data sheets but as I recall the variation in Qts was less than typically what you would measure with difference signal levels. There is a slight increase, due to an increase in the air mass load I believe.

On the other hand, if you are interested in open baffle performance you might look at this.

http://www.musicanddesign.com/A_B_C_Dipole.html

[rick57]

John,

Exciting to hear of your A, B, C, Dipole package - for folks outside the US (I'm in Australia postcode 3000), what is the cost of either the ABC or the special with both ABC  and the SoundEasy Design Guide ?

>Qts isn't effected much by a baffle.
Thinking that through, then is there an ideal/ target Qts or range for use on open baffles?
I recall your study on Q in sealed boxes www.geocities.com/kreskovs/Box-Q.html, concluding essentially IIRC that Q is not as important as the specific room it’s used in.

Personally I’m working on design where I’m now specifically interested in midbass, > 90-120 Hz, ie not deep bass extension.

If Q is not so important, and you have some low distortion low Q (0.33) drivers, if they are used with a tube amp with an output impedance that might be 15-20 times higher than a SS amp, any thoughts on how the resultant Q would perform?

Thanks

[johnk...]

Shipping anywhere is included.

As for Q, understand that in the context of ABC Dipole, for woofers with Qts<0.5 the woofer will be actively equalized to have a Q = 0.5 and further equalized for flat response. If Qts>0.5 only active eq for flat response is appled and the the system Qt and Fs remain at the driver's values (or driver + amp's effect). In this regard, low Q low Fs drivers are more flexible in that you can shift the cut off frequency around, but they may require greater equalization. Room interaction, in relation to the article you are referring to, isn't as much of a concern because dipole woofers do not benefit from room pressurization as sealed box woofers do.

[rick57]

Hi John,

Could you comment on high output impedance tube amps, wrt any assumptions within ABC Dipole?  (Not for the bass, but for the mids, ie +100 Hz).

With www.musicanddesign.com/A_B_C_Dipole.html

“To order, PayPal $10.95* US to sales@musicanddesign.com  and
request A, B, C, Dipole or contact Music and Design directly for
information on commercial license.
Introductory special! Order both A B C Dipole and the
SoundEasy Design Guide  for only $17.95.
Offer good until April 30, 2007”

What enhancements or changes to ABC Dipole are likely;
(and at what charge will the introductory people be able to acquire them)?

Thanks

[ttan98]

John,

I am also interested in the purchase (ie both A,B,C dipole and sound easy documentation), I am about to acquire the latest version of sound easy version 13.

Normally the purchaser of software will get free updates for a certain period(not sure: 12 month for eg.) does this applies to your software as well?

Please explain....

[rick57]

Hi John

I read the page on ABC Dipole.
I have two basic questions, re the XOs and woofers

Options for the XOs seem to me to include :
 - The need for CAD filter package - is to design a PCB? ~ for those who haven’t used one, and it’s likely to be a one off, maybe could you suggest a package on your web page,
 - though SE has this capability?
 - if you use are prepared to buy a DCX, it has enough power to handle all the filters? (I’d imagine so)
 - Or if you had access to a sufficient number of filter boards from BobEllis’s Active filter board group buy
 “Each board will be suitable for a one channel two way will contain provisions for:
   . input buffer that can include BSC
   . 4th order HP and LP
   . 2 Notch/Peak filters
   . one output buffer with variable gain
   . 2 All pass sections 
     www.diyaudio.com/forums/showthread.php?s=&threadid=74420&highlight= I’d also think that should work.

And for those who already have dipole woofers, W frame - could they be incorporated, probably with some experimentation on component values affecting their upper rolloff?

Thanks

[FlorianO]

I don't want to ruffle any purists feathers, but I would wager that double-blind testing would prove that most folks cannot hear a 1 or 2 db bump at an Fs below, say, 45 Hz.  I'd also wager that under double-blind tests most would choose clean dipole bass with a Qts as high as 1.0 over a .5 monopole producing either lean, or flabby bass (depending on room dimensions).

Hey, lots of purists also love tube amplification, which ironically "warms up" bass response via underdamping. 

I don't quite follow the argument ... If one uses a SET with a low damping factor than it would need to be balanced by employing "lean" drivers with small Qt since large Qt drivers will only compound "the problem".

Or am I missing smth ?

[johnk...]

With regard to tube amps, ABC Dipole design calculations are based on the driver's Fs and Qts. If high output impedance amps are used then appropriate adjustment should be made to the values of Qts. If you ignore the  adjustment some slight misalignment will occur depending on the actual output impedance of the amp.

No enhancements are currently planed. The offer for free updates until September 1, 2007 is for protection of initial purchasers, in good faith, in case some kind of bug in the initial program is discovered. That offer may be withdrawn at any time. The package deal for both the SE guide and ABC will remain until April 30, 2007.

In general updates, if any, will be at a fraction of what ever the current price is mostly to cover shipping and materials.


You do not need a PCB CAD designer. The circuits are all designed and laid out (schematics) in the Excel tool. The Excel tool allows you to correctly tune the circuits to your application and then computes the required component values. I have circuit boards available for building the circuits, but you are free to look elsewhere too. The board I am offering will come with complete instructions (though I haven't written them yet. )for implementing the designs. If you want to design your own boards that is another issue.

You would only need something like LspCAD or SoundEasy to design the midrange/tweeter crossover in a 3-way. But neither code can be used to design PCB's.

You do need to be able to measure SPL to correctly design the midrange equalization. I provide a detailed discussion of how to make the midrange measurements in the hypertext guide part of the package. You do not need to measure woofer SPL to design the filter of equalization. But near field woofer SPL measurements are required to tune a U-frame or to look at the possible need for a notch filter with an H (or W) frame woofer.

With regard to the DCX, it certainly should have the required number of filter stages and the midrange eq/HP filter should pose no problem. The only potential problem with the woofer is that correct equalization for a dipole woofer can require more than 15dB of gain which is the limit for the DCX. So I guess you would have to factor some eq stages into parts to accomplish the requires gain. Not being a DCX user I can't say for sure.

If Bob's boards support the required circuits they would be fine. On the other hand, the boards I am offering are designed for these circuits. One board per channel. My board do not include delay circuits as I do not like them. They aren't really the correct thing to do for a dipole system. http://www.musicanddesign.com/Dipole-offset.html

From the equalization perspective a W frame is no different than an H frame. You would follow the H frame design and then, if necessary, add a notch filter to kill any potential W frame resonance peak. While I didn't address W frames in  ABC (maybe I should?) I do address the use of notch filters for the woofer system, if required, and my PC boards have the provision for them.

I hope this answers your questions. The intent of ABC is to provide a straight forward, relibale dipole design procedure which will yield reliable, high quality results. But it's not the only possible approach.

[rick57]

Hi John

> I hope this answers your questions.

All except one not specific to ABC Dipole, so I’ve just the money . . 

the question  . .
> If high output impedance amps are used, then appropriate adjustment should be made to the values of Qts.

 . . which maybe anyone could chime in on:

Here’s an article which tells how to adjust Qts in a “Sealed box design” www.winebase.com.au/audio/index.html
Am I right that the math would be the same for driver Qts in a dipole?

Cheers

[johnk...]

Qts = Qms x Qes /(Qms + Qes)

To correct Qtc for high output impedance, or other resistance in series with a driver,:

Qes' = Qes x (Re + Rs)/Re,

where Re is the DC R of the driver and Rs is the series resistance (amp output impedance, etc).

Qts' = Qms x Qes' / (Qms + Qes')

And thanks for the order, Rick.

[rick57]

Hi John

Thanks for explaining how to correct Qtc for high output impedance, that had me baffled 

Cheers

[rick57]

Hi John

I intend to use Aurum cantus ribbon tweeters that I already have (larger ones that can cross c 1500 Hz) , are there any limitations using ABC Dipole or dipoles generally with ribbons?

And does ABC Dipole model rear firing tweeters?

Thanks