A minimalists baffle for the B200 (or any other 8" BB)

Rudolf · « **on:** 12 Mar 2007, 11:20 pm »

The biggest obstacle for a wider diffusion of Open Baffles in the home are those big sheets of wood protruding into lovely living rooms. Is there any way to keep OBs dear to the eye and near to the wall?
There is a guy who´s done it and it looks like this:

This Open Baffle of 30x120 cm size with the driver at ca. 75 cm height has to be placed no more than 30-10 cm from the wall. And it has been designed to be crossed at any HT-receivers X-over frequency to a subwoofer.

How is it done? To understand it´s working principle we have to look a little bit into dipole theory:

Any loudspeaker that close to a wall can´t be understood without it´s mirror image. In this case what the SOS (Small Open "Schallwand", as it has been named by the inventor) forms with it´s mirror image is a "longitudinal (or linear) quadrupole" (LinQ).
You can learn everything about "linear quadrupoles" from Dan Russell: http://www.kettering.edu/~drussell/Demos/rad2/mdq.html Look for the last demo down the page and be sure you recognized the near field directivity pattern shown at right.
For an in-depth explanation Dan Russell povides this: http://www.kettering.edu/~drussell/Publications/MDQSources.pdf.

So what is the catch in it? The radiation pattern of a LinQ in the nearfield (and in normally sized rooms any LinQ does - at least at low frequencies - work in the nearfield) is a double eight. So it radiates into the room - but along the wall too. And the distance conventional OBs need to the wall is needed by the SOS to objects along that wall.

Looking at my watch - it´s after midnight already. So I will close for now. Expect me to continue with some more relevant data in some hours.

Rudolf

SET Man · « **Reply #1 on:** 13 Mar 2007, 12:28 am »

Hey!

Fascinating

What is xover between the main drive to the sub anyway? With baffle that narrow it must be pretty high isn't it?

Oh! BTW... thanks for the link. I love the animated ripple demo graphics. Dude.... that is very psychedelic fantastic!

Take care,
Buddy

MJK · « **Reply #2 on:** 13 Mar 2007, 12:59 am »

Rudolf,

Have you run the simulation of the design? That would be an interesting result to see the SPL response of this narrow baffle and the interaction with the rear wall. Should be easy to calculate.

Martin

JohninCR · « **Reply #3 on:** 13 Mar 2007, 04:36 am »

Rudolph,

It appears that someone is pulling people's legs trying to call that setup a LinQ. Out of curiousity I set up a LinQ with a pair of 15's, and yes they exhibit the pictured dispersion pattern, but the output is very greatly reduced. That setup appears to be no more than prudent placement to use the rear wave to offset the early roll-off due to the narrow baffle of the main unit with a shallow U-baffle used for added bass augmentation. The only problem I see is the early reflections of the higher frequency content of the rear wave at 2 milliseconds or so. I've found that kind of placement can work with judicious rear wave damping.

Too bad a real linear quadrupole is so terribly inefficient, otherwise it would be a great path to a simple and very narrow open baffle woofer setup including push/pull and mechanical cancellation. Just put 2 drivers on a flat baffle wired out of phase with the edge directed at the listening position.

Rudolf · « **Reply #4 on:** 13 Mar 2007, 01:48 pm »

Quote from: MJK on 13 Mar 2007, 12:59 am

Have you run the simulation of the design? That would be an interesting result to see the SPL response of this narrow baffle and the interaction with the rear wall. Should be easy to calculate.

Martin,
a simulation for different distances from the front wall shows, that at least up from 150 Hz this narrow baffle close to the wall can hold its own:

Sadly the associated polar response diagram always shows a perfect dipole eight regardless of the distance to the wall

Would you mind to comment on that?

I add another view of the front wall to show the minimalism of this alignment to the full extend. Image a 2 m wide window extending the wall to the left and the wall extending 2 m into a staircase to the right. The baffles are slightly toed in.

MJK · « **Reply #5 on:** 13 Mar 2007, 01:56 pm »

Hi Rudolf,

Right now the polar response plot is in the free field so all three should look the same. To upgrade it to include room boundaries will be tricky. For example, how do I treat a response radius greater than the distance to the wall? I included the plot to give the user some idea of how much sound energy is being radiated in all directions. Creating the polar plot was an interesting math challenge. Like everything else, the polar response plot will evolve and future worksheets will increase the mount of information included in the polar calculation.

One interesting trick for a multi-way OB is to define the system lying on its side and then the polar response plot will produce the radiation pattern associated with the crossover.

Hope that helps

Martin

JohninCR · « **Reply #6 on:** 13 Mar 2007, 04:06 pm »

Rudolph,

The wall is more like 40-50cm away from the baffle. The base is deeper than it is wide. plus there's another 8-10cm from the base to the wall. It is an attractive setup, and I'm 100% in favor of getting rid of monster sized baffles. I wish we could get Martin on board with that idea too. His are so big that they're somewhere between IB and dipole.

MJK · « **Reply #7 on:** 13 Mar 2007, 04:32 pm »

Quote

I wish we could get Martin on board with that idea too. His are so big that they're somewhere between IB and dipole.

They are big and I have learned a lot designing and building them. The next pair of speakers I build will be smaller, but I am not sure if they will be an OB. All I can say is that they were a no compromise design and I have enjoyed their performance in particular the bass output.

scorpion · « **Reply #8 on:** 13 Mar 2007, 05:44 pm »

I did the same simus with fullrange B200 values as Rudolf with MJK's Worksheet. I also got the same result. Best with 20 cm (8") from wall 3 m (10') from listener. But I think it is also worth mentioning that the results are quite sensitive to these distances. Once you do any alteration the simu can change
quite a bit. So not only distance from the wall is critical but also distance to the listener.

PS. I also removed the carpet as in the picture.

/Erling

Rudolf · « **Reply #9 on:** 21 Mar 2007, 01:47 pm »

I got some time to carry on with this thread.

Since someone was curious: The fullrange driver shown in the SOS picture is a (discontinued) Visaton TL 8/C 50 F which was manufactured by Coral.

The bass driver in the minimalist OB is a Beyma 15K200 with modified dust cap.

The recommended position of the SOS relative to the wall would be like this:

Driver 10-30 cm from the wall.

The designer of the SOS recommends a correction network of a 1,5 mH coil parallel to a 15 Ohm/10 W resistor - both in line with the B 200. If crossing over to stereo subs ~200 Hz no additional baffle step compensation would be needed. For a conventional x-over @80 Hz an additional increase of 6-8 dB at 100 Hz would be necessary.

Rudolf
www.dipolplus.de

Rudolf · « **Reply #10 on:** 21 Mar 2007, 02:14 pm »

As a deeper investigation into the quadrupole character (possibly) of the SOS I did some more sims with MJKs OB worksheet. I simulated for different angles at 1 and 3 m distance:

Does anyone see any difference between both sims that suggests the proposed near/farfield difference in linear quadrupoles? http://www.kettering.edu/~drussell/Demos/rad2/mdq.html

In theory for nearfield conditions SPL at 75° should be much lower than at 90°. And in the farfield radiation at 75° should be much lower than would be expected for a regular dipole.

Please keep in mind that in this case the nearfield/farfield transition is not only a function of listening distance but of frequency too.

Any comments?

Rudolf

JohninCR · « **Reply #11 on:** 22 Mar 2007, 02:15 am »

Rudolph,

Don't try to make it more complicated than it is. It's just a dipole close to the wall, so the reflection has significant influence. Those sims are what I would expect, since in the nearfield the "D" of the rear wave and the reflection are quite significant in relation to the measuring distance. Doing simulations below 1khz of an OB at 1m are about as useful as measuring them at 1m, so why even bother? If you think those results unusual, you should come and measure my U-baffle sub that is very close the the wall. At 1m I think it may have more output at 90deg to the driver plane than directly in front.

scorpion · « **Reply #12 on:** 22 Mar 2007, 10:27 pm »

I have been a bit sceptical over Timo's SOS-design but the proof of the pudding would of course be to test oneself. That is not so hard. But the more fundamental question would be: How accurate are the simulations we produce ? We have a front (back) wall distance and a possibly damped floor, full speaker data and a flat bafflesize with specified speaker placement and not anything more. From that we try to estimate response at a certain point in (I suppose) a room. How accurate can that be ? Yet MJK's MathCad sheet is the best there is.

JohninCR, haven't you started measuring ? I was of the impression you had all the equipment. Myself I fixed a perfectly good measurement set up from a Panasonic WM-61A mic module and the free ARTA software. All in less than two hours from building start to measurement accomplished.

/Erling

JohninCR · « **Reply #13 on:** 23 Mar 2007, 12:02 am »

Erling,

No not yet. I didn't know there was something free and easy, so I was waiting for Thuneau's system that will tie in directly with the XO system (kinda measuring and XO for dummies). I've just been doing a bunch of building in the meantime. I'll probably check out ARTA in the meantime. Thanks for suggesting it.

I don't doubt the results of the setup, because with that kind of placement the rear wave reflection is going to enhance the midbass, so the augmenter doesn't have to run as high as it would out in the room. From the nearwall placements I've tried, I know I'd want to attenuate the HF stuff on the backside because I like good imaging. That's the whole reason I continue trying different things....I want to obtain the best possible imaging while retaining the most open OB sound (IOW the perfect speaker, at least for me, in my room).

Rudolf · « **Reply #14 on:** 23 Mar 2007, 10:54 am »

Quote from: JohninCR on 22 Mar 2007, 02:15 am

Rudolph,
Don't try to make it more complicated than it is. It's just a dipole close to the wall, so the reflection has significant influence.

Exactly, John! And the usual means to simulate reflections is a mirror source.

If the distance to the mirror source equals the dipole distance D, we have a linear quadrupole. As simple as that.
I could give the impression to be a "simulation guy", but I love simulations only because they work under predefined assumptions. So results are not disfigured by the side effects of real life. I would never expect to see my loudspeakers working exactly like the worksheet predict.

Regarding the driver-to-wall-distance Timo comments, that the driver has to be closer than 30 cm to the wall or farther than 80 cm away to avoid bad imaging. This is quite in line with the scientific research results for early reflections that I found on the net.

Fortunately the dipoles I build for myself behave according to theory - my latest effort being a good example:

It´s a 15" chinese PA driver in a 45x45x42 cm cube.

The 40x40 cm concrete slab as top of those H frames has been the most massive and effective "tranquilizer" I ever applied to a project.

JohninCR · « **Reply #15 on:** 23 Mar 2007, 09:05 pm »

Rudolph,

I like the concrete top. Have you tried any properly damped U-baffles? If so, what are the reasons you still go dipole?

Regarding the dipole near a wall, I simply can't buy into the quadrapole thing. That would mean virtually any dipole woofer in a typical placement is a quadrapole. I'll agree that they behave similar in some ways, and just agree to disagree otherwise. You don't call a monopole in a corner, a horn, do you? Also, I'm sure you wouldn't dream of running real a quadrapole full range.

I believe the idea works in the bass and midbass, but otherwise I share scorpion's skepticism. While reducing the reflection down to the 1ms time range may help imaging, it's still going to be a strong destructive influence similar to very early side wall reflections. If it sounds good now, I believe it will sound even better with a layer of polyfil batting and/or open cell foam over the back.

Rudolf · « **Reply #16 on:** 24 Mar 2007, 09:53 pm »

JohninCR,
I am not into this SOS with my hearts blood.

It was meant as a way to explore for those, who are radically limited in baffle width for whatever reasons. Which we both fortunately are not.

So there certainly is much truth in the limitations you are discussing for the SOS. And as I have learned, treating it as a linear quadrupole may be a nice academic exercise (for me), but with little impact and value for real life. So we probably can advance to other fields which are more at our hearts.

Like H- and U-frames.

I´m just contributing to some interesting new threads in german forums about cardioids and whether they should better be done as U-frames or as a dipole-monopole-combination. If we come along with a simple and effective design, I will surely test it for myself. But I don´t see U-frames as real "open baffle speakers" any more and I don´t see them as anything "better" than dipoles. Just different solutions for different problems.

BTW: If you should know any "properly damping" for U-baffles that covers more than 3 octaves with constant damping properties I would be extremely interested.

Rudolf
www.dipolplus.de

scorpion · « **Reply #17 on:** 24 Mar 2007, 11:15 pm »

Rudolf,

Good points. I know JohninCr has his 'Flintstones' so I think yours must be the 'Basstones'. Clever design.
But I mostly agreee with regard to the U-baffles. I do not think anyone is constructing dipolewoofers exclusively for subwoofer use. They can of course be but will have a hard time going into the sub frequencies with strong output. So dipolewoofers are mostly for music and as such should be useable up to at least 250-300 Hz whithout any markable sideeffects. At least that is my view to be able to crossover to any other speaker element I would like to. Landing in that conclusion only flat baffle and H-baffle has fullfilled this request for me so far. I am into testing the MJK type 'The Topless U-baffle' next. I would like to give a link to a wellknown Swedish audiocharacter and his OB-solution: http://www.svalander.se/anlag/baffel.htm (only in Swedish but from the pictures you will be able to evaluate the setup I think), quite minimalistic I would say !

/Erling

JohninCR · « **Reply #18 on:** 24 Mar 2007, 11:55 pm »

Rudolf,

I was just wondering if you tried U's and found problems. I wouldn't say better or worse, just smaller.

Regarding the term Open Baffle, I wish we could come to a consensus. Since there is no other term, I prefer to use it as the all encompassing term for all alignments where the front and rear output of the driver is exposed to the world, and the intent is to avoid changing the output through manipulation of the rear wave or create resonances like a BR, TL or RLH. Then we have 2 types of speakers, boxes and open baffles. The subsets within OB include everything from flat baffle dipoles to U baffles and folded alignments like W's and ripoles. If OB isn't the term, then we need to coin one, because they all share one thing in common, in freespace bass cancellation occurs below Fequal which is determined by "D" regardless of whether the baffle is flat or folded.

maxro · « **Reply #19 on:** 25 Mar 2007, 03:41 am »

Quote from: JohninCR on 24 Mar 2007, 11:55 pm

Regarding the term Open Baffle, I wish we could come to a consensus. Since there is no other term, I prefer to use it as the all encompassing term for all alignments where the front and rear output of the driver is exposed to the world, and the intent is to avoid changing the output through manipulation of the rear wave or create resonances like a BR, TL or RLH.

The rear wave is somewhat "manipulated" in a U baffle, whether intended or not.

So, how about "asymmetrical dipole" for U baffles and "symmetrical dipole" for flat and H baffles (I don't know where that leaves W and N baffles or Ripoles). It does seem more than a bit awkward though.