Suggestions of using dipole subs with electrostats?

[Capaciti]

Hi Folks,

here are my experiences and opinion about dipol-subs with ESL:

1- A dipol sub works very well with ESL, since the radiation doesn't "break" from dipol to monopol at the cross-over frequency.

2- An ESL need to be positioned 4-6 feet away from backwall for best performance and fortunately this is exactly the distance a dipol-sub needs to work properly. But be aware, that a dipol will miss 30-60 Hz range completely if not positioned perfectly.

3- I tried many dipol-subs and imo the most important thing is the problem of cavity resonances. If you want to gain efficiency at lower frequencies, you need to increase the size of the baffle. Most concepts like Linkwitz and ripol increase the width by back side extensions. This is more friendly regarding visual aspects, but will create resonances depending on the size of the extension.

4- Most concepts show resonances at 150 - 250 Hz with peaks up to 15db. Some people control that resonce by passive of active filters. Imo such means are just damping the resonance, but it is still there !

5- Such a concept should be used up to 70-90 Hz only, but than an ESL with large area and low fundamental resonance is required. Due to acoustic coupling the woofer will blow the membran into the stators even at moderate levels. This problem increases if you place the dipol-sub on the left or right side adjecant to the panel, its less if the panel is above the woofer. There should be a certain air gap ob the baffle between the woofer and the Panel.

6- A cross-over frequency of 200-250 Hz is preferrable, since the compliance of the ESL-membran can be much lower and thus reducing/avoiding acoustic coupling. In addition an ESL shows lowest distortions from 200-20000 Hz, while distortions increase significantly at < 150 Hz.

7- Doing this you should avoid any possible resonance of the dipol sub. There are two choices: Increase the width of your baffle, or keep it limited and loose efficiency

8- One of the best dipol-subs i ever heard is the "Gradient" for the Quad-ESL. The concept has been very consequent. Due to the width of the Quads,it allowed the implementation of two adjacent 12" woofers. One woofer is mounted with the magnet to the front, in order to reduce distortion.

9- Currently i am playing with dipol-subs as a line-source array positioned on the left or right side of the panel. One array is 15" width and 6feet height and includes eight 8" woofers. Driven with an powerful amp,the performance is simply breathtaking.

10- And the last advises are:
Use as most membran area and excursion you can get for a dipole-sub and select the type of woofer regarding distortion level at frequency range 30-300 HZ. The SLS 10 has a perfect relation between cost and low distortions, while cheap drivers just save money !!!!!!!
People tell you, that e.g 100 watts are sufficient to drive a dipole-sub. Forget about that ! Due to its principle a dipole requires area, excursion and power,power,power........

Regards,Capaciti

[Calvin]

Hi,

most Qs are answered already by Capaciti. Spares me some work

My curent project is quite similar to capacties. I´ll use 8 18cm drivers per channel in a slim tower (size: 20cm W, 20cm D, 180cm H). Concepted as 4 stacked modules, each with 2 drivers mounted to it, the footprint is very small, yet high. This should be advantageous, because this arrangement leads to a dipolar cylindrical distribution pattern quite similar to that of the panel. So one of the two reasons for bad integration of the bass with the ESL -change of distribution pattern- is omitted. The second reason -different acoustical filter character --> nonlinear group delay- is a question of filter design.
Hope to provide You with pics soon

jauu
Calvin

[nilsomat]

@Calvin: I hope i am not pushing but would politely ask for informations regarding your SLS12 design, in case you might have overlooked my question in my last post. Email would be very much appreciated!
Greetings,
Nils

nilsp (at) gmx (dot) net

[Brian Beck]

A random thought about using dipolar subs close to ESLs:

At any point in the plane of the cone of a dipolar sub, there will be tangential air flow as air rushes from one side (phase) of the driver to the other side. From push to pull, if you will. The closer to the sub, the greater the effect. The direction of air flow is normal or perpendicular to the observer. So there will be no far field sound pressure, due to dipolar cancellation. If such a sub is placed near to a dipolar ESL and coplanar with it, as is often deemed desirable, the tangential air flow from the sub will tend to push and pull the ESL diaphragm back and forth away from its center, effectively modulating it. This has the POTENTIAL for reducing dynamic range and increasing distortion in the ESL. Worst case would a large baffle with several large woofer drivers placed right next to an ESL. The farther the sub can be placed away from the ESL, the less this effect, but woofer-to-ESL integration at crossover may then be more problematic. I can’t say that I have noticed a problem with my Quads either way, but I ended up with closed box subs crossed over around 60 Hz (although variable as required). This combo seems to work well.

It is often said that one should match a dipolar ESL with a dipolar sub. In the region of frequency overlap, there is a lot of validity to this rule of thumb. But sound reproduced in a room from any kind of speaker can be analyzed into several regions where wave behavior is different from region to region. The way that a 20 Hz wave develops in a room is far different from the way a 200Hz wave propagates, or compared to 2000Hz, or to 20kHz, etc. What ultimately matters is sound pressure level at the listening position (both steady state and over time). When an ESL is capable of dipping into the sub-80Hz or sub-100Hz region, crossing over to a well-placed closed box subwoofer may be the more attractive solution, not only for bass depth and power, but to eliminate modulating the ESL diaphragm. If a dipolar woofer is used, try it farther away from the ESL.

Just a thought…

[Calvin]

Hi,

acoustic coupling works with all kinds of woofers. So the danger of blowing the diaphragm into the stators is always there. The trick is to have an airgap between panel and woofer. This lessens the problem much.
Wether You use a CB or dipole is rather a matter of taste. Both principles can work excellent. I prefer the dipoles, because it was most often easier to integrate the dipole with the panels. Too they have the advantage of playing better in rooms with strong roommodes.
Today I took pics of my new TowerSubs and made a new active crossover....I guess the system -apart from room probs- it´s now playing at 85% of its capabilities. Even so we discovered some new and very fine details on well heared discs. Have You ever heard, that in Music from Madonna, there´s someone knocking on a wooden stick every now and then?
The dipole towers add to this an outstanding clear bass with high resolution without any booming at all. Just great

jauu
Calvin

[Brian Beck]

Calvin, Nice looking speakers, both the bass columns and the planar! Can you describe your ESLs?

The only thing missing is tubes

acoustic coupling works with all kinds of woofers

True, to some degree. My point was that dipolar woofers which are placed close to other planars would cause more of this effect due to greater near-field velocities than would a nearby pressure driver woofer (CB). The size of the ESL baffle is relevant either way.

So the danger of blowing the diaphragm into the stators is always there.

Probably not too much with a pressure driver though.

The trick is to have an airgap between panel and woofer. This lessens the problem much.

Agreed, the bigger the better, which was my point.

Both principles can work excellent.

Agreed.

Music from Madonna

Now we don't agree

[paulspencer]

Getting back to the original question.

If I understand correctly, the 8" drivers are blown, hence you want to replace them with a dipole. How high were they crossed? 300 Hz?

That is probably too high for a design like the Linkwitz Phoenix woofer or for John K's Nao. I believe their top end is limited to around 120 - 150 Hz. This is covered on John K's site. It's related to resonance problems.

I'd expect the best option would be to choose some affordable woofers (not subwoofers - they aren't designed to perform well as high as you will probably want to use them) and put them on a dipole which is fairly wide (as wide as you can deal with) and with small wings perhaps.

So your system might look like this:
sub - monopole up to 40 Hz
phoenix dipole 40 - 120 Hz
open baffle dipole with hifi midbass woofers 120 Hz to crossover to electrostatics 300 Hz?

This might seem like overkill, however if you already have the XLS drivers for the Phoenix dipole W frame then it spares you having to buy enough woofers to move air down to 40 Hz.

If you haven't bought the XLS drivers then I'd have a look at the Rythmik audio directservo kit as the drivers are probably at least as good as the xls BEFORE the servo is used. On an open baffle you might be able to get high enough to cross to your main panels, and low enough to cover modes.

BTW if you look at the FAQ section on the Phoenix at Linkwitz site you will find his defence of using a low Qts driver - he claims it is in fact easier to eq than a driver considered to have the ideal Qts value for dipole use.

[Capaciti]

Hi Folks,

lets take a look at ML statement:

1- panel from 400-20000 Hz , large area, small stator-membran distance, low compliance
result: high sensitivity and output, low distortion, now issue with acoustic coupling

2- 6" woofers as a dipol line source (i guess 10 pcs. each) from 80-400 Hz adjecent to panels with little air gap between
result: low distortion, cylindrical wave down to 80 Hz

3- 6 big towers, closed box with 12 pcs. 12" woofers from 20-80 Hz

I had a chance to listen - its simply the best !!


capaciti

[moray james]

Have been reading the thread and find this very interesting. Would like to ask a few questions which I suppose should be obvious but alas are not.
1) could someone please explain the rational behind the reduction of driver cabinet volume and lowering Fs? This seems counter intuitive so I guess this has to do with the two drivers firing at each other in close proximity?
2) the cavity or space between the two driver, how is this determined? I would think that the area of the sides around the drivers (at least) should equal the piston area (of one or both drivers) ? If using a single pair of drivers with the sub sitting on the floor then three sides would need to equal piston area but what about a vertical collum of pairs? The pair on the bottom only have three sides from which to vent (front back and top) where the next pair up have the ability to vent all the way around (the driver). Does this need to be compensated for? Someone spoke about symetry.
3) What about placing the two drivers as close face to face so that the gap between the two surrounds equals the piston area (again for one or two drivers?) and then have the cabinet edges cut off at 45 degrees making a conical flare. Would that help to smooth turbulance of air movement in and out of the cones?
4) Just checking but asume that driver pairs are running in push push (in phase)?
There are probably many more questions but this is a very good start so thanks to all who reply. Best regards Moray James.