Open baffle vs ESL?

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hey all I wanted to throw this out there to see what people think on which of these technologies have the best all round performance.

I'm currently torn between two designs:

1) A 250mm by 1000mm curved ESL system
2) A 330mm by 1000mm tall open baffle system using two Alpair 10 full range drivers each side.

Both will utilise a sealed sub at the bottom. As much as I want to make both, I only have time/funds for one right now.

The open baffles will be powered at 25W per driver (i.e. 50W per side not including sub). My question is, what levels of volume will I be getting with esls this size in comparison to the open baffles? I will normally listen to them at comfortable volumes, but I will need to be able to crank it to high levels when doing things around the house. I can't really make anything wider than 1ft as that will start compromising the free space available around the baffle as well as WAF :warped:.

I'm not too worried about compromising sound between these two systems for practicality and reasonable spl ;).
Using two full range drivers on a baffle is not a good idea as the high frequency vertical directivity will be very listening angle dependant. Smooth directivity calls for an acoustically small source.

Acoustically small means that the dimensions of the source are to be preferably less than the wavelength at the highest frequency of interest. The wavelength at 10 kHz is 34.4mm, which explains why a 25mm dome tweeter is often used.

You will have two tweeters spaced around 11 inches? apart resulting in the vertical pattern deteriorating above about 1.5kHz. This will result in a very narrow listening window where you will have to find a chair that puts your ears at the mid point of the two drivers!

Of course these remarks apply equally to ESL's, the difference being that the ESL is to be viewed as a line source over most of its range. Thus the high frequency pattern will be consistent between the floor and the height dimension of the ESL. Sorry to be the bearer of bad news, but physics will have its way.

For open baffle I recommend the Tang Band 1808 full range driver in the upper 1/3 of a 28-32 inch wide baffle which may or may not be tapered. About 40-48 inches tall tilt back about 3-5 degrees. For arranging an integral open baffle sub woofer look at this V-2 - V-2

There are a lot of open baffle on the GR-Research site. Since hearing the Tang Band 1808 I think it is a very smooth and non fatiguing sounding driver, and really nice bass for an 8 inch OB. It is much less expensive than the coaxial driver GR-research sells.

As for ESLs I happen to love old Acoustats but they are a handful when there are problems with the panels and power interface. They also need a high current amp or to be driven with the Acoustat Servo Drive tube amps, in which case then the "care and feeding" goes up another few notches. Many are swearing by the DIY approach to ESL's and I don't discourage but that is a bit of work to get right. There are nevertheless plenty of threads here to get you through the rough spots.

Agree with Keith that two full rangers separated by a large amount would not be good. The single Tang Band will get your ears to bleed SPL wise. Only the lowest octave needs to be reinforced with a sub preferably an OB set like the link above. I heard the Tang Band in a 3 foot wide by 4 foot tall piece of MDF propped up with fir strips in my living room driven by a Nelson Pass J2 and it was nothing short of stupid good. $400 worth of drivers slapped into a board and thats it. The bass would be fine for all but the most demanding music in a modest sided room. Pretty incredible.
Last edited:
Thanks Keith, I didn't realise that would be a problem. I suppose one full range driver per side with some kind of sub will be fine. As for the Tang Band 1808 full range driver, any idea of availability and cost in the UK?

I have a variable ESL supply already and a couple 400W tripath amps that should get me up and running for now. I have built a set of small wire stator ESLs before and they worked but weren't the best design. I've got a new design that's based on the Martin Logan curved design and should be more promising. Need to find/make the transformers though.

I have no idea in what ESLs compare to open baffles in SPL. The Martin Logan ESLs spec around 90 - 94dB which is promising.

If I do an open baffle design, I don't want to go more than 1 foot in width. These speakers will have to sit between a chimney and a wall next to the mantle piece of a fire place and any wider will make the speakers too close to the adjacent walls. How much of a wing would I be able to add to keep the sound open enough?

As for the sub, I'm looking towards a sealed sub but would OB subs give enough for the size? What about a W frame dipole sub?

if done right, the ESL will outperform any dynamic driver.
In Your case this means though to accept some restrictions. A panel of the suggested size should be crossed over not below 300Hz. It´ll be no easy task to find a matching bass partner. As Keith already mentioned the distribution character of the panel differs greatly to a dynamic FR driver and the dynamic bass. In comparison the greater the listening distance the louder the panel will play. The dynamic OB will be loudest in close distance, quickly loosing SPL over distance. Same holds true for the bass, but matters are less serious here, because of the room acoustics somewhat counter the effect.

How much SPL You can get, depends on basically three factors:
1: Listening distance and room acoustics
2: Panel construction and distribution character
3: Impedance of the panel

1) A strip shaped thin and high panel is a longthrow device. The SPL-distribution over distance is far more even than with other shapes or drivers.
2) The panel needs to be constructed as eficient as possible, meaning smallest stator-stator-distances and highest mechanical diaphragm tension as possible. These are the most decisive factors in panel construction anyway. Forget FR-ESLs here.
3) Finding the right audio transformer is a compromise-game again.
Higher transformation factors don´t increase SPL much, but effect the load the amplifier needs to drive on a much greater scale. It´s more important to find a good match between the panel´s requirements and the amps capabilities.

If done right, You can ecpect >90dB@2.83V@4m for a panel of the suggested size. This doesn´t sound much at first glance, but theoretically You need to add 12dB compared to a standard measurement in 1m distance for a global distributing driver (pulsating sphere). In praxis 6-8dB are left. You can add a further 2-3dB in comparison to dynamic drivers, because of the absence of power compression/heat losses.
So You don´t need high wattage specced amps. 20-50W@8Ohms are more than sufficient if the amp works stable.

I think I will concentrate on ESLs then. I was planning on crossing over somewhere around 300Hz. The martin logan ESLs seem to cross over at 270Hz straight to the sealed bass. Know of any drivers that would be good for the low end? I suppose I'd need to do a lot of equalising and have a sturdy enclosure if I'm running it near 300Hz?

I'll be listening to these in a medium sized living room. So will get a good few meters of distance.

I was planning on having 1/16" gap each side from the diaphragm to the stators. I'll insulate the stators so I can really play around with the bias voltages to see what's good. As for the transformer, I read on another thread that you can use toroidal mains transformers? I have plenty of these. How will the iron core fare near the 300Hz point?
Sorry to be contrary but ESL's do not absolutely need to cross at such high frequencies. Listen to a 300Hz tone sometime. It is not bass by any stretch no matter how you slice it. ESL's can easily be run down to 80Hz with no excursion problems and probably lower. It is the lowest octaves that present a problem for a larger panel. By that I mean 6-8 inch wide and 45 inch tall effective area. Segmenting the panel is not necessary either. Even considerably conservative 100Hz is a better option. 300Hz is the lower midrange and it lies in a very critical and sensitive band of hearing. Many DIYers here cross their hybrid ESL's low instead of high (270-300) I am getting strong 40Hz bass out of a single Acoustat panel per side. Yes there are issues mostly with the lowest octave overdriving the panel. If I were to cut everything below 50Hz the Acoustats could do absolutely everything thrown at them with no problem. Also at 1/16" gap you may be better on efficiency but you really could afford to gap a bit bigger. 400 watts will do the job with a 1/8" gap depending on your charge voltage.
Last edited:
I plan on using a digital EQ so I can play around with these things.

I'm tempted to even go even wider, about 1ft. Will I need to use a thicker insulation between stators and diaphragm?

The stator material I plan on using is 60deg Perforated galvanised steel with 3mm dia holes, 1.6mm thick, 51% open area. Sound reasonable? I can go with more open area, but I can only find ones with 6mm holes, is this a better choice or are the holes too big?.
300Hz has got to be one of the nastiest resonances for dynamic woofers in a cabinet to deal with.

I thought it would be. I'm looking into the mini DSP systems. They look promising, you can even attach a USB DAC they sell. I plan on using the speakers on a media centre PC. The on board sound is a bit **** :rolleyes:. Good enough for surround but I could really do with a DAC built into a nice DSP
If you go 1 foot wide or more then you might want to employ silicone dot spacers on the middle of the diaphragm which sort of have a dividing effect but not as restrictive as segmenting the diaphragm like a ML. Others here can provide better info on how these dots are used but generally they are a hard insulating spacer that is the same thickness as your D/S gap and they touch the diaphragm and stator on both sides (one on each opposing side) opposite one another. It end up being a mechanical coupling, a sandwiching of stator, spacer dot. diaphragm, spacer dot, stator. on a 1 foot wide panel that was say 45 inches tall you could do two sets at the 6 inch position horizontally and evenly positioned/spaced apart in the vertical. I might try this myself.
Most folk here tend to want to restrict ESL diaphragm excursion more than I feel is necessary their opinion is not unsubstantiated. There is a sound quality penalty IMHO though when you make the ESL a deep range tweeter and not extend into the foundational bass enough to get the coherency that ESL's offer. Just one mans opinion.
I've done that before with small blobs of silicone. I want to use a curved panel so will that work?

No I don't think you can disrupt the continuity of the curve geometry of the diaphragm with dots. You would have to horizontally segment the stator/diaphragm assembly like a ML. I really think at the width your doing curving it is not a great advantage. You'd be surprised at how little beaming there is with one single 9 inch wide by 45 inch tall Acoustat panel. A little soundstage "shifting" and some high frequency roll off but the beaming is minimized with just one narrow panel. Have you ever attempted tensioning a thin membrane in the vertical dimension while maintaining an even wrinkle free curvature with minimal "saddling" in the center and then apply that to a segmented stator? That sounds like hell on earth to me. I know folks here do it but no thanks for me flat is where its at.:D

I think the requirements were:
1- curved panel of 25x100cm size, or at least less than 1ft width
2- low power needs
3- large dynamic range
4- closed box bass

1. Its always positive to have as much diaphragm area as possible. At least its preferrable to ´passive´ wings to increase baffle size. If You maximize width of the panel this asks for a slim but strong frame to hold the panel. In any case will the acoustic phase cancellation of a 10" wide panel take effect below app. 800Hz and will cost ~6dB at 300Hz and even more further down. The equalization of the drop in the amplitude response costs on dynamic range and asks for more power.
This collides with points 2. and 3.
Furthermore going deeper in xover frequency You need to increase the diaphragm-stator distance d/s to cope with the increased excursion needs. Also You need to decrease the mechanical diaphragm tension to allow for a lower resonance frequency. This leads to a lower upper limit of the HV-bias You can apply. All this reduces the efficiency and the possible dynamic range dramatically.

- silicon dots are probabely not the way to do with curved panels. I assume that it´ll be impossible to achieve the levels of precision required for small-d/s-highefficiency panels.
- 3mm holes are a good choice. It´ll be a bit difficult though to get thick enough insulation without reducing open area too much or closing a few holes alltogether.
- Try to get thinner sheets as 1.6mm, like 1mm, otherwise the upper bandwidth limit may reduce too much. The curving will supply for sufficient stiffness of the thin sheets.
- curving increases stability alot and it helps in slightly widening dispersion. We don´t want much dispersion, but a flat non-segmented panel of this width shows too much directivity to be really practical.
- a 8"-driver could be used as bass partner. There are many good drivers in every price class around that easily allow for xover frequencies higher than 300Hz. You might have a look at the Dayton RS-225.

Last edited:
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.