Two way active: ESL/Dynamic Line array

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am planning a 2 way active line array using 3 ESL 440 panels per side with 6 Seas excel 180mm woofers per side. Believe it or not I independently came up with the basic concept of an ESL line mated to an open baffle woofer tower. I then found a member "Calvin" who recommended this idea in this forum to other members. I also read everything on Calvin's website about his Blue Klaus Prototype and Purist ESL commercial speakers. What convinced me to pursue my project was reading Calvin's posts in this sub-forum, Thanks Calvin.

I plan to use these speakers for music and home theatre. I have various solid state A/B as well as digital icepower amps to utilise for the project.

The ESL panels are similar too the 505 mini panels from Eraudio. But are around 440mm in length. The woofers I have for the project are Seas W18E/V-NHT (H 1387-08ohms). These are similar to the Seas w18ex001 with magnesium cones but the NHT version has dust caps instead of phase plugs and are also magnetically shielded. Crossover and EQ will be handled by an existing miniDSP HD 2x4 ... I plan to design crossover using a UMIK microphone, REW and Rephase.

I plan to house the ESL panels in a simple narrow flat baffle.
Woofers will be in a narrow A frame open baffle.
Baffles will be CNC'ed from furniture grade plywood ... type of plywood and thickness TBD. I have a contact who will handle the CNC work.

Questions:

1) How many degrees of lean back would be ideal for this design? My listening position is around 3.5m both left and right speakers.

2) I would like to cover as ideally as possible both seated and standing listening positions. For this reason I'm thinking of making the speaker around 180mm tall ... that way the top of the speakers radiating surface is at ear height for good standing coverage. With the line being higher from the floor so I will loose some bass due to a lack of floor bounce. Is it feasible to solve this with EQ via miniDSP? Another poster in this forum claims to have achieved good bass response from 4 x Dayton RS180s U frame towers raised off the floor so I figure my 6 x 180mm drivers should be adequate? With the miniDSP I could limit the lowest frequencies to keep the driver excursion under control ... I ultimately plan to build subwoofers ...

3) Since I have access to CNC I am toying with the idea of mounting the woofers behind the baffle and machining a waveguide into the baffle. Part of the idea is to create a sort of H frame (wave guide at the front and A frame at the back) and also to narrow the dispersion of the woofer tower to better blend with the 440 panel. Mounting the woofers behind the baffle would also make decoupling the woofers from the baffle via magnet mounting easier. The waveguide may help with the Low QTS? Thoughts about the waveguide in general would be appreciated. I am leaning toward the usual flush mounted driver method presently ... since I think my waveguide idea is a bit to experimental ...

4) The QTS of the woofers is presumably about the same as the Seas w18ex001 (QTS = 0.26). This is a very low QTS for open baffle ... I plan to EQ by measuring and adjusting the built speakers via miniDSP ... should I also address the low QTS in other ways? suggestions?

5) Is there any reason why the ESL baffle should be the same width as the Bass tower? I'm planning to make both as narrow as the drivers/ESL panels allow.

Note: I'm not in a rush to start this project. I plan to fine tune all the details until I'm confident that I've scrutinised and made confident decisions about all aspects of the build.
 
Note: I'm not in a rush to start this project. I plan to fine tune all the details until I'm confident that I've scrutinised and made confident decisions about all aspects of the build.

Wannabee engineering rears its head again.

The alternative to believing you can engineer everything perfectly on paper is to build a mock-up or a breadboard of your concept and let reality teach you.

There's a respected poster (a name like "xr471") who builds amazing enclosures with styrofoam, mostly for R&D. Get yourself a soldering-gun styrofoam cutter, proper glue cartridge, and you're all set... if styrofoam makes sense for your concept. Or quarter-inch plywood and lots of space-filling types of glue or a pound of Blu-Tac.

Yes, Calvin is a smart fellow.

B.
 
Last edited:
Thanks for your reply bentoronto

I agree with much of what you write ... and I have considered building prototypes.

There are a couple of reasons why I feel this could be avoided ... to an extent.

I am basing my design on Calvin's ... so the general concept of what I'm attempting has been proven to an extent. The basic layout is an ESL panel line array with the panels laid out as close as possible and an A frame woofer line array tower with closest possible driver spacing. I've read several posts on this forum about integrating ESL and wideband ribbons with woofer towers very similar to what I describe. In these posts prototypes were not built and the DIYers seemed very happy with their results.

I do not intend to build a finished speaker straight up.

I think that the baffle that will house the ESL panels is entirely straightforward. A prototype baffle for the ESL panels I feel is not needed. I plan for the ESL baffle to be height and lean angle adjustable and the same width as the baffle/frame that the 440 panel comes installed into from the factory. The 440 panels will be spaced as close as possible.

For the woofer section I plan to have the main section of the front baffle machined to the minimum size needed to accomodate the drivers. The other sections of the woofer tower will be prototyped using some sort of scrap material. I think that flush mounting the drivers in a vertical line with closest possible driver spacing is a pretty proven approach so as long the core baffle section is just a module that will become part of the final speaker its pretty reasonable to CNC this from thick furniture grade plywood straight up.

Also I plan to have the woofer and ESL towers independent from one another. Calvin's design has a common base for the ESL and woofer tower. Keeping the towers seperate will allow me to make adjustments in spacing, angle, height based on listening and measurements. Also the MiniDSP will allow unlimited tweaking of the crossover design.

In my first post I mentioned my idea of mounting the drivers on the back side of the baffle and machining a wave guide into the front of the baffle. This is an idea that I will most likely abandon because I lack the knowledge to design a waveguide and I think that it would probably not work well without prototyping. I'm hoping that I will get some interesting feedback about this waveguide concept that will either encourage me to pursue it further or the opposite.
 
Unless panels can go down low (an ideal rarely seen), you do need mid-range drivers close by. But a sub in a corner is hard to beat. Most commercial speakers need to sell as one enclosure (like Calvin's) but you can tri-amp and put the sub in the corner or wherever is best. That greatly reduces the sonic challenge on the line array drivers and gets smaller, as you seem to want.

B.
 
…I'm hoping that I will get some interesting feedback about this waveguide concept that will either encourage me to pursue it further or the opposite….the idea is to create a sort of H frame (wave guide at the front and A frame at the back) and also to narrow the dispersion of the woofer tower to better blend with the 440 panel.
If you plan to crossover over in the typical range of 200Hz – 400Hz, the waveguide would need to be rather large to provide any narrowing of the dispersion pattern beyond what is inherent for a dipole. Since both the ESL dipole line source and woofer dipole line source will have similar width dimensions, the dispersion match between the two sources is already about as good as you will be able to achieve.
 
thanks for your input

Thanks to everyone who has contributed to this thread.

Just wanted to provide some more information about by build.

I have experimented with the miniDSP, 440 ESL panel and my floor standing Krix Euphonix speakers (vented design with 4x 5.25 woofers per side). With a single 440 panel I get pretty good results with a 4th order LR crossover at 300Hz. Based on this I expect to crossover somewhere between 200Hz and 300Hz.

Also these are the only specifications that I've found for the 440 ESL panel:
Frequency range: 200 Hz – 20,000 Hz
THD: < 1%
Polarisation voltage: Approx 2.5 kV
HV power consumption: <0.05 W
Impedance: Minimum 4 ohms
Sensitivity: 88 dB
Amplifier requirements: +/- 30 V or higher rails with peak current capability of 10 amps
Dimensions (W x H x D): 168 x 440 x 8 mm (approx 6.2 x 17.3 x 0.3 inches)
Weight: 440 g (approx 1 pound)

Question: Would it be a good idea to make the baffle for the ESL panels and the baffle for the woofers the the same width? This would mean making the ESL panel a bit wider than I'd originally planned.
 
.... 5) Is there any reason why the ESL baffle should be the same width as the Bass tower? I'm planning to make both as narrow as the drivers/ESL panels allow..

Hows the planning coming along? Did you work out the baffle width question?

One consideration planning the width of the bass tower is how low you want the OB bass to go. Hornresp allows H-frame and OB design modelling and might save sawdust.
 
Calvin's Purist speakers are gorgeous:


IMG_1559-1.jpg


Why the gap between the ESL and bass array? If they were joined the bass panel would be larger and have better bass extension. Is the gap for diffraction limiting, directivity or to allow the ESL membrane to sit in the bass dipole null so its not flapped around?
 
since the bass drivers ..... could blow the membrane into the stators if it weren´t for the slot.


Thanks for clarifying. Using he dipole null to protect the membrane is so smart. I wonder how wide the slot needs to be to allow a fully formed dipole bass null to envelope the membrane?


Did I say those Purist ESL's look superb to my eye and Ive had a long training in design aesthetics. I which I could hear them.


IMG_1563-1.jpg
 
..... it is not the dipole null at 90°, because that relates to the sound pressure.
Air particle velocity though is maximal at the pressure null.
So at 90° there´s a lot of air volume moving just where the panel is situated.

Thanks for sharing. Ive thought about what you've written but got confused. I used Hornresp to sim the dipole null with and without baffle but that didnt help. :eek:
 

Attachments

  • Dipole null with baffle.JPG
    Dipole null with baffle.JPG
    293.2 KB · Views: 119
  • Dipole null no baffle.JPG
    Dipole null no baffle.JPG
    270.9 KB · Views: 113
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.