[tnargs]

Rob from ER Audio he's given me a couple of hints that have had me thinking about my approach. So I've got a new design in my head now.

I can't wait to hear about the new design. I still haven't settled on a design.

[tnargs]

I've also learned a few things, and in a discussion with Rob from ER Audio he's given me a couple of hints that have had me thinking about my approach. So I've got a new design in my head now.

I'm still intrigued about the new design in your head. Though I can understand if you want to keep it under wraps and your audience on tenterhooks… 

[JohnR]

I'm still intrigued about the new design in your head. Though I can understand if you want to keep it under wraps and your audience on tenterhooks… 

Hi, no there's nothing under wraps. Sorry about the hooks tho I've learned a fair amount about the panels but since woodwork is such a pain for me I'm still experimenting to figure out where to put that next effort with the router. Plus other priorities.

Because of their size they are very directional at high frequencies so I think that will make subjective reports vary widely. Like other panel speakers you need to experiment... a lot. One tip Rob gave me was to vary not just the horizontal angle (toe in) but also the vertical (tilt back). This will drop the highs and also make the response at the standing position much more like the sitting position. About 10 degrees tiltback seems to be about the place to start, although I think it still needs a notch centered at around 3.5k in that case.

I've also been experimenting with a diffusor/reflector in front of the panel, which gave promising results, but I need to make two now as listening in mono is weird for me.

Also, I'm looking more to using these in a desktop setup now. The nearfield is more promising for a panel this size, as the dimensions turn out to be more optimal for that. I need to go back to Jim Griffin's line array paper to learn more about this. Also, putting a single pair in a large-ish (6.4x7.3 m) room is probably asking a bit much of them. The desktop configuration is what I originally had in mind when they piqued my interest.

Somewhat as an aside, I measured an electrostatic tweeter last night, which is part of the full-sized ER Audio kits. The radiating surface is only about 30 mm wide, the horizontal dispersion is excellent. So that also has me rethinking things a bit.

[tnargs]

Thanks John, yes they are very interesting and do exercise the mind!

Because of their size they are very directional at high frequencies so I think that will make subjective reports vary widely.

Yes I am sure you are right there.
 

Like other panel speakers you need to experiment... a lot.

That's interesting I didn't know it was common to panels, but I sure think it is true for these babies.
 

One tip Rob gave me was to vary not just the horizontal angle (toe in) but also the vertical (tilt back).

Bingo! this is starting to be a case of great minds think alike (I wish).
 

This will drop the highs and also make the response at the standing position much more like the sitting position.

True, although I was thinking more along the lines of it makes the effect common to a range of seated positions, whereas using horizontal off-axis positioning to get an effect will be highly varying for each seating position. I was noticing how many ML stats are tilted back a few degrees, too.
 

About 10 degrees tiltback seems to be about the place to start, although I think it still needs a notch centered at around 3.5k in that case.

I've also been experimenting with a diffusor/reflector in front of the panel, which gave promising results, but I need to make two now as listening in mono is weird for me.

I don't believe it! Bingo again, although I am thinking more of pointing the panels outside of the listening area and deflecting sound back towards the listeners. I am also wondering whether the deflector surface should be flat or have a 3D shape to scatter the HF more effectively.
 

Also, I'm looking more to using these in a desktop setup now. The nearfield is more promising for a panel this size, as the dimensions turn out to be more optimal for that. I need to go back to Jim Griffin's line array paper to learn more about this. Also, putting a single pair in a large-ish (6.4x7.3 m) room is probably asking a bit much of them.

I am getting quite reasonable SPL unless I heavily EQ the rising response - then they can 'hit the stops'. I still think I can use them for the whole room -- although your room's area is 90% more than mine. But I am chasing solutions as discussed above, which do not heavily use EQ.

I am starting to think they are 'special' speakers, i.e. they can do some things wonderfully but only for one or two seats, and well worth developing with that in mind. What they don't do is high-SPL neutral sonics over a wide listening area. But I am already having listening experiences with them that are not available with the more normal speakers I have.

[josh358]

Because of their size they are very directional at high frequencies so I think that will make subjective reports vary widely. Like other panel speakers you need to experiment... a lot. One tip Rob gave me was to vary not just the horizontal angle (toe in) but also the vertical (tilt back). This will drop the highs and also make the response at the standing position much more like the sitting position. About 10 degrees tiltback seems to be about the place to start, although I think it still needs a notch centered at around 3.5k in that case.

I've also been experimenting with a diffusor/reflector in front of the panel, which gave promising results, but I need to make two now as listening in mono is weird for me.

Also, I'm looking more to using these in a desktop setup now. The nearfield is more promising for a panel this size, as the dimensions turn out to be more optimal for that. I need to go back to Jim Griffin's line array paper to learn more about this. Also, putting a single pair in a large-ish (6.4x7.3 m) room is probably asking a bit much of them. The desktop configuration is what I originally had in mind when they piqued my interest.

Somewhat as an aside, I measured an electrostatic tweeter last night, which is part of the full-sized ER Audio kits. The radiating surface is only about 30 mm wide, the horizontal dispersion is excellent. So that also has me rethinking things a bit.

I'm thinking they could be magnificent on the desktop. You could pair them with the Maggie woofer and maybe the electrostatic tweeters and have an electrostatic Mini Maggie. Otherwise I think you're running into the fundamental (or almost fundamental) limitation of electrostatic design, namely, that excursion is limited by practical limitations on voltage. The figure I've seen is that a planar magnetic can give you on the order of ten times the maximum excursion of a stat, and of course dynamics can do even better.

This limitation runs up against the need to keep the size of a diaphragm smaller than wavelength to avoid beaming. A true ribbon tweeter can be made only 1/4" wide and have essentially perfect dispersion at 20 kHz. A dome tweeter can't be made quite so small but at say 3/4" it can still still have reasonable dispersion. But you can't keep an electrostatic element smaller than wavelength without imposing an SPL limit.

To overcome that, there are several options that I know of:

- Let them beam and revel in it. You can tame the highs with EQ. I've never been fond of this because I find the head-in-a-vice effect uncomfortable but I think that's a matter of personal habit.

- Line source with conventional crossovers. This is a partial solution that increases bass output, which is most problematic. Acoustats, Kings are examples. Typically, they still have dispersion/power response issues because a line source is most effective at low frequencies where driver output adds.

- Curved (Martin Logan) or faceted/curved (Sound Labs, RTR tweeters) diaphragm -- gives you controlled "pie wedge" dispersion good up to high frequencies but only when the diaphragm has is on the order of four times as wide as the wavelength being reproduced, so for it to work really well you need something the size of Sound Labs. It gets around the excursion constraint by allowing you to add more surface area without sacrificing dispersion. A good approach but not optimal since it suffers either from higher distortion (continuous curve) or spatial aliasing (faceted), and since it focuses the rear wave (though with acoustical treatment that could potentially be a plus).

- Electronically curved diaphragm -- the Quad approach, a phased array using a delay line. Quad emulated a point source but it could also be done with a line source and you'd get more output and better imaging. If I were designing an electrostatic, I think I'd do an electronically curved line source.

- Overlapping frequency range approach -- I think this is almost as good as the electronically curved diaphragm, it uses .5-way crossovers for good dispersion/power response combined with good power handling. The approach Magnepan uses in the 1.7 and 3.7. Peter Walker said the delay line approach is better and from the emulations I've seen of finite sized coaxial drivers he was right (not that I'd ever doubt Peter Walker in the first place). But it's still a good approach that doesn't require miles of cable, or multiple amps/transformers and digital delay.

Then there are esoteric approaches like the Beveridge or Dayton-Wright, or the folded woofer in the Quads. I think acoustic lenses and waveguides are problematic -- too hard to get right in a DIY project. Shading I think is also problematic in the case of electrostatics. You could put foam wedges on either side of the driver, the way Raal does vertically on their ribbons, but that decreases output which is already marginal in stats. The same would be true of electronic shading, in any case if you're set up to do that you're already set up to do overlapping crossover ranges or a phased array.

But I don't think any of these limitations apply near field, since you don't need as much excursion. So you could just do a side-by-side M-T arrangement as in the Mini Maggies, or better yet, a coaxial one (though I assume you can't do that with these). Or go without the tweeter and let them beam like the old Eminent Tech desktop planars. If you got good results, you could mate them with a Maggie woofer under the desk.

[Hank]

John, don' t give up on an open baffle line array - would probably need two of these 'stats stacked per side.    I'm thinking a line of 7" mid-bass ultra clean drivers with a couple of these 'stats stacked mid way beside them.  I've got a couple dozen Peerless HDS 830875 gems that I'm tempted to use in an OB line array of some kind.

[studiotech]

I'm imagining something like my BG Z-62, but it would work better due to a larger radiating surface area than the Neo10.  Kind of a planar coax, which would allow you to cross higher(say above 5KHz) than normal and still not have an odd polar pattern off to the sides.  You could use the Neo3 with the backcup on or take it off and place a nice piece if 1/2" wool felt or Bonded Logic in btw the drivers.  Or, if crossing high enough, the Dayton PT2 planar with the faceplate removed.  It's a little thicker than the Neo3, but narrower.  Now that could be a nice desktop system!

Ever since I've heard AJ's 1812 prototypes(12" pro coax driver based) and how even the image was as you moved around left/right and up/down I've been thinking of ways to do coaxial planars.  Kind of like Piega has invented.

Greg

[studiotech]

Erling,

Any news on your comparo?

Greg

[josh358]

I've been thinking of ways to do coaxial planars.

Me too. Wisdom Audio has done it using a modular approach. Unfortunately, the commercially-available drivers don't seem to be sized or optimized for it, e.g., if you put a Neo 3 or a true ribbon between a pair of Neo 8's or 10's, the assembly would beam at the crossover point. I've seen some BG speakers that get around this by forex mounting the tweeters in front of the mids, but that doesn't work if you want dipoles.

I understand that there's an OEM version of the RD drivers that allows you to drive the foil traces separately, so you can use the center trace for mid+highs and all three traces for mids.

One thing I'd love to try would be a triple ribbon -- a single 2" wide magnet assembly with a 1/4" central tweeter flanked by two 7/8" midrange ribbons, with thin aluminum dividers. But there's no way I could find time for a project of that magnitude.

Expense is also an issue. You could make a line source with a central line of Neo 3's and Neo8's on either side and then use foam to shade the Neo 8's for good dispersion. But you'd have to spend an awful lot on drivers. Maybe that would be more practical with an omni, but even if you made a lateral M-T-M arrangement what would you do about vertical dispersion? Chances are you end up using more than one driver and making a curved array or a little line source as in the Nolas.

[studiotech]

Me too. Wisdom Audio has done it using a modular approach. Unfortunately, the commercially-available drivers don't seem to be sized or optimized for it, e.g., if you put a Neo 3 or a true ribbon between a pair of Neo 8's or 10's, the assembly would beam at the crossover point. I've seen some BG speakers that get around this by forex mounting the tweeters in front of the mids, but that doesn't work if you want dipoles.

Hence the picture I've included above of the coaxially mounted Neo3 in front of the Neo10.  That idea used with these ER panels seems like a no-brainer to me to at least try out, but as you mentioned, not gonna be dipole in the highs.  I guess you could use another Neo3 suspended on back too for a dipole approximation.

John, I have a pair of non PDR Neo3 just lying around.  I'd be glad to send them along if you want to play around with the idea for  week or so.

My own idea is to use 4 Neo8-S version surrounding the OEM Raal, all bolted to a custom faceplate to hold them as close as possible.  Something like the faceplate I had made for the new studio monitors I just developed.  The top and bottom Neo8 would be turned horizontally and grouped as tightly as their frames allow.

Greg

[josh358]

The polar response on a single 8-S looks like it's good to 4K or so. So to something under 2K for the two Neos plus the Raal. I see the Raal uses a fourth-order crossover at 1.8 kHz. So it's a bit dicey from the crossover region power response perspective, but it looks like you could get away with it. If not, you could always use foam wedges to add some shading on the outer edges of the 8-S's. Do you know the width of the OEM Raal? I didn't see any specs on their site. I assume though that the 8-S's would be within 1/2 wavelength of one another in the crossover region.

[sfdoddsy]

My own idea is to use 4 Neo8-S version surrounding the OEM Raal, all bolted to a custom faceplate to hold them as close as possible.  Something like the faceplate I had made for the new studio monitors I just developed.  The top and bottom Neo8 would be turned horizontally and grouped as tightly as their frames allow.

Greg

What do you see the advantage of such a coax arrangement being?

[studiotech]

What do you see the advantage of such a coax arrangement being?

Well if it works out like a REAL coaxial driver, it would give you a more solid image as you moved off axis.  Right now, my system is more of the typical head-in-a-vise for perfect sound due to uneven, non-symmetrical off axis response.  When AJ brought his system over here, it was nice to be able to move off to either side or up and down and have a similar tonality and imaging.  One of the reason recording engineers have liked Tannoy coaxs over the years is the ability to be off axis while adjusting a compressor or EQ and not have to move back into a laser like sweet spot to hear what you just did. 

Greg

[studiotech]

I've been inspired by this design for years:

http://www.dnaudio.com/DNA%20Speaker.html

[JohnR]

I don't believe it! Bingo again, although I am thinking more of pointing the panels outside of the listening area and deflecting sound back towards the listeners. I am also wondering whether the deflector surface should be flat or have a 3D shape to scatter the HF more effectively.

What I tried is a deflector that is curved horizontally in front of the panel. It looks weird but measured response over a 30 degree window (horizontal) is actually pretty good. I'm not sure what other things I'll run into.

I am starting to think they are 'special' speakers, i.e. they can do some things wonderfully but only for one or two seats, and well worth developing with that in mind. What they don't do is high-SPL neutral sonics over a wide listening area. But I am already having listening experiences with them that are not available with the more normal speakers I have.

No arguments with anything there 

[JohnR]

I'm thinking they could be magnificent on the desktop. You could pair them with the Maggie woofer and maybe the electrostatic tweeters and have an electrostatic Mini Maggie.

Hi Josh, might be a while but I think I'll put the effort in that direction. However I don't think I'll go for DWMs. Here the DWM is $1000 and the whole mini-Maggie system is $1500. So there's not really much reason to do it with these panels and a DWM. To be honest I'm struggling a bit with Magnepan's pricing on the DWM.

[JohnR]

Hence the picture I've included above of the coaxially mounted Neo3 in front of the Neo10.  That idea used with these ER panels seems like a no-brainer to me to at least try out, but as you mentioned, not gonna be dipole in the highs.  I guess you could use another Neo3 suspended on back too for a dipole approximation.

Hi Greg, that's an interesting idea and I'll give it a try. Thank you for the offer but I do have a pair of Neo3 PDR with the back cups still on so I'll use those. I did mount a Neo3 PDR above the baffle and used a first-order passive crossover but didn't really like the result.

[josh358]

Hi Josh, might be a while but I think I'll put the effort in that direction. However I don't think I'll go for DWMs. Here the DWM is $1000 and the whole mini-Maggie system is $1500. So there's not really much reason to do it with these panels and a DWM. To be honest I'm struggling a bit with Magnepan's pricing on the DWM.

I see what you mean. Not exactly economically viable. And I've noticed that there's a dearth of DIY planar bass and midbass drivers.

Knowing Wendell, I'm sure he wishes they could have sold the Minis and DWM's for less. But as he once pointed out when the price of the DWM's came up, the cost of manufacturing speakers isn't measured in board feet. The savings on MDF, cloth, wire etc. due to its smaller size presumably don't amount to much. With its dual voice coils and crossovers, the R&D and parts and labor costs are I suspect much the same for a DWM as they are for a full-sized speaker of similar complexity.

[scorpion]

I am warming up in the Open Baffle circle.

/Erling

[rick57]

Seems like a good candidate for eg a miniDSP, or more accurately  vice versa