NEW PLANAR SPEAKERS

[Stuart Polansky]

Modifying or replacing my Magnepan Tympani IV midrange/tweeter panels.

Bass panels will remain as-is, but be braced.

Midranges will be four or six Bohlender Graebener Neo8 drivers per channel in a line array. Array will probably be about 96-99dB/W/M.

Two options for panels: 1) put midrange drivers in existing slot of Maggie MDF panel (yes, they fit) or; 2) build new, pretty panels out of nicer material (and hopefully better sounding material). I want as much LF extension as possible from the mids, because I want to cut off the bass panels as close to 100Hz as possible. I realize that goal may be severely compromised to as high as 400Hz. They work to over that now, so it's not impossible to work them that high, it's just that I think a faster driver would be better in that range.

The remaining issue is tweeters. Maggie has 2.2R 5' direct drive ribbons with 1R series resistors to match the low sensitivity of the rest of the drivers. Any informed opinions as to the viability and relative performance of the Maggie ribbons versus other tweeter options?

An array of BG Neo3?
Raal 70-10D
Aurum Cantus G3
Scan-Speak D2608/9130
SEAS Prestige 27TBFC/G

Some other tweeter?

Or is the Maggie ribbon in the same league with these? Almost as good? Junk? Better?

If I go with option 1, and reuse the existing MDF panels, then the Maggie ribbon is the obvious choice. With option 2, which I really want to do, I have a decision to make.

The Neo8s are suggested to crossover (LP) at ~5-6kHz.

As an investigative/developmental tool, I'll be using a Behringer DCX2496 to determine XO points, slopes, relative drive levels, etc. The system WILL be triamped, unless the MR/Tweeter setup is close in efficiency and can be passively crossed over. Still, I'm leaning towards triamping, period.

Ultimately, the Behringer would be replaced by a combination line stage/crossover using tube circuitry.

Stuart

[FerdinandII]

IMHO.....clarified, qualified, and legalified...

The Maggie is certainly a top contender in that bunch.
They're all good, although I think the Raal might be the best.

My suggestion is to take the time, money, and energy you would have spent replacing a perfectly good tweeter, and put it elsewhere in the system.

[Stuart Polansky]

IMHO.....clarified, qualified, and legalified...

The Maggie is certainly a top contender in that bunch.
They're all good, although I think the Raal might be the best.

My suggestion is to take the time, money, and energy you would have spent replacing a perfectly good tweeter, and put it elsewhere in the system.

Saving time, money and energy?!?!?

Hey FerdinandII, I love you man!!

Thanks for the advice. the answer I was hoping for (saving $$$)

So, I think, pretty wood panels and the Maggie ribbons with the Neo8 mids.

Stuart

[tomp]

Stuart:

I have a suggestion. Instead of multiple Neo 8s, why not use the BG RD40. There is a real problem of using stacked arrays of mid and high frequency drivers in that they do not act as a true line array. Since the vertical spacing between the drivers becomes significant vs the wavelengths reproduced you wind up with lobing in the radiation pattern, not a good thing. The advantage of the RD40 is that the total length is one diaphragm which is a true line source with no lobing. They are $489 at Parts Express vs almost $600 for six Neo 8s. I have been using the RD75s for years and they are really great. In a normal room, one of the most important things is control of radiation pattern which is greatly helped using a true line array.

Tom

[Stuart Polansky]

Stuart:

I have a suggestion. Instead of multiple Neo 8s, why not use the BG RD40. There is a real problem of using stacked arrays of mid and high frequency drivers in that they do not act as a true line array. Since the vertical spacing between the drivers becomes significant vs the wavelengths reproduced you wind up with lobing in the radiation pattern, not a good thing. The advantage of the RD40 is that the total length is one diaphragm which is a true line source with no lobing. They are $489 at Parts Express vs almost $600 for six Neo 8s. I have been using the RD75s for years and they are really great. In a normal room, one of the most important things is control of radiation pattern which is greatly helped using a true line array.

Tom

Tom,

I did give the RD idea some thought. Though I would have made the same driver choice as did you, the RD75.

My reasoning, faulty as it may be, was based on specification (oh no!), and suggestions by an Audio Asylum member who has actually performed the swap and is able to measure his room. He highly recommends this approach.

1) Efficiency. I am tired of the need to build BIG amplifiers to get effortless high SPL sound. the RD series are spec'd at 88dB/W/M (much better than the Maggie 84-86dB/W/M, but), a single Neo8 at 93dB/W/M. An array of four should be ~99dB/W/M. The AA user reports a tad over 100dB/W/M for his array of six/channel. That's more like it.

2) Beaming. The width of the RD series tell me they MUST beam at HF. Switching to the much narrower Maggie tweeter at ~5-6KHz should fix that.

3) I'm a little skeptical of the one-graph-fits-all spec from BG for the Radia series. A cavity resonance isn't apparent at all. None. Yet the users report a peak for which a notch filter must be used. The LF extension is shown as the same for any of the series. Can't be. Maybe I'm reading it wrong. The rolloff of the Radia over 12KHz (or so, I'm interpolating from a not so fine graph) bugs me. Users report using active equalization to correct this. No thanks. I'll stick to the dedicated tweeters I have. Otherwise, I think more experimentation with these would be needed in my application. The VERY attractive feature of the RD series is the LF extension, which would make crossing over to my bass panels at a lower frequency much easier. But again, that 88dB/W/M

4) Flexibility. The RD is 6 ohms. With the Neo8, 4 ohm driver, I have the choice of 2, 4 or 16 ohms. If I use six drivers: 1, 2.7, 6, or 24 ohms! The 16 and 24 ohm options are very attractive to me, since I'll be using tube amps, and using all of the copper in a winding should be more efficient. It's also a better option with OTL amps. And an OTL amp is on my bucket list.

The fact that I am repeating an effort, made by another, with his support and advice, gives me a little more confidence in the potential success of the project. I'm sure the RD75 is astonishing and look forward to hearing it some day.

Stuart

[tomp]

Stuart:

I have a suggestion. Instead of multiple Neo 8s, why not use the BG RD40. There is a real problem of using stacked arrays of mid and high frequency drivers in that they do not act as a true line array. Since the vertical spacing between the drivers becomes significant vs the wavelengths reproduced you wind up with lobing in the radiation pattern, not a good thing. The advantage of the RD40 is that the total length is one diaphragm which is a true line source with no lobing. They are $489 at Parts Express vs almost $600 for six Neo 8s. I have been using the RD75s for years and they are really great. In a normal room, one of the most important things is control of radiation pattern which is greatly helped using a true line array.

Tom

Tom,

I did give the RD idea some thought. Though I would have made the same driver choice as did you, the RD75.

My reasoning, faulty as it may be, was based on specification (oh no!), and suggestions by an Audio Asylum member who has actually performed the swap and is able to measure his room. He highly recommends this approach.

1) Efficiency. I am tired of the need to build BIG amplifiers to get effortless high SPL sound. the RD series are spec'd at 88dB/W/M (much better than the Maggie 84-86dB/W/M, but), a single Neo8 at 93dB/W/M. An array of four should be ~99dB/W/M. The AA user reports a tad over 100dB/W/M for his array of six/channel. That's more like it.

As far as sensitivity, the other day David Berning brought over a 6W/channel SET he built and we ran it on the RD75s. We were both astonished how well it worked at very respectable levels. Go figure. I normally run it with a Crown Macro Reference which will put out about 1Kw per channel. The big difference is that if the Crown ever did clip you would run out of the room screaming. The SET was very benign when pushed into never never land.

2) Beaming. The width of the RD series tell me they MUST beam at HF. Switching to the much narrower Maggie tweeter at ~5-6KHz should fix that.

Beaming in a small room is good and desirable if the line source is long enough to cover the whole listening area in the vertical dimension. Floor and ceiling early reflections which are most obnoxious are eliminated. The RDs seem to actually be narrower in width than the Neo 8s which would give a wider dispersion in the horizontal direction. The off axis response on the RDs has been very good. In fact, I use a narrow baffle on the insides of the speakers and a wide curving one on the outsides. That cancels toward the inside and extends the coverage toward the outsides which widens the image.


I'm a little skeptical of the one-graph-fits-all spec from BG for the Radia series. A cavity resonance isn't apparent at all. None. Yet the users report a peak for which a notch filter must be used.

There is a cavity resonance. BG provides a passive trap for that but I use an active crossover and compensate for it there. BTW, all planars I have ever seen have a cavity resonance. It is the nature of the physical structures surrounding the active element.

The LF extension is shown as the same for any of the series. Can't be. Maybe I'm reading it wrong.

The low frequency extension should be approximately the same for both. What is not the same is the linear volume displacement which limits amplitude of a low frequency signal you can produce with the 40 compared to the 75. The low frequency capability will also depend on the crossiver slope. I use 48 db/octave.

The rolloff of the Radia over 12KHz (or so, I'm interpolating from a not so fine graph) bugs me.

The rolloff with my 75s starts around 16K. At my age that is not a problem.

Users report using active equalization to correct this. No thanks. I'll stick to the dedicated tweeters I have. Otherwise, I think more experimentation with these would be needed in my application. The VERY attractive feature of the RD series is the LF extension, which would make crossing over to my bass panels at a lower frequency much easier. But again, that 88dB/W/M


4) Flexibility. The RD is 6 ohms. With the Neo8, 4 ohm driver, I have the choice of 2, 4 or 16 ohms. If I use six drivers: 1, 2.7, 6, or 24 ohms! The 16 and 24 ohm options are very attractive to me, since I'll be using tube amps, and using all of the copper in a winding should be more efficient. It's also a better option with OTL amps. And an OTL amp is on my bucket list.

Again, refer to the discussion above about David's 6W SET.



The fact that I am repeating an effort, made by another, with his support and advice, gives me a little more confidence in the potential success of the project. I'm sure the RD75 is astonishing and look forward to hearing it some day.

Before you go too far down this road I suggest you come listen to the RD75s. It can't be before the holidays but after that will be fine. Let me know.

Stuart

[Stuart Polansky]

Thanks Tom,

I'll take you up on that offer. I can't get really serious about the speakers until I finish the new bass amps anyway!

Stuart

[Stuart Polansky]

As I'm attacking this project again, some details are being worked out.

It will be three-way with a shallow sealed enclosure on the inboard side for four Dayton RS180-4 midbass drivers, 60Hz-~1kHz. The line array of 6 Neo8s will take over from ~1kHz to ~6kHz, the Maggie ribbon, robbed from the Tympanis will operate from ~6kHz up.

Because of efficiency and impedance differences, this is definitely an active crossover project. Since the Neo8's can be wired as 2.7 ohms, they actually match the Maggie tweeter without the attenuating resistor. Hoping to make that XO point work passively, we shall see.

Bass below 60Hz will be in a separate enclosure.

Big questions are:

1) How close to the edge of the panel can I mount the ribbon? does panel strength alone control this, or is there a sonic issue? OB guys (Roscoe?) please feel free to comment here.

2) Is there a detriment to using a shallow enclosure behind the midwoofers? They are 7" and will be in a sealed 2.2 cubic foot box, 6' tall, and about 11" wide, 5" deep (internal). They will be heavily braced and treated with Acoust-X.

All 0.75" birch plywood. Panel thickness is doubled. Box will be braced on 7" centers with horizontal braces, each containing two rounded edge 4" holes. Top and bottom panels will be attached to the first brace with a 1" dowel in the center of the brace.

Attachments

RS180-Neo8-Maggie-ribbon.pdf

(22.45 KiB) Downloaded 904 times

[Roscoe Primrose]

Crossed over that high, I don't think there's any non-structural limit to how close to the edge you can put the ribbon...

Having speakers of different impedances on a passive crossover gets tricky if the amplifier output impedance is >0....

Roscoe

[tomp]

As I'm attacking this project again, some details are being worked out.

It will be three-way with a shallow sealed enclosure on the inboard side for four Dayton RS180-4 midbass drivers, 60Hz-~1kHz. The line array of 6 Neo8s will take over from ~1kHz to ~6kHz, the Maggie ribbon, robbed from the Tympanis will operate from ~6kHz up.

Because of efficiency and impedance differences, this is definitely an active crossover project. Since the Neo8's can be wired as 2.7 ohms, they actually match the Maggie tweeter without the attenuating resistor. Hoping to make that XO point work passively, we shall see.

Bass below 60Hz will be in a separate enclosure.

Big questions are:

1) How close to the edge of the panel can I mount the ribbon? does panel strength alone control this, or is there a sonic issue? OB guys (Roscoe?) please feel free to comment here.

Probably what is more important here is the shape of the edges of the panel. In my situation I use a long gently curved baffle on one side and a short but rounded edge on the other. In either case, sharp edges are to be avoided but the radius of the curve depends on the frequency. A small radius is OK for high frequencies but as the frequency decreases the radius must get larger.

2) Is there a detriment to using a shallow enclosure behind the midwoofers? They are 7" and will be in a sealed 2.2 cubic foot box, 6' tall, and about 11" wide, 5" deep (internal). They will be heavily braced and treated with Acoust-X.

I would not use an enclosure behind the midwoofer. With the Neo 8s being used in a dipole arrangement it is better is you can keep the radiation pattern the same by using it in a dipole configuration. The best results with speakers has always happene when all drivers have an acoustic radiation pattern that is the same or at least the same in the crossover regions. My configuration is dipole for the RD75s and mid bass drivers with the subs in a sealed box. BTW I saw that Parts Express says that they can no longer get the Neo 8s. Que Pasa?

All 0.75" birch plywood. Panel thickness is doubled. Box will be braced on 7" centers with horizontal braces, each containing two rounded edge 4" holes. Top and bottom panels will be attached to the first brace with a 1" dowel in the center of the brace.