Types Of Planar/ribbon Type Speakers

I'd love to see a pinned topic about the different types of electrostatic, ribbon and planar speakers, and anything similiar.

Anyone else interested in this?

I think one of our moderators is a sort of guru in this type of speaker dabbling...

Very well, Paul, here's installment #1. It's just the basics, of course, but hopefully it'll provide you with what you're looking for.

Electrostatic

Electrostatic speakers are one of the oldest speaker designs, but didn’t see much widespread use in higher fidelity reproduction until 1957 when the famous Quad ESL-57 came out. Rather than a typical electrostat that uses the entire surface to radiate sound simultaneously, the designer saw the advantages in the then 10 year old Tannoy dual-concentric design and so attempted to reproduce a point source with his panel by having concentric rings of radiation fed by consecutive signal delay. This design choice made for a wider than normal (to panel speakers) “sweet spot” by reducing the effects of beaming. The design was a hit ultimately leading to other TOTL designs such as the KLH Nine, the Quad ESL-63, the Infinity Servo-Static that attempted to improve on both and added a servo-controlled subwoofer, ultimately Sound Lab with their huge sectional panels and Martin Logan with their supermodel looks.

The electrostatic speaker works on a very basic principle, it uses electrostatic attractance and repulsion to move a charged diaphragm between two charged stators driven by the amplifier’s signal. The diaphragm is typically very thin (only a few microns for Sound Lab, a couple dozen for Quad) Mylar that is either treated with rubbed in carbon (old Quad design) or coated with vapor deposited aluminum. This treatment is to make the other wise insulative Mylar conductive to electricity, though at a very high resistance. This is then suspended taught inside a frame and heated to 300 degrees Fahrenheit – at this temperature Mylar has the wonderful quality of either tightening itself to the proper tension if too loose or loosening itself a bit if too tight. Attached to either side of this frame are the stators. These typically consist of perforated metal sheets with an insulated coating (Quad, Martin Logan) or an array of fine wires run through insulating, reinforcement sleeves (Sound Lab). The insulation being to protect the end user and his/her children or pets from the high voltage necessary to operate this kind of speaker.

Here is a diagram showing the rough construction of this panel:


To this device, a power supply capable of producing several kilovolts is attached. This is used to charge the diaphragm of the panels and keep them charged so that the amplifier’s influence of the charge may create movement of the diaphragm. To this end, an electrostatic loudspeaker requires an impedance matching transformer to match the 4-8ohm desired load of the amplifier to the several kilo ohms impedance of the panel while ensuring no voltage goes back from the speaker to the amp and thus release the magic blue smoke that makes it work.


Derivations of this design:

The ESL principle has seen many variations in its decades of existence. The oldest models were flat that resulted in severe beaming. That is, as you progress higher in frequency, the panel becomes larger as far as the sound waves are concerned and so the dispersion angle (the angle at which the sound will radiate out from the panel) becomes smaller and smaller. In the treble region, this can make for a narrow beam of sound only a foot or two wide that will subsequently make for difficult placement and a one-person wide “sweet spot”. To correct this issue, the QUAD ESL-57 used the time delay/dual concentric principle that tends to have its own drawbacks. The next generation used several smaller flat panels set into a frame such that each was angled outwards slightly. This helped a little bit. The generation afterwards used a multi-faceted treble panel (the Sound Lab A-2 uses a mid/treble panel that is half a hexagon in cross-section) in addition to angled bass panels to again improve dispersion. Today, panels are curved with single, large elements like the Martin Logan designs curved a few degrees or the large Sound Labs with their multi-sectional design (to reduce issues with resonance) curved into a 90 degree arc and thus making the entire room the sweet spot.

An interesting vintage aside regarding this design, the 1970s saw an explosion of consumer grade audio equipment. The general consensus was that electrostatic speakers were the ones to beat, especially with regards to midrange and treble. However, they’re also expensive to build and a cheaper design was needed. The Radio Shack brand, Realistic, came out with one such design that was a hybrid in a box. It consisted of a conventional woofer run up to, perhaps, too high of a frequency and then crossed over to single-ended electrostatic tweeter elements. Rather than using two stators, this design had a single stator with the diaphragm positioned in front of it. I believe that to prevent a hearty shock for touching it, the diaphragm had its conductive coating on the backside, towards the single stator, so that the Mylar itself acted as the insulator. Yes, it was electrostatic, but it did not give electrostatic performance.

This is what it looked like in cross-section:


Performance and Limitations:

Overall, an electrostatic loudspeaker is typically very revealing and transparent in sound and is thus capable of reproducing very low level detail. The sound tends towards being very lifelike. Unfortunately, ESLs in general tend to lack dynamics or impact nor do they tend to possess the capability to achieve high sound pressure levels. This is largely due to the limited excursion capability of the diaphragm, which also presents another weakness of this design – drive the panels too hard and the diaphragm will contact the stator, creating a spark due to the high voltage and burning a hole in the Mylar thus creating an expensive (at times in excess of $2000) to repair. The diaphragm need not even make contact for this kind of damage to occur for having the bias voltage set too high (the higher the voltage, the more sensitive the panel, typically) can cause an arc from the stators to the diaphragm thus burning the Mylar at limited excursion. Another problem is that with the dipole design that most ESLs, below a certain frequency dependent upon the baffle (or panel) size, experience cancellation of the output when the backwave effectively curves around the baffle, canceling the front wave. This makes for limited bass response and means that either larger panels are necessary to lower the frequency of this cancellation for bass reproduction to occur or a hybridized design with a conventional cone subwoofer driver is necessary.

The principle tends to be very insensitive and subsequently need very large power amplifiers capable of handling sharp impedance dips (your typical large panel can present a load in excess of 30 ohms in the bass region, but less than a 1 ohm load in the treble region). The electrostatic design is also very capacitive, which can be hard on amps, and performs best when fed with high wattage/voltage, low current designs.

- JP

Up Next: Magnetoplanar/Magneplanar/Magnetostatic Design

Excellent post JP. I look forward tot he next post. I have lots of nakeds of the Maggies if need be.

- some are here http://k-dogg.smugmug.com/gallery/1088197/2

Kelly

Sorry about the delay, I'll try to get the other installment up tomorrow.

- JP

No apologies needed. I really appreciate you making the effort to educate me on these. I dig the burn of the learn, you know?

If I was living on the upper west coast, between you, your speakers and your big brain, and Kelly, his gear, and his bikes, I think you'd be hard-pressed to shake me.

I dig you cats the most-est!

And Dingus, methinks you too!

thanks Paul,

both Kelly and JP have a big leg up on me as far as the techincal aspects of anything (especially JP). i've been lucky enough to find more than my fair share of decent gear at bargain rates, so i guess i do have a nose for that (not on par with bolly, but then who is?).

Bolly? Well, that cat is, shall we say, full of the black magic, full of the voodoo, inhuman, I say, unnatural in his magical abilities...

Hmm, everyone calls Bolly the master of incredible deals, but I do believe I've done consistently bette with regards to speakers.

Anyways, here's the edited down post. 'Twas four times longer when it dawned on me that you just wanted a basic comparison and not the technical junk behind it that tends only to confuse.

Magnetoplanar/Magneplanar/Magnetostatic

Magneplanars are an interesting speaker when compared against electrostatics in that they are an odd duality of oppositions. While they use a completely different mechanism for operation, the design was modeled on electrostats.

Jim Winey, founder of Magnepan, was a big fan of the KLH Nine, but was put off by their deficiencies, such as the fragility of the panels, the inability to play loudly, and the cost. So, he set about in his spare time designing a magnetic equivalent. (This is why you’ll sometimes see this type of speaker referenced as magnetostatic.) The resultant design was essentially flattening a conventional dynamic driver voice coil, securing it to a Mylar diaphragm, and instead of using an electrically charged stator, a bed of magnets was used instead.

Jim Winey next to his original functional prototype:


The basic magnetoplanar design is rather simple. The base layer is a perforated metal screen (to let the music out) to which is affixed row after row of strip magnets. Over this a Mylar diaphragm is stretched with the voice affixed to the opposing side. The earliest designs from Magnepan used magnet wire glued to the Mylar with a polyurethane based compound called Miloxane. Newer variants use an etched aluminum voice coil trace that’s basically aluminum foil glued to the Mylar. For both of these, the wires are set so that they line up with the channel between each row of magnet so that the current from the amplifier generates a magnetic field to interact with the two magnetic fields.

The older style with a diagram of how the principle operates:


The next generation etched foil design:


Whenever you stretch what effectively amounts to a piece of plastic, so that it hangs relatively free, you’ll encounter resonances of which the frequencies are determined by tension of the material, its geometry and size. To spread the frequencies out of these resonances so there won’t be a particular peak in the resonance, Magnepan uses a differentially tensioned diaphragm so that the taughtness of the material increases from the bottom towards the top. This tends to make the panel seemingly have more bass towards the bottom and more midrange towards the top sometimes tricking new planar listeners into believing a two-way design is actually three. This effect is accentuated thanks to a brace towards the top between the Mylar and the magnets to, again, reduce resonances and to minimize “flapping” – when the diaphragm loses some linearity of control and over-excursions to the point of colliding with the magnets due typically to intense low-frequency bass. Unlike with electrostats where this could cause expensive damage, it’s only an annoying sonic artifact with magneto planars.

Apogee, essentially a Magnepan built to a higher price point, chose a more expensive alternative route to minimizing resonances by rotating the magnets 90 degrees so that the rows are horizontal rather than vertical and then gradually reducing the length of the magnets and width of the diaphragm from the bottom towards the top.



Derivations of this design:

The basic planar-magnetic principle makes it available for very wide variations upon the design. The most common variation is the use of a push-pull magnet design vs. the original single ended design. In this, two beds of magnets are used sandwiching the diaphragm between. Typically the smaller examples of this design will use spacers so small sections move independently rather than as one to again reduce issues with resonances and completely avoid issues of “flapping.” The famous Infinity Electromagnetic Inductive Midranges/ Tweeters (EMIM/EMIT) used just such a design, but with rare Earth Samarium-Cobalt magnets to create a more intense magnetic field.

Infinity EMIM:


Other variants use a tall, narrow magnetoplanar element to take advantage of the strengths without the beaming inherent to planar designs. The best known example is the Carver Amazing that combines a 4’ tall push-pull magneto planar element with four conventional woofers to provide the bass.

Performance and Limitations:

The magnetoplanar design is an incredibly well-rounded design. The best can nearly rival electrostats in most areas and excel at power handling. The typical saying is that anyone who’s had the chance to hear a pair of Magnepans set up and powered properly will have an extremely difficult time being satisfied with regular box speakers ever again. There are no box colorations with these speakers and the dipole design lends itself to a very open sound and 3D soundstage. Of course, no speaker is perfect and there are several drawbacks.

Magneto-planars suffer from the typical detractions of dipole planars, they are difficult to position, are fairly insensitive* (low to mid 80s dB/w/m for most), and they are bass limited requiring very large panels to make low bass. This design also suffers from veiling of low level detail in the music and subsequently is not as transparent as other designs can be detractions that are combined with the general nature of these speakers needing higher SPL levels to “open up.”

Whereas electrostatic designs are extremely capacitive, magneto-planars are the exact opposite and are nearly purely resistive. That is to say these speakers have nearly flat impedance curves and so are much easier for a well-designed amplifier to drive than the ratings would imply. These speakers benefit from a high-current amplifier.

*Something to remember is that most of these planar speakers act as line arrays. This means that the sound level falls off half as fast for the distance as conventional cone speakers, so an 86dB sensitive magneto-planar speaker can easily sound louder at the sitting position than a 90dB cone speaker for the same power.


Up Next: False Ribbons, Quasi-Ribbons, and True Ribbons

- JP

I look forward to the next installments with eager anticipation...

Thank you for taking the time to write this, JP.

I'll hopefully be able to get my printer working again- to print this out and digest it along with a beer and a burger will surely please me greatly!

Sorry about the delay in this last installment, life got in the way a bit.

Fake Ribbons, Quasi-Ribbons, and True Ribbons

Fake Ribbons

“Ribbon” is easily one of the most abused terms in audio and is often the speaker equivalent of PMPO power ratings on consumer electronics. Odds are if the marketing department advertises some ribbon driver on a speaker, it isn’t a ribbon or anything close. The reason is simple, true ribbons sound fantastic as such the word “ribbon” has become synonymous with excellent sound in most enthusiasts minds, so the marketing departments picked up on this and applied it to just about any flat driver that’s also long and narrow. This has actually been going on for decades ever since Jim Winey (yes, him again) invented the true ribbon tweeter in the late 1970s.

For the record, planar drivers are not ribbons, nor are Linnaeums. Sadly, Apogee is the greatest culprit in the abuse of the term “ribbon” applying it to their magneto-planar, quasi-ribbon, and true ribbon hybrids by proclaiming their products as the first “true full-range ribbon loudspeakers” when in fact the only true ribbon in their speakers crossed in at several kHz. Following their success, other companies adopted a broader definition of “ribbon” to remain competitive and recently even Magnepan has started calling their designs full-range ribbons. (An act directly in consequence of Apogee copying Winey’s patented designs and calling them ribbons.)

True Ribbons

A True Ribbon, as they have come to be known to differentiate them from the fake ribbons, is a very simple design. It consists typically of two columns of very powerful rare earth magnets creating a magnetic field strength in between of 0.5 Tesla or greater (upwards of in excess of 1 Tesla) in the space in between. [For comparison, the Earth’s magnetic field is only 0.00005 Tesla whereas the most powerful MRI devices create a field strength of 6 Tesla. – There was actually a humorous incident of the ribbons in my Flatline Designs being stuck to my truck bed and the subsequent difficulty I had in removing them. ;-) ] These magnets are typically mounted in a very sturdy steel frame to keep the magnetic attraction from collapsing the contraption.

In the middle of these magnet columns is suspended a metal ribbon typically constructed of very thing Aluminum foil. At the very top and bottom it is affixed to the frame, but is otherwise held in free space between the magnets. To allow for movement forwards and backwards and prevent stretching when a signal is applied, the ribbon is slightly accordion folded throughout its length. The added advantage of this folding is that slightly more material is present throughout the length, which subsequently allows for a slightly sensitivity and higher impedance – more about this later.

Here is Magnepan’s drawing of a True Ribbon’s cross-section:


The operating principle is as simple as the construction. When the amplifier’s signal is passed through the ribbon via connections at each end, the current produces a slight magnetic field that interacts with that of the permanent magnets. (The same principle as conventional dynamic drivers and magneto-planar designs.)

The large nearly full-range units are fairly rare due to their great expense in manufacture and relative fragility. So, most true ribbons are used only as tweeters with small elements, such as the famous Raven.

Raven exploded diagram:


View from the top:


A similar assembled unit:


In this last image, you can see the huge, powerful magnets and the subsequent support structure to maintain that small gap in the bottom where the ribbon will be suspended. True Ribbon tweeters are typically designed to mount in a conventional box-type speaker and so the back wave needs to be absorbed in the felt-lined back chamber within the driver to maintain consistently good sound without interference from the backwaves of the other drivers.

What you may have noticed is the presence of a small matching transformer in the back of these images. True ribbons are, as noted, essentially strips of aluminum foil and so have very limited inherent resistance. As such, the impedance on even a 175CM long element can be just barely above 1 ohm. For true ribbon tweeters such as the Raven that use an element only a few centimeters long, the amplifier would essentially see a short were one connected directly. Therefore, a step-up transformer is typically required on the driver to allow the amplifier see a typical 4 or 8 ohm impedance load.

Performance and Limitations:

The strength of this design comes from the miniscule moving mass of the ribbon element that can be as light as the diaphragm in an electrostatic speaker, but without the resonance disadvantages cause by the Mylar being secured at all ends and held to a particular taughtness. As such, the low level detail and transient response of true ribbon drivers is incredible, perhaps surpassing even the best ESLs. Most of the larger units (uncommon due to the high expense of manufacture) are true dipoles thus gaining that added advantage of a “boxless” sound. Simply put, true ribbons are some of the least colored, most accurate, and most transparent/revealing drivers out there, bar none.

The weaknesses, however, tend to be found in the fragility of the element. Most true ribbon elements, the larger units at least, are very fragile due to the extreme thinness of the metal and so breakages are common. At higher SPL output levels the ribbon, if not broken, can at least stretch slightly reducing the overall quality of performance. The true ribbon tweeter in the Magnepan MG-20.1 comes with a recommendation of element replacement once a year due to this issue.

The other weakness, as has been noted, is the very low impedance inherent with these drivers. Smaller units can suffer from transformer colorations, though some prefer this sonic effect. The larger units that go without a transformer can often tax an amplifier far beyond the point where even competently designed models go up in smoke. Due to this, the recommended amplifier for a near full-range true ribbon is one capable of handling near-short loads, such as the Krell KSA-100S, an amplifier designed with no protection circuitry and thus capable of handling down to 0.2 ohm loads. (If anyone has one available for trade, I’d really love one for my Flatline Designs that dip down nearly this low.)


Quasi-Ribbons

Due to the expense of construction of True Ribbons, a cheaper alternative was made, the Quasi-Ribbon. Whereas the True Ribbons use just a thin aluminum element, the Quasi-Ribbon borrows from the magnetoplanar design. The element in these drivers is typically a strip of Mylar upon which a voice coil is etched in a foil trace is glued. This element is then suspended between two columns of very powerful magnets much like a True Ribbon, with only the short ends attached and the length in the center free. The advantages to this approach are that the etched voice coil allow for regular impedance loads for the amplifier while the multiple runs also allow for a slight increase in sensitivity (typically, ribbon sensitivity is mostly determined by the strength of the permanent magnets’ field.) Thanks to Mylar possessing a higher tensile strength than steel at these scales, the quasi-ribbon tends to be less prone to breakage and stretching than the true ribbons and so will offer up years of consistent performance. Despite these advantages, however, a quasi-ribbon has not yet been made that can equal the true ribbon in performance due to the much greater mass of the Mylar/foil combination.

It is fairly easy to visually identify a quasi-ribbon as the element obviously has several vertical metal traces on a Mylar substrate. All but the highest end Apogee/Perigee/Analysis speakers use these as their “ribbon” element (the TOTL models do use a true ribbon for the tweeter above 10kHz), so if you see one in person, have a look and you’ll be able to tell.

Due to the nature of Quasi-Ribbons, the term is fairly loose in its definition. A true quasi-ribbon (that sounds odd) is only attached at the ends and suspended between two columns of magnets. However, magneto-planar designs using push-pull magnet arrangements and the foil traces have also regularly been called quasi-ribbons as well.

I was only able to find this one poor quality picture of a quasi-ribbon; oddly, most people with quasi-ribbon speakers don’t take close-up pictures of the elements:


Performance and Limitations:

Quasi-ribbons often allow a budget approximation of true ribbon sound while being more robust and allowing for cheaper amplifiers due to fewer impedance issues. The weakness lies in their much higher moving mass compared to true ribbons and slight plastic or “credit card” coloration.


If you come up with any questions, just ask and I'll try to answer them.
- JP

iv been wondering how these speakers work for quite sometime, im still not quite sure this is like rocket science too me very nice and detailed topic, when im smarter il come back and reread it