Calvin said:
Dealers are too far away for me - -besides I'll end up with having to listen to the gibberish from some slimey sales person with a wallet radar.
I agree - being an ex Quad Electrostat guy -- quite right - they are not what one would expect for the money - for the money Quad are asking they should be "exceptional" , but they never quite make it.
They do put my KEF 104/2 reference and my Cambridge Audio R50's and my "Icebricks" (Isobariks) to shame in the midband - treble and lack of colouration.
Thanks to all that have given me some nice linkies to read -- quite a lot of those have given me some ideas.
I think I may very well go for my own home made electrostats - so expect me to be playing with Mylar PP film - foil and drinking the coating fluid 😀
Time to dust off my Bruel & Kjaer and Ferrograph test equipment
Calvin said:
I would have to dis-agree that the mag field on a true ribbon is non-linear when excursion is in the field , anyway yes it can be non linear if not designed correctly or used in an improper way , but it is ten times better than the acoustic phase issues of ESl's with a stator in front and back of it's diaphragm. There is so much smearing from an electrostatic ( compared to a ribbon , with ESL's i tend to favor open stators ) due to it's configuration , recording detail would have to be of low importance to live with that . ESL's do have an advantage in Speed ( very audible) but the smearing and gigantic big mouth sound all the time leave me cold ( my subjective opinion of course) with lack of detail and poor imaging it does a poor job of recreating time and space. The sound is open ( planers or big panel speakers do this well ) and READ not flaming anyone who likes it , just giving my objective and subjective opinion.
Calvin,
I do agree on the quad statement , do not agree on maggies ( poor ribbon speaker , good tweeter ) and do agree that you would only be able to find a handful of Commercially available ESL and Ribbons that are acceptable. I have heard really fantastic versions of both with the strengths and weakness described previously...............
As to the cost of magnets for RIbbons , a few years a go i had a fantastic supplier for magnets i can check to see if he is still in business , if so his prices back then where very reasonable, 120 typically to build a 4 ft ribbon driver..................
Just wonder if anybody has ever build narrow (~1") ESL line source as tweeter only? More robust than ribbon.
Polar response is like ribbon and if long enough, SPL shouldn´t be that big issue >1kHz? Any comments? Experiences?
Polar response is like ribbon and if long enough, SPL shouldn´t be that big issue >1kHz? Any comments? Experiences?
ABJensen said:Elektrotats are too wide 20 cm.Planars (my diy) is 2 cm wide
An externally hosted image should be here but it was not working when we last tested it.
Would love to see more details on your ribbon speaker , could you please post more pics......
Hi,
if something as ´acoustic phase smear´, lack of detail or gigantic mouth occurs it aint been a proper ESL or proper positioning of the device!
I don´t really know what You are talking about anyway?
ESLs can be built extremely open and very flat...so there´s no room for ´phase issues´ whereas the size of the magnets of a Ribbon can lead to early reflections ahem...phase smear.
Anyway both systems are capable of acoustic phase linearity superior to the classical speaker.
But being a ´massless´ system the ESL outperforms Ribbons easily with regard to resolution of details. The high proportion of direct sound vers radiated sound leads to a extremely precise localisation even of the tinyest sonic events. But You have to carefully control what the planar radiates to the backside. Otherwise You get the too big soundstage quickly.
@AbJensen:
If You design the ESL properly, You will have a perfectly homogenous field in which the diaphragm can move. Finite element methods are very helpful to show whats going on and how to design the stator.
jauu
Calvin
if something as ´acoustic phase smear´, lack of detail or gigantic mouth occurs it aint been a proper ESL or proper positioning of the device!
I don´t really know what You are talking about anyway?
ESLs can be built extremely open and very flat...so there´s no room for ´phase issues´ whereas the size of the magnets of a Ribbon can lead to early reflections ahem...phase smear.
Anyway both systems are capable of acoustic phase linearity superior to the classical speaker.
But being a ´massless´ system the ESL outperforms Ribbons easily with regard to resolution of details. The high proportion of direct sound vers radiated sound leads to a extremely precise localisation even of the tinyest sonic events. But You have to carefully control what the planar radiates to the backside. Otherwise You get the too big soundstage quickly.
@AbJensen:
If You design the ESL properly, You will have a perfectly homogenous field in which the diaphragm can move. Finite element methods are very helpful to show whats going on and how to design the stator.
jauu
Calvin
I don't know why but a lot of people are under the impression that you must spend a lot of money and devote massive engineering resources to designing and building an ESL. Finite element analysis and EM simulations make wonderful topics of conversation but are also completely unnecessary. Quad certainly didn't use either in designing the ESL57s or 63s.
Some seek nirvana in diaphragm coatings, others in bias supply topology. In most cases the obsessive attention to these minor quibbles has greatest impact in the mind of the obsessor.
If you understand and keep in mind the simple physics that determines the performance of an ESL you can make a great sounding speaker comparable to any commercial ESL but at much lower cost. For example, you simply can't get low frequency response from any practical sized ESL unless you're making headphones. The low frequency response limit is inversely proportional to the size of the speaker.
The coating is there to allow charge to be placed on the diaphragm. High resistance, low mass, and above all, adhesion to the diaphragm are characteristics that matter.
The bias supply simply supplies charge to the diaphragm. Any high voltage DC supply is fine, but safety is a concern so a very small HV supply should be used.
The stators exist to establish an electric field that will move the diaphragm. Any perforated conductive material can be used. The stators also usually provide some mechanical rigidity to the speaker.
Insulators can be any insulating material, but they too usually provide some mechanical rigidity. They should be kept thin to make a speaker that is easy to drive with common power amps.
The trickiest stuff is the adhesives that hold the tensioned diaphragm to the insulators and the transformers that drive the ESL. Neither of these are well kept secrets.
Low voltage toroidal power supply transformers that are often available as cheaply as $5 each can be used to drive the ESLs with results that match those of expensive, purpose-wound audio transformers.
Construction methods for ESLs can be found on numerous web sites and in books. I prefer simple methods with proven reliability over some of the complex schemes found at some web sites, but you're free to choose the methods that appeal to you. Just keep in mind that everything that costs more in either cash or effort should justify itself by improved performance. Even using the simple and inexpensive techniques and materials yields results that are very hard to improve upon.
I_F
Some seek nirvana in diaphragm coatings, others in bias supply topology. In most cases the obsessive attention to these minor quibbles has greatest impact in the mind of the obsessor.
If you understand and keep in mind the simple physics that determines the performance of an ESL you can make a great sounding speaker comparable to any commercial ESL but at much lower cost. For example, you simply can't get low frequency response from any practical sized ESL unless you're making headphones. The low frequency response limit is inversely proportional to the size of the speaker.
The coating is there to allow charge to be placed on the diaphragm. High resistance, low mass, and above all, adhesion to the diaphragm are characteristics that matter.
The bias supply simply supplies charge to the diaphragm. Any high voltage DC supply is fine, but safety is a concern so a very small HV supply should be used.
The stators exist to establish an electric field that will move the diaphragm. Any perforated conductive material can be used. The stators also usually provide some mechanical rigidity to the speaker.
Insulators can be any insulating material, but they too usually provide some mechanical rigidity. They should be kept thin to make a speaker that is easy to drive with common power amps.
The trickiest stuff is the adhesives that hold the tensioned diaphragm to the insulators and the transformers that drive the ESL. Neither of these are well kept secrets.
Low voltage toroidal power supply transformers that are often available as cheaply as $5 each can be used to drive the ESLs with results that match those of expensive, purpose-wound audio transformers.
Construction methods for ESLs can be found on numerous web sites and in books. I prefer simple methods with proven reliability over some of the complex schemes found at some web sites, but you're free to choose the methods that appeal to you. Just keep in mind that everything that costs more in either cash or effort should justify itself by improved performance. Even using the simple and inexpensive techniques and materials yields results that are very hard to improve upon.
I_F
Hi,
there are probable millions of ´great sounding´ speakers out there.
And indeed its not that tricky to built a great sounding ESL. But isn´t the aim here to built something which is way above average?
If so, You have to pay attention to the smallest Detail.
Maybe because God is in the details ;-)
Than its just not so simple math (never wondered why there are nearly no usable sim-progs out there usable for ESLs?), not any coating, not any stator, not any insulator and definitely not any size and dimension.
To be able to build SOTA ESLs means slightly more than simply to put 5 parts together. 😉
jauu
Calvin
there are probable millions of ´great sounding´ speakers out there.
And indeed its not that tricky to built a great sounding ESL. But isn´t the aim here to built something which is way above average?
If so, You have to pay attention to the smallest Detail.
Maybe because God is in the details ;-)
Than its just not so simple math (never wondered why there are nearly no usable sim-progs out there usable for ESLs?), not any coating, not any stator, not any insulator and definitely not any size and dimension.
To be able to build SOTA ESLs means slightly more than simply to put 5 parts together. 😉
jauu
Calvin
HI Calvin
I am asking myself the usual product design questions as I set out my design parameters.
What size? - physical shape?
What materials choices do I have?
Quality of materials?
Basically, what ingredients do I make the cake from ?
Power handling?
What Impedence structure (both mechanical and electrical) can I expect?
How many elements?
What are the knowns?
What can I mathematically work out or predict from experience?
What facts will I have to discover as I go through my build journey?
What are the unknowns at this stage?
What innovation (if any) could I introduce to my variants?
I'm getting excited again -- this is the reason I exist....Science 🙂
I am asking myself the usual product design questions as I set out my design parameters.
What size? - physical shape?
What materials choices do I have?
Quality of materials?
Basically, what ingredients do I make the cake from ?
Power handling?
What Impedence structure (both mechanical and electrical) can I expect?
How many elements?
What are the knowns? What can I mathematically work out or predict from experience? What facts will I have to discover as I go through my build journey? What are the unknowns at this stage?What innovation (if any) could I introduce to my variants?
I'm getting excited again -- this is the reason I exist....Science 🙂
Calvin said:
Calvin ,
Thanks for the response , Acoustic phase smear , LOL ,..................
Yes the ribbon does suffer from these reflections , ( all speaker's to some degree ) far less so than a ESL. Now an ESL with a closed grid is the worst (logans for eg.) , open grid is much better .
Now all speakers do suffer from various strengths and weakness's , the combined totals is what leads to the end results .
Speaking objectively an ESL is not a mass less speaker ( there is no such animal, quite possible you are being tongue in cheek, here ) and relative, speaker to speaker, a pure ribbon is still very light and is driven enough through it's length for the extra mass relative to an ESL, is negated by it's positive.
A Pure Ribbon is much more open and it's less obscure radiating area gives it a big advantage over the right in front of the diaphragm grid plates that affect an ESL. Again this is what makes audio enticing to the hobbist and Anathema to a designer /manufacturer . Technically sound does not make it universally loved.
Calvin i would love to hear how to build an ESL that is open and free of it's grid , the best i have ever done had a wired open grid stator and measurements still showed it had too much reflection vs a pure ribbon element ., subjectively it was audible also............
One last tidbit , the purely resistive load of the Ribbon , with very little phase angle changes make for a much easier load to drive than the low impedance capacitor of an ESL , this goes along way in how it affects the sound of an audio amplifier..
Now not saying Ribbon is Better than an ESL , Objectively and subjectively to me i prefer RIbbons , Hybrid Ribbons...
A.Wayne
a.wayne said:
When I heard the Magneplanar 3.6 -- I was very impressed with the punch it was able to produce -- for fun I chose the Dark Side of the Moon by Pink Floyd to see how it coped with the heartbeat at the intro.
There are so many other planars I have not had the experience of like the Martin Logans.
ABJensen said:
I guess the diaphragm weight is much more important than the magnet type. The Analysis magnetostats have ceramic magnets and are worlds more dynamic than maggies. BTW, does anyone have schematic diagrams how Apogee and Analysis magnetostats are built?
a.wayne said:
Yes its a planar. Ribbons is way too easy too make .And too my ears it sound more relaxed. And paper gives a nice warm sound.
worth considering...
PLL you should take a look at the Superior Essex list of magnet wire a numberr of dual coat versions which have breakdown voltages over 12Kv http://www.superioressex.com/magnetwire.aspx?id=1390
The Dutch ESL groups have for the most part gravitated toward insulated wire stator construction as there are many benefits over perf metal. If you want to gow with perf metap I would suggest having someone like Russ Knotts at just ral music coat you panels to your size. That way you know that they will work and that they will all be the same. The average powdercoat shop does not have the experience to do this kind of coating and you only get one attempt per panel so it can be an expensive learning curve.
Go for the thinnest mylar C or Hostaphan RE you can find 2.5 - 3.5 micron. You might see if you can find low micron PEN Teonex Q72 film which is even stronger that the PET films (Milinex, Mylar and Hostaphan RE). Teonex is made by DuPont.
The standard Acoustat panel has an overall size of 9"x 46" and an active driven area of 6.75"x 43". Each diaphragm has a 1/2 inch wide peripherial non driven gap of diaphragm. I know that under the right conditions four such panels per channel can make astounding bass so I dont think that you would need to have more panel area than that. That's it for the moment.
PLL you should take a look at the Superior Essex list of magnet wire a numberr of dual coat versions which have breakdown voltages over 12Kv http://www.superioressex.com/magnetwire.aspx?id=1390
The Dutch ESL groups have for the most part gravitated toward insulated wire stator construction as there are many benefits over perf metal. If you want to gow with perf metap I would suggest having someone like Russ Knotts at just ral music coat you panels to your size. That way you know that they will work and that they will all be the same. The average powdercoat shop does not have the experience to do this kind of coating and you only get one attempt per panel so it can be an expensive learning curve.
Go for the thinnest mylar C or Hostaphan RE you can find 2.5 - 3.5 micron. You might see if you can find low micron PEN Teonex Q72 film which is even stronger that the PET films (Milinex, Mylar and Hostaphan RE). Teonex is made by DuPont.
The standard Acoustat panel has an overall size of 9"x 46" and an active driven area of 6.75"x 43". Each diaphragm has a 1/2 inch wide peripherial non driven gap of diaphragm. I know that under the right conditions four such panels per channel can make astounding bass so I dont think that you would need to have more panel area than that. That's it for the moment.
Hi,
@wayne
I admit I had some difficulties to understand your text! But I hope I got it all right that the answers will be ok ;-)
A ESL is of course not a ´massless´ speaker , but its generally referred to as such, because its diaphragm mass doesn´t play a significant role in any respect -and in no formula- throughout the audio band. Mass related rolloff for examples lies way above 20kHz for typical materials like Mylar of less than 6micron thickness.
I seriously doubt that You could measure any possible influence at all...if so, I´d like to know how You possibly have done it! The dimensions (hole sizes, wire distances et al) are so small that You couldn´t proof what You claim with typical microphone capsules and measurement setups.
Audible comparisons are in this case imho useless because they underly greatly the personal taste, require enormous listening experience together with profound technical knowledge of both principles, a usually lacking number of test specimen, and the sound of different specimen of the same principle varies vastly even with minor design differences.
It makes for example remarkably large sonic differences what coating is applied to an ESL-membrane. How will You be able than to tell of what influence a stator-construction is? Spacer distance and thickness, panel size....and there are many more variables that affect the sound of a ESL.
What makes You sure that You simply did listen to a rather badly designed panel or a different sonic effect?
I appreciate You saying that Your impression is subjective....so is mine 😉 I heard Maggies, I heard Analysis. And while I admit that I liked especially the Analsysis very much and she outperformed some of the better known ESLs it was no match to a tweaked ML, a Capaciti or my own Panels.
You are completely right with regard to the ease or difficulty of drive needs. But a amplifier just needs to be designed apropiately to drive an ESL without any signs of stress. There are not many on the market though....but than again.....isn´t this here not all about good stuff?? 😉
The Kronzillas are prime examples of an fine tubeamp that doesn´t care what is connected to it and there are some decent class-d amps around that have equally no problem of driving the ESLs.
@loopy
If I count the words I´ve written in forums about ESLs over the years, I´d been better off if I´d written a book instead 😀
Well there are imo no good short answers as long as there´s no sharp image of what You want to achieve and the amount of effort You´re willing to put in such a project.
I´d recommend to read copies of Mr. Sanders´s, Peter Baxandall´s and Ron Wagner´s texts to get a basic understanding of the functioning, design and used materials of ESLs. Searching this forum will reveal lots of ´secrets´ too.
Just one recommendation: As a first project do Yourself a favour and don´t start with a FR-ESL!
jauu
Calvin
@wayne
I admit I had some difficulties to understand your text! But I hope I got it all right that the answers will be ok ;-)
A ESL is of course not a ´massless´ speaker , but its generally referred to as such, because its diaphragm mass doesn´t play a significant role in any respect -and in no formula- throughout the audio band. Mass related rolloff for examples lies way above 20kHz for typical materials like Mylar of less than 6micron thickness.
I don´t agree with that. You regard the grid as an obstacle with a negative effect. But that is rather not the case. It could instead be far more regarded as a flow resistance that introduces some kind of damping, which would otherwise be completely absent with all the negative side effects. Some Ribbons miss out any form of mechanical damping (accompanied with the always present danger of destroying the delicate thing)
I seriously doubt that You could measure any possible influence at all...if so, I´d like to know how You possibly have done it! The dimensions (hole sizes, wire distances et al) are so small that You couldn´t proof what You claim with typical microphone capsules and measurement setups.
Audible comparisons are in this case imho useless because they underly greatly the personal taste, require enormous listening experience together with profound technical knowledge of both principles, a usually lacking number of test specimen, and the sound of different specimen of the same principle varies vastly even with minor design differences.
It makes for example remarkably large sonic differences what coating is applied to an ESL-membrane. How will You be able than to tell of what influence a stator-construction is? Spacer distance and thickness, panel size....and there are many more variables that affect the sound of a ESL.
What makes You sure that You simply did listen to a rather badly designed panel or a different sonic effect?
I appreciate You saying that Your impression is subjective....so is mine 😉 I heard Maggies, I heard Analysis. And while I admit that I liked especially the Analsysis very much and she outperformed some of the better known ESLs it was no match to a tweaked ML, a Capaciti or my own Panels.
You are completely right with regard to the ease or difficulty of drive needs. But a amplifier just needs to be designed apropiately to drive an ESL without any signs of stress. There are not many on the market though....but than again.....isn´t this here not all about good stuff?? 😉
The Kronzillas are prime examples of an fine tubeamp that doesn´t care what is connected to it and there are some decent class-d amps around that have equally no problem of driving the ESLs.
@loopy
If I count the words I´ve written in forums about ESLs over the years, I´d been better off if I´d written a book instead 😀
Well there are imo no good short answers as long as there´s no sharp image of what You want to achieve and the amount of effort You´re willing to put in such a project.
I´d recommend to read copies of Mr. Sanders´s, Peter Baxandall´s and Ron Wagner´s texts to get a basic understanding of the functioning, design and used materials of ESLs. Searching this forum will reveal lots of ´secrets´ too.
Just one recommendation: As a first project do Yourself a favour and don´t start with a FR-ESL!
jauu
Calvin
Re: worth considering...
Calvin,
There is a reason why ultimate esl are made with open wired stators vs perf metal ( as also expressed by james moray post) and yes the difference and it's influence can be measured. Your comments on ribbon drivers tend to discuss them as
full bandwidth drivers , that they are not and i would never use a ribbon driver below 250 Hz. as there are better choice's for this .
Anyway grab your perf/metal esl and measure its frequency response vs an open grid wired stator esl of the same dimn.
you will see there are far less reflections. Ironic that as you point out and technically I'm in agreement with you about planer magnetic drivers being non linear , yet i have never heard any ESl of any dimn out perform an Apogee in the bass , the power from the planer is fantastic and is the best part of that speaker , of course takes a life time to get it right in the room , but when you do , fantastic.
As far as ESL's go i have listen to many , tested a few and A/B comparison too many to count, Soundlabs, Logans,( including the statement) Acoustat, Quads, single stacked , triple stacked, Bozak,
ohh gosh so many over the years most recently one called FInal, barely acceptable that one .......
moray james said:
Calvin said:
Calvin,
There is a reason why ultimate esl are made with open wired stators vs perf metal ( as also expressed by james moray post) and yes the difference and it's influence can be measured. Your comments on ribbon drivers tend to discuss them as
full bandwidth drivers , that they are not and i would never use a ribbon driver below 250 Hz. as there are better choice's for this .
Anyway grab your perf/metal esl and measure its frequency response vs an open grid wired stator esl of the same dimn.
you will see there are far less reflections. Ironic that as you point out and technically I'm in agreement with you about planer magnetic drivers being non linear , yet i have never heard any ESl of any dimn out perform an Apogee in the bass , the power from the planer is fantastic and is the best part of that speaker , of course takes a life time to get it right in the room , but when you do , fantastic.
As far as ESL's go i have listen to many , tested a few and A/B comparison too many to count, Soundlabs, Logans,( including the statement) Acoustat, Quads, single stacked , triple stacked, Bozak,
ohh gosh so many over the years most recently one called FInal, barely acceptable that one .......
Regarding damping....
A.Wayne: I agree with Calvin on this one and not just for ESL's but for Ribbons too. Bith ribbons as well as ESL's require additional resistive damping to controll the diaphragm as the local air load is not enough to do the job and the higher the diaphrgm mass the more so required. I you read about the Rall Tweeters you will see such statements. For ESL's I think that the optimum open area is around 30 - 35 % open stator to provide enough air load on the diaphragm. Most people shoot sor 40 - 50% open and I think that this is too high. I know that a chap by the name of Dave Lang did some fine studdies and published them in speaker builder some years back and he was surprised to find that a perf metal plate of 30 - 35 % open was in fact more efficient than a aluminym screen stator with a much more open area with wires every 1/16 of an inch in each direction. Dave's expectation had been the reverse so he was very surprised.
A.Wayne: I agree with Calvin on this one and not just for ESL's but for Ribbons too. Bith ribbons as well as ESL's require additional resistive damping to controll the diaphragm as the local air load is not enough to do the job and the higher the diaphrgm mass the more so required. I you read about the Rall Tweeters you will see such statements. For ESL's I think that the optimum open area is around 30 - 35 % open stator to provide enough air load on the diaphragm. Most people shoot sor 40 - 50% open and I think that this is too high. I know that a chap by the name of Dave Lang did some fine studdies and published them in speaker builder some years back and he was surprised to find that a perf metal plate of 30 - 35 % open was in fact more efficient than a aluminym screen stator with a much more open area with wires every 1/16 of an inch in each direction. Dave's expectation had been the reverse so he was very surprised.
- Status
- Not open for further replies.