ESL vs Planar Magnetic

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
what can be said objectively about the tradeoffs between the two technologies ?

please note i am not asking which one is "better" just what are some of the differences ?

for example i have heard that magnepan has stronger bass than martin logan - can such a comment be traced back to the phyics of the technologies involved somehow ?

does any one of the technologies tend to produce less measurable distortion ?
 
most planar magnetic designs (full range like maggies) are single ended. This means the drive is only on one side of the diaphragm and the available linear drive range is very limited. Once the voice coil is out of the small linear range of gap the driving force either gets progressively weaker or stronger depending on diaphragm direction. Bass is a spin off bnefit due to voice coil mass combined with diaphragm tension so as with box speakers the bass is resonant in nature.
ESL's on the other hand have in comparrison a virtually massless diaphragm with balanced drive on both sides. With an ESL the driving forces (both pushing and pulling) are always balanced regardless of the position of the diaphragm between the two stators. This results in distortion levels which can only be matched by a ribbon driver and closely approached by a properly designed horn loaded driver (horn loading permits the driver to stay in its linear range). If you want the lowest distortion possible use either a ribbon or an ESL in conjunction with a sub. Both are suited to DIY construction though at the end of the day an ESL (if well planned) is in my estimation an easier and less expensive project. As an example a used set of Acoustat's can be purchased (in N.A.) for under 5- hundred dollars. This provides you with a full transformer step up interface along with your high voltage supplies. You can then concentrate on building your own panels and even compare them to the stock panels to judge your work. there are many construction threads on this forum for both ESL's as well as ribbons. Hope this helps. PS: the single ended aspect of planar magnetic drivers is not a problem as it results in second harmonic distortion which sounds good in reasonable quantities (as do maggies) but you asked about the lowest distortion and that is in the relm of ESL's and ribbons.
 
most planar magnetic designs (full range like maggies) are single ended. This means the drive is only on one side of the diaphragm and the available linear drive range is very limited. Once the voice coil is out of the small linear range of gap the driving force either gets progressively weaker or stronger depending on diaphragm direction. Bass is a spin off bnefit due to voice coil mass combined with diaphragm tension so as with box speakers the bass is resonant in nature.
ESL's on the other hand have in comparrison a virtually massless diaphragm with balanced drive on both sides. With an ESL the driving forces (both pushing and pulling) are always balanced regardless of the position of the diaphragm between the two stators. This results in distortion levels which can only be matched by a ribbon driver and closely approached by a properly designed horn loaded driver (horn loading permits the driver to stay in its linear range). If you want the lowest distortion possible use either a ribbon or an ESL in conjunction with a sub. Both are suited to DIY construction though at the end of the day an ESL (if well planned) is in my estimation an easier and less expensive project. As an example a used set of Acoustat's can be purchased (in N.A.) for under 5- hundred dollars. This provides you with a full transformer step up interface along with your high voltage supplies. You can then concentrate on building your own panels and even compare them to the stock panels to judge your work. there are many construction threads on this forum for both ESL's as well as ribbons. Hope this helps. PS: the single ended aspect of planar magnetic drivers is not a problem as it results in second harmonic distortion which sounds good in reasonable quantities (as do maggies) but you asked about the lowest distortion and that is in the relm of ESL's and ribbons.

i know BG has push pull planar magnetic drivers. i thought the top of the line maggie was also push pull ? well perhaps they're not, but they *could* be made that way in theory. i think the linearity of a push-pull electrostat would still be greater than that of push pull planar-magnetic though.

you know from your description one would think that planar magnetic speakers have no benefit at all - but they have their following. isn't there something they do better ?
 
Last edited:
please do not read your own bias into my reply. you asked which technology had the lowest distortion and I did my best to answer your question. I do not wish to get into a mud slinging match here. I have owned and enjoyed both styles of drivers and have a pair of each at present. Personally I like ESL's and I think they offer the easiest approach for the DIY enthusiast but that is just my opinion.
As I said planar magnetic designs have drive issues in terms of linear drive range but like any dynamic driver if kept within their linear range they are just that linear. It matters not if you double end a planar magnet design as far as this issue goes though you can increase the range of linear throw with a double sided magnet structure but once past that point the same issues are there. You can not however equal the truly balanced drive of an ESL. This does not mean an isodynamic driver cannot be made to have linear throw in its available range nor that it cannot sound good. I am only trying to make clear what one design has as differences Vs the other.
I did mention several positive aspects of single ended isodynamic drivers which you seem to have overlooked or ignored, that being the production of second harmonic distortion when driven beyond thier linear range. This gives the driver a sweet round sound which is very pleasing, just listen to an old SMG or an MMG and see for yourself they sound excellent. The other benefit of isodynamic drivers is that of the mass of the voice coils which when tweaked along with the diaphragm tension yields bass output which you could not achieve from the same size panel when driven as an ESL. Debates could be had over the quality of this resonant bass but as I have said maggies do sound good so the debate is somewhat moot. I hope that I have been able to make more clear this time what I attempted to tell you in my first post. Just so you still dont think I have a negative bias against isodynamic drivers I codesigned a planar magnetic full range loudspeaker which was taken to market many years ago and was first sold through Sumo in California. That was a planar diaphragm driven by a single 1.25 inch four ohm voicecoil via a very good motor structure. This design went on to be sold under the Museatex bannar as the Melior loudspeaker (there were a number of versions). There is a good discription of this design in an early ninties issue of Audio if you are interested to look. Regards Moray James.
 
good analysis !

what about maximum SPL ?

i am VERY concerned about maximum output.

i ideally would like dance club sound pressure levels - can either or both technologies do this ? is one inherently more or less efficient ?

i am somewhat concerned about efficiency but mostly about ultimate output capacity. if it takes a 10 kilowatt amp - so be it, just as long as it can deliver the goods ?

is efficiency of electrostats frequency dependent ? is it a function of the transformer ?
 
Last edited:
It can be done with ribbon mids and tweets.
An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

http://www.transmissionaudio.com/
My attempt, 11x2000cm ribbon:
URL]
 
Last edited:
Hi,

ohh, it can be done but it´s quite hard to achieve such SPLs with ESLs. You have to restrict on a hybrid ESL working >200Hz with considerable membrane area and add a suitable bass.
One problem regards the distribution of SPL over distance and angle.
A tall and wide panel (with sufficient membrane area) will limit the distribution angle to a small value. You can fudge a bit by placing several panels (curved panels in preferance) in parallel, angled to each other.
If the panel is tall it´ll throw a dipolar cylindrical wave into the room which is at first gaining in SPL (up to 3-4m) and then starting to fall off again. As such they are longthrow devices, keeping the SPL up much longer than a point source. The distribution of SPL over distance is more even. This allows to ´drive´ the panel less hard than a point source. But it is more difficult to define a louder dancefloor-area and a softer lounge-area.
Example: A panel distributing in a dipolar cylindrical fashion and which is measured to deliver a SPL of 110dB@4m is equivalent to a point source delivering 122dB@1m (standard measurement distance, where the panel probabely delivers only ~100dB). But while the panel delivers 107dB@8m the point source drops to 104dB. This is a difference of 8dB in SPL-variation (122-104)|(110-100)over a distance of just 7m!
Even though You can reach very high levels of SPL the ESL might not sound as loud, because with very clean signals the ear is less stressed and reacts with less distress (tinnitus). You can stand those high levels longer and with less fatigue (nonetheless are those levels very unhealthy).
The ESL can take high levels of input power indefinetly time because it doesn´t heat up like a voicecoil-driven speaker. The increase in SPL isn´t compressed thermally/powerwise, because there´s no resistance in the drive system! The needed power can be quite low. The cited 110dB@4m were reached with an amplifier output voltage calculating to 50W@4Ohms.

jauu
Calvin
 
Last edited:
Hi,

ohh, it can be done but it´s quite hard to achieve such SPLs with ESLs. You have to restrict on a hybrid ESL working >200Hz with considerable membrane area and add a suitable bass.
One problem regards the distribution of SPL over distance and angle.
A tall and wide panel (with sufficient membrane area) will limit the distribution angle to a small value. You can fudge a bit by placing several panels (curved panels in preferance) in parallel, angled to each other.
If the panel is tall it´ll throw a dipolar cylindrical wave into the room which is at first gaining in SPL (up to 3-4m) and then starting to fall off again. As such they are longthrow devices, keeping the SPL up much longer than a point source. The distribution of SPL over distance is more even. This allows to ´drive´ the panel less hard than a point source. But it is more difficult to define a louder dancefloor-area and a softer lounge-area.
Example: A panel distributing in a dipolar cylindrical fashion and which is measured to deliver a SPL of 110dB@4m is equivalent to a point source delivering 122dB@1m (standard measurement distance, where the panel probabely delivers only ~100dB). But while the panel delivers 107dB@8m the point source drops to 104dB. This is a difference of 8dB in SPL-variation (122-104)|(110-100)over a distance of just 7m!
Even though You can reach very high levels of SPL the ESL might not sound as loud, because with very clean signals the ear is less stressed and reacts with less distress (tinnitus). You can stand those high levels longer and with less fatigue (nonetheless are those levels very unhealthy).
The ESL can take high levels of input power indefinetly time because it doesn´t heat up like a voicecoil-driven speaker. The increase in SPL isn´t compressed thermally/powerwise, because there´s no resistance in the drive system! The needed power can be quite low. The cited 110dB@4m were reached with an amplifier output voltage calculating to 50W@4Ohms.

jauu
Calvin

if overdriven how will ESL fail ? will there be sparks flying between the stators ? is the transformer insulation going to fail ? what will actually happen ? what is the limiting factor ?
 
Pretty Impressive steve , could you post a pic of your setup ?

It doesn't look all that impressive to me, but I have been living with it for many years. I no longer take pictures as it is too difficult for me these days. My speakers are Magnepan T-4As being driven by a pair of 1200W mono blocks. I have removed the fuses for the speaker mids and tweeters as I don't worry much about the amps clipping.
 
Hi,

how the ESL behaves depends on its design and includes several factors.
The limits are:
- arcing because of the air breaking down as isolator
- membrane touches the stators
- arcing within the cabling or Audio transformer

The ESL should be designed such that the mechanical limit is the one which comes first. It´s just a ´warning´ noise when the diaphragm hits the stators, but it doesn´t harm.
Arcing between stator and diaphragm is usually not very harmful with highly resistive membrane coatings. It will only burn a small hole in the membrane at max. It may be fatal though with low low resistance coated membranes, but those are seldomly used nowadays.
Arcing within the cabling or even worse within the transformer should be avoided, because it destroys the tranny.
You can use some safeguard devices to protect the speaker from damaging overload voltage conditions.
Electrostatics simply can't do the excursion necessary due to arcing.
is not at all true. The excursion has nothing to do with arcing in first place. Keeping to a frequency range where ESLs rule (>200Hz) the needed excursions are small, very small.
Ths is one of the few real limitations a "good" ESL has to accept -- no bass! If You obey this limitation You can easily design ESLs with a very large dynamic range, comparable even to Horns..

jauu
Calvin
 
IIRC, the larger "professional" series Acoustat 6's and 8's were good for over 120dB in a large room -- assuming you had the current available. Acoustat claimed a -3dB point in the mid to high 20's. The radiating area was considerable in either version -- something like 20-30 square feet per speaker, and since they were floor to ceiling tall, they had the advantage of significant reinforcement. Unfortunately, imaging was definitely "head in a vice."
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.