maybe your assumptions are wrong - transducer efficiency ultimately has a trade off against bandwidth - although most speaker diver's efficiency is way below that limit
also electrostatic drivers look horribly inefficient when driven from conventional amps because they are a mostly reactive load - if you used a bidirectional switching topology amp you could recover the stored charge each cycle
at a single frequency you can cancel the esl reactance with a series inductor for good electrical drive efficiency at that frequency
horns have limited bandwidth, possible nonlinearity due to high pressures in the throat
also electrostatic drivers look horribly inefficient when driven from conventional amps because they are a mostly reactive load - if you used a bidirectional switching topology amp you could recover the stored charge each cycle
at a single frequency you can cancel the esl reactance with a series inductor for good electrical drive efficiency at that frequency
horns have limited bandwidth, possible nonlinearity due to high pressures in the throat
Last edited:
but you can probably improve efficiency by optimising the transformer for the application and not using standard audio standards. (maybe) I like the idea of designing custom systes with the whole chain in mind, including the room. I guess I don't understand where the 4-16 Ohm standard for speakers came from. It doesn't seem to be optimised for any amplifier technology ever used.
the amplifier switch topology needs to support bidirectional current flow at both positive and negative output V - bjt switches won't
there is an old JAES paper on a multi-section esl driving a horn - possibly this could be upped in vertical dimension to a line source AES E-Library Horn-Loaded Electrostatic Loudspeaker
there is an old JAES paper on a multi-section esl driving a horn - possibly this could be upped in vertical dimension to a line source AES E-Library Horn-Loaded Electrostatic Loudspeaker
I don't understand peoples interest in line arrays they seem to amplify the shortcommings of the drivers used. I.e. limited dispersion, frequency cancelations, etc, what exactly are the advantages? It seems like a horn with one wide frequency driver far exceeds the performance of a line array with like 8 of them.
you misunderstood the dimension I was suggesting you "stretch"
the planar drivers use elements the height of your room - the horn becomes "2-d", flaring only in the horizonal plane
this avoids the interference/comb filtering with multiple drivers
floor, ceiling reflection just look like an extension of the line source
with more radiating area, slower roll-off with distance you can get higher spl with less displacement = better linearity
a inherent problem is that cylindrical wave propagation is dispersive although you are probably listening "near field" so frequency balance would be fine
if you can afford the space for 2 coffin sized speakers in your room line sources may be a good option
the planar drivers use elements the height of your room - the horn becomes "2-d", flaring only in the horizonal plane
this avoids the interference/comb filtering with multiple drivers
floor, ceiling reflection just look like an extension of the line source
with more radiating area, slower roll-off with distance you can get higher spl with less displacement = better linearity
a inherent problem is that cylindrical wave propagation is dispersive although you are probably listening "near field" so frequency balance would be fine
if you can afford the space for 2 coffin sized speakers in your room line sources may be a good option
Last edited:
Here's my fantasy-physics: fabulous idea.
A strength of ESLs is that the impedance of the Saranwrap diaphragm in making sound waves is vastly closer to air than shaking big blocks of cardboard (AKA cones) is. With a horn transforming this a lot further, aberrant behaviour of the ESL diaphragm begins to vanish* and only the motive power (which is clean electric waves coming from the amp and transformers (if any)), defines the very very clean sound output.
Yes conventionally, the parameters of horns and ESLs don't seem to work together. But the concept sure seems worth exploring.
*Really due to negative feedback from the air load.
A strength of ESLs is that the impedance of the Saranwrap diaphragm in making sound waves is vastly closer to air than shaking big blocks of cardboard (AKA cones) is. With a horn transforming this a lot further, aberrant behaviour of the ESL diaphragm begins to vanish* and only the motive power (which is clean electric waves coming from the amp and transformers (if any)), defines the very very clean sound output.
Yes conventionally, the parameters of horns and ESLs don't seem to work together. But the concept sure seems worth exploring.
*Really due to negative feedback from the air load.
Last edited:
Efficiency is the acoustic output power divided by the electrical input power.
Sensitivity is the acoustic output power divided by the electrical input voltage.
They are related, often confused, but not the same.
Compared to dynamic drivers, ESLs are very efficient, but generally a bit less sensitive.
When you hear people say that ESLs are less efficient that usually mean they have to turn up the volume to obtain the same acoustic output level as a dynamic speaker. Since the volume knob is controlling the output amplitude of a voltage source, this is lower sensitivity, not efficiency.
A related limitation that is unique to the ESL is that of maximum output. ESLs are limited in the amount of force they can generate per unit area of diaphragm. Trying to exceed this limit results in ionizing the air in the gap between stator and diaphragm. Depending on the construction method and materials arcing, corona, and ozone generation may result. What this means is that if you want it louder, you can’t just keep increasing the voltage levels. The only way to increase the peak SPL output capability is to add area, or operate the ESL in a gas that has a higher ionization threshold.
Last edited:
The Dayton-Wrights were designed to be capable of powering movie theaters, at least with multiple sets.
If I am not mistaken, the same arithmetic that magically multiplies the peak power capability of the amps in a music system when you bi-amp also applies when you divide the tone compass with speakers.
Having said that, playing my D-W speakers 5 octaves wide, I can still blast over 100 dB in my modest-sized music room.
I understand the difference well I guess I had just always heard of people saying that their ELSs needed a 500w amp to really sing, I guess I assumed that this meant they were inefficient. I suppose the situation could just be due to the fact that amp designed for low voltage high current output were just not capable of the voltages necisarry to get the volume they are looking for.
Right. The place to make improvements is in the amp. Very clear improvement using a high voltage direct-drive amp rather than a 1:100 step-up transformer... although dangerous.
But that doesn't change the voltage breakdown issues. For that, either welder's gas or live in a very dry climate.
I still think if a horn helps a Rice-Kellogg gizmo, it should help an ESL too.
Thanks ,Steve,you brainiac,as I tried to say that but you do it so much better,he,he
I haved pushed my panel so hard to these limits that it all holds true!
And those were The answers that I was trying to find when I first started my project.
There are limitations!
But sinceraly you can't beat the low distortions that this technology has to offer!
jer
I agree,I am still working on that DDamp.
But with all of the limitations the trick is to get 25kv p-p.
jer
I used tubes and very direct drive (no caps in path) with a B+ of 2400 volts (which adds to the negative bias, of course), building on classic Sanders amp. Worked quite flawlessly 1980-95, roughly, when I sold it. Loud but not super loud with D-W wide-gap panels in fresh air.
These are the sort of posts I was talking about in my intial post. If all you have to offer is reasons you think it won't work, don't bother posting. You can certainly say to make it work one would have to overcome this problem, or deal with this limitation, but that is not what you have done.
Any coupling efficiency short of 100% is low, and can be improved. The coupling efficiency of an electrostatic is nowhere near 100% so your point is moot. I'm not saying this is the way to do it but I think much evidence has been shown to indicate that it is an avenue worth exploring. If you disagree than work on your solution, on your thread.
Is this due to the high presure or the high voltage? If it is because of the high voltages this seems like another possible advantage to horn loading.
The max SPL limitation of ESLs is due to voltage gradient across the air gap.
Unlike dynamic drivers, the mass that the electrostatic forces have to accelerate is made up of essentially just the air. That is why people keep saying that the coupling is already close to ideal. This is completely different from dynamic drivers where moving mass is dominated by the cone and voice coil.
In my experiments, when ever I increased the airload on an ESL with a horn, waveguide, AMT slot, etc the amplitude of the ESL diaphragm motion was reduced by the same ratio that the airload was increased. So, you wind up with no gain in sensitivity.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.