I have noticed if you don't login and read the posts you were sent email notification about while logged into your account you sometimes stop getting additional emails until you do. Perhaps this was the case here.
Line sources are advantageous for ESLs because they allow you to use larger areas you need for efficiency while retaining narrower widths needed for good horizontal dispersion. If you can make your line source floor to ceiling in length, mirroring makes the line act acoustically like an infinitely long line source which removes the issues of response ripples from frequency cancellation in finite length line sources.
I have a dream.
Have you increased the airload on both sides of the diaphram i.e. a compound loaded horn? Even so it seems to me that if would be a worthwhile endevor simply for the better high frequency dispertion (using a front horn and phase plug) and decreasing the low frequency cancelations with a longer rear horn. The design I am currently envisioning (rife with design flaws to be sure) is an 8" circular diaphram with a short front horn boosting the range from about 400Hz to 1.6K, with a rather elaborate phase plug assembly to give the desired balance between dispersion and imaging, along with a rear horn helping with the 80Hz to 400Hz region, with the whole thing powered by a tube amp with it's high voltage output directly coupled to the stators to elliminate the need for expensive transformers. Add a tapped horn subwoofer powered by an inexpensive Class T or D amp for the ultimate in low distortion megadynamic audio. Go ahead now, burst my bubble.
Have you increased the airload on both sides of the diaphram i.e. a compound loaded horn? Even so it seems to me that if would be a worthwhile endevor simply for the better high frequency dispertion (using a front horn and phase plug) and decreasing the low frequency cancelations with a longer rear horn. The design I am currently envisioning (rife with design flaws to be sure) is an 8" circular diaphram with a short front horn boosting the range from about 400Hz to 1.6K, with a rather elaborate phase plug assembly to give the desired balance between dispersion and imaging, along with a rear horn helping with the 80Hz to 400Hz region, with the whole thing powered by a tube amp with it's high voltage output directly coupled to the stators to elliminate the need for expensive transformers. Add a tapped horn subwoofer powered by an inexpensive Class T or D amp for the ultimate in low distortion megadynamic audio. Go ahead now, burst my bubble.
No I actually clicked the link on my email and it gave me a message that said you are no longer subscribed to this thread.
If you clink on the link at the top of the email you are linked to the thread.
If you click on the link at the bottom of the email you are unsubcribed.
Yes. With multiple AMT slots you increase airload on both sides of the diaphragm and again, diaphragm motion and SPL is reduced by the ratio that airload has been increased. The ESL doesn't care which side of the diaphragm the load comes from.
Increasing high frequency dispersion is a worthy goal. With ESLs, you have the ability to easily decrease the radiating area with increasing frequency by segmentation so you can modify the radiation pattern at will. The native response of an ESL dipole point source slopes upward with increasing frequency at about 6dB/octave. So by segmenting the ESL diaphragm and reducing radiating area with increasing frequency you not only improve dispersion, you can flatten the frequency response as well. Attached is a response plot I took many years ago of a uniformly driven 8” x 8” square ESL dipole at 1m. I believe the two curves are slightly different off axis angles, but can’t remember exactly. Your 8” circular ESL would have a similar response.
Note that with a transformer step-up ratio of 100:1 the sensitivity of your 8” ESL will be pretty close to 86dB/2.83Vrms at 10kHz. So, if you want an ESL that matches the output capability of dynamic drivers at lower frequencies, you will need to extend the segmentation concept and add additional ESL area to your 8” ESL as you go down in frequency.
As you suggest, a long rear horn would help some with phase cancellation at low frequencies. But adding a simple baffle edge extension around the ESL will give you the same result since at low frequencies its really only about front-to-back path length. In general, if you keep the baffle extensions no longer than 1/2 the diameter of your ESL you won’t have problems with a midrange diffraction notch appearing in your on-axis response.
I don’t have that much experience with DD tube amps but would suggest that they tend to be more expensive than step-up transformers.
Attachments
I still haven't hear any explaination as to why this happens
I would imagine this is because the load is basically a capacitor I wonder what effect connecting the two stators with a large resistor and or a coil would have?
What step up transformer?
Wha't is a DD tube amp?
Last edited:
Hmmmm...I thought I had explained it well. I will try one last time.
Sound pressure is caused by pressure waves in the air which are created by accelerating the air.
F=MA, so acceleration = Force / mass
If you increase mass, acceleration is reduce by the same ratio.
If it was air we were accelerating, SPL would be reduced by this ratio as well.
Think of a dynamic woofer with moving mass of 200grams. Some of this moving mass is airload, but it is a small portion. Stick 200Grams of clay to the cone and measure the midband response, it will have fallen 6dB.
ESLs don't have any moving mass other than the air to accelerate.
So, if you double the airload with the same driving force, the acceleration of the air will be halved, and SPL will drop by 6dB.
The sloping response is caused by 2 acoustic phenomenon.
1) For low frequencies, the response falls 6dB/octave for frequencies lower than those that have wavelength > ESL diameter due to dipole phase cancellation
2) For high frequencies, the response rises 6dB/octave for frequencies higher than those with wavelength < ESL diameter due to directivity gain. (ie beaming)
Placing a resistor inline to the stators will cause the voltage applied to the stators to fall at 6dB/octave above the frequency where the resistance = the capacitive impedance. The result is the response will be flat above this frequency.
No particular transformer, just any that has a step-up ratio of 100:1 with adequate bandwidth.
DD = direct drive, basically the amplifier you described with tube output directly coupled to stators.
Simply not true! "Very light weight" does not equal "no mass". Compaired to the air the diaphram is still quite massive.
Both of which phenomenon my design eliminates.
Exactly what I'm shooting for.
Okay what part of "direct coupled" are you missing here?
Why would eliminating 4 of the most expensive parts make a system more expensive to build?
They keywords are 'high voltage' and 'worth doing'.
high voltage: over 1000V part prices explode
worth doing: it should be as safe and reliable as a transformer, and since it costs way more it should at least outperform the transformer by a significant margin. Not an easy job to do!
Your sarcasm is uncalled for.
Imho, you could probaby build a DD amp for less than that providing you know what you ar doing. Include R/D costs and you hit the roof pretty quick.
Now, alot of the discussion in this thread is over my head and I'll try to go back and re-read it all for a better grasp.
However I have a simple question. (Simple questions somehow tend to have the most complex answers, I don't know why?)
What would the result be if one were to make a "replica" of a Beveridege speaker or a design inspired by the Beveridges but instead of the acoustic lens use a horn?
I'm still not 100% sure where the line is drawn between horn and wave guide?
The reason why I pose the question is for two major reasons
1) How to get uniform dispersion
2) How to avoid beaming
Part two!
Take the speaker above and remove the cabinet behind the panel, what would happen now?
Small musings after the fact...
Part one.
Would it be possible to adjust chamber for symmetrical air load to the diaphragm?
Part two.
What will the difference in air load to the diaphragm actually do?
Now, alot of the discussion in this thread is over my head and I'll try to go back and re-read it all for a better grasp.
However I have a simple question. (Simple questions somehow tend to have the most complex answers, I don't know why?)
What would the result be if one were to make a "replica" of a Beveridege speaker or a design inspired by the Beveridges but instead of the acoustic lens use a horn?
I'm still not 100% sure where the line is drawn between horn and wave guide?
The reason why I pose the question is for two major reasons
1) How to get uniform dispersion
2) How to avoid beaming
Part two!
Take the speaker above and remove the cabinet behind the panel, what would happen now?
Small musings after the fact...
Part one.
Would it be possible to adjust chamber for symmetrical air load to the diaphragm?
Part two.
What will the difference in air load to the diaphragm actually do?
It's offtopic, but being the one who's went this way:
Would you dare to make such an amp and put your vallet against possible injury. Have you been shocked with DC! voltage well above 1kV?
Even series cap as Kalvin suggested multiple times makes such an event quite unpleasant. It is fare to say that caps usually fail open circuit.
Well known Beveridge system has an amp confined inside the box... Suitable tubes are hard to come by and schematics IS indeed quite uncoventional.
I would try to kill backwave using sound absorbent materials instead of horn... though
Dangerous indeed, although I did raise three kids in a house with a DD amp... and no caps. Amp was housed on floor below, suspended from ceiling, above child height and used fancy Dayton-Wright cables and connectors between.
Just think of the cost (and size) of the oil-filled caps you need for filtering the B+ power supply. BTW, I used a choke in filter section too.
Horn, wave guide, lens, narrowing driver as freq is raised, curved diaphragm, multiple panels on an arc (Dennesen), lotsa ways to address beaming. Not to mention, it is no problem if you need just one chair in the sweet spot.
Just think of the cost (and size) of the oil-filled caps you need for filtering the B+ power supply. BTW, I used a choke in filter section too.
Horn, wave guide, lens, narrowing driver as freq is raised, curved diaphragm, multiple panels on an arc (Dennesen), lotsa ways to address beaming. Not to mention, it is no problem if you need just one chair in the sweet spot.
Last edited:
Did not mean to imply that Mylar was less dense than air, only that the weight of the Mylar is not as significant compared to the weight of the air it is throwing around as the weight of a cone and voice coil is. Several people had mentioned this, but you didn't seem to be picking up on it. So I tried a simplified explanation.
Don't read more into it than that.
Mylar is roughly 1000 times denser than air. So the weight of a 1/4 mil diaphragm is equivalent to about 1/4" thick layer of air on the diaphragm. At mid and lower frequencies, the slug of air the diaphragm is throwing around is equivalent to 0.85 x the radius of the diaphragm, so about 3.4" thick. This would be roughly 13.5 times heavier than the diaphragm. (See chapter 5 of Acoustics by Beranek) Suddenly the diaphragm doesn't look so massive does it
BTW, did you read the AES paper pointed out to you in post #8 and #25?
A method of getting around this issue is discussed in great technical detail.
Nobody suggested otherwise. People are just trying to let you know that you should not be expecting large gains in sensitivity you normally see when adding a horn to a dynamic driver.
I'm not missing anything.
Just providing data necessary for you to get a feel for typical ESL sensitivity compared to dynamic drivers since you had indicated high efficiency/sensitivity was one of your primary goals, and stated you did not have a feel for their sensitivity other than vague references to inefficiency.
As others already mentioned, you would have to build a high voltage tube amplifier capably of swinging 5kV - 10kV into a capacitive load to use instead of the transformers which could easily be driven with an amplifier you probably already have. If you already have a high voltage tube amplifier lurking in your closet, then sure it would be cheaper.
But buying the parts necessary to build one instead of buying transformers would most likely not be cheaper.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.