Seperate panels for bass?
Is there any point in biamping/triamping diy electrostatic speakers?
I am going to build 8'-3' electrostatic loudspeakers.
My plan was to break up the speaker into three panels. A small high freq with a small spacing, a bigger mid with a bit bigger spacing, and the rest bass with a large spacing. I was planning on triamping this.
Pointless? What do you think i should do?
Question #2 What happens to your system if you just keep adding things. Eg. Whole bunch of amps, heil tweeters, electrostatic mid, woofers, etc.
What would happen to the sound?
Is there any point in biamping/triamping diy electrostatic speakers?
I am going to build 8'-3' electrostatic loudspeakers.
My plan was to break up the speaker into three panels. A small high freq with a small spacing, a bigger mid with a bit bigger spacing, and the rest bass with a large spacing. I was planning on triamping this.
Pointless? What do you think i should do?
Question #2 What happens to your system if you just keep adding things. Eg. Whole bunch of amps, heil tweeters, electrostatic mid, woofers, etc.
What would happen to the sound?
I suspect if you needed to build a very fast bass to complement an existing set of ESL speakers, why not.
The purpose of splitting up the panels would be because of placement issues else a FR design (single panel) would be easier.
I have also toyed with the idea of just building an ESL for bass only as dynamic speakers usually are too slow to mate up with electrostats.
The purpose of splitting up the panels would be because of placement issues else a FR design (single panel) would be easier.
I have also toyed with the idea of just building an ESL for bass only as dynamic speakers usually are too slow to mate up with electrostats.
Hi,
the impression of slowness has nothing to do with velocity!
The same sonic could happen with a ESL too. Since it´s quite difficult to control the Q-factor properly there is nothing pro for an ESl-Sub.
Dynamic subs are smaller, more efficient and when You use the right working principle outperform the ESl-bass easily!
jauu
Calvin
the impression of slowness has nothing to do with velocity!
The same sonic could happen with a ESL too. Since it´s quite difficult to control the Q-factor properly there is nothing pro for an ESl-Sub.
Dynamic subs are smaller, more efficient and when You use the right working principle outperform the ESl-bass easily!
jauu
Calvin
I tend to disagree with the logic that dynamic speakers can keep up with an ESL bass.
One of the key elements of using an ESL for bass is the huge membrane size that can be constructed to move the air. The actual travel the ESL has to make to produce sound is very small in comparison with a dynamic speaker. Sure you can build a bass array of dynamic speakers but that also has it own problems and can be quite costly in comparison to an ESL.
The choice for using dynamic speaker in the bass is to save on space and make placement easier. The quality of the bass produced is debatable as in most cases you are compromising on so many levels.
Listen to a large FR ESL and then you’ll understand how seamless the bass blends into the midrange. Try mating it up with a dynamic bass and you will soon realize that the two perform very differently.
At the end of the day your hearing will determine what sounds best and I suspect an ESL bass will come thru as clean and uncoloured as the original recording.
One of the key elements of using an ESL for bass is the huge membrane size that can be constructed to move the air. The actual travel the ESL has to make to produce sound is very small in comparison with a dynamic speaker. Sure you can build a bass array of dynamic speakers but that also has it own problems and can be quite costly in comparison to an ESL.
The choice for using dynamic speaker in the bass is to save on space and make placement easier. The quality of the bass produced is debatable as in most cases you are compromising on so many levels.
Listen to a large FR ESL and then you’ll understand how seamless the bass blends into the midrange. Try mating it up with a dynamic bass and you will soon realize that the two perform very differently.
At the end of the day your hearing will determine what sounds best and I suspect an ESL bass will come thru as clean and uncoloured as the original recording.
Full Range ESLs
A couple of problems to overcome will be resonances in the bass region and the large capacitance of the big array. On the high end capacitance will put a big load on the amp and possibly cause frequency response issues with the driving transformer. A two way setup with an 8" center panel for midrange/treble flanked by two 14" panels for bass might be good.
Good luck with your project.
A couple of problems to overcome will be resonances in the bass region and the large capacitance of the big array. On the high end capacitance will put a big load on the amp and possibly cause frequency response issues with the driving transformer. A two way setup with an 8" center panel for midrange/treble flanked by two 14" panels for bass might be good.
Good luck with your project.
Hi,
i would not recommend to realise a real 2- or 3-way ESL-system.
1. I agree with calvin that it will sound like a 2 or 3-way system. You need to implement low pass crossovers for the mid and high Level panels - in result your step response will look like those filters !
2. I think it will be a never ending story, to develop 3 different stepup transformers, perfectly designed for each panel of the system.
3. Even if you damp the fundamental resonance of the mid-high by filters, your bass panel will cause interaction.
better to make a 1 - 1/2- system. Make two ESL panels with identical size and resonance. One of it will be the fullrange panel - in best case with electrical segementation to achieve a flat freqeuncy response. The second panel is just working in the lowrange. It will be connected to the first panel stators, by adding high value resistors in serie with the stators.
If your argument is efficieny, this is even the better way to go. Imo most fullrange ESL do not suffer for high frequency efficiency but in the low range. If you design specific mid an highrange panels, their membrane area won't add sound pressure in the lowrange.
Capaciti
i would not recommend to realise a real 2- or 3-way ESL-system.
1. I agree with calvin that it will sound like a 2 or 3-way system. You need to implement low pass crossovers for the mid and high Level panels - in result your step response will look like those filters !
2. I think it will be a never ending story, to develop 3 different stepup transformers, perfectly designed for each panel of the system.
3. Even if you damp the fundamental resonance of the mid-high by filters, your bass panel will cause interaction.
better to make a 1 - 1/2- system. Make two ESL panels with identical size and resonance. One of it will be the fullrange panel - in best case with electrical segementation to achieve a flat freqeuncy response. The second panel is just working in the lowrange. It will be connected to the first panel stators, by adding high value resistors in serie with the stators.
If your argument is efficieny, this is even the better way to go. Imo most fullrange ESL do not suffer for high frequency efficiency but in the low range. If you design specific mid an highrange panels, their membrane area won't add sound pressure in the lowrange.
Capaciti
I am torn between using a horn and an ESL for my midrange though I believe that I am leaning towards horn because of SPL issues. still, i think a three way with a large maybe pro dynamic driver, sealed and LT'ed to 50hz up to 250hz where the esl panel can beam its perfect little square waves up until diaphragm mass roll off. then at ten khz or so add a ribbon super tweeter. I might even make the bass and midrange dipole at that point.
The point is I really like the idea of using esl panels and for that manner FR speakers as WR sound producers. A high crossover accommodates the inability of a sizeable midrange source to truely shine up high, and because the sound energy increases rapidly as you go lower especially in voices, a crossover much lower than 200 is trouble.
My spl solution as been to use a front horn loaded ten inch down to seventy hz, a large format compression driver for midrange and the tweeter I haven't decided upon yet.
I really believe that the keys to a nice system are low energy storage and dynamic capability. Distortion is actually mostly not noticeable because it is a comparative effect, whereas energy storage can be downright painful. An esl is distorting hugely at low and high frequencies and at high volumes but its energy storage levels across the band are very very low. that is why ESl loudspeakers sound darn good.
Phasing is bunk. I do not believe in phasing. the human ear cannot hear quickly enough to tell the difference in time arrival or absolute phase between sources. phasing leads to frequency aberrations and those aberrations are always negative. this leads to holes in the response which are hugely less noticeable than peaks. I pay no attention to phasing because whatever "synergy" i may have lost by having my midrange source six feet in front of my midbass source and my subs two feet behind my head on the couch, is vastly overshadowed by the fact that I can have niel pert warming up in my living room, which to me is way more desirable.
The point is I really like the idea of using esl panels and for that manner FR speakers as WR sound producers. A high crossover accommodates the inability of a sizeable midrange source to truely shine up high, and because the sound energy increases rapidly as you go lower especially in voices, a crossover much lower than 200 is trouble.
My spl solution as been to use a front horn loaded ten inch down to seventy hz, a large format compression driver for midrange and the tweeter I haven't decided upon yet.
I really believe that the keys to a nice system are low energy storage and dynamic capability. Distortion is actually mostly not noticeable because it is a comparative effect, whereas energy storage can be downright painful. An esl is distorting hugely at low and high frequencies and at high volumes but its energy storage levels across the band are very very low. that is why ESl loudspeakers sound darn good.
Phasing is bunk. I do not believe in phasing. the human ear cannot hear quickly enough to tell the difference in time arrival or absolute phase between sources. phasing leads to frequency aberrations and those aberrations are always negative. this leads to holes in the response which are hugely less noticeable than peaks. I pay no attention to phasing because whatever "synergy" i may have lost by having my midrange source six feet in front of my midbass source and my subs two feet behind my head on the couch, is vastly overshadowed by the fact that I can have niel pert warming up in my living room, which to me is way more desirable.
so, sound energy falls off quite rapidly above 500hz?
That makes sense. so, i was thinking that 500hz or so is an OK crossover point because it is high enough that the energy levels are falling but it leaves the majority of harmonics etc coming from a single source. those three eights are in a FLH, is it a straight horn, how were they wired? was your esl diy? if so what were the stator spacings and bias voltage.
Was the midbass horn fairly long, and was it time aligned with the esl? I like the idea of using an esl because they beam and so you can mount it far and above behind a largeish midbass horn without terrible diffraction effects from the lip of the horn. Also the beaming keeps the efficiency up.
etc... surely all things that have crossed your minds.
btw, could beaming be adressed with vertical slats on the top and bottom with the middle section left alone. that might create an MTM like effect with little hf coming forwards from the top and bottom but being redirected to the sides.
That makes sense. so, i was thinking that 500hz or so is an OK crossover point because it is high enough that the energy levels are falling but it leaves the majority of harmonics etc coming from a single source. those three eights are in a FLH, is it a straight horn, how were they wired? was your esl diy? if so what were the stator spacings and bias voltage.
Was the midbass horn fairly long, and was it time aligned with the esl? I like the idea of using an esl because they beam and so you can mount it far and above behind a largeish midbass horn without terrible diffraction effects from the lip of the horn. Also the beaming keeps the efficiency up.
etc... surely all things that have crossed your minds.
btw, could beaming be adressed with vertical slats on the top and bottom with the middle section left alone. that might create an MTM like effect with little hf coming forwards from the top and bottom but being redirected to the sides.
I think you missed the whole point of ESLs. They reproduce a wide frequency range at very low distortion. They also produce very sharp imaging. But they won't do any of those things very loudly.
When you start mucking about with horns, crossovers, and to a lesser extent, bass drivers, you'll destroy the imaging completely.
Why go to the trouble of using an ESL for a tweeter? There are plenty of competent tweeters available and they are usually much lower cost and a lot less trouble to use and much lower maintenance than an ESL. You'll end up with a loud, mediocre speaker that has an ESL tweeter just like some of the old commercial speakers that used ESL tweeters as a marketing gimmick.
If you want high volume, don't use an ESL. That isn't what they do well. Now if you were to devote some energy to figuring out a way to make an ESL produce high volume along with all its other wonderful qualities, THAT would be special...
I_F
When you start mucking about with horns, crossovers, and to a lesser extent, bass drivers, you'll destroy the imaging completely.
Why go to the trouble of using an ESL for a tweeter? There are plenty of competent tweeters available and they are usually much lower cost and a lot less trouble to use and much lower maintenance than an ESL. You'll end up with a loud, mediocre speaker that has an ESL tweeter just like some of the old commercial speakers that used ESL tweeters as a marketing gimmick.
If you want high volume, don't use an ESL. That isn't what they do well. Now if you were to devote some energy to figuring out a way to make an ESL produce high volume along with all its other wonderful qualities, THAT would be special...
I_F
mid bass horn
My mid bass horn was a straight exponential horn about 24 inches square at the mouth and around 36 inches long. The three 8s were mounted vertical on a board and I think about a 2 inch slot for the throat. I wired the drivers in parallel, so about 3 ohms.
No crossover is really necessary since the horn has built in mechanical crossover. I remember the top end rolled off fairly fast all by itself. I mounted the diy ESL panels on top of the horn and could not tell any differnce with moving the position of the panels.
My panels were made of the cain metal about 60 thousands thick with 3KV bias. My system could blow away anything commercially made in the 70's.
My mid bass horn was a straight exponential horn about 24 inches square at the mouth and around 36 inches long. The three 8s were mounted vertical on a board and I think about a 2 inch slot for the throat. I wired the drivers in parallel, so about 3 ohms.
No crossover is really necessary since the horn has built in mechanical crossover. I remember the top end rolled off fairly fast all by itself. I mounted the diy ESL panels on top of the horn and could not tell any differnce with moving the position of the panels.
My panels were made of the cain metal about 60 thousands thick with 3KV bias. My system could blow away anything commercially made in the 70's.
I was going to use tensioned wire stator so i could decrease spacing and bias in the tweeter section.
I would think a 8"x84" panel with 2mm spacing could be very loud to 300hz. If the mylar was close to touching the stator at 300hz the displaced air would be equal to 7.6mm excursion of a 15" dynamic. At 300hz 8mm of excursion on a 15" woofer would be loud, very loud.
I would think a 8"x84" panel with 2mm spacing could be very loud to 300hz. If the mylar was close to touching the stator at 300hz the displaced air would be equal to 7.6mm excursion of a 15" dynamic. At 300hz 8mm of excursion on a 15" woofer would be loud, very loud.
Hi,
2mm is far too much for coupling from 300Hz on with such a big panel! 1mm is enough and with a safety margin of +50% because of acoustic coupling You end up at 1.5mm maximum! It is a very common mistake and it seems each and every beginner falls into this trap, that the spacing is chosen too wide! You don´t improve dynamics with a greater d/s, but it´s just the other way round!!...You loose! You need much more drive voltage, drive power, higher transformation factor, etc, etc. All this degrades the capabilities of the panel seriously technically and audibly! To build a panel thats capable to quite hornlike dynamics its of prime importance to understand, that You need a efficient panel as possible. What the panel can´t supply for, You won´t achieve with more drive power! Best example are imo the Quads. They are relatively easy to drive, thats true. But the panels themselves are terribly inefficient and are driven with trannies featuring factors well above 1:200(!!). The result is that they perform poorly with regard to dynamics.
If You want it seriously loud, rather use more panel width (10"-15") and less height (50" -70"). Than You have more reserves at the lower end. Set the resonance as high as possible (>150Hz) by using strong mechanical tension and approriately placed diaphragm supports (with a flat panel of the suggested size and d/s You can e.g. place horizontal spacers every 100-120mm. Such a panel would present a capacity of ~2nF (80kOhms@1kHz, 4kOhms@20kHz). So a transformer with something between 1:50 and 1:75 could be ok for fullsize driven panels. With segmented panels You could go up a bit in transformation factor (~1:100).
1.5mm will allow for a max. signal-voltage of ~3kVp-p (1kVrms). So the amp should be able to produce at min. clean and stable 20Vrms (100W@4Ohms).
As transformers You could than use a pair of simple power toroids 6V/230V/ >50VA, giving ~1:65-68). For a segmented wire panel You could use a quad of 4.5/230V (giving ~1:90-1:95). This way SPLs of >110dB@4m will be possible above 300Hz with very low distortion. This should be enough to satify even horn afficionados.
jauu
Calvin
2mm is far too much for coupling from 300Hz on with such a big panel! 1mm is enough and with a safety margin of +50% because of acoustic coupling You end up at 1.5mm maximum! It is a very common mistake and it seems each and every beginner falls into this trap, that the spacing is chosen too wide! You don´t improve dynamics with a greater d/s, but it´s just the other way round!!...You loose! You need much more drive voltage, drive power, higher transformation factor, etc, etc. All this degrades the capabilities of the panel seriously technically and audibly! To build a panel thats capable to quite hornlike dynamics its of prime importance to understand, that You need a efficient panel as possible. What the panel can´t supply for, You won´t achieve with more drive power! Best example are imo the Quads. They are relatively easy to drive, thats true. But the panels themselves are terribly inefficient and are driven with trannies featuring factors well above 1:200(!!). The result is that they perform poorly with regard to dynamics.
If You want it seriously loud, rather use more panel width (10"-15") and less height (50" -70"). Than You have more reserves at the lower end. Set the resonance as high as possible (>150Hz) by using strong mechanical tension and approriately placed diaphragm supports (with a flat panel of the suggested size and d/s You can e.g. place horizontal spacers every 100-120mm. Such a panel would present a capacity of ~2nF (80kOhms@1kHz, 4kOhms@20kHz). So a transformer with something between 1:50 and 1:75 could be ok for fullsize driven panels. With segmented panels You could go up a bit in transformation factor (~1:100).
1.5mm will allow for a max. signal-voltage of ~3kVp-p (1kVrms). So the amp should be able to produce at min. clean and stable 20Vrms (100W@4Ohms).
As transformers You could than use a pair of simple power toroids 6V/230V/ >50VA, giving ~1:65-68). For a segmented wire panel You could use a quad of 4.5/230V (giving ~1:90-1:95). This way SPLs of >110dB@4m will be possible above 300Hz with very low distortion. This should be enough to satify even horn afficionados.
jauu
Calvin
My first panel arrangement was 20 inches wide and 50 inches tall. Ran them from around 450hz up. With .60 panel thickness they played good and loud. Later with the mid bass horn I rearranged the panels to about 36" square amd set them on top of the horn. That worked very well too. Calvin is correct that more panel width will improve the low end response. Been a while since I did the math on all this stuff.
I was experimenting in the 70's following the Hermeyer/Sanders articles in the Audio Amateur. Lots of fun.
I was experimenting in the 70's following the Hermeyer/Sanders articles in the Audio Amateur. Lots of fun.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.