For hybrid ESL panels operating above 175 hz, a pair of inexpensive 50VA 230V/2x6V toroids on each panel sound wonderful. Unfortunately, you can't find them in the US anymore so you have to buy from across the pond. However, I did try out a tandem of 2x115V/2x6V toroids for a short while, which are available in the US, and they sounded fine too. If you go to my web page, you will find the wiring diagrams for both.
The capacitance of the panels plays into the equation but the toroids give wonderful highs coupled to my 12x48 panels. In fact the highs from the toroids are much better than I was getting from the 100:1 EI core trannys I was using previously and those trannys were wound specifically for ESL's (the toroids were less than 1/2 their cost too) . I'm rather certain that if the capacitance of your panels is not significantly higher than mine, the highs would be all there.
I should add that I can't personally vouch for all 230v/2x6v toroids-- only the Farnell/Multi-Comp trannys shown on my blog page, which I have personally used with very positive results. However, I'm not the only person that's had positive results using toroids.
Check this post:
http://www.diyaudio.com/forums/plan...-rec-step-up-transformers-esl.html#post725814
Here is a must read thread on a study about using power torrids for stepup transformers,
http://www.diyaudio.com/forums/planars-exotics/161485-step-up-transformer-design.html#post2093314
and here is a short summary of my findings,
http://www.diyaudio.com/forums/planars-exotics/181041-my-2-nd-esl-attempt-2.html#post2520508
and one more with much theory revisited,
http://www.diyaudio.com/forums/planars-exotics/186011-diy-bass-transformer-esls.html#post2520206
jer
http://www.diyaudio.com/forums/planars-exotics/161485-step-up-transformer-design.html#post2093314
and here is a short summary of my findings,
http://www.diyaudio.com/forums/planars-exotics/181041-my-2-nd-esl-attempt-2.html#post2520508
and one more with much theory revisited,
http://www.diyaudio.com/forums/planars-exotics/186011-diy-bass-transformer-esls.html#post2520206
jer
Last edited:
I'm still flirting with the idea of trying to make a set of these, but I have one question.
I have lots of subs around here and I usually use crossover points at around 80Hz. How large would I have to make one of these panels to get down to about that range? Would they be big enough to need a ribbon as well as the panel?
I have lots of subs around here and I usually use crossover points at around 80Hz. How large would I have to make one of these panels to get down to about that range? Would they be big enough to need a ribbon as well as the panel?
How big? That's an interesting question and I'm not qualified to answer it, since I've only built hybrids and you would need basically a full range ESL with big surface area and big trannys to play that low. Maybe (I hope) someone with full range ESL experience will jump in and lend an opinion on that.
I'm probably going to catch some grief from the full range guys for saying this but.... at least for those like myself with limited experience, I think full range ESL's are impractical for most builders. I'm thinking that even a panel that only has to play down to 80 hz would still need to be quite large and driven by expensive large-core trannys to get there-- and then you might need to segment the panel with a separate, narrower treble section with closer d/s spacing-- just seems to me that cost and complexity increases exponentially the lower you go and I'm wondering if it's worth it when you can get really thumping bass in a smaller package and maybe less money with a hybrid crossed over at 200-300 hz. Not to mention that a panel playing down to 80 hz would have to be plagued [and likely savagely degraded] by the diaphragm's fundamental resonance. Wheres, with a hybrid you can work around the diaphragm resonance by crossing the panel over to the woofer with a 48db slope at least one octave above resonance (two octaves above resonance if using a 24db slope).
Here's an interesting idea:
Mating a 12x48 ESL panel to a pair of 12" low-mass mid-bass woofers in a compact Ripole configuration for big sound and low bass in a small package. Any thoughts on that?
Last edited:
DrDyna> I'm just beginning my own ESL project and I've asked myself that exact question many times. Well pretty close anyway... My dream target would be something like 120-150Hz X-over but thinking and reading a lot has got me convinced that Charlie and most other experienced builders are very much right. Crossing in the 300 +/-50Hz area seems like the best compromise, at least for us n0obs. Going low will make everything much harder and more expensive. I'd be interested to hear an answer to your question as well but I fear I won't like the answer. :/
Charlie> Yes... The ripole is fascinating. It would be awsome if you could use a ripole with a panel. Between you and me, you're the one who's buildt one. What do you think about it? Could you make it work? Would it be a good match?
Charlie> Yes... The ripole is fascinating. It would be awsome if you could use a ripole with a panel. Between you and me, you're the one who's buildt one. What do you think about it? Could you make it work? Would it be a good match?
Last edited:
I have no excuse but being lazy for not already trying out that idea of mating a stat panel to a Ripole, since I already have both the stat panels and a pair of Ripoles that I currently use for subs (crossing over from my TL mid-bass woofers at 50hz). I've been reluctant to do it because it's taken so long to get my system tweaked in sounding just right and I haven't wanted to change anything and have to go thru the re-tuning process again. But I'll get bored one day and do it.
I had talked to Roger Sanders a few times about this and that and he convinced me to buy a digital crossover with 48db slope capability so's I could then lower my hybrid's crossover down to 175 hz without excitng the diaphragm's fundamental resonance. The idea is to get the woofer's output below the critical portion of the midrange where the electrostat excels. This worked great and sounded wonderful at low to moderate volume but, as it also required more EQ on the panel (panel was now playing lower with greater excursion) to compensate the dipole rolloff-- at high volume peaks the diaphragms were driven into the stators (makes a frightening noise when that happens) so I had to bump the crossover back up to 200 hz. That's when I realized that higher diaphragm tension would have let me go down lower with the crossover (on my blogpage, I then increased the recommended diaphragm tension from 1.25% to 1.5% elongation).
Good luck with your project MarkusA!
Last edited:
Awesome, you've given me something to chew on. I've got a spare pair of Peerless 831759's I'm sure would probably work well and cover the gap.
I'm thinking about making a small, cheap test panel just to make sure I've got the theory and "anatomy" behind it laid out properly.
Thanks for the info, and thanks for the Jazzman page as well, I'm sure I'll be re-re-reading it over the next few weeks!
I'm thinking about making a small, cheap test panel just to make sure I've got the theory and "anatomy" behind it laid out properly.
Thanks for the info, and thanks for the Jazzman page as well, I'm sure I'll be re-re-reading it over the next few weeks!
Even my little panels get good bass but it can only be expreienced nearfield due to dipole cancelations.
And the diagphram resonance becomes a real probelm at higher levels and is around 70hz to 90hz depending on how much tension there is, as I use heat to tension them.
It is very hard to dampin it with out taking away from the overall detail and the time envovled I spent trying just wasn't worth it because the quality of sound became mediocre.
As Charlie stated the D/S is a big factor when you start to go below 200hz.
On my little panel (3.25" wide) I am using a D/S of 200:1 ,and, on my bigger panel (7.75" wide) I had to use at least 150:1 D/S in order to prevent excursion clipping even though I have not used them since 2003 when I made them.
This is not an issue when crossing over around 300hz-400hz and above.
In order to get decent lows one must use a rather large width of at least 18" and more.
Because of the dipole cancelation the f3 is a 1/4 wave length of the lowest frequency = the width of the panel (hope I said that right).
But the wider the panel beaming becomes more of an issue.
This is where segmentation and or using many smaller width sections to create a large width panel helps.
In order to keep a spherical wave front the driver must be less than 1/4 wavelengh of the frequency it is reproducing.
As it was stated somewhere in another thread that there is a reason that woofers are so big (beside displacment issue to create good bass) and midranges are 4"dia. and tweeters are 1"dia.
I have been contemplating a threeway panel using a 1"-2" stripe for the treble and a 3"-4" stripe for the mid's on both sides and 7"-11" stripe for bass on each side aswell.
But this would be a rather large width panel and would not fit well in my current room,although I'm sure that it would perform well.
Since I like listening to them nearfield anyhow I am cosidering using the same design minus the bass panels and using 2or3 8" subs in an H-Dipole.
But that is just my preference and the way my room is setup and if I really need it rockin, I can always fire up the other two stacks of woofers I have in each corner of the room.
I completely understand where you are coming from.
The cool thing is that you can build these things just about any size you want.
But, Their characteristic sound is something that is hard to describe and the only word that I can come up with no matter what size they are is ,Breathtaking!
My little panels sound the same as my bigger panels with the exception of a little more bass extension due to the increased width.
My suggestion would be to build a smaller panel first this way you won't have a whole lot invested in materials as you work on getting the electronics configured as this is what cost the most.
And, to learn and get used to your own construction method as a smaller panel is much easier to handle and will allow you to move it around easily in order to understand the hows and whys of their charecteristics.
The main reason I choose my constrution method was #1 cost and #2 I wanted to be able to take them apart in order to change out the diagphram,frames and coatings.
But alot has changed since 2003 and all of that research is available to everyone here on DIYAudio.
The main reason I haven't done any thing with them in the past 7or 8 years was the issues of the step up transformers and that one got solved last year and I have been very happy since.
But like Charlie stated I too have been lazy and I need to get going on my system and finnish it and then move on to the next design as I have several.
Happy building,Cheers! jer
P.S. In reference to post #62 of this thread.Also I did build my little panels before I built the larger ones aswell.
And the diagphram resonance becomes a real probelm at higher levels and is around 70hz to 90hz depending on how much tension there is, as I use heat to tension them.
It is very hard to dampin it with out taking away from the overall detail and the time envovled I spent trying just wasn't worth it because the quality of sound became mediocre.
As Charlie stated the D/S is a big factor when you start to go below 200hz.
On my little panel (3.25" wide) I am using a D/S of 200:1 ,and, on my bigger panel (7.75" wide) I had to use at least 150:1 D/S in order to prevent excursion clipping even though I have not used them since 2003 when I made them.
This is not an issue when crossing over around 300hz-400hz and above.
In order to get decent lows one must use a rather large width of at least 18" and more.
Because of the dipole cancelation the f3 is a 1/4 wave length of the lowest frequency = the width of the panel (hope I said that right).
But the wider the panel beaming becomes more of an issue.
This is where segmentation and or using many smaller width sections to create a large width panel helps.
In order to keep a spherical wave front the driver must be less than 1/4 wavelengh of the frequency it is reproducing.
As it was stated somewhere in another thread that there is a reason that woofers are so big (beside displacment issue to create good bass) and midranges are 4"dia. and tweeters are 1"dia.
I have been contemplating a threeway panel using a 1"-2" stripe for the treble and a 3"-4" stripe for the mid's on both sides and 7"-11" stripe for bass on each side aswell.
But this would be a rather large width panel and would not fit well in my current room,although I'm sure that it would perform well.
Since I like listening to them nearfield anyhow I am cosidering using the same design minus the bass panels and using 2or3 8" subs in an H-Dipole.
But that is just my preference and the way my room is setup and if I really need it rockin, I can always fire up the other two stacks of woofers I have in each corner of the room.
I completely understand where you are coming from.
The cool thing is that you can build these things just about any size you want.
But, Their characteristic sound is something that is hard to describe and the only word that I can come up with no matter what size they are is ,Breathtaking!
My little panels sound the same as my bigger panels with the exception of a little more bass extension due to the increased width.
My suggestion would be to build a smaller panel first this way you won't have a whole lot invested in materials as you work on getting the electronics configured as this is what cost the most.
And, to learn and get used to your own construction method as a smaller panel is much easier to handle and will allow you to move it around easily in order to understand the hows and whys of their charecteristics.
The main reason I choose my constrution method was #1 cost and #2 I wanted to be able to take them apart in order to change out the diagphram,frames and coatings.
But alot has changed since 2003 and all of that research is available to everyone here on DIYAudio.
The main reason I haven't done any thing with them in the past 7or 8 years was the issues of the step up transformers and that one got solved last year and I have been very happy since.
But like Charlie stated I too have been lazy and I need to get going on my system and finnish it and then move on to the next design as I have several.
Happy building,Cheers! jer
P.S. In reference to post #62 of this thread.Also I did build my little panels before I built the larger ones aswell.
Last edited:
CharielM already hit on all the key points.
But, since I spent the last 2 years experimenting and building ESLs that can handle the 2 octaves that separate most hybrids from being able to cross over with subwoofers at 80Hz...I'll restate his points and add some comments.
How Big?
This depends on how loud you want it to be able to play and how high a voltage you want to deal with. If you use a large gap, you can get away with smaller areas. But, you will need to use higher bias voltage and step-up ratio transformers which are more difficult to design . If you stick with typical 2mm - 3mm D/S and want to be able to hit 105dB - 108dB then you will need about 6ft^2 - 8ft^2 depending on how well you are able to control the diaphragm resonance and how large a baffle you decide to put around the ESL.
Complications
As CharlieM already mentioned, when you move from a crossover of 300Hz down to 80Hz things get a lot more complicated. You will have to deal with 3 things.
1) Diaphragm resonance will now be near the pass band rather than being far below it where it can be easily removed with crossover filters and ignored. You will have to deal with it somehow, damping with notch filters, resistive cloth, or silicon dot dampers.
2) Transformers will need to have at least 4 times the saturation capability which means you will have to design and build your own, or be willing to pay for the expensive ones that are available.
3) Directivity control and equalization of the ESL response becomes a larger task now that the panel is bigger and beamier, and the bandwidth was extended by 2 octaves.
Is it worth it pursuing?
Definitely in my opinion. Being able to keep the same radiation pattern all the way down to 70Hz-80Hz where you are well in to the room mode region adds a coherence I have never heard with a hybrid crossing (250hz-350Hz) to a point source woofer. Sometimes you think you got it all worked out with a hybrid, and then you compare to a full range ESL like Acoustat or Soundlab and the difference is obvious.
There is one other alternative I would mention, and that is using the configuration Calvin and Martin Logan have used.
Cross the ESL panel at 250hz-350Hz to a line array of small dipole woofers to handle the range down to 70Hz-80Hz.
http://www.diyaudio.com/forums/plan...-suitable-x-over-frequency-8.html#post2284109
Should first time builders consider it?
Unless you are building from a kit with good manufacturer support, I would recommend build a hybrid first to get your feet wet. This gives you an ESL to experiment with and learn the different issues involved.
Hello CharlieM,
A couple questions on your experience with running into problems when lowering your crossover frequency. You have nearly 4 ft^2 of diaphragm area and D/S=1/16", I would think crossing at 200Hz should not be a problem.
1) You mention hearing frightening noises. Is there any chance the transformer cores were saturating? or was it obviously the sound of diaphragms slapping stators. Most likely it was the diaphragms flapping around uncontrollably at resonance, but thought I would ask.
2) What was the resonant frequency of your panels? Do you use your electronic crossover to notch out the resonance? or just rely on the crossover slope to take care of it. What was the slope? 24db/oct? or the 48dB/oct you mentioned Sanders recommending.
The reason I ask is that the diaphragm resonance generally boosts the diaphragm motion by 20dB or more. So, with 24dB/oct slopes, if your resonance isn't at least 2 octaves below resonance the crossover by itself may not be lowering the drive signal at resonance enough to avoid problems.
3) How much does your woofer excite the ESL resonance? Test by disconnecting audio signal to your ESL while playing music thru the woofer at a typical volume level and watch reflection on the diaphragm to see how much the woofer output is coupling to the ESL. It could be the combination of 2) & 3) that is causing the problem.
4) You mention increasing tension to try and solve the problem. Did you try the lower crossover setup(200Hz) with increased tension? Increasing tension raises the resonant frequency moving it closer to the crossover point so the HP filter will be cutting the drive signal less at this higher resonance. So, increasing tension may not be the right answer. It sounds counter intuitive, but it makes sense once you think about it. Now, if the majority of diaphragm motion at resonance is coming from coupling with the acoustic output of the woofer, then increasing tension to raise resonance will help.
crossing over at 200 hz was no problem.
When I lowered the crossover to 175 hz I would occasionally hear a sharp, loud "pop" on the high energy peaks-- I couldn't imagine what else it could be other than the diaphragm slapping a stator, but I'm not sure of that. I'm also not sure I would know what the trannys saturating would sound like. I'm using a tandem of 50VA trannys (100VA total per speaker).
I've never measured the diaphagm resonance-- I just assumed based on the size and spacing that it's somewhere around 60-70 hz--not sure about that either. I've got the crosover setup as 3-way with subs, mid-bass woofers and stat panels on L/M/H bands, respectively, and I'm using 48db slopes. Since the stat panel is on the crossover "H" band with it's own amp and each band has it's own EQ that acts only within it's band, I don't see how I could use the EQ on the H band to affect fequencies outside the band. Anyway, the idea was to keep the crossover point at least one octave above the diaphragm resonance and I figured 175hz with a 48db slope should be fine.
Good idea... there's a Ripole sub sitting beside each speaker too and it's moving a lot of air and I can see from the light reflecting off the diaphragms that they are definitely coupled to the bass and moving in rhythm with the bass notes. Still-- I didn't have a problem until I dropped the crossover frequency to 175hz.
Using 3-M foam tape for the spacers makes panel assembly fast and easy but once the panel is together, there's no taking it apart to adjust anything inside without rebuilding the whole panel (replacing the spacers, diaphragm & charge ring)-- so I haven't actually increased the tension-- rather, Im just going to live with limiting the crossover frequency to 200hz or above. If I ever rebuild those panels, I'm thinking I should bump up the tension a bit.
BTW... you're right about a full range design being worth pursuing, it's just that very few builders would have the know how to get there.
Aah, would help if I read more carefully. 200Hz = Ok. 175Hz = problem.
Transformer saturation can sound different depending on the capability of the amplifier driving it. Usually just sounds like grungy gross distortion during bass transients. You can also hear noise eminating from the transformer. With 48dB/oct HP crossover, I doubt that it would be a problem. Arcing usually makes a snapping noise. Popping is probably diaphragm hitting stators as you thought.
With long strips, resonance is mainly dependent on the width which is about 3.5" in your case right? So, with high tension, resonance is probably pretty close to 100Hz. If interested in determining it and don't have a signal generator handy, here is a website where you can download a simple signal generator called Tone. Click on sinusoid and then the play button at the top. There are two sliders at the bottom that let you change the frequency. Just sweep upward from 50Hz and you will have no problem locating resonance of undamped ESL panels.
TMH KTH :: Music software by Svante Granqvist
48dB/oct..Nice
Some digital crossovers have independent parametric EQ settings that can be used in conjunction with any band.Understand. Rebuilding foam tape panels always took me longer than building them in the first place. The adhesive from the tape is tough to remove completely. For some reason I thought I remembered you mentioned in another thread that you had completed a rebuild.
200Hz is a much better answer than rebuilding.
Hi,
I run my 1.2 x 0.2 meter active diaphragm area panels full range and found them to be very satisfactory. The bass is deep and tight. It sounds very different from that of a closed box woofer. However, with the kind of music I listen to, I very much prefer it this way. I use 2 mm spacer and bias voltage of around 3KV.
Yes, transformers make a lot of differences. I have mine custom made here locally. The step up ratio is around 1:90. Each transformer must weigh at least 7 kilograms.
Wachara C.
I run my 1.2 x 0.2 meter active diaphragm area panels full range and found them to be very satisfactory. The bass is deep and tight. It sounds very different from that of a closed box woofer. However, with the kind of music I listen to, I very much prefer it this way. I use 2 mm spacer and bias voltage of around 3KV.
Yes, transformers make a lot of differences. I have mine custom made here locally. The step up ratio is around 1:90. Each transformer must weigh at least 7 kilograms.
Wachara C.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.