MJ Dijkstra said:
Martin Jan.
Martin Logan does (with plate stators and with 1000 of panels in a production process)....
I know that it is done by diy ers also....
But lets be honest to each other..... all of us , more experienced diy builders, know also that the change of building successfully a curved wire stator, with high efficiency, thus thight tolerances, by someone totally new in diy ESL building, is very small.
I just wanted to help Skyseeker , by pointing out what is really important in the building process of a good efficient stator.
It is even so that Martin Logan curves its stators with a reason of structural stiffness, more ,than the reason of the limited wider dispersion pattern.
A wire stator can be segmented to get the same effect.
No need to try to get a curved one.
Anyway , just my opinion in this of course
Geert
Thanks guy's, you are a big help!
Geert: I already have "a few" years of experience in DIY. I am not afraid of a curved design.
The reason for wanting to build a curved design is that I expect a better balance between the highs and the mids without EQ'ing, as well as having the highs available to the whole couch (2 seater) instead of just one. I know that the stereo image will be compromised though for one of them.
Since you are right about being new to ESLs, I have decided to first build a half hight flat panel. That should get me up to speed. Next I will build a half hight curved panel. I will measure the two and compare them. If I find that curved does indeed bring the expected advantage, I will then build a curved full hight panel.
Calvin: My current room is 2.4m high. Even with a 1.8m and 0.3m blanks on top and bottom I will only just cover the whole room hight. Besides, 8 speakers of 8" (0.2m) with just a little space in between also need 1.8m.
I got a very nice quote from partsexpress for the speakers, converted into euro's its almost a laughable price
Btw: If there is someone in Holland/Germany/Belgium who also wants to order from partsexpress, I would be willing to combine shipments. You can use the PM feature. I want to order next monday, so you will need to be quick. To convert USD into EUR, be sure to add approx 22% taxes. Shipment costs will be split according to ordered value as long as you do not order something prohibitively big and/or heavy.
Geert: I already have "a few" years of experience in DIY. I am not afraid of a curved design.
The reason for wanting to build a curved design is that I expect a better balance between the highs and the mids without EQ'ing, as well as having the highs available to the whole couch (2 seater) instead of just one. I know that the stereo image will be compromised though for one of them.
Since you are right about being new to ESLs, I have decided to first build a half hight flat panel. That should get me up to speed. Next I will build a half hight curved panel. I will measure the two and compare them. If I find that curved does indeed bring the expected advantage, I will then build a curved full hight panel.
Calvin: My current room is 2.4m high. Even with a 1.8m and 0.3m blanks on top and bottom I will only just cover the whole room hight. Besides, 8 speakers of 8" (0.2m) with just a little space in between also need 1.8m.
I got a very nice quote from partsexpress for the speakers, converted into euro's its almost a laughable price
Btw: If there is someone in Holland/Germany/Belgium who also wants to order from partsexpress, I would be willing to combine shipments. You can use the PM feature. I want to order next monday, so you will need to be quick. To convert USD into EUR, be sure to add approx 22% taxes. Shipment costs will be split according to ordered value as long as you do not order something prohibitively big and/or heavy.
gvy said:
Martin Jan.
Martin Logan does (with plate stators and with 1000 of panels in a production process)....
I know that it is done by diy ers also....
But lets be honest to each other..... all of us , more experienced diy builders, know also that the change of building successfully a curved wire stator, with high efficiency, thus thight tolerances, by someone totally new in diy ESL building, is very small.
I just wanted to help Skyseeker , by pointing out what is really important in the building process of a good efficient stator.
It is even so that Martin Logan curves its stators with a reason of structural stiffness, more ,than the reason of the limited wider dispersion pattern.
A wire stator can be segmented to get the same effect.
No need to try to get a curved one.
Anyway , just my opinion in this of course
Geert
Hi Geert,
I agree with you, building a curved one is not a good start. Making a good flat one is difficult enough for most diy.
I don't agree about the statement that curving is mainly done for structural stiffness. I've build a small curved wire esl (+/- 25 degrees dispersion, so 50 degrees total, closely matching ML) and the sound is not comparable with a flat one. The flat one being shouty and spitting, the curved one smooth. Unfortunately construction of the curved one is troublesome indeed and I have to solve several problems before it is good, but I like that kind of a challange as I've build several flat ones already.
"The flat one being shouty and spitting, the curved one smooth"
Bingo. That would match my expectation on theoretical grounds. And more, I believe that the effect slightly offset from the optimum position would be even bigger.
The more I think about this, the more I get convinced that a flat ESL is actually one of the worst possible shapes for a sound source, UNLESS it is going to be used as a "headphone at a distance". In the later case flat is the way to go, and close up the the right listening position.
Be aware: this is just conjecture on my part. Not having build an ESL jet.
Btw MJ, 50 degrees is a BIG opening. Have you done that with a single diaphragm?
Bingo. That would match my expectation on theoretical grounds. And more, I believe that the effect slightly offset from the optimum position would be even bigger.
The more I think about this, the more I get convinced that a flat ESL is actually one of the worst possible shapes for a sound source, UNLESS it is going to be used as a "headphone at a distance". In the later case flat is the way to go, and close up the the right listening position.
Be aware: this is just conjecture on my part. Not having build an ESL jet.
Btw MJ, 50 degrees is a BIG opening. Have you done that with a single diaphragm?
SkySeeker said:
Hi,
It's not that big compared with ML!
It's a single sheet of Mylar® as well.
The trick is to tension only vertically (very high tension) and almost zero in horizontal direction.
ML has a horizontal spacer each 12 cm (sequel2) but you could reduce that distance and use thinner Mylar. I am planning to use 4 micron with 3 to 4 cm distance between the spacers. Spacing will be around 1,5 mm or slightly more.
Hi,
having built both, flat wire as well as curved sheet metal successfully I don´t see why it should be more difficult to build a curved stator than a flat one.
It is a lot of work to get all (!) the wires perfectly straight.
There is the risk to tear the wire apart or to get a failure point in the insulation while pulling it straight. Next possible points are the numerous glue joints (its quite a task to get here a precision of 0.1mm)
that have to be of equal thickness, etc. etc.
Building a curved stator is just ordering the stator sheets, have them insulated and build a simple curved stretching jig for the membrane.
Everything else is just prescision handcrafting.
There are small sonic differences between segmented flat wire panels and curved sheet metal panels and those ly within the dynamic performance which is slightly superior with the curved panel (tonality is equal when both types are equed precisely and in the correct way). Efficiency is highest with curved panels too. The only ´drawback´ against flat panels is their restriction to the midhighs (i.e. Hybrid-ESLs)
jauu
Calvin
having built both, flat wire as well as curved sheet metal successfully I don´t see why it should be more difficult to build a curved stator than a flat one.
It is a lot of work to get all (!) the wires perfectly straight.
There is the risk to tear the wire apart or to get a failure point in the insulation while pulling it straight. Next possible points are the numerous glue joints (its quite a task to get here a precision of 0.1mm)
that have to be of equal thickness, etc. etc.
Building a curved stator is just ordering the stator sheets, have them insulated and build a simple curved stretching jig for the membrane.
Everything else is just prescision handcrafting.
There are small sonic differences between segmented flat wire panels and curved sheet metal panels and those ly within the dynamic performance which is slightly superior with the curved panel (tonality is equal when both types are equed precisely and in the correct way). Efficiency is highest with curved panels too. The only ´drawback´ against flat panels is their restriction to the midhighs (i.e. Hybrid-ESLs)
jauu
Calvin
Hi,
I had my panels powdercoated with the powder I chose before.
They looked perfect, but I feared the thickness of the coating to be a bit too low (it was safe when testing with the nomnal expected voltages, but I liked to have reserves in this special case).
So I added several layers of a clear 2-component PU-laquer since the coating reached the thickness I wanted. PU has very favourable features for this kind of insulation.
My panels work very well with this insulation and are absolutely safe within their intended voltage range and with lots of reserves above that range. (otherwise I very probabely couldn´t wrte these words ;-))
jauu
Calvin
I had my panels powdercoated with the powder I chose before.
They looked perfect, but I feared the thickness of the coating to be a bit too low (it was safe when testing with the nomnal expected voltages, but I liked to have reserves in this special case).
So I added several layers of a clear 2-component PU-laquer since the coating reached the thickness I wanted. PU has very favourable features for this kind of insulation.
My panels work very well with this insulation and are absolutely safe within their intended voltage range and with lots of reserves above that range. (otherwise I very probabely couldn´t wrte these words ;-))
jauu
Calvin
Woofers almost finished
Now that I have assembled my woofer for the first time, I though to post a picture.
Thanks to Calvin for helping me out with a few idea's when I needed them
Btw: I first needed to build the woofers, so that I can check the stator spacing on the ESL design. Now that that is finished I will build the amplifier next (from HiFi-Akademie) then the pre-Amp then the stretcher then the ESL prototype and finally the big ESL. Still quite a long way to go.
Woofer spec: 8x Dayton RS225S-8. Total height about 2.18m, width 29.5cm, open baffle type A, with sides angled outward at 10 degrees.
Now that I have assembled my woofer for the first time, I though to post a picture.
Thanks to Calvin for helping me out with a few idea's when I needed them
Btw: I first needed to build the woofers, so that I can check the stator spacing on the ESL design. Now that that is finished I will build the amplifier next (from HiFi-Akademie) then the pre-Amp then the stretcher then the ESL prototype and finally the big ESL. Still quite a long way to go.
Woofer spec: 8x Dayton RS225S-8. Total height about 2.18m, width 29.5cm, open baffle type A, with sides angled outward at 10 degrees.
Attachments
Looks great! I have 12 of the same woofers still packed in their boxes as I work on plans for my own OB design to go with a flat, segmented, stretched wire ESL. I have a few questions about your woofer array, if you don't mind:
1) What sort of series/parallel wiring scheme did you adopt for the woofers?
2) Did you do any comparison between a simple flat panel and the final form you built, perhaps while in the prototype stage? I'm leaning toward a flat panel at the moment but am open to other ideas.
3) Do you have any measurements of the response of the 8-speaker array?
Keep us posted on your continued progress. It looks like you're at the beginning of a very nice project.
Few
1) What sort of series/parallel wiring scheme did you adopt for the woofers?
2) Did you do any comparison between a simple flat panel and the final form you built, perhaps while in the prototype stage? I'm leaning toward a flat panel at the moment but am open to other ideas.
3) Do you have any measurements of the response of the 8-speaker array?
Keep us posted on your continued progress. It looks like you're at the beginning of a very nice project.
Few
I tend to think you have to consider speaker ranges as a basic issue. If organ music is a favourite, as we are told, then a truly fine low woofing (esp. a mixed sub-bass in the corner) is essential.
Then, the two sides need good midranges but not necessarily profound bass. That makes it easier to get speakers to handle that limited range well. And you don't need as many.
I play Dayton-Wright ESLs (but I have done a fair amount of DIY with raw drivers from D-W too... and will be doing more at some point because I have a secret stash of panels). They cover with ease a wonderfully broad 5 octaves up the middle - nearly all the music. That's great. But in olden times, the two-way Janzen ESLs crossed in like at 1000 or higher and was cone below; yet you'd get almost as much ESL quality as you could want and right up to supersonic.
I know a high cross-in for the the ESLs sounds unambitious. But it makes construction a lot less demanding, eases the ESL layout compromises because it limits the bandwidth, and might make good sense here.
Then, the two sides need good midranges but not necessarily profound bass. That makes it easier to get speakers to handle that limited range well. And you don't need as many.
I play Dayton-Wright ESLs (but I have done a fair amount of DIY with raw drivers from D-W too... and will be doing more at some point because I have a secret stash of panels). They cover with ease a wonderfully broad 5 octaves up the middle - nearly all the music. That's great. But in olden times, the two-way Janzen ESLs crossed in like at 1000 or higher and was cone below; yet you'd get almost as much ESL quality as you could want and right up to supersonic.
I know a high cross-in for the the ESLs sounds unambitious. But it makes construction a lot less demanding, eases the ESL layout compromises because it limits the bandwidth, and might make good sense here.
> 1) What sort of series/parallel wiring scheme did you adopt for the woofers?
I am planning on 2 drivers in series, and the 4 groups of 2 in parallel. The simple calculation then yields 4 Ohm (since each individual speaker is rated 8 Ohm)
This is a bit of a guess I have to admit. The impedance is a complex value and I am unsure what the final impedance will be. A fall-back solution is to operate all drivers in parallel wich should be possible since I am building a class-D amplifier that should easily cope with really low impedances.
> 2) Did you do any comparison between a simple flat panel and the final form you built, perhaps while in the prototype stage?
Nope.
My reasoning was as follows: A long thin flat panel is difficult to stabelize at high power low frequencies. Thus some kind of side support is likely to be necessary. But having two parallel sides could induce resonances from the nearby ESLs.
Secondly, the resonance frequency from the drivers is a tad on the high side. Having a little impedance from the sides helps in lowering that by a few Hz.
All in all though, its a leap of faith.....
For me the bigger issue is line-source or point source and open baffle versus closed.
> 3) Do you have any measurements of the response of the 8-speaker array?
Not yet. I will make some eventually, but I still have to build the amplifier first. It will probably have to wait until I have the first ESL prototype ready, still a "couple" of weeks away
> bentoronto
I do like Organ music, a lot...
I hope (and trust) that the performance of these woofers will be sufficient. But if not, I could indeed always add a subwoofer for the sub 30 Hz frequencies.
I doubt that I will do that though. My wife is quite understanding, but a line will have to be drawn somewhere
Btw: As you can see from the thin line, I build them in two parts. An upper four and a lower four. This is because of weight issues, or rather my-back :-( I build them from 22mm MDF and the drivers weigh in at 4 Kg each. Total weight (without counterweight) is about 57 Kg. Split over the upper part of 23.5 Kg and the lower part of 33.5 Kg. Which is in fact still to heavy for my back, so I need to dismount the drivers from the frame for every transport.
Btw2: The power amplifier will be placed on the feet of the "L" and the counter weight will be fit inside the feet of the "L". The amplifier I estimate will be about 3 KG, the counter weight probably somewhere between 10 and 20 KG.
Still I worry about toppling, so I will probably attach the top to the wall with a short cord.
Last point: I have not yet fit the "grille". I need to protect the drivers from flying objects (my kids are still young, old enough not to damage it on purpose, but too young not to throw something now and then). I am still trying to figure out how to best do that. I am thinking about a mesh of fishing line between two rods, one on each side of the drivers.
I am planning on 2 drivers in series, and the 4 groups of 2 in parallel. The simple calculation then yields 4 Ohm (since each individual speaker is rated 8 Ohm)
This is a bit of a guess I have to admit. The impedance is a complex value and I am unsure what the final impedance will be. A fall-back solution is to operate all drivers in parallel wich should be possible since I am building a class-D amplifier that should easily cope with really low impedances.
> 2) Did you do any comparison between a simple flat panel and the final form you built, perhaps while in the prototype stage?
Nope.
My reasoning was as follows: A long thin flat panel is difficult to stabelize at high power low frequencies. Thus some kind of side support is likely to be necessary. But having two parallel sides could induce resonances from the nearby ESLs.
Secondly, the resonance frequency from the drivers is a tad on the high side. Having a little impedance from the sides helps in lowering that by a few Hz.
All in all though, its a leap of faith.....
For me the bigger issue is line-source or point source and open baffle versus closed.
> 3) Do you have any measurements of the response of the 8-speaker array?
Not yet. I will make some eventually, but I still have to build the amplifier first. It will probably have to wait until I have the first ESL prototype ready, still a "couple" of weeks away
> bentoronto
I do like Organ music, a lot...
I hope (and trust) that the performance of these woofers will be sufficient. But if not, I could indeed always add a subwoofer for the sub 30 Hz frequencies.
I doubt that I will do that though. My wife is quite understanding, but a line will have to be drawn somewhere
Btw: As you can see from the thin line, I build them in two parts. An upper four and a lower four. This is because of weight issues, or rather my-back :-( I build them from 22mm MDF and the drivers weigh in at 4 Kg each. Total weight (without counterweight) is about 57 Kg. Split over the upper part of 23.5 Kg and the lower part of 33.5 Kg. Which is in fact still to heavy for my back, so I need to dismount the drivers from the frame for every transport.
Btw2: The power amplifier will be placed on the feet of the "L" and the counter weight will be fit inside the feet of the "L". The amplifier I estimate will be about 3 KG, the counter weight probably somewhere between 10 and 20 KG.
Still I worry about toppling, so I will probably attach the top to the wall with a short cord.
Last point: I have not yet fit the "grille". I need to protect the drivers from flying objects (my kids are still young, old enough not to damage it on purpose, but too young not to throw something now and then). I am still trying to figure out how to best do that. I am thinking about a mesh of fishing line between two rods, one on each side of the drivers.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.