well yes, but the dispersion of a curved panels is not even that great. and what Roger says in the youtube video about dispersion is very personal ,some people would love to watch a movie or listen to music with more then one person. my opinion is that a bit of dispersion is a must. hate to say to people yeah they rock but you must sit there or it sound like pooo.
Indeed. Curved panels are somewhat effective in spreading out the top octave, but the coverage is pretty ragged and necks back down in the midrange. It doesn't provide anything look the broad, smoothly controlled directivity of a segmented ESL.
Directivity plots comparing un-segmented, curved, and segmented panels of the same width were posted in another build thread here:
http://www.diyaudio.com/forums/planars-exotics/246846-first-time-esl-builder-14.html#post4163636
I read Sanders article and makes me wonder why Martin Logan doesn't make segmented ESLs.... then again I wonder why they don't use active crossovers also... and for that matter, I really don't understand why high end speaker makers don't use active crossovers - they are vastly superior.
yep... well i guess because then there is no need to buy really expensive caps and coils. and it does not stroke with the audiophile way of thinking and products
Permanently bent. The wire I will use is 20 awg single strand copper with .010 XLPVC insulation. It's O.D. is .052" and the spacing will be 11 wires/inch, giving about 43% open area. This equates to .039 (1mm) spacing between wires. So I purchased 1mm steel (not stainless) dowel pins from McMaster Carr.
I did a quickie test with a 1mm pin mounted in a piece of hardwood (not the AL plate)-- stretching (2) 48" lengths of 20 awg copper wire. I figured the mounting hole in the hardwood might elongate but, to my surprise, the hardwood held and the pin bent (permanently).
So, I got some 1/16" pins and I'm good to go. If I had used wire with standard .016 insulation, the O.D. would be larger; hence spacing between wires would need to be larger and so too the pin diameter.
Last edited:
As mentioned recently in another thread, Martin Logan soon to be offering an active ESL hybrid with DSP crossovers and on-board ICE module amplification.
MartinLogan Renaissance ESL 15A Demo at CES 2016 - AVSForum.com
MartinLogan ?Reinvents? the Hybrid Electrostatic | Sound & Vision
MartinLogan Renaissance ESL 15A Demo at CES 2016 - AVSForum.com
MartinLogan ?Reinvents? the Hybrid Electrostatic | Sound & Vision
I think Sanders makes some valid points regarding the advantages of narrow dispersion (flat panels are indeed magical in the sweet spot). I also think he down plays the head-in-vise effect a bit too much in his Youtube video-- of course, I might do that too if I were selling unsegmented flat panels.
.. I would add ...I have seen Audiostat.wire segmented panels...you can move there wire looks like 14-5 gage...can move 1/4"....thay do use a lot of segments....I have played with only 3pr.............Audiostat...can play as loud as any esl I have ever owned...but felt my old flat Acoustat....gave a better sound..wish they play as loud...not
Martinlogan been selling speaker for over 10 year with active xover on bass drivers....but leving the ESl passive ...to be driven with tube amps...or what ever amp you may have -like ...all so active can sound very good... but I have heard some....less jest say....not so high end...sounding active xover
an passive can have it magic....to my old ears..
....but I well be going active with old Apogee bass ribbons panels.....keep up the geart work....thanks for all the info...
Martinlogan been selling speaker for over 10 year with active xover on bass drivers....but leving the ESl passive ...to be driven with tube amps...or what ever amp you may have -like ...all so active can sound very good... but I have heard some....less jest say....not so high end...sounding active xover
an passive can have it magic....to my old ears..
....but I well be going active with old Apogee bass ribbons panels.....keep up the geart work....thanks for all the info...
Sanity Check
So I want to make sure I understand everything and have it right (appreciate any input).
Here is what I have for my panels (from Bolserst's spreadsheet)...
h: 45.00 in height
w: 14.25 in width
d: 0.0625 in gap
r: 5.0000 m distance
A: 0.4137 m^2 area (panel)
h: 1.1430 m height
d: 0.0016 m gap
Vpol: 3,175 volts Vbias (optimal)
Vsig: 6,350 volts Vsignal (max)
N: 12 # sections
fL: 250.00 Hz LFbreak pt
fH: 144000.00 Hz HFbreak pt
R: 46.00 Kohms Feed Resistance
Ctot: 1153.18 pF Cpanel
C: 96.10 pF Csection
w(sec): 1.19 in width section
So as I understand it....
I will have 23 individual segments .6 inches in width starting in the very center and then going side to side (as the picture shown demonstrated).
What I really need help in understanding is, what step up ratio do I need for the transformers?
If it helps, here are the specs for the amp I am using: https://www.outlawaudio.com/products/7100.html
If I read that right, I have 28V max voltage coming out of that... so do I divde 6350V/28V to get the ratio (227:1)? That doesn't seem right.
Thanks!
So I want to make sure I understand everything and have it right (appreciate any input).
Here is what I have for my panels (from Bolserst's spreadsheet)...
h: 45.00 in height
w: 14.25 in width
d: 0.0625 in gap
r: 5.0000 m distance
A: 0.4137 m^2 area (panel)
h: 1.1430 m height
d: 0.0016 m gap
Vpol: 3,175 volts Vbias (optimal)
Vsig: 6,350 volts Vsignal (max)
N: 12 # sections
fL: 250.00 Hz LFbreak pt
fH: 144000.00 Hz HFbreak pt
R: 46.00 Kohms Feed Resistance
Ctot: 1153.18 pF Cpanel
C: 96.10 pF Csection
w(sec): 1.19 in width section
So as I understand it....
I will have 23 individual segments .6 inches in width starting in the very center and then going side to side (as the picture shown demonstrated).
What I really need help in understanding is, what step up ratio do I need for the transformers?
If it helps, here are the specs for the amp I am using: https://www.outlawaudio.com/products/7100.html
If I read that right, I have 28V max voltage coming out of that... so do I divde 6350V/28V to get the ratio (227:1)? That doesn't seem right.
Thanks!
Hi bengel
three factors to consider for step up ratio
1. Speaker sensitivity: If you look at the paper you will be able to cobble together an equation looking like
SPLo=20log(N.Vin.Vpol.sqrt(C/R)/d/h/sqrt(2.pi.r.c)/Po)
where
N is step-up ratio
Vin = 2.83 V (1W into 8 ohms)
Vpol = polarising voltage
C = segment capacitance (for basic model, asymmetric T line)
R = segment resistance(" " ")
d= stator diaphragm spacing
h = speaker panel height
r is listener distance (normally 1 m, more later)
c = speed of sound
Po = 2x 10^-5 Pa, reference sound pressure
This will help you to a comparison against conventional speakers. Moderate efficiency is about 85 dB, high efficiency 90 dB. However, this is a line source, which behave differently versus distance from a conventional point source. To get an accurate comparison, you should set r = 1/5m, the equation will then give you the efficiency of the equivalent point-source speaker giving the same SPL at 5 m.
2. Bandwidth. There are 4 factors that will limit the bandwidth (I) the ESL capacitance, the amplifier output inductance x step-up ratio squared - the inductance is not shown in any of the outlaw manuals but 10 uH would become a problem, the leakage inductance of the transformer, and the winding capacitance of the transformer. The calculations for the final bandwidth are not simple, but be assured that a step up ratio of 200 is high - near the upper limit.
3 breakdown voltages. Its nice not to have a protection circuit in the ESL, which is Ok if Vout,max x N is less than the breakdown
Keep in mind too that you can add transformers.
I noticed that Multicomp make a series of small toroids, 7VA and smaller. They claim that the 230V primary is a single winding tapped at 110 V. If so, the 7VA ones may be OK. I may see if I can get some samples and measure them to see if they would be OK.
regards
Rod
three factors to consider for step up ratio
1. Speaker sensitivity: If you look at the paper you will be able to cobble together an equation looking like
SPLo=20log(N.Vin.Vpol.sqrt(C/R)/d/h/sqrt(2.pi.r.c)/Po)
where
N is step-up ratio
Vin = 2.83 V (1W into 8 ohms)
Vpol = polarising voltage
C = segment capacitance (for basic model, asymmetric T line)
R = segment resistance(" " ")
d= stator diaphragm spacing
h = speaker panel height
r is listener distance (normally 1 m, more later)
c = speed of sound
Po = 2x 10^-5 Pa, reference sound pressure
This will help you to a comparison against conventional speakers. Moderate efficiency is about 85 dB, high efficiency 90 dB. However, this is a line source, which behave differently versus distance from a conventional point source. To get an accurate comparison, you should set r = 1/5m, the equation will then give you the efficiency of the equivalent point-source speaker giving the same SPL at 5 m.
2. Bandwidth. There are 4 factors that will limit the bandwidth (I) the ESL capacitance, the amplifier output inductance x step-up ratio squared - the inductance is not shown in any of the outlaw manuals but 10 uH would become a problem, the leakage inductance of the transformer, and the winding capacitance of the transformer. The calculations for the final bandwidth are not simple, but be assured that a step up ratio of 200 is high - near the upper limit.
3 breakdown voltages. Its nice not to have a protection circuit in the ESL, which is Ok if Vout,max x N is less than the breakdown
Keep in mind too that you can add transformers.
I noticed that Multicomp make a series of small toroids, 7VA and smaller. They claim that the 230V primary is a single winding tapped at 110 V. If so, the 7VA ones may be OK. I may see if I can get some samples and measure them to see if they would be OK.
regards
Rod
Your understanding is correct.
To answer your previous question on sensitivity, with transformer having step-up ratio of 100:1 your sensitivity for 2.83Vrms @ 1m would be 85dB. Since your ESL behaves as a line source in its operating bandwidth it will effectively gain 3dB sensitivity over its monopole woofer for each doubling of distance. So @ 2m it would match sensitivity with a monopole woofer rated at 88dB (2.83Vrms @ 1m)
I’m a bit overloaded at the moment, but still plan on updating the spreadsheet to include this calculation. (give me a couple days)
Other additions that have been requested and I plan to add:
- option for selecting desired units (in, ft, mm, cm, m etc) for all parameters.
- parameter inputs auto-generated for ESL_SEG_UI that can be pasted into the input blanks to get off-axis response for Symmetric Configuration 1 and 2.
On the [Notes] tab, the stator to stator voltage Vsig is identified as peak voltage. So you would need to divide by sqrt(2) = 1.414 to get Vrms. In your case 4490Vrms/28Vrms => (160:1). Again, this is for the optimum ratio of bias voltage to stator voltage that yields maximum SPL output. To some extent, you can compensate for lower step-up ratio by increasing bias voltage without impacting maximum SPL too much. More details here: http://www.diyaudio.com/forums/planars-exotics/264972-measured-my-esl-attempt-5.html#post4163640
It is easy to say I “need” a step-up ratio of 160:1 to get maximum output, it is altogether a different matter to design a transformer that will operate over the desired bandwidth with that step-up ratio; I see golfnut covered this in his previous posts.
On the topic of transformers, what do you consider a reasonable $ amount to spend?
The M299B transformers sold by Just-Real-Music would work great for a hybrid like yours.
ESL Electronics
Oh...If interested, the [Notes] tab of the spreadsheet also goes thru the derivation for the pieces that go into the equation golfnut posted.
Last edited:
Thanks for the info... though I have to admit alot of this is getting over my head. I kinda understand but I am not really putting it all together.
Just want to make sure of this though (i.e. I am reading the schematic correct), these toroids are no good because the primary winding is NOT center tapped.... it is double wrapped next to each other, correct? http://www.antekinc.com/content/AN-01XX.pdf
Thanks... Admittedly, I am a bit overwhelmed with the electronics knowledge here so I will take your word for it and plan for something like 88
. I was thinking of using 2 of my adire extremis drivers in a dipole config in an attempt to blend woofers into the ESL panels better (don't know if this will really work like I think though).They are 86dB each so 2 would give me 89db..... The active crossover I plan to have the output buffer with a 2X adjustable range so I should be able to get them "lined up" in SPL.
Correct. The Antek transformers are not optimal for stacking lots together for high step-up ratio because of the high winding capacitance. ( > 1000pF)
Antek Toroidal power transformer for Step-up, Measurements
However, if step-up ratio of ~ 75:1 works for you, they can be used although you may need to use 4 instead of 2 for higher power amplifiers to avoid potential corona/ozone generation and subsequent failure of the windings if driven too hard.
Antek Transformer...secondary arcing test
Last edited:
Hi bengel
To be clear (i.e. I may have misunderstood your comment), if you build each ESL as two panels 36" high, stacked vertically, the SPLs do not add. You can see this in the formula - C doubles, R halves, and h doubles - SPL stays the same. This is a line source which produces a cylindrical wavefront, adding another panel above simply extends the cylinder.
OK- so why do I need two panels you might ask ? It turns out that at low frequencies the panel behaves like a point source and the bass OP falls away more quickly than the treble. The frequency where this happens is r.c/h^2, where r is listening distance, c is speed of sound = 340 m/s, and h is the height of the panel - actually the figure that should be used is twice the height of the panel to account for the reflection from the floor.
for a 1m high panel, at 5 m the transition occurs at 5x340/4=425 Hz
If you double the height of the ESL by using two 36"panels, that point will drop to just above 100 Hz at 5 m.
...more to digest
regards
Rod
To be clear (i.e. I may have misunderstood your comment), if you build each ESL as two panels 36" high, stacked vertically, the SPLs do not add. You can see this in the formula - C doubles, R halves, and h doubles - SPL stays the same. This is a line source which produces a cylindrical wavefront, adding another panel above simply extends the cylinder.
OK- so why do I need two panels you might ask ? It turns out that at low frequencies the panel behaves like a point source and the bass OP falls away more quickly than the treble. The frequency where this happens is r.c/h^2, where r is listening distance, c is speed of sound = 340 m/s, and h is the height of the panel - actually the figure that should be used is twice the height of the panel to account for the reflection from the floor.
for a 1m high panel, at 5 m the transition occurs at 5x340/4=425 Hz
If you double the height of the ESL by using two 36"panels, that point will drop to just above 100 Hz at 5 m.
...more to digest
regards
Rod
No, I was going to do a single 45" inch high panel (2 total L/R)...
I was referring to the woofer box (which I haven't put too much thought into yet) if I use 2 dynamic drivers I would get double the output/sensitivity (+3dB).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.