Hello All
In an 8" (20cm) wide ESL would two 4" ( 10cm) segments give you better sound quality
even though the D/S spacing allows the full width? Would silicone dots in the middle of
the wide panel be the same as a segmented panel? I've read here that a single panel can sound harsh.Or does that have to do with having a different DS spacing for two panels?
I know panels with different DS is better in free range but I hope to avoid that problem
playing > 200hz.
Lucius
In an 8" (20cm) wide ESL would two 4" ( 10cm) segments give you better sound quality
even though the D/S spacing allows the full width? Would silicone dots in the middle of
the wide panel be the same as a segmented panel? I've read here that a single panel can sound harsh.Or does that have to do with having a different DS spacing for two panels?
I know panels with different DS is better in free range but I hope to avoid that problem
playing > 200hz.
Lucius
Hi Lucius,
segmenting isn't just a matter of segmenting the stator areas by additonal spacers or silicone dots, but dividing the entire stator width into electrical independent stripes of defined width.
This is to compensate for the 3dB/octave increase of level, which is inherent in planar dipol speakers.
To set the right size and appropriate frequency range of each segment will result in a flat frequency response and an improved high frequency dispersion.
Capaciti
segmenting isn't just a matter of segmenting the stator areas by additonal spacers or silicone dots, but dividing the entire stator width into electrical independent stripes of defined width.
This is to compensate for the 3dB/octave increase of level, which is inherent in planar dipol speakers.
To set the right size and appropriate frequency range of each segment will result in a flat frequency response and an improved high frequency dispersion.
Capaciti
Im curious what you are basing this on? Is it from seeing so many measurements of planars that you noticed a pattern, or are there some equations that reflect this phenomenon? If you have any links, or have read it any books I would like to see them.
BTW Im not doubting the validity of your statement Im just curious about it, I have some Neo8's that have a sloped response and I eventually want to build some tall narrow planars.
Segmented esl
Hi All,
My name is Marco Petrolesi and I'm an italian employer
in my free tyme i'm writing a book (in italian) about ESL.
About 3dB compensation, this is still right because more streep
play the same frequencies with different slopes.
This make possible to compensate the for phase cancellation and
improve the dispersion range of High frequencies.
If the segmentation are calculated right this is OK.
Sorry for my english...
...Marco
Hi All,
My name is Marco Petrolesi and I'm an italian employer
in my free tyme i'm writing a book (in italian) about ESL.
About 3dB compensation, this is still right because more streep
play the same frequencies with different slopes.
This make possible to compensate the for phase cancellation and
improve the dispersion range of High frequencies.
If the segmentation are calculated right this is OK.
Sorry for my english...
...Marco
Hi,
You say that it is possible to calculate segmentation.
Could you please give a short description of how it is done ? I understand that ideally , the section width must be less than wavelength it reproduces.How to calculate the frequency , from which the rise of high frequencies starts ? I assume that there must be multiple segments for the best results.
Is it possible to simulate this with any kind of software ?
Regards,
Lukas.
You say that it is possible to calculate segmentation.
Could you please give a short description of how it is done ? I understand that ideally , the section width must be less than wavelength it reproduces.How to calculate the frequency , from which the rise of high frequencies starts ? I assume that there must be multiple segments for the best results.
Is it possible to simulate this with any kind of software ?
Regards,
Lukas.
Hi,
directivity calculations depend on the size of the speaker.
Therefore the so called k-factor was introduced with k=omega/c = 2*pi*f/c.
With c=340m/s or 34.000cm/s k becomes: k=f*1.848*10^-2(1/m) or f*1.848*10^-4 (1/cm).
With a as radius of a circular speaker the cone becomes directional when ka=2
So f =2/(a*1.848*10^-x) x beeing 2 or 4 depending on units chosen for the radius a
Example: an 8"cone with a diaphragm radius of 10cm will become directional above a frequency of f=2/(10*1.848*10^-4)= 1084Hz
Since the frequency and the radius vary oppositely linear You can easily estimate that a diphragm of 5cm radius will start to direct at a frequency of ~2.1kHz and a diaphragm of 1cm radius will start to direct at a frequency of ~10.8kHz
jauu
Calvin
directivity calculations depend on the size of the speaker.
Therefore the so called k-factor was introduced with k=omega/c = 2*pi*f/c.
With c=340m/s or 34.000cm/s k becomes: k=f*1.848*10^-2(1/m) or f*1.848*10^-4 (1/cm).
With a as radius of a circular speaker the cone becomes directional when ka=2
So f =2/(a*1.848*10^-x) x beeing 2 or 4 depending on units chosen for the radius a
Example: an 8"cone with a diaphragm radius of 10cm will become directional above a frequency of f=2/(10*1.848*10^-4)= 1084Hz
Since the frequency and the radius vary oppositely linear You can easily estimate that a diphragm of 5cm radius will start to direct at a frequency of ~2.1kHz and a diaphragm of 1cm radius will start to direct at a frequency of ~10.8kHz
jauu
Calvin
hi,
it is right that say calvin.
you must consider only the width on one single cell.
to calculate the segmentation you can make a division
of esl in more strips that have the dimension wavelenght of 500hz (for start) and make more little strip at step by one octave,
until 20Khz.
the results is 7 strips but for pratice you can make only 5...
the other 2 won't sound because of their width...
bye marco!!!!!
it is right that say calvin.
you must consider only the width on one single cell.
to calculate the segmentation you can make a division
of esl in more strips that have the dimension wavelenght of 500hz (for start) and make more little strip at step by one octave,
until 20Khz.
the results is 7 strips but for pratice you can make only 5...
the other 2 won't sound because of their width...
bye marco!!!!!
Hi,
Say , the panel's width is ~= 11 cm.
The segments might be 6 , 3 , 1.5 , 0.75 cm.
The directivity for 6 cm section starts at ~1.7 kHz.
Lets say , it has a capacitance of 100pF. Therefore the needed resistor will be f=2*pi*f*c ~= 900 kOhms. Is it correct ?
How do i calculate the next resistor (for 3 cm section) ? By the way , i do connect the segments in series , right ?
Regards,
Lukas.
Say , the panel's width is ~= 11 cm.
The segments might be 6 , 3 , 1.5 , 0.75 cm.
The directivity for 6 cm section starts at ~1.7 kHz.
Lets say , it has a capacitance of 100pF. Therefore the needed resistor will be f=2*pi*f*c ~= 900 kOhms. Is it correct ?
How do i calculate the next resistor (for 3 cm section) ? By the way , i do connect the segments in series , right ?
Regards,
Lukas.
Hi bazukaz,
it is 3,4 Khz for 6cm width !
Forget about a 0,75cm segment. There will be no output from such a segment. 2 cm is the least width of a segment.
Yes 900K is correct - you need 450K for each half to put in serie to the segmented stator.
some people say its better to connect all segments is serie, some say connect them in parallel. The advantage of serie is that resistor value of the first segments adds to the value of the following. On the other hand, serie connection also puts the capacity of the segments in serie, probably causing phase shifts.
capaciti
it is 3,4 Khz for 6cm width !
Forget about a 0,75cm segment. There will be no output from such a segment. 2 cm is the least width of a segment.
Yes 900K is correct - you need 450K for each half to put in serie to the segmented stator.
some people say its better to connect all segments is serie, some say connect them in parallel. The advantage of serie is that resistor value of the first segments adds to the value of the following. On the other hand, serie connection also puts the capacity of the segments in serie, probably causing phase shifts.
capaciti
Hi,
Won't you get phase shifts anyway? I am not an electrical engineer, but putting resistors for a capacitor is a low-pass RC-crossover and any crossover will introduce phase shifts, right?
Since the ESL is a fairly ideal (air-) capacitor, the audible effects of the 'cross-over' may be less than normal crossovers with off the shelf capacitors (dynamic speakers), but it still is a crossover.
Most of the segmentation is done at several kHz, very close or within the most sensitive range of the audible spectrum. Hmmmm.
By, MartinJan
Won't you get phase shifts anyway? I am not an electrical engineer, but putting resistors for a capacitor is a low-pass RC-crossover and any crossover will introduce phase shifts, right?
Since the ESL is a fairly ideal (air-) capacitor, the audible effects of the 'cross-over' may be less than normal crossovers with off the shelf capacitors (dynamic speakers), but it still is a crossover.
Most of the segmentation is done at several kHz, very close or within the most sensitive range of the audible spectrum. Hmmmm.
By, MartinJan
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.