Yes, maybe on some ESL's but not in these with the Plitrons. They make normal cone floorstanders green from the low THD numbers at the bass.
Congrats on the successful rebuild.
Like Calvin, I am used to seeing THD rise significantly at lower frequencies approaching and below resonance where there is significant diaphragm excursion. Had you done similar distortion measurements on your dynamic dipole woofers for comparison? Do you know what the resistance is of your diaphragm coating?
Have you done any near field measurements to see the behavior of the resonances of the different sections? Perhaps this would give some insight in to the admirable distortion performance.
Attached are some sectional near field measurements I took of a Sound Lab A-1 full range ESL.
Attachments
Thanks.
Yes I have measured the woofers, one of them plays at approx 0,3% at 95dB at the upper bass and 1% at 105dB if i rememeber correctly. The one model with linear xmax of 8mm, can throw +/- ~10mm before 10% THD at Fs of higher frequencies when measured near field free air. The wooders are BMS 18S430v2, 18N850 and 18N860. The motor is pretty high tech, but the spider suspensions are little too progressive for home use imo. I have seen Klippel measurements of two of their lesser 15" models, both Le(x) and Bl(x) are top notch, but Cms(x) is average due to the high progression. I will do THD measurement for the whole system after I get to remembrane the other panel also.
I will measure the sections of the panel soon. I have listened to them near the panel, the smaller sections (13-15cm) definitely have a higher resonance to them judged by the ear. The membrane has not settled yet though, It will take at least 200hrs I think.
Last edited:
It's quite hard to get the resonances to show as my mic is omni and the room has concrete walls (room modes).
Panle widthwise in the room:
Panmel lenghtwise in the room:
Panle widthwise in the room:
An externally hosted image should be here but it was not working when we last tested it.
Panmel lenghtwise in the room:
An externally hosted image should be here but it was not working when we last tested it.
Hi,
Legis, the response looks fine. I assume that to be a farfield measurement at 4m(? or what) distance.
Obviously the 20dB difference in SPL falling from 500Hz to ~150Hz is only in part due to acoustic phase cancellation. It´d be interesting how much of the drop is left in a nearfield measurement. If there´d also a be distinct drop in SPL it could explain the surprising absence of THD-increase below 200hz.
I also tried the BMS 18N850. The build and design look exceptional well. But same as You Legis, I disliked the all too progressive suspension, which hinders the driver to reach the specified excursions values. The spider is way too small in diameter to allow for the claimed stroke capability.
Besides that, BMS couldn´t supply for these drivers for several months (because of probs with the spiders, as I was told).
So I finally turned these drivers down and used some different.
Still have one 18N850 lying around, apart from some measurements unused and in mint condition. If anybody wants to make an offer, I´ll listen
jauu
Calvin
Legis, the response looks fine. I assume that to be a farfield measurement at 4m(? or what) distance.
Obviously the 20dB difference in SPL falling from 500Hz to ~150Hz is only in part due to acoustic phase cancellation. It´d be interesting how much of the drop is left in a nearfield measurement. If there´d also a be distinct drop in SPL it could explain the surprising absence of THD-increase below 200hz.
I also tried the BMS 18N850. The build and design look exceptional well. But same as You Legis, I disliked the all too progressive suspension, which hinders the driver to reach the specified excursions values. The spider is way too small in diameter to allow for the claimed stroke capability.
Besides that, BMS couldn´t supply for these drivers for several months (because of probs with the spiders, as I was told).
So I finally turned these drivers down and used some different.
Still have one 18N850 lying around, apart from some measurements unused and in mint condition. If anybody wants to make an offer, I´ll listen
jauu
Calvin
Hi,
Legis, the response looks fine. I assume that to be a farfield measurement at 4m(? or what) distance.
Obviously the 20dB difference in SPL falling from 500Hz to ~150Hz is only in part due to acoustic phase cancellation. It´d be interesting how much of the drop is left in a nearfield measurement. If there´d also a be distinct drop in SPL it could explain the surprising absence of THD-increase below 200hz.
I also tried the BMS 18N850. The build and design look exceptional well. But same as You Legis, I disliked the all too progressive suspension, which hinders the driver to reach the specified excursions values. The spider is way too small in diameter to allow for the claimed stroke capability.
Besides that, BMS couldn´t supply for these drivers for several months (because of probs with the spiders, as I was told).
So I finally turned these drivers down and used some different.
Still have one 18N850 lying around, apart from some measurements unused and in mint condition. If anybody wants to make an offer, I´ll listen
jauu
Calvin
Yes the upper bass drop is due phase cancellation/room interaction since I cannot measure it near field.
The THD measurement was done at 2m with 1st order high pass at 150hz (Plitron would have saturated otherwise). I forgot to mention it. From this picture you can see the responses with and without XO.
Last edited:
I made some full range* THD measurement at near field (15cm away from the membrane) as high level as Plitron takes (*with only the 60µF series capacitor, xo-point at ~40Hz). The drive level is the same in both bias settings.
Low bias (~2,3kV):
High bias (~4,6kV):
THD from high bias setting:
60Hz: 0,303%
100Hz: 0,131%
200Hz: 0,178%
300Hz: 0,106%
400Hz: 0,100%
By the way, any idea why the THD rises towards the high frequencien (I mean this measurement)? Because of the voltage driven stat and the radiating area gets (from which the signal sums up to mic) smaller as the frequency increases?
Low bias (~2,3kV):
An externally hosted image should be here but it was not working when we last tested it.
High bias (~4,6kV):
An externally hosted image should be here but it was not working when we last tested it.
THD from high bias setting:
60Hz: 0,303%
100Hz: 0,131%
200Hz: 0,178%
300Hz: 0,106%
400Hz: 0,100%
By the way, any idea why the THD rises towards the high frequencien (I mean this measurement)? Because of the voltage driven stat and the radiating area gets (from which the signal sums up to mic) smaller as the frequency increases?
Last edited:
Very nice measurement set. Still puts dynamic drivers to shame
Distortion rising trend at LF is what I am used to seeing.
The distortion rise at HF you mention is often due to limits of driving & measurement electronics.
Have you tried a loop back test from output of amplifier while driving ESL to see what the best case distortion/noise levels you could expect?
How far away from the panel sections was the microphone placed when taking the sectional measurements shown in post#23?
If I remember correctly, the section heights varied by nearly a factor of 2.
I would have expected to see greater variation in the fundamental resonance between the smallest and largest sections.
Very nice measurement set. Still puts dynamic drivers to shame
Distortion rising trend at LF is what I am used to seeing.
The distortion rise at HF you mention is often due to limits of driving & measurement electronics.
Have you tried a loop back test from output of amplifier while driving ESL to see what the best case distortion/noise levels you could expect?
How far away from the panel sections was the microphone placed when taking the sectional measurements shown in post#23?
If I remember correctly, the section heights varied by nearly a factor of 2.
I would have expected to see greater variation in the fundamental resonance between the smallest and largest sections.
Yes the measuring and driving equipment could explain some. Here's a the measurement of the driving monos, 10V into 4R load made from wirewound power resistors: http://i817.photobucket.com/albums/zz95/LegisActio/ShengYa%20PSM-300%20Monos/ShengYa_10V_4R-lowbias.png I quess the THD at the HF rises quite a bit with an ESL load but the amps should be quite adequate for low impedances with their 12 pairs of BJTs. I will measure the THD with the ESLs as the load once I get new low-thd soundcard (the prevous broke
), current measuring equipment cannot measure 0,000x thd levels of these balanced Class A monos.The mic almost touched the front stator during the sweeps. Maybe the crosstalk between the sections with an omnipolar mic or the room messes my measurements. How did you measure the Soundlabs?
Last edited:
The omni mic will certainly pic up some crosstalk from adjacent sections, but it drops off pretty quickly with distance.
Measurements for the Soundlabs were with omni mic 1" away from diaphragm.
I used FFT analysis of pulsed pink noise excitation and no smoothing.
This seems to give the best visualization of exactly what the diaphragm resonance modes are.
Measuring with steady tones results in a more damped looking response, I'm not exactly sure why.
Your plotted curves appear to have smoothing applied.
Do you have non-smoothed versions available?
Here are unsmoothed REW sweeps (I made them again):
... and 1/24oct smoothed, ungated made with periodic noise with ARTA (cannot overlay unsmoothed):
I had forgotten that I had put 5dB/3,2Q parametric eq at 62Hz, it affected the last measurements but not these.
An externally hosted image should be here but it was not working when we last tested it.
... and 1/24oct smoothed, ungated made with periodic noise with ARTA (cannot overlay unsmoothed):
An externally hosted image should be here but it was not working when we last tested it.
I had forgotten that I had put 5dB/3,2Q parametric eq at 62Hz, it affected the last measurements but not these.
Last edited:
Hi,
the nearfiled response cleared things up, thanks.
As expected the base resonance is still quite high ~+15dB.
The resonance distribution can only be really effective if the dimensions and sizes of the single segments differ considerably.
But if one wants to preserve efficiency within reasonable limits, the allowable differences are way to small to show a significant effect. As such the claims made for this technique are rather marketing blabla.
In the interest of one of the most important parameters -highest efficiency- therefore same sized segments would still be best.
Higher mechanical tension would reduce the Q of the resonance, but also raisen the frequency of Fs.
Mechanical damping with a flow resistance would be much more effective, but typically omits with the clean, proper and good looks of a panel abd would probabely result in acoustic artefacts.
Another Q regarding the measurement setup.
What are the devices used and the parameters of the measurement setup?
As the THD of the measuring devices needs to be taken into account, one has to have an eye especially on the behaviour of the driving amplifier and the microphone. The Plitron toroids will probabely present the amplifier a horric load (electrical phase wise) which will lead to amplifier THD. In addition the impedance minimum will be at the upper end of the frequency range (I assume ~12-15kHz) and of below-1Ohm value. Any amplifier will show an increase in THD with such low impedance values.
Do You happen to know the THD of the mic? It must be a really good device.
I got similar results when Aachen University measured my panels and they used a freshly calibrated, several thousand of Euros costing, B&K capsule.
After B&K one should expect a THD of ~0.1-0.2% at SPLs >100dB.
Our own calibrated Mic, MB550, already costing a couple of hundreds of Euros, gives at least +0.1 to +0.2% higher THD-values than the B&K.
In any case, as Bolserst mentioned, the THD values of a well made panel put every other speaker and many amplifiers to shame. Well done Legis
jauu
Calvin
the nearfiled response cleared things up, thanks.
As expected the base resonance is still quite high ~+15dB.
The resonance distribution can only be really effective if the dimensions and sizes of the single segments differ considerably.
But if one wants to preserve efficiency within reasonable limits, the allowable differences are way to small to show a significant effect. As such the claims made for this technique are rather marketing blabla.
In the interest of one of the most important parameters -highest efficiency- therefore same sized segments would still be best.
Higher mechanical tension would reduce the Q of the resonance, but also raisen the frequency of Fs.
Mechanical damping with a flow resistance would be much more effective, but typically omits with the clean, proper and good looks of a panel abd would probabely result in acoustic artefacts.
Another Q regarding the measurement setup.
What are the devices used and the parameters of the measurement setup?
As the THD of the measuring devices needs to be taken into account, one has to have an eye especially on the behaviour of the driving amplifier and the microphone. The Plitron toroids will probabely present the amplifier a horric load (electrical phase wise) which will lead to amplifier THD. In addition the impedance minimum will be at the upper end of the frequency range (I assume ~12-15kHz) and of below-1Ohm value. Any amplifier will show an increase in THD with such low impedance values.
Do You happen to know the THD of the mic? It must be a really good device.
I got similar results when Aachen University measured my panels and they used a freshly calibrated, several thousand of Euros costing, B&K capsule.
After B&K one should expect a THD of ~0.1-0.2% at SPLs >100dB.
Our own calibrated Mic, MB550, already costing a couple of hundreds of Euros, gives at least +0.1 to +0.2% higher THD-values than the B&K.
In any case, as Bolserst mentioned, the THD values of a well made panel put every other speaker and many amplifiers to shame. Well done Legis
jauu
Calvin
Hi,
the nearfiled response cleared things up, thanks.
As expected the base resonance is still quite high ~+15dB.
The resonance distribution can only be really effective if the dimensions and sizes of the single segments differ considerably.
But if one wants to preserve efficiency within reasonable limits, the allowable differences are way to small to show a significant effect. As such the claims made for this technique are rather marketing blabla.
In the interest of one of the most important parameters -highest efficiency- therefore same sized segments would still be best.
Higher mechanical tension would reduce the Q of the resonance, but also raisen the frequency of Fs.
Mechanical damping with a flow resistance would be much more effective, but typically omits with the clean, proper and good looks of a panel abd would probabely result in acoustic artefacts.
Another Q regarding the measurement setup.
What are the devices used and the parameters of the measurement setup?
As the THD of the measuring devices needs to be taken into account, one has to have an eye especially on the behaviour of the driving amplifier and the microphone. The Plitron toroids will probabely present the amplifier a horric load (electrical phase wise) which will lead to amplifier THD. In addition the impedance minimum will be at the upper end of the frequency range (I assume ~12-15kHz) and of below-1Ohm value. Any amplifier will show an increase in THD with such low impedance values.
Do You happen to know the THD of the mic? It must be a really good device.
I got similar results when Aachen University measured my panels and they used a freshly calibrated, several thousand of Euros costing, B&K capsule.
After B&K one should expect a THD of ~0.1-0.2% at SPLs >100dB.
Our own calibrated Mic, MB550, already costing a couple of hundreds of Euros, gives at least +0.1 to +0.2% higher THD-values than the B&K.
In any case, as Bolserst mentioned, the THD values of a well made panel put every other speaker and many amplifiers to shame. Well done Legis
jauu
Calvin
Hi, the mic is budget version, uncalibrated Behringer EMC8000. The impedance minimum/resonance is somewhere at ~18kHz according to a simulation with 2700pf panel capacitance and impedance is a good way under 1 ohms with 0,5-0,6R series damping resistor. The reason you assume resonance lower might be that there is phase cancellation at the upper end depending the angle from which I measure the panels. The response goes almost straight to 20kHz from right angle like this: http://i817.photobucket.com/albums/zz95/LegisActio/lhikentt.png Curved panels usually give the smoothest result when measured quite near the edge, at 1/3 - 1/4 of the width of the panel.
Hi,
jauu
Calvin
The results are then really astounding
no, I roughly interpolated the values I know from my panel with a capacitance of 2.2nF and a U of 1:50, giving a imp-minimum at ~19kHz. Since Your panel is considerably bigger I assumed more capacitance. I nearly forgot though that Your d/s is nearly twice the d/s I use. So 18kHz might fit with Your U of 1:75.
jauu
Calvin
Has anyone tried mixing silicone dots and strips?
I've been thinking to try out an arrangement (with big curved FR panels) that I would add silicone dots only on the back stator to support the membrane from behind.
The dots I have are 3,5mm and the spacers are 3mm so the membrane would be constantly against the dots even when the membrane is moving quite a bit. Curved membranes also tend to flatten ~0,5mm towads the back stators by nature.
My goal is:
1) to be able to use highest possible biases with 3,5µm membrane without a fear that the membrane is going to get sucked into the back stator and prevent membrane slapping the stators near it's resonant frequency (which moves the membrane quite much because the Fs is ~35-45HzHz and resonance's Q is quite high due to low-ish tension that 3,5µm is capable to support in big panel).
2) Make the membrane's resonant behaviour better/smoother, hoping that silicone dots would smooth out the panel resonances quite a bit. I predict silicone dots would dampen/distribute the main resonance and also smooth the horizontal/vertical standing waves that each segment has..
I'm going to try antistatic spray + acrylic lacguer combo for the membrane coating (now I have graphite boosted Licron Crystal but it's resistance is little on the low side), as it feels good on my sample piece, and I though I could also try out new spacing technique if that would yeald some benefits.
Any thoughts?
I've been thinking to try out an arrangement (with big curved FR panels) that I would add silicone dots only on the back stator to support the membrane from behind.
An externally hosted image should be here but it was not working when we last tested it.
The dots I have are 3,5mm and the spacers are 3mm so the membrane would be constantly against the dots even when the membrane is moving quite a bit. Curved membranes also tend to flatten ~0,5mm towads the back stators by nature.
My goal is:
1) to be able to use highest possible biases with 3,5µm membrane without a fear that the membrane is going to get sucked into the back stator and prevent membrane slapping the stators near it's resonant frequency (which moves the membrane quite much because the Fs is ~35-45HzHz and resonance's Q is quite high due to low-ish tension that 3,5µm is capable to support in big panel).
2) Make the membrane's resonant behaviour better/smoother, hoping that silicone dots would smooth out the panel resonances quite a bit. I predict silicone dots would dampen/distribute the main resonance and also smooth the horizontal/vertical standing waves that each segment has..
I'm going to try antistatic spray + acrylic lacguer combo for the membrane coating (now I have graphite boosted Licron Crystal but it's resistance is little on the low side), as it feels good on my sample piece, and I though I could also try out new spacing technique if that would yeald some benefits.
Any thoughts?
Last edited:
Seeing your proposed dot layout reminded me of the long-time defunct David Lucas Co. who had plans for a curved ESL using dots (1/2" squares actually of foam tape) in a 2' x 3' panel divided into 6-7 horizontal segments. Each segment, about 5"-6" wide, had seven dots. The Three dots in the center were spaced more closely together than the others. Don't know if this worked very well since I haven't made curved panels, but the idea of altering the spacing of the center dots seems interesting. Good luck.
Hi,
35-45Hz is way too low for a curved panel. Btw. Is there really a 10Hz variation in fs?? +-1Hz should be the aim. Using also a 3.5um in largish panels I always got fs>>120Hz, up to 180Hz, applying the old ratio of 70:1-100:1 of free membrane area length over d/s. Adding dots You might increase the distance of the strips, because going lower than 70:1 compromises on efficiency.
Adding spacers on a single side, the spacers need to stick well to the membrane.
Silicone dots would ditribute resonance but don't damp. They stick quite well to the membrane and they exhibit hydrophobic qualities which helps prevent the buildup of moisture related dirt films that lead to leakage.
As waldtraut suggested, You could position the dots unevenly over the free membrane area. Not only varying in distance to each other, but also not laying in a straight line.
jauu
Calvin
35-45Hz is way too low for a curved panel. Btw. Is there really a 10Hz variation in fs?? +-1Hz should be the aim. Using also a 3.5um in largish panels I always got fs>>120Hz, up to 180Hz, applying the old ratio of 70:1-100:1 of free membrane area length over d/s. Adding dots You might increase the distance of the strips, because going lower than 70:1 compromises on efficiency.
Adding spacers on a single side, the spacers need to stick well to the membrane.
Silicone dots would ditribute resonance but don't damp. They stick quite well to the membrane and they exhibit hydrophobic qualities which helps prevent the buildup of moisture related dirt films that lead to leakage.
As waldtraut suggested, You could position the dots unevenly over the free membrane area. Not only varying in distance to each other, but also not laying in a straight line.
jauu
Calvin
Hi,
Some more comments about silicone dots related to application techniques.
I have found it is really important to gently squeeze each dot just after application so all silicone that is very close to film actually bonds to it. Otherwise buzzing noise can be expected at higher excursions and and low frequencies(<200Hz or so). The dots tend to acquire very shallow angle very close to the film if not pressed. Also they must be round and have no "tails". That means it's easier to apply them after having the film glued, usually.
So, be careful with them!
By the way, different grades of silicone behave very different. They adhere from poor to very excellent to mylar so I suggest testing before use.
Regards,
Lukas.
Some more comments about silicone dots related to application techniques.
I have found it is really important to gently squeeze each dot just after application so all silicone that is very close to film actually bonds to it. Otherwise buzzing noise can be expected at higher excursions and and low frequencies(<200Hz or so). The dots tend to acquire very shallow angle very close to the film if not pressed. Also they must be round and have no "tails". That means it's easier to apply them after having the film glued, usually.
So, be careful with them!
By the way, different grades of silicone behave very different. They adhere from poor to very excellent to mylar so I suggest testing before use.
Regards,
Lukas.
Last edited:
When you add silicon dots within a panel section, you will increase the effective stiffness of the diaphragm so the resonance Frequence, Fs, of that section will increase. To keep the same Fs you will need to make the section wider than if no dots are used. So, for the same Fs, the gain in stability with dots may not be as much as you desire.
As Calvin mentioned, trying to make curved ESLs with Fs < 100Hz is problematic. The best approach is probably to use facets similar to Soundlab. Attachment #1 shows a layout you proposed back in post#10. You had mentioned that with 3mm spacers, the larger 11-12cm wide sections had the diaphragm positioned within 1.5mm of the rear stator. This was because you were using some amount of side-to-side tension which resulted in flat facets rather than curved sections.
Attachment #2 shows that if you kept the same layout, but used 4mm spacers for the rear stator and 2mm spacers for the front stator you would position the diaphragm closer to the middle of the gap even with the side-to-side tension.
As Calvin mentioned, trying to make curved ESLs with Fs < 100Hz is problematic. The best approach is probably to use facets similar to Soundlab. Attachment #1 shows a layout you proposed back in post#10. You had mentioned that with 3mm spacers, the larger 11-12cm wide sections had the diaphragm positioned within 1.5mm of the rear stator. This was because you were using some amount of side-to-side tension which resulted in flat facets rather than curved sections.
Attachment #2 shows that if you kept the same layout, but used 4mm spacers for the rear stator and 2mm spacers for the front stator you would position the diaphragm closer to the middle of the gap even with the side-to-side tension.
Attachments
I have remembraned the first panel, it has worked ok. The difference in Fs of the sections is more pronounced than without the dots when they all seemed to resonate nearly at the same frequency for some reason. Now I get nearfield resonances between 41Hz and 100-120hz and the panel's far field response is more even than without the dots. Because of those high-ish resonances, the bass sound is a tad more drum skin like, but this effect has mitigated in burn in before.
Some time ago I wrote about membrane modulation problem. The membranes got modulated from the woofers output when playing loud because they were at the close proximity of the panels. The dots seem to mitigate this behaviour making the membranes more stable, but I cannot say any conclusive yet. But the first impressions are positive.
They are actually polyurethane dots when I cheked the material, 3M Bumpon.
I cannot finish the speakers yet because I'm changing the mylar coating. I will be trying Calvin's Tesa glue + water + hi-carbon india ink -recipe this time (india ink is this VERY conductive ink: http://www.diyaudio.com/forums/planars-exotics/162971-dropping-efficiency-coating-3.html#post2121564). For some reason the Licron crystal is not fully satisfactory for me when wiped on. It's not transparent, one layer stays optimally conductive for only 3-4 months and even then I'm not absolute sure about the HF quality compared to some very good HF-coatings like non-hardening antistatic spray (but it has shortcomings elsewhere and cannot be used). I get the ink and the glue in 1-3 weeks.
I also got new 8-8,5kV-capable HV-supplies from Crescendo Systems, powered by Twisted Pear Placid HD shunt regulated power supplies, which I will be installing after I get the transformers. It's interesting to hear how they affect the sound and also to see how high the bias can be set. 4,5kV is not a problem with my current supplies.
Steve, the flattening towards the rear stator is quite minimal, maybe 2,5mm/rear and 3,5mm/front at the worst parts.
Some time ago I wrote about membrane modulation problem. The membranes got modulated from the woofers output when playing loud because they were at the close proximity of the panels. The dots seem to mitigate this behaviour making the membranes more stable, but I cannot say any conclusive yet. But the first impressions are positive.
They are actually polyurethane dots when I cheked the material, 3M Bumpon.
An externally hosted image should be here but it was not working when we last tested it.
I cannot finish the speakers yet because I'm changing the mylar coating. I will be trying Calvin's Tesa glue + water + hi-carbon india ink -recipe this time (india ink is this VERY conductive ink: http://www.diyaudio.com/forums/planars-exotics/162971-dropping-efficiency-coating-3.html#post2121564). For some reason the Licron crystal is not fully satisfactory for me when wiped on. It's not transparent, one layer stays optimally conductive for only 3-4 months and even then I'm not absolute sure about the HF quality compared to some very good HF-coatings like non-hardening antistatic spray (but it has shortcomings elsewhere and cannot be used). I get the ink and the glue in 1-3 weeks.
I also got new 8-8,5kV-capable HV-supplies from Crescendo Systems, powered by Twisted Pear Placid HD shunt regulated power supplies, which I will be installing after I get the transformers. It's interesting to hear how they affect the sound and also to see how high the bias can be set. 4,5kV is not a problem with my current supplies.
Steve, the flattening towards the rear stator is quite minimal, maybe 2,5mm/rear and 3,5mm/front at the worst parts.
Last edited:
Wow, your diaphragm tension must be extremely light if you are able to get resonance of 41hz with dots added to the sections.
Are you saying that the lowest resonance for some segments is 100Hz and others 41Hz? Or, that you measure resonances between these two extremes across the panel. It is possible that the added dots are accentuating some of the higher diaphragm resonance modes for sections that alos have a lower fundamental resonance.
I understand that you have minimal flattening to the rear with your current horizontal sectioning configuration. I was suggesting a way to that you could return to the vertical sectioning your mentioned in post#10 for which you were having the flattening problem.
http://www.diyaudio.com/forums/planars-exotics/125246-dots-strips-spacers.html#post3081536
Using vertical sections with flat diaphragm facets allow the use of uniform tension which works better at LF where the diaphragm is moving a large % of the gap.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.