What causes panels to wear out?

[Chippieboy]

What causes panels to wear out? (sorry for the long winded posting but the more info the better I always say)


Back story first, I have a pair of Sequel II’s I purchased in 2017 and replaced the panels in 2019. Then rebuilt the crossovers in 2021. I did both since I was striving for the best sound from those speakers as possible.


They sounded dull in the beginning so I did the shower bath thing, they sounded a bit better but I opted for new panels since they were still available. I was not sure how long panels would be available since they were originally built in 1989.


A few years back I came across a used pair of Request for a pretty good deal. I opted for those for what I thought would be a bit better bass in my room. I was thinking 10” vs 12” woofers.


The bass IS better. I was allowed to audition them and they still, from what I heard sounded pretty crisp and clear. The previous owner must have took care of them but I don't know how since you can’t beat father time. I now know the sound difference since my experience with the bad Sequel’s. It is my understanding that when panels “wear out” (for lack of a better term), they begin to sound dull. These Request did not sound dull and still do not.


A few weeks ago, I swapped out the Sequel’s for the Request. They were not as bright like the Sequel’s (I also heard that the Sequel’s kinda squawk at 2.?KHZ.) They also had a better sound stage (could be the room) and they were less fatiguing to listen to. I really like the ReQuest. I looked up the serial numbers this past weekend and saw that these are the original panels from 27 years ago. I my mind, they have to be worn out because of the age but they do not sound like it unless ReQuest sounded clearer than Sequel’s in the beginning. New panels are still made as of today. I am now wondering if I will get the “Return on Investment” sound wise with new panels. I really don’t want to allocate the new panel budget toward a new set of speakers. The recent purpose of changing out to the ReQuest is to purge the unused set and travel lighter. My options are either get new panels or listen to them like they are or reinstall the newly rebuilt Sequel’s.


Also, do any of the Cap’s in the crossover wear out in time like in older radios? I know old radios had a different electrolyte that contributed to the failure not like modern components but you never know.

 

[amey01]

Are you talking the panel, or the XO/caps/et al?

Purely for the panel - it is primarily pollutants / dust / other airborne debris.

The panel works by having a high electrostatic charge held over a very light panel membrane.

Pollutants such as dust/smoke/cooking oils/etc are attracted to the panel through the electrostatic charge. When they stick to it or cannot be removed, that is where problems start.

They weigh down what should be a light membrane. And they inhibit electrostatic charge building up. So you have a panel which is heavier and concurrently has less charge to move it. Output suffers. High output suffers first because high frequencies require this now-heavy and underpowered membrane to move back and forth at a faster rate.

Obviously there are many other reasons why they can have problems or fail - but I'd bet that there would be very few premature failures due to aging if all panels operated in pollutant free environments.

 

[DanR]

What causes panels to wear out? (sorry for the long winded posting but the more info the better I always say)


Back story first, I have a pair of Sequel II’s I purchased in 2017 and replaced the panels in 2019. Then rebuilt the crossovers in 2021. I did both since I was striving for the best sound from those speakers as possible.


They sounded dull in the beginning so I did the shower bath thing, they sounded a bit better but I opted for new panels since they were still available. I was not sure how long panels would be available since they were originally built in 1989.


A few years back I came across a used pair of Request for a pretty good deal. I opted for those for what I thought would be a bit better bass in my room. I was thinking 10” vs 12” woofers.


The bass IS better. I was allowed to audition them and they still, from what I heard sounded pretty crisp and clear. The previous owner must have took care of them but I don't know how since you can’t beat father time. I now know the sound difference since my experience with the bad Sequel’s. It is my understanding that when panels “wear out” (for lack of a better term), they begin to sound dull. These Request did not sound dull and still do not.


A few weeks ago, I swapped out the Sequel’s for the Request. They were not as bright like the Sequel’s (I also heard that the Sequel’s kinda squawk at 2.?KHZ.) They also had a better sound stage (could be the room) and they were less fatiguing to listen to. I really like the ReQuest. I looked up the serial numbers this past weekend and saw that these are the original panels from 27 years ago. I my mind, they have to be worn out because of the age but they do not sound like it unless ReQuest sounded clearer than Sequel’s in the beginning. New panels are still made as of today. I am now wondering if I will get the “Return on Investment” sound wise with new panels. I really don’t want to allocate the new panel budget toward a new set of speakers. The recent purpose of changing out to the ReQuest is to purge the unused set and travel lighter. My options are either get new panels or listen to them like they are or reinstall the newly rebuilt Sequel’s.


Also, do any of the Cap’s in the crossover wear out in time like in older radios? I know old radios had a different electrolyte that contributed to the failure not like modern components but you never know.

Excellent post bud! And I’ll show you long winded😂. And BTW long winded in my opinion is a post that drones on and on and says a whole lot of nothing your post has a lot of good points in it.
Take this for what it’s worth it’s my opinion, I have a pair of 25 year old Odysseys that I replaced the panels with new ones on and went to active crossover on the woofers which I also replaced with drivers of my choice and specs and am completely satisfied.
In your situation I would think the only way you’ll know which speaker you’ll prefer in your room is to get the new panels for the requests and then listen to both pairs in your room to see which works better to your liking and then sell the other set. I went with this same dilemma after I bought my Classic 9s I then found the Odysseys at a good price and re paneled them and compared the two speakers in my room and preferred the Odysseys by a big margin so the 9s are still up for sale.
If you think about the justification of re paneling the Requests which in my opinion are a great speaker, if you take the price you paid for them used add the price of the new panels add them together and lets say its 4k that’s what it was for my Odysseys. Now look at new speakers any brand and style you want to new or used for that price and I pretty much assure you that you WILL NOT find a speaker that comes anywhere close to the sound of those re paneled Requests.
As for the caps in the crossover yes they do “wear out” over time some quicker than others and from what I’ve seen in my Odysseys Logan doesn’t use real high quality caps in their crossovers so that’s something you’ll have to decide if it’s worth the cost if you keep the Requests. How’s that for long winded😂

 

[Jeff Zaret]

FYI, new panels for the Sequels are a bit over $1400. I know because I got a quote for my Sequel ii's. I use mine for rear speakers and they still sound OK. Would new panels make they better, probably, but, they have been moved to rear speakers in a surround environment.
Enjoy the journey!

 

[Jazzman53]

What causes panels to wear out?

Assuming the problem is within the panels (not up-stream);

Panels lose volume when their diaphragms can no longer hold a full charge. These conditions can cause a panel to lose charge:

1) Contaminants creating shorting paths to the stators and draining away charge, in which case shower cleaning can restore volume.

2) Corroded connection where the bias supply wire contacts the diaphragm, which is not likely but possible. A corroded connection can be side-stepped by inserting a new wire thru the panel edge to contact the periphery charge ring, and then connecting the bias supply to this wire. If this doesn't work, the problem wasn't a corroded connection.

3) Deterioration of the diaphragm's conductive coating, which typically occurs slowly, over time. The only fix for this is replacing ($$$) or rebuilding ($) the panels.

 

[Chippieboy]

1st There is no problem with the speakers. That is why I posted here. Super curious as to what makes the sound/go bad. Just with the passage of time they should be shot. But they are not.
2nd I noticed for the first time last night that there is a small LED at the bottom that illuminates when on. That tells me they shut off when no music signal is present. I'm guessing when they are not being listened to, they have no charge present to attract pollutants. Perhaps that is why they still sound great. Maybe only a year of actual listening time.
3rd I did the math and the cost of the speakers and new panels will be less than 4K. That means a go on new panels for me.
4th I was thinking of replacing the caps in the crossover. I do have the ability to use a MiniDSP I have unused on a shelf somewhere. Not sure what direction to go there. Please advise. I'm not sure how to separate the xover components from the power supply components if using an active xover. I have not looked inside yet the speakers yet.
5th I was thinking of doing a side by side test. A B comparison is too hard. I will use the Sequel on one side and a Request on the other. Use Mono mode and listen to how each panel sounds. I am guessing the Request will be superb but the Sequels are completely rebuilt. I can't wait to hear the differance.
6th Amp choice. I know this is an age old question. Currently I am running a newly refurbished Sunfire. I also have in my fleet, Bryston 3B and a Mark Levinson Model 23. I do like the ML but it is a room heater that can play music. Any thoughts?
BYW...Thanks for all the replies from the forum.

 

[amey01]

2nd I noticed for the first time last night that there is a small LED at the bottom that illuminates when on. That tells me they shut off when no music signal is present. I'm guessing when they are not being listened to, they have no charge present to attract pollutants. Perhaps that is why they still sound great. Maybe only a year of actual listening time.

That's precisely what it is; and that is precisely what it is for.

Note though - while the charge accelerates the attraction of pollutants - this process happens regardless of charge. That's why you see speakers advertised as "from dedicated listening room", "always covered when not in use", "vacuumed regularly", etc.

 

[JonFo]

4th I was thinking of replacing the caps in the crossover. I do have the ability to use a MiniDSP I have unused on a shelf somewhere. Not sure what direction to go there. Please advise. I'm not sure how to separate the xover components from the power supply components if using an active xover. I have not looked inside yet the speakers yet.

Well, if you already have a miniDSP, you are halfway there. Active XO is soooo much better than the factory passives, it's like having a brand new speaker. With some EQ, you might even correct the lack of high-end from the old panels. That's what I do with my 31-year-old Sequel rear speakers.

Driving the woofer is a piece of cake; bypass all the passive stuff and directly connect to the woofer. for the panel, same, bypass the passive XO components and wire up straight to the step-up transformer. See my SL3XC build thread for examples and process.

I also highly recommend looking at a woofer replacement, one to get a modern-quality woofer that can handle higher crossover (320Hz) and play cleaner. I assume you have a subwoofer, so the driver does not need to go below 60Hz. So no 'sub' type drivers, just a nice, clean woofer. We've discussed options for these 12", and both Goldwood and Peerless have decent options. The Dayton Audio DSA315 is also a good option.

 

[john65b]

The ML Panels have an infused coating that charge up the panels, but this coating will deteriorate over many years.

Another problem I see with ML panels is as the panels get old, they separate along edges. The Bias Ring - basically a copper tape that is connected to the red bias wire that is run along the inner edge of panel - supplying "charge" to the diaphragm coating, gets detached from diaphragm. Since detached, the charge supply gets diminished to coating on panels...

Charge Leaks from Bias Ring or Diaphragm to stators are also a problem. The supply bias charge gets "drained" to the stators.

Quads have a Neon lamp that lights up when a charge dose is distributed to the panel. On a healthy Quad panel, the Neon lamp will light once every 15 - 20 seconds. I put a Neon lamp (between panel and interface on bias wire) on an old CLS interface to see if I had a charge leak (many flashes, low output), or just weak coating (few flashes, low output).

 

[Jazzman53]

I believe that ML added the music sensing circuit to mitigate the panel fading problem by de-energizing the HV bias supply when no music signal is detected.

Whereas on earlier models the bias supply remained energized 24/7 (whenever its AC power cord was plugged into a wall outlet).

If so, then it's the charge on the diaphragm (not humidity or contaminants) that progressively deteriorates the diaphragm's conductive coating over time.

Even with the music sensor circuit extending life span, the panels are charged when playing, and it follows that panels that are played more will fade sooner than panels that are played less.

All that assumed; panel life is then determined by time-played rather than calendar age.

I'm sure ML continues to work on this and maybe their latest speakers have fully reliable coatings, but I wouldn't bet on it.

The one bullet proof coating I know of was the carbon-black coating used on the old Acoustat panels, some of which are 40 years old and still playing fine. The Acoustats were covered with a grill sock so it didn't matter aesthetically that the coating was opaque black.

Among DIY ESL builders, finding a reliable diaphragm coating is a big concern. Various coatings have been used with varying results, including rubbed-in graphite, wiped on dish washing liquid, concoctions made from dryer sheets, and various ESD floor coatings like Staticide, Stat-Clear and Licron Crystal.

I use the spray-on Licron Crystal coating, which is only 2 microns thick, and it has worked fine so far after seven years of use, but I have no idea how long it will last.

My homebuilt ESLs don't have a music sensing circuit, so I have the bias supplies plugged into switched outlets which are only energized when the amps are on.

 

[john65b]

Interesting...

The coating Quad used also are near bulletproof, but may not be as durable as Acoustat...

Acoustat's coating is very thick compared to Quad, ML, etc. I saw old pics from Acoustat showing them applying the coating with what looked like a paint roller...we talk of ML panels accumulating dust and grime that ends up weighing down the diaphragm, but that Acoustat thick coating, and on 12 micron mylar (CLS is 6 micron), it just seems like Acoustat panels should not sound as good as they do...but maybe that's why they run at 5kv bias instead of 3.5kv bias?

I tried Licron Crystal on a set of re-mylared CLS and it turned cloudy (I like that the CLS panels are clear). It was applied in winter, so humidity was not the cause....I ended up clearing the Licron Crystal off with acetone and just using liquid Dish soap (Sodium Lauryl Sulfate), and just rubbing more on after I believed the panels were fading. I should have just left the Licron Crystal on...that stuff is NOT cheap.

Anyway, my next "CLS experiment" is using electrically non-conducive Magnet strips as spars and industrial strength Velcro on edges , to allow easy re-opening to recoat (when needed) to have perpetual repairable panels. Also running the copper Bias strip completely around edges, and on (between) the vertical and horizontal magnet spars, to be sure the charge gets on the panel...

Jazzman - Let me know what you think...

 

[Robert D]

Interesting...

The coating Quad used also are near bulletproof, but may not be as durable as Acoustat...

Acoustat's coating is very thick compared to Quad, ML, etc. I saw old pics from Acoustat showing them applying the coating with what looked like a paint roller...we talk of ML panels accumulating dust and grime that ends up weighing down the diaphragm, but that Acoustat thick coating, and on 12 micron mylar (CLS is 6 micron), it just seems like Acoustat panels should not sound as good as they do...but maybe that's why they run at 5kv bias instead of 3.5kv bias?

I tried Licron Crystal on a set of re-mylared CLS and it turned cloudy (I like that the CLS panels are clear). It was applied in winter, so humidity was not the cause....I ended up clearing the Licron Crystal off with acetone and just using liquid Dish soap (Sodium Lauryl Sulfate), and just rubbing more on after I believed the panels were fading. I should have just left the Licron Crystal on...that stuff is NOT cheap.

Anyway, my next "CLS experiment" is using electrically non-conducive Magnet strips as spars and industrial strength Velcro on edges , to allow easy re-opening to recoat (when needed) to have perpetual repairable panels. Also running the copper Bias strip completely around edges, and on (between) the vertical and horizontal magnet spars, to be sure the charge gets on the panel...

Jazzman - Let me know what you think...

Great idea with the velcro. Let us know how that works out.

 

[Jazzman53]

Interesting...

The coating Quad used also are near bulletproof, but may not be as durable as Acoustat...

Acoustat's coating is very thick compared to Quad, ML, etc. I saw old pics from Acoustat showing them applying the coating with what looked like a paint roller...we talk of ML panels accumulating dust and grime that ends up weighing down the diaphragm, but that Acoustat thick coating, and on 12 micron mylar (CLS is 6 micron), it just seems like Acoustat panels should not sound as good as they do...but maybe that's why they run at 5kv bias instead of 3.5kv bias?

I tried Licron Crystal on a set of re-mylared CLS and it turned cloudy (I like that the CLS panels are clear). It was applied in winter, so humidity was not the cause....I ended up clearing the Licron Crystal off with acetone and just using liquid Dish soap (Sodium Lauryl Sulfate), and just rubbing more on after I believed the panels were fading. I should have just left the Licron Crystal on...that stuff is NOT cheap.

Anyway, my next "CLS experiment" is using electrically non-conducive Magnet strips as spars and industrial strength Velcro on edges , to allow easy re-opening to recoat (when needed) to have perpetual repairable panels. Also running the copper Bias strip completely around edges, and on (between) the vertical and horizontal magnet spars, to be sure the charge gets on the panel...

Jazzman - Let me know what you think...

Good stuff John,

Most Acoustats were full range ESLs, which need about twice the diaphragm-to-stator gap (d/s) to accommodate the greater excursion for bass, so the bias voltage had to be at least 5kV to keep the efficiency up.

The breakdown voltage of air is about 100V per mil, and a good rule of thumb for efficiency is to push the bias voltage to about half the breakdown voltage-- close to but not greater than 50 volts per mil of d/s.

I saw the same Acoustat video of the carbon black coating being applied... and it sure looks to be quite a bit thicker than what's used today.

The Licron Crystal is expensive (I'm guessing about $70 now, with shipping) and temperamental too. It tends to blush on humid days, and even on non-humid days if applied a bit heavy, and it seems that every can sprays differently, depending on the pressure in the can and the angle of tilt and mabye the phase of the moon(?). It still works though, even when it blushes.

I have good results when I orient the panel upright so that I can hold the can upright while spraying. This reduces sputtering, and I apply one thin, "just wet" coat and walk away for 8 hours.

I love the idea of the magnetic strips and velcro. That would make it so easy to access the diaprhagm, should it ever need to recoating or replacing-- too bad the manufactures make it impossible to open the panel without destroying the diaphragm and charge ring.

Charlie

 

[Brandon Hartwick]

My Aeries I speakers are from 1989 and as far as I know still have the original panels, they still have the stickers on the back from London Audio in Ontario Canada where they were purchased new. When I bought them used the panels were crazy dirty and washed them with Simple green organic cleaner and warm water in the shower, that was 2 years ago and they still work fantastic.

As far as your question about capacitors goes, only electrolytic capacitors dry out with age while film caps last basically forever. I don't believe the panel section of the crossover uses any electrolytic caps but Martin Logan has made MANY different crossover variants from what I've seen so I can't speak for what's inside your speaker.

I made a walkthrough a while back detailing what you need to do to convert your speakers to active crossovers. You should give it a read when you have a chance, it really does make the biggest difference out of anything you can possibly do to your speakers. A big reason for this is because the panel is a line source and the woofer is a point source, these both behave very different depending on your listening distance. As a test, place your ear against the panel and slowly walk backwards and you'll notice the volume doesn't really change even as you get further away from the panel, now do the same with the woofer and you'll understand how they do not operate in the same manner at distance. With active crossovers you can use steeper slopes that stop the woofer from bleeding into the panel frequencies and mudding up the sound, gain controls allow you to dial in the woofer vs panel balance to get it just the way you want for your listening position.

I need to update this but it's a great reference
https://www.martinloganowners.com/threads/how-to-active-bass-section.19237/

 

[Chippieboy]

Wow!!!!! Much information. I am a bit overwhelmed. Everyone answered my question on how panels wear out. I am wondering since my Sequel's have new panels did Martin Logan do anything different with replacement panel building to increase the lifespan.


A funny thing, a few years back (after the panel replacement) one of my Sequels woofer had torn and I replaced both with a Scanspeak Discovery 26W/4534G 10” Woofers. From what I remember, those were the closest to the factory ML woofer specs. I hated the sound. That is when I purchased the Mini DSP kit to correct the mid/low bass issues. Before I setup the Mini DSP I stumbled on the ReQuest. I connected the ReQuest but there was something about the ReQuest the just didn’t sound right. Crystal clear and enjoyable to listen to but I was not quite sure. Then along came a fellow ML Forum person and stated he had a set of the OEM Sequel woofers. He gave them to me for free. I just paid for shipping. I am forever grateful for that person. I removed the Scanspeak’s and installed the OEM woofers. The sound I loved was back. After that, I chose to rebuild the crossovers. That was a few years ago. Now I am reaching retirement and wanted to thin the Hi-Fi stuff. (less for my heirs to throw away) I put the ReQuest back in the rotation hence the first post a few days ago. I wanted to keep either the Sequel’s or the ReQuest. All logical signs point to keeping the ReQuest but sound wise, I still have not decided. I did do the comparison last night, that made it harder. I kinda liked and disliked the sound of each model. So now it is suggested that I change out the woofers on the ReQuest and eliminate the passive crossover. I do have a bit of PTSD with the previous woofer experience but I am leaning to what the Forum suggests. After all, isn’t that why we are all here anyway. I am concerned with the panel and woofer being out of phase. It would be great to find a schematic of the crossovers. I do have the ML Abyss sub in the equation. I guess I don’t need to worry if I have bad Caps in the Crossover since I am going active anyway.


So the plan, thanks to everyone here, is to change out the woofers, setup the DSP and then purchase the new panels and install them. Sounds like a great winter time project to me. I hope that does it and I love the sound. If not, somebody will enjoy a rebuilt pair of Sequel’s and ReQuest. Then it is back to the Speaker drawing board. Any ideas of alternative speakers if the plan does not work out?


Any ideas on the Amp suggestions I mentioned earlier? The Sunfire would be the easiest with the DSP since it is 5 channel. Thanks everyone and happy listening.

 

[john65b]

too bad the manufactures make it impossible to open the panel without destroying the diaphragm and charge ring.

Yup, prolly same reason Chevrolet had a 60,000 mile design life on the Vega. Planned Obsolescence - i.e. More Sales / Money?

I will start a separate thread on my rebuild....do not want to clutter up this thread...just one more question - what is target mylar tension on a CLS? I have never been able to determine best tension to use / they use. Trade secret perhaps?

As you know, replacing the mylar is a ton of work. What tension do you regularly use in your builds? I tried as little as 2 pounds, as much as 14 pounds...

 

[Jazzman53]

Yup, prolly same reason Chevrolet had a 60,000 mile design life on the Vega. Planned Obsolescence - i.e. More Sales / Money?

I will start a separate thread on my rebuild....do not want to clutter up this thread...just one more question - what is target mylar tension on a CLS? I have never been able to determine best tension to use / they use. Trade secret perhaps?

As you know, replacing the mylar is a ton of work. What tension do you regularly use in your builds? I tried as little as 2 pounds, as much as 14 pounds...

I can tell you how my MkIII speakers diaphragms are tensioned but it would not be applicable to an ML panel, because my panel configuration is so different.

I tension my speaker's 6-micron Mylar diaphragm to 1% -1.25% elongation (verified with a simple scale measurement), which results in a drum-head resonance of about 100Hz - 110Hz.

I then set the crossover frequency 1.5 octaves above the drum-head resonance, using at least a 24db/octave sloped Linkwitz-Riley digital filter. A 24db filter is easier for tuning but a 48db filter more effectively mitigates the drum-head resonance.

If I were to change the span between diaphragm supports (spars), the same tension would produce a different drum head resonance (resonance goes higher if span is reduced, or lower if it's increased). So; when it comes to tensioning the diaphragm, one size doesn't fit all.

ML doesn't share their tensioning or design specs but I suspect the ML panels are tensioned to at least 200Hz, as measured across the widest span between spars.

ML also sections the diaphragm into multiple unequal spans in order to break up what would otherwise be a single, nasty-loud drum-head resonance peak, into multiple less-loud resonance peaks spread across a wider bandwidth (i.e. "distributed resonance").

Whereas my speakers are desiged to avoid exciting the drum-head resonance by tensioning the diaphragm to resonate significantly below the crossover frequency; I surmise that ML intentionally uses the distributed drum head resonance energy to enhance the upper bass and increase efficiency without having to add a bunch of EQ boost (boost which would not be available with a single-amp/passive crossover setup anyway).

I once rebuilt an older ML Theater Center Channel speaker. Before the rebuild, I measured the drum head resonance as 200Hz. I then used a simple DIY deflection gauge to determine the tension on the diaphragm. When I replaced the diaphragm in the rebuild, I tensioned it to the same measured deflection, in order to reproduce the 200Hz resonance frequency.

Here's a link to my write-up on that ML Theater Panel rebuild:

BTW; ER Audio in Australia offers rebuild kits for some ML speaker models, which includes all materials for a rebuild and instructions for using their spring-scale method for tensioning the diaphragms. The ER method does not require building a jig. My method doesn't either, but a jig made the job much easier.

 

[john65b]

Yea, I remember reading about about CLS % elongation on mylar (tensioned length-wise only), but I do not believe it was from ML....

I guess I can make a few panels with different tensions and run some REW tests...or just build a DIY deflection meter like you did...problem is the only real unmolested ML CLSIIz panels I have are intact and not separating, and I dont wanna chance destroying them just to measure deflection (BTW - thanks Jeff!!).

Sigh....this is what made me stop playing around with these panels 10 years ago!! Spring Scale method would be easiest, just need to know value....

On the other hand, the Dupont Mylar C 6 micron should shrink 2% in Machine Direction, and 1.2% in Traverse Direction with 300F heat applied for 30 minutes...this is what I last worked with years ago...but I cannot remember what result was!

Mechanically...

Elongation = F*L/A*E

F= force (what we are looking for)
L= Length
A= Cross sectional area mylar
E= Modulus of elasticity (using Mylar 710kpsi, cannot find Mylar C value)

Tensioning 36 pounds should get 1% elongation. I still think this is too high tension.......maybe shooting an email to ER Audio is in order...

 

[Jazzman53]

The heat-shrink method isn't the best choice for curved panels because it tensions the diaphragm equally in all directions, whereas a curved panel should be tensioned predominantly lengthwise, with only enough widthwise tension to pull out any wrinkles (tension across the curve tends to pull the diaphragm into the rear stator).

I have no clue what the tension should be for a CLS panel.