Thanks, I guess that means it might at least be worth a try.
I doubt I'll be throwing any rock concerts in my living room any time soon.
So... 1.1mm d/s test run.
Panel size: 1650x250mm
Active diaphragm size: 1600x200mm
Extra vertical spacer from top to bottom monted at the center line (100mm).
What width should I choose for the spacer?
Tapes are available in 25mm and 19mm Width. (I got 19mm already and I'll be buying 25mm). Splitting them in half I can get 12.5mm and 9.5mm widths. (Splitting them might include some irregularities?)
I doubt I'll be throwing any rock concerts in my living room any time soon.
So... 1.1mm d/s test run.
Panel size: 1650x250mm
Active diaphragm size: 1600x200mm
Extra vertical spacer from top to bottom monted at the center line (100mm).
What width should I choose for the spacer?
Tapes are available in 25mm and 19mm Width. (I got 19mm already and I'll be buying 25mm). Splitting them in half I can get 12.5mm and 9.5mm widths. (Splitting them might include some irregularities?)
Since you reference a vertical center spacer, I assume you will be building a flat panel.
My slightly wider flat panel (30.5cm) uses 19mm on the edges and two 9.5mm vertical support spacers. Since your panel will have only one vertical support spacer, I would use 19mm on the edges and 12.5mm for the vertical center spacer.
Last edited:
You're absolutely correct Charlie. 
It'll be a flat panel. I might be crazy but there's got to be a limit to my madness. It's my first panel and flat seams to be the easiest route.
I think I'll try heat shinking as well to get the diaphragm tension up, unless there a good reason not to?
If I go for 19mm at the edges I'll drop the 25mm tape all togeather and in that case it'll either be 19mm or 9.5mm center spacer for obvious reasons.
Maybe 19mm edges is the way to go?
It'll be a flat panel. I might be crazy but there's got to be a limit to my madness. It's my first panel and flat seams to be the easiest route.
I think I'll try heat shinking as well to get the diaphragm tension up, unless there a good reason not to?
If I go for 19mm at the edges I'll drop the 25mm tape all togeather and in that case it'll either be 19mm or 9.5mm center spacer for obvious reasons.
Maybe 19mm edges is the way to go?
Hi,
1.1mm is good for a maximum SPL-value of 110dB@4m for f>200Hz with my small panel.
Sounds sufficient eyyh?
jauu
Calvin
Have you measured that loudness from one of your panels (at what distance and room??) or is that the theoretical excursion that is needed for 110dB at 200 Hz?
You're absolutely correct Charlie.
I think I'll try heat shinking as well to get the diaphragm tension up, unless there a good reason not to?
If I go for 19mm at the edges I'll drop the 25mm tape all togeather and in that case it'll either be 19mm or 9.5mm center spacer for obvious reasons.
Maybe 19mm edges is the way to go?
Either tensioning method would work. I've not tried heat-shrinking myself but many builders prefer it. You can get much higher tension mechanically and higher tension is preferable when using very high bias voltage (over 2.5kv, let's say) with close d/s. On the other hand, heat shrinking is easy and fast and the lower tension gives a lower resonant frequency and that's a good thing too (we want the resonance at least low enough to fall below the crossover frequency). Again, either method works and each has its tradeoffs-- so it's a matter of personal preference and matching the results to the application.
19mm tape is sufficient to secure the diaphragm to the stator and leaves more active radiating area on the panel. I think 9.5mm would be fine for the center support spacer if your stators are pretty flat. Otherwise, I would go with the wider 12.5 mm to better stick and hold the stators parallel.
Hi,
Markus, you may rethink the vertical dimension of Your panel.
The increased height increases the capacitance of the panel without beeing of good use, because SPL-wise you don´t gain when the height exceeds 1m-1.3m. The higher the panel the farther away comes the transistion from nearfield to diffuse field conditions (which gives the point of highest SPL).
This also means that as long as You listen in nearield You hardly will get the impression of a realistic soundstaging, but rather the impresssion of listening to a oversized headphone. If You want to listen in a seat as well as in standing position just raise a smaller panel in mounting height.
A flat panel forms laser-beamlike highs. Already small variations in head position lead to varying sonic character. Curved panels are a bit more forgiving. Think about additional bracing of the panel to increase stiffness.
Flat panels will easily rattle and resonate.
You won´t gain high dynamics if You tension the membrane just thermally. Tension mechanically and add a heat treatment after.
The higher the tension the more fixating area you need. 19mm could be sufficient for thermal treatment , but I doubt that it would withstand mechanical tensioning for long. I´d opt for 30mm for high tension.
The segmenting spacers may be silicon dots with a flat panel. This leaves the most active area left. With curved panels -or generally if using strip shaped spacers- use the smallest width (6mm for 3M). Some Tapes may be a pain to be cut by hand. Rather use 2 or 3 different widths instead.
The cited 110dB@4m is a real value measured by Prof. A. Goertz in the anechoic room of Aachen University using MonkeyForest and freshly calibrated Bruel&Kjär Mic-capsules. The measurement was repeated to confirm the values. Signal voltage calculated to an equivalent power of 50W into 8Ohms.
jauu
Calvin
Markus, you may rethink the vertical dimension of Your panel.
The increased height increases the capacitance of the panel without beeing of good use, because SPL-wise you don´t gain when the height exceeds 1m-1.3m. The higher the panel the farther away comes the transistion from nearfield to diffuse field conditions (which gives the point of highest SPL).
This also means that as long as You listen in nearield You hardly will get the impression of a realistic soundstaging, but rather the impresssion of listening to a oversized headphone. If You want to listen in a seat as well as in standing position just raise a smaller panel in mounting height.
A flat panel forms laser-beamlike highs. Already small variations in head position lead to varying sonic character. Curved panels are a bit more forgiving. Think about additional bracing of the panel to increase stiffness.
Flat panels will easily rattle and resonate.
You won´t gain high dynamics if You tension the membrane just thermally. Tension mechanically and add a heat treatment after.
The higher the tension the more fixating area you need. 19mm could be sufficient for thermal treatment , but I doubt that it would withstand mechanical tensioning for long. I´d opt for 30mm for high tension.
The segmenting spacers may be silicon dots with a flat panel. This leaves the most active area left. With curved panels -or generally if using strip shaped spacers- use the smallest width (6mm for 3M). Some Tapes may be a pain to be cut by hand. Rather use 2 or 3 different widths instead.
The cited 110dB@4m is a real value measured by Prof. A. Goertz in the anechoic room of Aachen University using MonkeyForest and freshly calibrated Bruel&Kjär Mic-capsules. The measurement was repeated to confirm the values. Signal voltage calculated to an equivalent power of 50W into 8Ohms.
jauu
Calvin
I would second Calvin's advice on using mechanical tensioning. Heat tensioning can only provide so much tension, and it's more difficult to control. Personally I used heat after gluing one stator just to even out some small wrinkles. Just be careful to keep moving the hot air source around without pausing, to avoid burning holes in the film. Mylar has exceptional mechanical properties (it will eat scissors for breakfast ) but heat is its enemy.
) but heat is its enemy.Make a small panel and just DO IT. Can you now see that some important issues can be best resolved by experiment not engineering contemplation?
Being a biker has immeasurably improved the strategic balance in my life between contemplation and action. Started when I was a grad student. 50 years ago.
Being a biker has immeasurably improved the strategic balance in my life between contemplation and action. Started when I was a grad student. 50 years ago.
*Drawing a deep breath*
Calvin, you rock my world and I guess it's back to the drawing board once again.
I can work with a smaller panel, especially if there's no real downsides to it?
Mechanical tension! I guess I'll have to build myself a jig?
Curved panels? Since it looks like I'll be building a jig anyway I might just go all out if it's worth the extra work?
Let's lay out the ground rules then, shall we?
It might look complicated when I line up the questions like that but in my head it's a fairly straight forward process as long as I get the details figured out.
I have a fair understanding of the crafts involved and the underlying material science.
bentoronto> I've worked as a mechanic myself and know that theory seldom mimics the real world. I try to minimise the "fixing it on location" kind of problems before hand. I know there'll always be stuff to correct but you can try keeping it to a minimum.
Calvin, you rock my world and I guess it's back to the drawing board once again.
I can work with a smaller panel, especially if there's no real downsides to it?
Mechanical tension! I guess I'll have to build myself a jig?
Curved panels? Since it looks like I'll be building a jig anyway I might just go all out if it's worth the extra work?
Let's lay out the ground rules then, shall we?
- What curvature should I choose? (20 degrees or what?)
- Will I need to roll the stators before hand? Or will it be enough to just bolt them to the fixture?
- Going for curved stators, am I right in assuming horizontal spacers is the way to go?
- The perf metal comes in 1000x2000mm sheets, practical matters put the absolute maximum width at 250mm (possibly somewhat less due to cut away scraps). What are my target dimensions? 1300x250mm total panel area? (1250x200mm radiating area minus horizontal spacers?)
- How important is it to line up the holes in the front/rear stators?
It might look complicated when I line up the questions like that but in my head it's a fairly straight forward process as long as I get the details figured out.
I have a fair understanding of the crafts involved and the underlying material science.
bentoronto> I've worked as a mechanic myself and know that theory seldom mimics the real world. I try to minimise the "fixing it on location" kind of problems before hand. I know there'll always be stuff to correct but you can try keeping it to a minimum.
Last edited:
Hi Markus,
I have just finished my first panels following Charlie's blog and they sound fantastic.
Thanks Charlie!
At my first attempt I used 1.1mm 3M VHB tape and heatgun for tensioning the mylar (instead using Charlie's jig). Soon I realised that I can not get even tension because of the size of the panel and the diamater of the heatgun, and it is very easy to melt holes. However I finished the panels, hooked up and they worked but I couldn't get good sound level because coupling and arcing. After several hours of listening and testing I got to the final conclusion I need to rebuild my panels using high mechanical tension and thicker spacing. So I did and now I am happy with the result. Very happy.
I also used the licron spray but the spary cap is such a rubish I find it impossible to make an even coat so after a thin spay I used a brush to spread it.
If I would start again I would go to a shop to get my perforated shets cut by them and get it rolled flat because the evenness is everithing. Wichever process you chose make sure you can repeat it at least one more time to get equally sounding panels.
So I would go for high mechaninal tension, shop cut and rolled metal and powder coating. All worth it.
I have just finished my first panels following Charlie's blog and they sound fantastic.
Thanks Charlie!
At my first attempt I used 1.1mm 3M VHB tape and heatgun for tensioning the mylar (instead using Charlie's jig). Soon I realised that I can not get even tension because of the size of the panel and the diamater of the heatgun, and it is very easy to melt holes. However I finished the panels, hooked up and they worked but I couldn't get good sound level because coupling and arcing. After several hours of listening and testing I got to the final conclusion I need to rebuild my panels using high mechanical tension and thicker spacing. So I did and now I am happy with the result. Very happy.
I also used the licron spray but the spary cap is such a rubish I find it impossible to make an even coat so after a thin spay I used a brush to spread it.
If I would start again I would go to a shop to get my perforated shets cut by them and get it rolled flat because the evenness is everithing. Wichever process you chose make sure you can repeat it at least one more time to get equally sounding panels.
So I would go for high mechaninal tension, shop cut and rolled metal and powder coating. All worth it.
Before finalizing your decision to maximize diaphragm tension you might read over this thread. Capaciti reports better control of resonances, and cleaner cumulative spectral decay (CSD) results, when using lower tension. That said, it's definitely a matter of finding the right compromise. Tension that is too low comes with its own set of problems.
Few
Few
A beautiful thing about using 3M foam tape for the spacers is that it also sticks and holds the panels together quite well-- so you won't need any fasteners. For example:
Charlie, saw your page, wondering why you bothered with beam spliter with hybrid design? If dipole cancellation occurs around a wavelength thats 4*panel width, then this corresponds to a freq of 281.5Hz (using 12" width). I believe you xover at 300Hz to the bass module. Since you put the lows in a TL theres no back field to worry about for the lows.
Why not use the corner of the room, with sound panels?
Do you know any details about the magnet strips that Jonus was using. McMaster Carr, has these
McMaster-Carr
The magnet strips is a great idea, it builds into the design that eventually your esls will fail and need replacing. Not only that, it allows for cleaning if needed, mylar retensioning, recoating, expermenting, etc with relative ease. Thanks for sharing.
Enjoyed reading your pages, very helpful.
Take Care
B
Hello B,
Actually, I didn't need or even want a beam splitter, for the reasons that you pointed out. The sole reason was to enable using a TL bass in the same enclosure/frame:
My goal was to have transmission line bass with the woofer under the stat panel in a single speaker (rather than having the bass & stat panels as separate speakers standing side-by-side). The problem was what to do with the TL's huge box volume!
The only means I could devise to merge a huge TL box and a stat panel into the same enclosure was to extend the box volume vertically and place it behind the stat panel. Of course, you can't place a flat/parallel box directly behind the stat panel or the box would bounce the rear wave back into the diaphragm. Ultimately, my solution required forming the bass box into a beam splitter.
I recall that I exchanged email with Jonas about those magnets and the conclusion was that the magnets strips from McMaster Carr should work.
Charlie
Last edited:
Charlie,
Thanks for your reply. I know your a fan of Roger, in his TL hybrid he puts more channels, more folds in the box. But, if my memory serves me right, his box, the one with detailed drawings and measurements was never meant to be stacked, and as you say was probably meant to work next to the esl panel.
I think you tuned your TL down to 30Hz, which would be a very long wavelength, do you think pushing the res. freq up more would be better\worse?
I dont recall rogers input, but, do you recall what he says about xovering the lows. He said something that you should avoid putting the xover below (x?)Hz because this is where the human ear is most sensitive.
Love your setup, with the active low level xovers, eqs, and amps. nice job!
In one of your photos, it looked like you had 3 or 4 equalizers? one was the nice digital one.
You would only need one eq. though, and the digital one you have would do a better job of discreatizing over the audible band then those analog ones (which has a more 'courser' discreatization, 1kHz intervals). any reason.
Great, happy tpo know that about the magnets, there very reasonable. looking at those rolls they have two sets of polls on one roll, so i guess i dont need as much magnetic strips as i need 3m double side. For the magnetic strips, I think this means I would need only 1/4 the needed length of 3M double side.
Do you have Jonus's email
Thanks,
Actually, I think a vertical array of multiple dipole woofers would blend best with an ESL but the cost for all those woofers and amps to push them was prohibitive for me a the time so I chose the TL option. I am a fan of Roger Sanders for his pioneering work to promote DIY ESL's. That said, I think his book is a great guide for the basics but now rather dated. I used his guidelines in my blog because I'm no expert myself and I wanted to keep it simple and cheap for the benefit of other newbies like myself.
The best TL would be a straight pipe with no folds at all. Sanders' Eros speaker TL uses multliple folds to acheive it's 8ft line-length in a minimal height to fit the entire box volume under the stat panel (unlike mine, which is mostly behind the stat panel and much larger in volume). I think most TL builders would say that Sanders' Eros line has too little volume to work efficiently but he's using a servo-controlled woofer in it with tons of power and apparently he makes it work.
As I recall, Sanders recommends to keep the crossover below 400hz, as it's less audibly objectionable to have a lower crossover. It's a trade off, though, as with narrow panels, the lower the crossover the more EQ is needed to compensate the dipole phase cancellation.
The photos of my system on the blog show more gear than I'm actually using. In fact, my rack contains two complete systems:
The main system with the ESL's uses just two of those (4) Carver TFM-25 amps, the C1 preamp, Behgringer digital EQ, Carver CD decks, Carver tuner and DBX crossover and I left one of the Audio Control EQ's in the tape monitor loop (but bypassed) just because I love its light show.
The Carver TFM-42 amp will power some ripole subs I'm currently building. The other system with Carver CT-7 tuner/preamp, one of the remaining TFM-25 amps and the other EQ are in the system I use for TV and AM radio. The remaining TFM-25 is a backup.
I don't have Jonas' email but you can PM him here, as I did.
The best TL would be a straight pipe with no folds at all. Sanders' Eros speaker TL uses multliple folds to acheive it's 8ft line-length in a minimal height to fit the entire box volume under the stat panel (unlike mine, which is mostly behind the stat panel and much larger in volume). I think most TL builders would say that Sanders' Eros line has too little volume to work efficiently but he's using a servo-controlled woofer in it with tons of power and apparently he makes it work.
As I recall, Sanders recommends to keep the crossover below 400hz, as it's less audibly objectionable to have a lower crossover. It's a trade off, though, as with narrow panels, the lower the crossover the more EQ is needed to compensate the dipole phase cancellation.
The photos of my system on the blog show more gear than I'm actually using. In fact, my rack contains two complete systems:
The main system with the ESL's uses just two of those (4) Carver TFM-25 amps, the C1 preamp, Behgringer digital EQ, Carver CD decks, Carver tuner and DBX crossover and I left one of the Audio Control EQ's in the tape monitor loop (but bypassed) just because I love its light show
. The Carver TFM-42 amp will power some ripole subs I'm currently building. The other system with Carver CT-7 tuner/preamp, one of the remaining TFM-25 amps and the other EQ are in the system I use for TV and AM radio. The remaining TFM-25 is a backup.
I don't have Jonas' email but you can PM him here, as I did.
Last edited:
Thanks Charlie.
For those that use the bike tube jig for tensioning the mylar, do you think it would be possible to build such a jig, but keep the tire always around the film and build into the esl mount. This way, if over time and the many hours of playing it we needed to retension the mylar we could be inflating the tire more? However, do you need to have a solid back and then it would not be 'acousticaly transparent'? I see the tire jig working in that it only has to stretch the film to 1-2% elongation, which you get when the tire inflats.
If not then if we only use double side tape to secure it, will that alone be sufficient to keep the tension for some time? If not, then the only way to retension down the road is to start over?
Also, anyone here study RLC oscillations in an ESL before. If we model the step-up transformer as non-ideal then it would have its own built in cap, res, and inductance. If we reduce the circuit as much as possible to a simplified RLC circuit then there would be RLC oscillations that would have some resonant freq.the resonance frequency of an LC circuit with no resistance R, undamped is fres=(1/2Pi)*(1/(LC)^0.5). the driven res freq, would take more work to determine, but if there is some freq in the circuit that doesnt correspond to the audio it could give distortion.
Thanks
Bryan
For those that use the bike tube jig for tensioning the mylar, do you think it would be possible to build such a jig, but keep the tire always around the film and build into the esl mount. This way, if over time and the many hours of playing it we needed to retension the mylar we could be inflating the tire more? However, do you need to have a solid back and then it would not be 'acousticaly transparent'? I see the tire jig working in that it only has to stretch the film to 1-2% elongation, which you get when the tire inflats.
If not then if we only use double side tape to secure it, will that alone be sufficient to keep the tension for some time? If not, then the only way to retension down the road is to start over?
Also, anyone here study RLC oscillations in an ESL before. If we model the step-up transformer as non-ideal then it would have its own built in cap, res, and inductance. If we reduce the circuit as much as possible to a simplified RLC circuit then there would be RLC oscillations that would have some resonant freq.the resonance frequency of an LC circuit with no resistance R, undamped is fres=(1/2Pi)*(1/(LC)^0.5). the driven res freq, would take more work to determine, but if there is some freq in the circuit that doesnt correspond to the audio it could give distortion.
Thanks
Bryan
one more thing, using magnets and double side tape, what strength magnets do you think would be sufficient. Mcmaster-Carr sells flexible magnets that range from 4-24 lbs/ft
McMaster-Carr
Found local powder coater, will cost $110 and he can perform flash tests on the finished panels.
Bryan
McMaster-Carr
Found local powder coater, will cost $110 and he can perform flash tests on the finished panels.
Bryan
I think a bike tube in the frame would add unnecessary complexity and the tube would leak down and not hold constant pressure anyway.
If you use a bike tube jig and tape to tension and hold 6 micron film and you stretch to let's say 1.5% elongation, you will see only a small amount of creep in the tape and corresponding small reduction in tension when the pressure is released from the tube jig, as the diaphragm tension then loads the tape in shear. But not to worry; the majority of the tension remains and you will have no problems provided the distance between diaphragm support spacers is within 70 -100 x the diaphragm-to-stator spacing. I can verify 3 years of trouble-free service with mine so far and I believe FEW has some speakers still playing fine after 10 years. I you use tape, you want to use it within 2 years of it's manufacturing date-- so you would want to buy from a source that sells a lot of it (keeps their stock rotated and has fresh tape on hand).
About RLC oscillations in the trannys, you might find some info if you do a search; especially of Calvin's posts.
Last edited:
ok, i'd be using the magnetic strips to hold tension, and this would not be a shear force, but a normal force. The tape has the ability to prevent the mylar from slipping since it adds quite a bit of friction, which is a shear force.
So, what im trying to consider is how strong the magnets need to be so that I hold the tension at least as well as using only the double side tape.
Your probably right about the bike tube in the jig. I think that if you get a reliable tube and you operate under room temp that the tube should hold for sometime. If needed you could always pump more air into it.
I have to find Jonus on here and see if had issues with the magnetic strips holding the tension for sometime.
Thanks
Bryan
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.