Still researching before starting a ESL project. Have a question that I have not found yet.
How do you go about matching the panels? I can set the tension the same on both and adjust the voltage for the same db output.
Is that all there is to it besides making sure all of the components are the same?
Thanks for all the great input on this board!
Jeff
How do you go about matching the panels? I can set the tension the same on both and adjust the voltage for the same db output.
Is that all there is to it besides making sure all of the components are the same?
Thanks for all the great input on this board!
Jeff
Panel Matching ...
Short of some construction that allows diaphragm tension to be adjusted after assembly, no, there is no way ...
Most panels are assembled as a sandwich with glue and clamps, with the diaphragm already tensioned. Once set, the only way to "re-tune" the diaphragm is to rebuild it ! ...
You may have seen photos of condenser microphone capsules with a
ring full of adjustment screws which allows just such diaphragm tuning on the capsule, but they are small enough to handle. An ESL
panel would be to large and cumbersome to do the same. But it is food for though and if you come up with a solution, do let us know !
Good luck !
Short of some construction that allows diaphragm tension to be adjusted after assembly, no, there is no way ...
Most panels are assembled as a sandwich with glue and clamps, with the diaphragm already tensioned. Once set, the only way to "re-tune" the diaphragm is to rebuild it ! ...
You may have seen photos of condenser microphone capsules with a
ring full of adjustment screws which allows just such diaphragm tuning on the capsule, but they are small enough to handle. An ESL
panel would be to large and cumbersome to do the same. But it is food for though and if you come up with a solution, do let us know !
Good luck !
I wonder if a jig could be made so that the Mylar is wrapped around the spacer in such a fashion that the tension could be modified before the assembly is glued up. Then when the system is certified, and glued, the panel is removed from the jig.
This would probably look like the condenser mike adjustment jig, in fact I have an idea from pictures of ESL 57 jigs that i have seen.
I have to figure that the bigger manufacturers do not hand match the panels and have some sort of test Jig that they use to setup the parameters of the panels. Probably is some sort of an open table with mikes on at least 4 places on the top and bottom that they can use to adjust the response with.
I have a bunch of servo motors laying around, I wonder if I can get those to work instead of changing tension at certain parts of the panel, and letting it cool down (hi-voltage) then adjusting again.
This would probably look like the condenser mike adjustment jig, in fact I have an idea from pictures of ESL 57 jigs that i have seen.
I have to figure that the bigger manufacturers do not hand match the panels and have some sort of test Jig that they use to setup the parameters of the panels. Probably is some sort of an open table with mikes on at least 4 places on the top and bottom that they can use to adjust the response with.
I have a bunch of servo motors laying around, I wonder if I can get those to work instead of changing tension at certain parts of the panel, and letting it cool down (hi-voltage) then adjusting again.
Hi,
I have always tensioned and glued, the membrane of two matching ESL's together out of 1 sheet of mylar.
first I have used a mechanical stretching table, where 2 panels could be placed next to each other.
Now I have made myself a pneumatic stretching table (with bicycletires), where also 2 panels can be placed next to each other.
My panels have identical tension and resonance frequency.
greets
Geert
I have always tensioned and glued, the membrane of two matching ESL's together out of 1 sheet of mylar.
first I have used a mechanical stretching table, where 2 panels could be placed next to each other.
Now I have made myself a pneumatic stretching table (with bicycletires), where also 2 panels can be placed next to each other.
My panels have identical tension and resonance frequency.
greets
Geert
Hi,
its always positive to use as much mechanical tension as possible.
When using a mechanical stretcher very high values of tension can be reached. When You start tensioning You´ll notice that at first the resonance frequency rises app. linearly with growing tension. From a certain point (depending on the film You use and often found around 1%-2% elongation) the resonance frequency keeps nearly constant with further rising tension. Tensioning up to this point (simple measurement of elongation) You will reach a good enough consistency.
Using thermal tensioning the film settles always at the same value of tension. So consistency isn´t really a problem with this method. Rather the relatively low value of tension could present a problem. But that depends on the concept and design of Your panel.
jauu
Calvin
its always positive to use as much mechanical tension as possible.
When using a mechanical stretcher very high values of tension can be reached. When You start tensioning You´ll notice that at first the resonance frequency rises app. linearly with growing tension. From a certain point (depending on the film You use and often found around 1%-2% elongation) the resonance frequency keeps nearly constant with further rising tension. Tensioning up to this point (simple measurement of elongation) You will reach a good enough consistency.
Using thermal tensioning the film settles always at the same value of tension. So consistency isn´t really a problem with this method. Rather the relatively low value of tension could present a problem. But that depends on the concept and design of Your panel.
jauu
Calvin
Hi Folks,
i disagree that ESL are consistent in performance when comparing two to each other. Most constructions i have seen from DIYers suffer from constructional aspects.
especially when perforated stell is used on large ESL the eveness of the plates is the critical point. a stator-membrane spacing of e.g. 2mm makes deviations of the stators, which can be up to 1mm easily, very significant.
I do not want to put good speach on my owm work, but my panels are one of the most precise regarding construction. Nevertheless deviations in frequnecy response between two panels are at least 1dB. Lower than 1dB is a lucky case.
There is a second issue, which most builders ignore. I never listended to an ESL, which showed its final performance before 3 month break in time. Mechanical tensioned membranes even take longer. So what you measure or listening directly after building is not comparable to the situation after break in.
The right tensioning method considers the time based relaxing properties of mylar and those are strongly related to the bidirectional tensioning you apply. This bidirectional tensioning must be optimized in relation to the width-height-ration of the membrane area.
Capaciti
i disagree that ESL are consistent in performance when comparing two to each other. Most constructions i have seen from DIYers suffer from constructional aspects.
especially when perforated stell is used on large ESL the eveness of the plates is the critical point. a stator-membrane spacing of e.g. 2mm makes deviations of the stators, which can be up to 1mm easily, very significant.
I do not want to put good speach on my owm work, but my panels are one of the most precise regarding construction. Nevertheless deviations in frequnecy response between two panels are at least 1dB. Lower than 1dB is a lucky case.
There is a second issue, which most builders ignore. I never listended to an ESL, which showed its final performance before 3 month break in time. Mechanical tensioned membranes even take longer. So what you measure or listening directly after building is not comparable to the situation after break in.
The right tensioning method considers the time based relaxing properties of mylar and those are strongly related to the bidirectional tensioning you apply. This bidirectional tensioning must be optimized in relation to the width-height-ration of the membrane area.
Capaciti
Capaciti: By bidirectional tensioning do you mean to say that you recommend different tensions along the long and short dimensions of the panel? Is this assuming a particular style of stator-diaphragm spacers (arrays of dots, or perhaps several parallel horizontal or parallel vertical linear spacers)?
Hi Few,
yes i mean that different tension is applied for width and length. The only way to avoid the impact of tension on the longer side is to tension a mylar sheet which is 3 times longer than the height of the ESL itself. But this is very uneconomical.
I think nobody would prefer this, so you need tension on both directions. It is nearby impossible to predict the time based creeping of the molecules, so i had to go the long trial and error way. My personal experience tells that the final behaviour, especially at fundamental resonance, is very dependent from the tensioning ratio of a specific ESL-geometrie.
e.g. if the tensioning procedure is at its optimized level, fundamental resonance show Q-factor lower than 1,5 and even the suckout in the upper bass range is less significant.
Capaciti
yes i mean that different tension is applied for width and length. The only way to avoid the impact of tension on the longer side is to tension a mylar sheet which is 3 times longer than the height of the ESL itself. But this is very uneconomical.
I think nobody would prefer this, so you need tension on both directions. It is nearby impossible to predict the time based creeping of the molecules, so i had to go the long trial and error way. My personal experience tells that the final behaviour, especially at fundamental resonance, is very dependent from the tensioning ratio of a specific ESL-geometrie.
e.g. if the tensioning procedure is at its optimized level, fundamental resonance show Q-factor lower than 1,5 and even the suckout in the upper bass range is less significant.
Capaciti
Hi,
In my opinion, this works indeed good.
And because I can place 2 stators next to each other, they have te same tension too.
Geert
In my opinion, this works indeed good.
And because I can place 2 stators next to each other, they have te same tension too.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Geert
Well,
It was a really big bike
No serious,
I simply cut two inner tubes open with a knife and glued them together to one long tube with rubberglue. You know, the one you use to repair your inner tube in case of a hole.
It worked fine. I just had to find a trick to do the glue job well.
I put one rubber tube end in the other and layed over the above.
Then I put glue on the inner and outer and let dry the glue.
Then I rolled back the outer one over the glued inner and pressed it together like you do when you repair a tire.
I build 4 panels (hybrids) for my home theater.
Before I used a mechanical stretching table.
With the pneumatic table you can easily get a very even tension on the membrane.
You just have to be carefull not to go to far, because it can create a massive tension force (and you don't want your mylar to tear)
I also learned to lay the mylar very loose and with a lot of wrinkles on the table and let the tube and air do de job.
The first time, I tried to get the membrane already a bit tensioned and without wrinkles on the table. That wasn't a good idea, because once I started pumping , there was no room anymore for the membrane to search it best position and in some corners the tension got directly to high, while in the other, there were still wrinkles.
So if you start with a very loose membrane, it finds very easily its way to divide the tension very evenly. A tube can expand very much.
On some places, the tube gets in a real big bubble, but that is OK. The air divides that way a really even tension on the membrane.
Geert
It was a really big bike
No serious,
I simply cut two inner tubes open with a knife and glued them together to one long tube with rubberglue. You know, the one you use to repair your inner tube in case of a hole.
It worked fine. I just had to find a trick to do the glue job well.
I put one rubber tube end in the other and layed over the above.
Then I put glue on the inner and outer and let dry the glue.
Then I rolled back the outer one over the glued inner and pressed it together like you do when you repair a tire.
I build 4 panels (hybrids) for my home theater.
Before I used a mechanical stretching table.
With the pneumatic table you can easily get a very even tension on the membrane.
You just have to be carefull not to go to far, because it can create a massive tension force (and you don't want your mylar to tear)
I also learned to lay the mylar very loose and with a lot of wrinkles on the table and let the tube and air do de job.
The first time, I tried to get the membrane already a bit tensioned and without wrinkles on the table. That wasn't a good idea, because once I started pumping , there was no room anymore for the membrane to search it best position and in some corners the tension got directly to high, while in the other, there were still wrinkles.
So if you start with a very loose membrane, it finds very easily its way to divide the tension very evenly. A tube can expand very much.
On some places, the tube gets in a real big bubble, but that is OK. The air divides that way a really even tension on the membrane.
Geert
Hi Geert,
Perhaps I'm missing something, but I thought Capaciti specified different tensions for length and width, whereas the bicycle tyre approach gives one uniform tension in all dimensions? So could someone please show me a practical way to get different tensions in different dimensions? Thanks.
Perhaps I'm missing something, but I thought Capaciti specified different tensions for length and width, whereas the bicycle tyre approach gives one uniform tension in all dimensions? So could someone please show me a practical way to get different tensions in different dimensions? Thanks.
Capaciti said:
I suppose this would depend for a large part on the material used. If you had for example a TD/MD of 40/60 this would require different tensioning as with 40/40. Correct?
Did you find a predictable relation between geometry / film properties and the tension ratio? Like a rule of thumb where to start?
Hi wolf-moses
I have build a stretching jig that allows just that.
It concains to 2 rails, one one a long side and one a short side. Each rail contains short blocks in MDF that can slide sideways. All the blocks are linked together with a rubber band.
Each rail can be pulled to stretch with a number of threated wires (one turn is a displacement of 1 mm).
I put double sided tape on the rubberstrip and use this to fix the membrane. The rubber strips on de rails can get longer while beeing pulled aside when the other rail is displaced.
The complete jig was made from one panel of MDF 18 mm (+ rubber strips of 6 mm thickness).
After prestretching the film to remove the (few) wrinkles I have by fixing the membrane, I measure the distance between the double sided tape and calculate how many turns I must apply to get a preset elongation.
This jig permits to use a different ratio in lenght and width, and works quite well.
For the moment I have used it a number of times with a film called WINWRAP of 9 micron. It was ment as a test to check the consistancy of the rest of my frames. It has been playing rather well for six months, and has stabilised quit well. I hope soon to put the real Mylar to use.
I'll try to find some photo's we must have made once while tensioning.
Edwin
I have build a stretching jig that allows just that.
It concains to 2 rails, one one a long side and one a short side. Each rail contains short blocks in MDF that can slide sideways. All the blocks are linked together with a rubber band.
Each rail can be pulled to stretch with a number of threated wires (one turn is a displacement of 1 mm).
I put double sided tape on the rubberstrip and use this to fix the membrane. The rubber strips on de rails can get longer while beeing pulled aside when the other rail is displaced.
The complete jig was made from one panel of MDF 18 mm (+ rubber strips of 6 mm thickness).
After prestretching the film to remove the (few) wrinkles I have by fixing the membrane, I measure the distance between the double sided tape and calculate how many turns I must apply to get a preset elongation.
This jig permits to use a different ratio in lenght and width, and works quite well.
For the moment I have used it a number of times with a film called WINWRAP of 9 micron. It was ment as a test to check the consistancy of the rest of my frames. It has been playing rather well for six months, and has stabilised quit well. I hope soon to put the real Mylar to use.
I'll try to find some photo's we must have made once while tensioning.
Edwin
Hi Folks,
sorry, up to now i did not find a way to predict the rigth tensioning depending on the panel geometrie, but you need to go the trail and error way.
It took me years to find the right one for my current serie which is 20cm in width and 137cm in height for the mebrane area. Even there is different tensioning along the height in order to accomodate for the different mass load of air (more on the bottom less towards the top of the speaker.
By optimizing the excisting panel i have been able to improve maxoutput by about 3dB, having identical frequency response. At the very beginning of development the membrane touches the stators at specific locations first, when reaching maxexcursion, while other sections showed reserve. Now the whole membrane area reaches its limits very consistent.
Capaciti
sorry, up to now i did not find a way to predict the rigth tensioning depending on the panel geometrie, but you need to go the trail and error way.
It took me years to find the right one for my current serie which is 20cm in width and 137cm in height for the mebrane area. Even there is different tensioning along the height in order to accomodate for the different mass load of air (more on the bottom less towards the top of the speaker.
By optimizing the excisting panel i have been able to improve maxoutput by about 3dB, having identical frequency response. At the very beginning of development the membrane touches the stators at specific locations first, when reaching maxexcursion, while other sections showed reserve. Now the whole membrane area reaches its limits very consistent.
Capaciti
Hi Capaciti,
Thanks for your continued contribution to this thread. When you say:
Is this what you have found in practice or am I misinterpreting something?
-Few
Thanks for your continued contribution to this thread. When you say:
do you mean there is more tensioning at the bottom of the panel, more air load on the diaphragm at the bottom of the panel, or both? My guess is that there is more of an air load at the bottom of the panel (due to floor reflections) so the membrane excursion is less in that region. This would allow the diaphragm tension to be less where the panel is closest to the floor.
Is this what you have found in practice or am I misinterpreting something?
-Few
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.