Think I am going to switch to using hot glue (specifically, Steinel 04043 GF 260 glue sticks) to glue the wires down. The E6000 is too much of a pain to work with.
Any concerns with hot glue? I presume it would be plenty strong enough as I don't think the forces on the wires would be that high.
The E6000 stuff is just to hard to put down without it being "raised" and sit on top of the wires. It seems to be too viscous.
I figure with hot glue, if the glue doesn't flow completely flat I can iron it down (as someone in this thread pointed out) or go over it with a heat gun.
Any concerns with hot glue? I presume it would be plenty strong enough as I don't think the forces on the wires would be that high.
The E6000 stuff is just to hard to put down without it being "raised" and sit on top of the wires. It seems to be too viscous.
I figure with hot glue, if the glue doesn't flow completely flat I can iron it down (as someone in this thread pointed out) or go over it with a heat gun.
The early ER audio panels like mine used plastic louvre and perf metal stators but cut out a few pieces at regular intervals. Would be pretty easy with a jigsaw and save a lot of gluing.
kffern
BTW, Definitely seeing the appeal of stretched wires now
As I recall, it was a bitch gluing down the welding rods every 1/2 inch.
On my current wood/wire panels, making the jig to stretch the wires was a bit challenging but assembling the stators was rather easy. Gluing the wooded support lattice onto the wires, in the jig worked out pretty well-- wax paper under the wires prevented the glue from sticking to the jig and the resulting glue lines were perfectly flush to the wires.
Yeah, I learned that lesson last night. I did end up melting the plastic a bit but it is mostly an aesthetic problem. I probably shouldn't of set the gun to 1000 degrees
. This stator will definitely be on the back side on one of my panels . The heat gun got the hot glue completely level with the wires which was good. I think the iron approach might work better though. I will see about using my temperature probe to set the iron temp to just above the glue melting point.
The hot glue is nice to be able to work through building out a stator without stopping/starting waiting for glue to dry...but going back over it to make it "flush" may suck even more. When applied, the hot glue just sits right on top of the wires. Probably because the wires are a heat sink.
I did briefly entertain the idea of energizing the wires with low voltage, high current to make them hot.... but decided it wasn't worth the hassle.
OK, so update on the glue front. Hot glue doesn't seem to work well for me. While using the "iron over it technique" (with parchment paper) made the glue flat, it also spread it too "wide" so it was covering part of the holes in the louver. I tried to put down the bare minimum of hot glue so it wouldn't spread too much but just couldn't get it to work well consistently. With the panel sizes I am doing, it would be too much of a pain to try and get it to work... (and I am uncertain how well it holds, not even close to as good as e6000).
......BUT.......
I did notice/discover one thing. Apparently the E6000 glue comes in high viscosity and medium viscosity forms. I was using the high viscosity version (didn't know better). This probably explains why it tended to sit "on top" of the wires rather than seeping down between them flat.
For the most part the e6000 glue I have on there is pretty flat. You can feel it if you run your hand across it. It is probably somewhere around .01 to .03 in thickness on top of the wires. So I hope I am not screwed and will have the diaphragm banging into the glue stripes.
So I got the medium viscosity version on order.
Also, I wanted to note, that you can get the e6000 glue in caulking tube form. This makes life much easier and it is even easier if you use an el cheapo pneumatic caulking gun to put it in. It is much easier to maintain a consistent "bead" width and flow rate (and saves some pain and suffering for your hands.... you just have to dial in the right amount of pressure from your compressor...... this what I was using/doing with the high viscosity stuff.
......BUT.......
I did notice/discover one thing. Apparently the E6000 glue comes in high viscosity and medium viscosity forms. I was using the high viscosity version (didn't know better). This probably explains why it tended to sit "on top" of the wires rather than seeping down between them flat.
For the most part the e6000 glue I have on there is pretty flat. You can feel it if you run your hand across it. It is probably somewhere around .01 to .03 in thickness on top of the wires. So I hope I am not screwed and will have the diaphragm banging into the glue stripes.
So I got the medium viscosity version on order.
Also, I wanted to note, that you can get the e6000 glue in caulking tube form. This makes life much easier and it is even easier if you use an el cheapo pneumatic caulking gun to put it in. It is much easier to maintain a consistent "bead" width and flow rate (and saves some pain and suffering for your hands
.... you just have to dial in the right amount of pressure from your compressor...... this what I was using/doing with the high viscosity stuff.Need some advice here....
So I tried the medium viscosity e6000 glue and it is still "sitting" on top of the wire stators.
Basically, the glue is sitting on top on the wires in the stators by .007 in (or a bit more than 3 thicknesses of paper).
Should I worry about this effective "loss" of excursion by the diaphragm or just ignore it?
Just wondering if I should abandon this approach and try a stretched wire setup now before make 3 more stators that may have issues.
Thanks
So I tried the medium viscosity e6000 glue and it is still "sitting" on top of the wire stators.
Basically, the glue is sitting on top on the wires in the stators by .007 in (or a bit more than 3 thicknesses of paper).
Should I worry about this effective "loss" of excursion by the diaphragm or just ignore it?
Just wondering if I should abandon this approach and try a stretched wire setup now before make 3 more stators that may have issues.
Thanks
I think you're good to go with .007. If there are any taller nibs you might try shaving them off with a razor blade. E6000 glue is really stubborn.
Thanks.... I kinda figured I was good but there isn't a whole lot of room for the diaphragm to vibrate so just wanted to be sure.
One more question....
Since my stators are longer than the tig wires, I need to add short segments to the end (as noted in prior posts).
Rather than tediously solder together every wire I was just going to but them together and maybe solder together one or 2 wires in the segment at the butt joint to carry the signal to the short part of the segment. The end of the short segment would still be all soldered together (at the "top"). i.e this saves me from having to run a dedicated wire to the top of each segment.
Is this a good/bad idea? I presume the currents involved are pretty small so I don't think ampacity of using a couple of tig wires in the segment to carry the signal to the rest of the segment is a problem.
The other question is, with the wires butting up to each other instead of each being soldered; do I have to worry about arcing between them? e.g. across any gap that might between the "long" and "short" wires? I think I will be ok with not having arcing with the diaphragm since I plan to strategically place one of the spacers right across where the wires butt up to each other. Just want to make sure the wires won't be arcing between each other where they butt together.
That will work. You can join two rods at the butt joint and then solder across the rods at the panel ends and you will have continuity to all rods thru the end connections.
Assuming that all rods in each group are connected at both ends of the panel and at least one rod in each group is soldered at the butt joint, then all rods will be energized to the same voltage and no arcing will occur at the remaining (unsoldered) butt joints.
To get this, just follow the electrical pathway: From the power input end, thru the (1) soldered butt joint, continuing thru that rod to the soldered connections at the opposite end, to the remaining rod halves that weren't soldered at their butt joints.
Last edited:
No, not between wires in a given group for the reasons CharlieM described.
However, there can be considerable voltage between wires of adjacent sections, especially for the first few sections. I still don't think it would be much of an arcing or corona concern. But if you plan to solder 2 wires in each segment anyways, you could choose to do the outer two so there are no sharp edges from wire ends near adjacent sections.
Here is a link to a spice model of the voltages on the individual sections of a segemented design that I was working on and it shows the relationship of the voltages on each section at Four different frequency's at 20, 200hz,2Khz and 20Khz.
It also shows the power dissipation required from the resistors as well, Note this is for test tones and would be the most extreme values that would be needed compared to a music signal.
I did these using Circuitmaker2000,
http://www.diyaudio.com/forums/plan...tor-esl-simulator-esl_seg_ui.html#post2912820
I hope this help you.
Cheers!!
jer
It also shows the power dissipation required from the resistors as well, Note this is for test tones and would be the most extreme values that would be needed compared to a music signal.
I did these using Circuitmaker2000,
http://www.diyaudio.com/forums/plan...tor-esl-simulator-esl_seg_ui.html#post2912820
I hope this help you.
Cheers!!
jer
Interestingly enough, I just ordered the resistors for my resistor banks (all ~500 of them!). I am using the PR02 (2W) version of these: http://www.vishay.com/docs/28729/pr010203.pdf
With 6 resistors per bank, I think I should be OK - not completely confident though. If I am doing the math right, the 1st bank will see the full ~3.3kV load (peak).... I must be misunderstanding how ohms law applies here because the power dissipation requirement is pretty amazingly high
I have halted work on my stators because "thelensguy" company seems to have gone dark. I made a mistake and didn't order enough louvers so I had to order one more.... that was 1 month ago and I can't get them to respond to me at all.... so not sure what it going on but I may have to figure something else out if I can't get the extra louver I need.
On another topic... wondering if you guys can help me understand something that has been gnawing in the back of my mind....
On Bolserst spreadsheet, there is a typical "Q" parameter comment that states the Q at resonance would be something in the 10-20. So that would be ONLY at resonance (say 75hz) - as I understand it..... but what would be the average Qts of an ESL like this at 250hz and above?
I would tend to think it would be well damped (Qts <.5) on larger panels just because the mylar is like a huge ultra lightweight sail with the air resisting change (thus damping it).
On Bolserst spreadsheet, there is a typical "Q" parameter comment that states the Q at resonance would be something in the 10-20. So that would be ONLY at resonance (say 75hz) - as I understand it..... but what would be the average Qts of an ESL like this at 250hz and above?
I would tend to think it would be well damped (Qts <.5) on larger panels just because the mylar is like a huge ultra lightweight sail with the air resisting change (thus damping it).
Here is one more post I wanted to show you earlier, I just found it!!
It has detailed closeup pictures of how I mounted the rods and shows a test of my conformal coating at 14KV.
http://www.diyaudio.com/forums/planars-exotics/246846-first-time-esl-builder-2.html#post3766238
Also here are the cross sectional blueprint details,
http://www.diyaudio.com/forums/plan...ut-segmented-wire-stator-esl.html#post4184450
Enjoy !!
jer
It has detailed closeup pictures of how I mounted the rods and shows a test of my conformal coating at 14KV.
http://www.diyaudio.com/forums/planars-exotics/246846-first-time-esl-builder-2.html#post3766238
Also here are the cross sectional blueprint details,
http://www.diyaudio.com/forums/plan...ut-segmented-wire-stator-esl.html#post4184450
Enjoy !!
jer
Last edited:
The resonance is the only thing you need to worry about as long as you cross them over above that.
After that point they are inherently flat!!
The Bias voltage also dampens it as well, In fact raising the bias voltage also lowers the resonate frequency.
Bolserst did a sudy on that in another thread.
If you measure them with a good mic close to the diaphragm they will be ruler flat throughout their range except for the diaphragm resonance.
The 6db slope you get from far field measurements is due to DIpole Cancelations this is the one of main reason also why we should be using Electrical Segmentation for equalization of the panel.
Not to mention even horizonal dispertion of all of the frequency's it produces.
jer
After that point they are inherently flat!!
The Bias voltage also dampens it as well, In fact raising the bias voltage also lowers the resonate frequency.
Bolserst did a sudy on that in another thread.
If you measure them with a good mic close to the diaphragm they will be ruler flat throughout their range except for the diaphragm resonance.
The 6db slope you get from far field measurements is due to DIpole Cancelations this is the one of main reason also why we should be using Electrical Segmentation for equalization of the panel.
Not to mention even horizonal dispertion of all of the frequency's it produces.
jer
Last edited:
Hi bengel
For a ESL strip radiator, the Q is the ratio of the reactive part of the radiation resistance, to the resistive part of the radiation impedance - at the resonant frequency. It is not meaningful to talk about the Q at any other frequency.
The attached picture shows the radiation impedance for a strip radiator like an ESL. The horizontal axis is ka which is proportional to frequency - more later.
Green curve: the reactive part of the impedance. Below the knee near ka =1, the reactance is proportional to frequency so the ESL behaves as though it has a constant mass of air attached to the membrane
Red curve: is the resistive part of the impedance. For the ESL strip radiator, it falls in proportion to frequency cubed - quite fast.
Orange curve: is magnitude.
ka is the product of k = 2.pi.f/c so is proportional to frequency.
2a is the width of the ESL strip.
so the knee at ka =1 corresponds to about 220 Hz for a 500 mm wide ESL, so most of the time, the resonance is below ka=1
The resonant frequency is determined by a combination of the radiation impedance, the tension in the membrane, the dimensions of the panel, and also a little on the bias voltage. The tighter the membrane and the smaller the panel, the higher the resonant frequency.
So.... the Q is given by the ratio of the green curve to the red curve at the resonant frequency, and can be quite high, especially for low resonant frequencies.
On the plus side - with the resonant frequency quite low - perhaps 50 Hz or so, it takes very little extra damping to bring the Q down to about 2, which is what you will probably need to extend the frequency response of the panel.
For a ESL strip radiator, the Q is the ratio of the reactive part of the radiation resistance, to the resistive part of the radiation impedance - at the resonant frequency. It is not meaningful to talk about the Q at any other frequency.
The attached picture shows the radiation impedance for a strip radiator like an ESL. The horizontal axis is ka which is proportional to frequency - more later.
Green curve: the reactive part of the impedance. Below the knee near ka =1, the reactance is proportional to frequency so the ESL behaves as though it has a constant mass of air attached to the membrane
Red curve: is the resistive part of the impedance. For the ESL strip radiator, it falls in proportion to frequency cubed - quite fast.
Orange curve: is magnitude.
ka is the product of k = 2.pi.f/c so is proportional to frequency.
2a is the width of the ESL strip.
so the knee at ka =1 corresponds to about 220 Hz for a 500 mm wide ESL, so most of the time, the resonance is below ka=1
The resonant frequency is determined by a combination of the radiation impedance, the tension in the membrane, the dimensions of the panel, and also a little on the bias voltage. The tighter the membrane and the smaller the panel, the higher the resonant frequency.
So.... the Q is given by the ratio of the green curve to the red curve at the resonant frequency, and can be quite high, especially for low resonant frequencies.
On the plus side - with the resonant frequency quite low - perhaps 50 Hz or so, it takes very little extra damping to bring the Q down to about 2, which is what you will probably need to extend the frequency response of the panel.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.