Molding stators with plastic like ESL 57's.
Dear ESL constructors,
I was looking at an original broken Quad ESL 57 stator. Why more people don't use the very simple Quad ESL 57 way of making stators? Maybe its because DIY folk started doing it the Sauders way and just continued.
It looks relatively easy compared with using perforated sheet or wire mesh with all the resonance/kinking issues involved, especially if you want to make accurate stator gaps of 1 mm or less over a large area.
From looking at the construction its simply a thermoplastic in with indentations made by compressing the thermoplastic in a simple ring mold while heating. This way you just clamp the film to the stators and no separators are needed. You could even mold them from clear acrylic, though its harder to drill, The conductive elements are simply attached to the surface of the plastic. I guess Aluminium foil and some varnish would be easy and simple, but maybe you use wire mesh in the mold.
With the Quad ESL 57 attached the conductive material inside for the high frequencies and outside for the bass (making bass panels almost indestructible). Attaching the conductor (probably not to hard) and the mold is the only hard part now, but you only need to make one.
The mold could be made very easily with different thicknesses of MDF, or even the same thickness and depth adjusted with some spacers under the center board. All you would need is a jigsaw and a couple of clamps on a a compression ring on the outside while in the oven, when the clamps loosen take out of over and tighten clamps and repeat until molded. I imagine some thick low density polythene sheets about 5mm think or more would be ideal.
Looking at the final result from a damaged stator, it looks like the holes where drilled before attaching the surface or molding.
The Quad ESL 57 does not sound so bad as to go with more complex solutions? I accept that segmented PCB's in nice injection molded stator holders like the ESL 63 would be better but its a lot harder even than the metal grill approach.
I will be making headphones and will probably use PCB material as its small but for larger drivers the approach taken by the Quad ESL - 57 seems very practical and molding would provide good uniformity of stators, which I imagine is key to getting the stereo imaging I love about electrostatic speakers.
Making a mold of sufficient precision just for the few DIY panels? Drilling countless holes, glueing conductive strips, insulate conductive areas, etc, etc. doesn´t sound all too easy to me.
The easiest and probabely most rewarding for DIY will still be the wired stator as materials are easyly and cheaply available and the chance of success is high.
Yes, making it by using mold is perhaps a quite difficult task.
But if one's has an access to CNC or perforation machine and sufficiently large PCB sheets then it could be quite viable IMO.
Searching through google shows typical PCB material FR-4 has a dielectric constant of about 4. So 1.5mm thick material would seem like some 0.3-0.4mm of air which is more or less acceptable.
However this has some quite serious issues with safety, as copper would be exposed to outside(same problem as with bare metal stators). The need for grounded conductive screen(s) complicates everything a lot.
After doing some experiments with sheet vs wire stators I never looked back to sheets of any kind(except I want to make a curved system in some future).
I read about a build some few years ago of a DIYer made molds and poured his own acrylic stators.
He said that they turned out very nice and worked well and used them for mix down monitors in the studio where he worked.
But after some time he said they yellowed a bit and started to warp and lost their precision shape.
I have used Plastic lighting grid and it works well for small width panels but for anything larger it takes extra support in order to have a stator the doesn't bow in or out.
An 8" wide panel can be flexed quite a bit and still work but it took a lot of tensioning to make them stable and this caused each side to bow out alot by as much as .010" to .030" (much to much) or more sometimes.
I know one was pretty bad although it still worked okay and I don't remember the exact value as it was 10 years ago.
And that is using a rigid plastic grid material.
Plastic sheet that has no lateral support can be much worse!!
I recall one DIYer that had to find a new Tweeter panel for one of his 57's becuase it was warped and only slightly.
On my latest build I have gotten within +/-5 mil or so using TIG wire and I am still having a bowing issue with the PCB stator support.
It is not to bad but it is there and it irritates me!
I will correct this when I build the next one.
I am not sure if it was caused by heat from having to resolder and reposition the rods causing them to have tensioning on the PCB material and/or if the PCB's warped from being reheated so many times.
I did have to replace one end due to a few pads that lifted off from the heat.
My new frames are made of standard sheet plastic and have no rigidity to them at all and I was scared the this was going to be an issue.
But the TIG rods gave the panel some very good rigidity in the end and enough to support the mylar being sandwiched at 5" intervals with bolts.
Thanks for the discussion points. I think making the mold will be the easiest part, as I even made a similar mold when I was 15 at school (we had a lesson in Plastic memory and molding). All you would need would be 3 sheets of MDF. One to make a ring around the molding "top", one to act as the "bottom" of the mold. The "bottom" would have holes cut on it where a similar thickness of MDF would be placed in the holes. These filled holes would cause the raised bumps by putting some spacers under the filler blocks, and placed on the molds floor so making nice simple smooth indents in the plastic.
I believe this part of the project could be done in an hour or two.
Calvin's point "Drilling countless holes, glueing conductive strips, insulate conductive areas, etc, etc. doesn´t sound all too easy to me." is an important issue. geraldfryjr made some interesting points about structural issues, which are very relevant too, I guess these issues of structural issues are the reason that Quad ESL 57 almost has two tweeters next to each other when looking inside the stator, this might be just to solidify the structure.
Getting slightly off topic and discussing PCB material, Luka's point about PCB material being a good dialectric is interesting. I personally would like to know how to design 3 layer PCB's cheaply so only the middle layer had copper.
Thankyou guys for your thoughts.
Wires must be stretched to straighten them(second photo is the stretching jig) and then somewhat released prior to gluing as no reasonable frame will handle the summed axial load of many wires.
PCB stators have been discussed here at diy before afaik.
You may look at older Beveridge Patent US4,533,794 or Sanders US2005/0094833.
The Beveridge uses basically a single-sided PCB-Material with Nylon lamination insulation to the copper side, while Sanders is pure PCB-material and cnc-milling.
One is only restricted to the typical size limit of PCB-manufacturers for a pure PCB-stator.
The PCB support is a small issue and will be corrected in the next one with a proper building jig to hold and space the rods while I am tacking the ends.
The second stator when together flawlessly and was perfect.
There are a few reason why I didn't use some common wire,
First there is the issue stretching and straightening the wire so that there are absolutely no kinks or ripples in them.
Not impossible to do, but very laborious!!
Therefore the rigidity of the steel rod is a big plus.
Although selecting rods that are completely straight can take some time as well.
Second I want to push the voltages to the limits with this panel as I have done so with my last one.
I can get nearly as much as 2Kv per mil of insulation factor using off of the shelf clear acrylic spray can paint.
Although my rod diameter is 1/16" a thinner wire and/or rods can be used as my last one used just window screen that was powder coated and then later painted seven years later.
I was able to push that panel with 10kv of bias and had as much as 25KV P-P across the stator's before the stator coating and sharp wire points caused it to fail.
These new stator's withstand 14KV (maybe more as that is the limit of my HV supply) each with no leakage problem with only 8 to 10 mil of coating thickness.
Third, I have tested some wire and it takes a pretty thick insulation to handle these types of voltages and it doesn't come on a thin gauged wire.
Kynar coated wire wrap wire is a great candidate and I have tested it at 10kv before. But, I believe that is about the limit for the common standard 30ga. stuff.
The cost of copper wire has gone through roof these days and it would still be quite costly to use it for even a small panel let alone a large panel!!
Steel has gone up as well but not quite as bad you can get a 10lb box of TIG rod for about $40 to $50.
There is about 35 rods per lbs.
And I already have some left over from the three 10.5" X 36" that I have finished.
I think I paid about $3 lbs 8 years ago or less.
It took 17 rods to make one panel.
So the cost of the pair of panels is about is less then $15 for everything including the paint.
10 years ago these cost probably would have been about half that.
I wanted to use common materials found down at the local hardware store and try not to revert to any special order items or exotic materials with the simplest method of construction that I could come up with.
Although I do have licron that I use for a diaphragm coatings, it has been reported that there are some materials commonly found on store shelves that may be some viable alternatives.
I got very good performance out of my last panels and it was by far the cheapest method of construction only I was pushing the voltages past what is I had designed it to do.
The stator coating was the biggest issue and I have perfected my method even using the window screen.
I just need to make a new pair of screens as I still have the frames for the diaphragms for them.
I agree that copper wire is not exactly cheap but it yields good results and is relatively simple to build larger panels this way.
The insulation is robust in both electrical and mechanical terms and can take a lot of abuse(including impacts with tools, etc). Breakdown of PVC is about 20kV/mm so typical PV3 or H07 wire should handle up to 10-13kV or so.
I have doubts if you can achieve similar levels of performance by using paint for real sized panels like 0.2-0.5m^2 square easily.
Also how did you test the insulation strength? Did you place the whole panel against an aluminum foil so there is no air gap and apply 10kV between it and stator?
I am not convinced that making a small panel with very high voltages is the best way to achieve good performance :
a) High bias will either need lots of tension or many supports to prevent the film from collapsing to a stator. This will rise fundamental mode a lot; other modes will be shifted to even higher frequencies
b) The need for very high step-up ratios makes life for the amplifier more difficult
c) Hissing/cracking noises during varying humidity levels can easily arise when air breakdown voltage is exceeded and air starts to ionize
d) Dipole cancellation for small panel is much more prominent limiting it to mid range & treble use, although this can be fought by using a baffle.
e) A line source of lets say 1.5m high or so would provide good vertical dispersion pattern which a small square panel would not
f) Long-term reliability test?
|All times are GMT. The time now is 06:12 PM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio