[chinsettawong]

I use contact cement (rubber glue).

 

[arnaud]

And forget about magnets while you're at it, these are stators 

 

[Salm0n]

Thanks for the advice Chinsettawong, I'll look into it.
 
Haha, perhaps one day I'll get into e-stats, likely when I have the knowledge and money to build an amp to go with them, but 278 NdFeB cubes sitting on my desk tells me that this project is going to be about planars. If I can get a camera I may post a build diary to the DIY forum, but I'll stop with off topic posts to this thread for now, just couldn't think of a better place to ask.

 

[wakibaki]

Thanks for the good wishes Wachara C. I can't figure out which is your family name, and which is your given name.
 
I fixed the electronics for my CNC mill, so got started trying to make some things.
 
Here's my first attempt to cut a spider-type stator:-
 

 
I think it needs slight changes to the dimensions, some mounting holes and some contact points, but it's definitely progress.
 
Arrow sent me the IXYS depletion-mode mosfets eventually, but they charged me $44 for the shipping, the parts only cost $22 for 8.
 
Here's the current state of the amplifier PCB design, they're 100*100mm, the PSU board (red tracks) sits on top of the amplifier board (blue tracks), the copper groundplanes are in between the tracks, on the bottom of the top board, and on the top of the bottom board, the supplies (680V and +/- 15V) drop down vertically to where they are required on the amplifier.
 
The mosfets are driven by NE5332 opamps, and are arranged as a CCS-loaded long-tailed pair, with the IXTP01N100's doing duty in the CCS's as well as being the active devices.
 
The amplifier has asymmetric voltage feedback taking advantage of the -15V negative rail, as in the Broskie 6SN7 electrostatic amp (http://www.tubecad.com/november99/page9.html). This maintains the bias conditions as well as providing distortion-cancelling feedback in the conventional sense.
 

 
Simulation shows plenty of bandwidth and 0.08% THD at 250Vpk-pk differential, and although I have been stuck with using the Supertex DN2540 and hacked-about IXCP10M90 spice models, I think the sim. is probably not unrepresentative.
 

The amplifier is AC coupled, hence the huge Wima polyester capacitors, but these are recommended as being distortion-free by Doug Self, and I see Broskie is not averse to using AC coupling in his 6SN7 electrostatic amplifier
 
The power supply uses just 2 transformers driven from a 12V AC wallwart, a 240~12V and a 240~15+15V. Roughly 680V is derived by using a conventional voltage-doubler.
 
It's probably best not to say any more about the amplifier until I actually get one working. Perhaps that should be if, rather than until...
 
w

 

[jcx]

for coupling to the ES stators I'd try NP0/C0G ceramic - say 2x 22nF 630 Vdc in series - better dielectric properties than Mylar and less chance of polarization induced piezo behavior - especially compared to the "squashed" oval wound film types which have varying tension, can "sing"
 
for the IXTP01N100D I use the LTspice triode symbol and have a hand fitted gm subcircuit with fixed internal C, R0 from datasheet >25 Vds, ~10 mA Id curves
 

 
 
 
 
hm - when did we lose file attachments?

 

[chinsettawong]

Hi Wakibaki,
 
That ring looks really nice.  Have you measured the thickness of the wire mess that you're going to use?  What kind of glue will you be using to glue the wire and the ring together?  Epoxy?  

 
I have really very little knowledge about electronics, so I can't give any comments on your amp design.  I'm just really looking forward to seeing your finished amp.  Good luck.
 
By the way, my name is Wachara.  Chinsettawong is my surname.  Thai people call each other by their first name.  

 
Wachara C.

 

[spritzer]

Same thing here but in Iceland we don't have surnames in that sense.  It makes checking into hotels very odd when nobody has the same last name... 

 

[chinsettawong]

Quote:

Same thing here but in Iceland we don't have surnames in that sense.  It makes checking into hotels very odd when nobody has the same last name... 

 
So there is no no family name in Iceland?  Then how can you tell that people are from the same family?
 
 
I have just made two more pairs of diaphragm using the same technique I discussed before, and they turned out pretty good.  When you have enough tension on the diaphragm, if you press the headphone cup against your ear, you'll hear the noise of diaphragm collapsing to the stator and bouncing right back.  It's similar sound to that of the famous Stax's fart. 

  If not enough tension, then the diaphragm would be stick to the stator.
 
Wachara C.

 

[wakibaki]

OK Wachara
 
I have 2 sheets of stainless mesh, one is specified as 25 micron, the other is 30 micron as per this advert:- http://www.ebay.co.uk/itm/Super-Fine-400-E-Cig-Stainless-Steel-Mesh-0-3mm-A4-210-x-300mm-Bargain-/151102693460?pt=UK_Crafts_Other_Crafts_EH&hash=item232e6c2054.
 
I don't know whether this is the spacing between the wires or the thickness of the wires, but they are like a fine cloth. Quite heavy in the hand but feeling like silk. I think they will be pretty much transparent to audio. I need to pretension them before attaching them to the stator.
 
Of course there are many types of mesh available, some of which are stiff and self-supporting, and I could easily obtain a different type.
 
I was thinking of attaching the mesh to the copper using carbon conductive glue, like this:- http://www.ebay.co.uk/itm/Conductive-Wire-Glue-Paint-NO-Soldering-Iron-Gun-Solder-/320599571856?pt=UK_Home_Garden_PowerTools_SM&hash=item4aa5393d90 which I have already. I bought it to see if I could use it for making vias in doublesided PCBs but it didn't work.
 
Thanks for the suggestions re. the capacitors jcx. 
 
I've already bought these 0.1u 1250/600V Wima Polypropylene caps http://www.maplin.co.uk/polypropolene-capacitors-with-metal-foil-98162 since I have 1M resistors loading the amplifier outputs (on the schematic). I felt that the 1M res would be necessary to provide a defined ground return path for the bias voltage. I could of course change these to a higher resistance. I bought the caps on the basis of the tests in Doug Self's 'Small Signal Design' book.
 
I do have some 0.1u 1kV ceramic disk caps on order, but these are intended for use as PSU decoupling for the amplifier proper, and they are not NP0/C0G.
 
Broskie did complain about his output caps 'singing' which forced him to make a change. I wasn't aware, however, that NP0/C0G types were available in nanofarad ranges. I will probably proceed with the build using the caps I have in hand, but if I order the boards from a manufacturer rather than milling them myself, I will be sure to use a dual footprint to enable a change if it turns out to be necessary.
 
Fred

 

[spritzer]

Quote:

 
So there is no no family name in Iceland?  Then how can you tell that people are from the same family?

 
Our last names are the first name of our fathers with son (son) or dóttir (daughter) added to the back.  There are some family names but they have always been frowned on. 
 
As for the family, we have been isolated on this rock for 1000 years so pretty much everybody is related to some extent...

 

[chinsettawong]

Quote:

 
Our last names are the first name of our fathers with son (son) or dóttir (daughter) added to the back.  There are some family names but they have always been frowned on. 
 
As for the family, we have been isolated on this rock for 1000 years so pretty much everybody is related to some extent...

 
That's so interesting.  I hope to visit Iceland someday.  

 

[chinsettawong]

Quote:

OK Wachara
 
I have 2 sheets of stainless mesh, one is specified as 25 micron, the other is 30 micron as per this advert:- http://www.ebay.co.uk/itm/Super-Fine-400-E-Cig-Stainless-Steel-Mesh-0-3mm-A4-210-x-300mm-Bargain-/151102693460?pt=UK_Crafts_Other_Crafts_EH&hash=item232e6c2054.
 
I don't know whether this is the spacing between the wires or the thickness of the wires, but they are like a fine cloth. Quite heavy in the hand but feeling like silk. I think they will be pretty much transparent to audio. I need to pretension them before attaching them to the stator.
 
Of course there are many types of mesh available, some of which are stiff and self-supporting, and I could easily obtain a different type.
 
I was thinking of attaching the mesh to the copper using carbon conductive glue, like this:- http://www.ebay.co.uk/itm/Conductive-Wire-Glue-Paint-NO-Soldering-Iron-Gun-Solder-/320599571856?pt=UK_Home_Garden_PowerTools_SM&hash=item4aa5393d90 which I have already. I bought it to see if I could use it for making vias in doublesided PCBs but it didn't work.
 
Thanks for the suggestions re. the capacitors jcx. 
 
I've already bought these 0.1u 1250/600V Wima Polypropylene caps http://www.maplin.co.uk/polypropolene-capacitors-with-metal-foil-98162 since I have 1M resistors loading the amplifier outputs (on the schematic). I felt that the 1M res would be necessary to provide a defined ground return path for the bias voltage. I could of course change these to a higher resistance. I bought the caps on the basis of the tests in Doug Self's 'Small Signal Design' book.
 
I do have some 0.1u 1kV ceramic disk caps on order, but these are intended for use as PSU decoupling for the amplifier proper, and they are not NP0/C0G.
 
Broskie did complain about his output caps 'singing' which forced him to make a change. I wasn't aware, however, that NP0/C0G types were available in nanofarad ranges. I will probably proceed with the build using the caps I have in hand, but if I order the boards from a manufacturer rather than milling them myself, I will be sure to use a dual footprint to enable a change if it turns out to be necessary.
 
Fred

 
Hi Fred,
 
Looking at the picture, I think the mesh could be too fine.  But since you have it already, please try it.  

 
By the way, if you think about it, the surface of wire mesh stator can't be absolutely flat due to the weavings of the wires.  The unevenness and irregularity of the electrostatic field might somewhat influent the sound.  Do you think that if we can find a piece of thin copper sheet and perforate it, won't it make better stators?  I have been thinking about this for a while and I have got a sheet of 0.1 mm copper on hand.  If time allows, I might just try it out.  

  What hole size do you think is good?  I wonder what the hole size is on SR009.
 
Wachara C.

 

[wakibaki]

I think some of the Stax designs use wire mesh. Perhaps somebody could tell us the wire gauge and spacing. I'm thinking Stax must have done some research, and they should know what is best.
 
I have some copper and brass sheet in various thicknesses, from ~0.1mm to ~0.5mm, I use it for making solder stencils for SMT soldering in a toaster oven. I use toner transfer and etch it to get the holes.
 
I will certainly be trying a few different stator types when I get better at programming the CNC machine. I think I might make the first one PCB with holes like the many you have done so I can get started with something that other people have tried first, so I can get some advice if I have problems. The spider was just my first try at converting a dxf file.
 
Fred

 

[arnaud]

I may be missing something but I don't see how a fin wire mesh could yield a proper electrode once mounted on a cart.

In particular, I think you need a minimum of rigidity for the stators (at least some orders of magnitude stiffer / heavier than the film you're driving) so as not to cause distortion from stators vibration.

I believe it's a reason why some this stuff remains a bit of an art because all the design parameters compete with each other. You want an acoustically transparent stator but you need it rigid as well (that was a driver for the redesign if the electrode for the sr009). You want a fin diaphragm to more easily damp it and increase bandwidth but it hits back at you with additional resonances to deal with or fall at frequencies where they're least welcome.

 

[spritzer]

The SR-X Mk1 and the SR-Omega (maybe SR-2 but it has never been confirmed as if officially doesn't exist) used very tightly wound copper mesh for the stators.