diyAudio (
-   Planars & Exotics (
-   -   ESL Panel Efficiency - First Build (

SyBorg 17th February 2013 05:12 PM

ESL Panel Efficiency - First Build
6 Attachment(s)
I have started down the road of constructing my first ESL system. I used an insulating lacquer paint on the 12"x12" panels and they passed the "spark test" with no issues. Used double sided foam tape 1/16" as the spacer, 6um mylar tensioned to 1.5% using a bicycle tire stretcher. Using Antec toroidal transformers and a HV supply board I designed (multiple taps for varying HV supply voltages). Using Licron spray as the coating.

The whole thing went together over the course of a couple of weeks, thanks mostly to the loads of useful advice on this forum. I had some issues along the way with too much Licron, then overtensioning the mylar, then a couple wrinkles in the mylar. These things happen, no big deal, clean off the panels and try again. Finally got a panel that looked correct, had a resonance in the right range, and so it was time to fire it up.

When I run an audio signal through it, I barely get a peep out of it, even with driving the panel fairly hard (using a 200W amp). In a quiet room, the best I can do is the level of soft background music. What is there sounds great, but nowhere high enough for normal listening.

Any advice would be much appreciated.

CharlieM 17th February 2013 06:18 PM

Even though I think your panel is much too small for a speaker intended to fill a normal size living room with sound, it should still play quite loud if all the elements are functioning. I would first check to see if the bias supply is supplying charge to the diaphragm and then check for a leak path that might be pulling charge off the diaphragm. In one of your stator photos I think I see a fairly large blob of solder at the connection to the copper foil charge ring. If so, that blob of solder could be embedding itself deep enough into the foam spacer tape to short the charge to the stator when the two stators were mated together. That's a definite possibility!

Even with a 20 Mohm charging resistor at the output of the bias supply, it should still pass enough current to arc across a quarter inch or so gap ( to it's opposite lead).
Assuming the panel is receiving charge from the bias supply, it should not only play loud, but if you then pull the power chord supplying the bias supply, the panel should continue playing for some few minutes using only the charge accumulated on the diaphragm--- if the sound diminishes quickly after unplugging bias supply, that would indicate a leak path is pulling charge off the diaphragm. And if the diaphragm were leaking charge faster than than the bias supply can replenish the loss, then the panel would never play loud at all.

Another possibility is a reversed connection between transformers. If the connection polarities look good, I would still not rule out that one of the leads could have been reversed or mis-color-coded at the factory. As I recall, those Anteks do not come identified with polarity dots on their winding diagrams. And if one of the windings/leads were reversed, you would then only be getting 1/2 the normal voltage output... and very little sound from the speaker.

When I posted the winding polarities for those Antek AN-506's on my blog page, the only way I could confirm it was accurate was to get my techie friend Martin to come over with his dual trace oscilliscope and we compared the input to output winding polarities with an AC current applied. I know that those cheap Antek transformers give excellent high output sound if all connections are good and the diaphragm is receiving and holding charge.

CharlieM 17th February 2013 06:33 PM

Note that I just edited my post so if you already read it, please read it again now.

geraldfryjr 17th February 2013 08:07 PM

From the looks of your picture it shows that only one of the transformers LV (primay) is hooked up!!
You must have at least one LV(6V) winding from each core being supplied from your amplifier in order for it to work.

And yes as Charlie stated, If one core is out phase you will get nothing or very little no matter how high your amplifier is set at.
I have done this a few times myself.

I called Antek on this a year ago and it seems as they don't give a hoot about there data sheets and connection polarity's.
Infact they didn't even know what i was talking about!!!
Tisk,Tisk,tisk, Antek!!! He,he,he,he
You should still be getting some good sound and output of a panel that size.


jer :)

SyBorg 17th February 2013 08:18 PM

Thanks Charlie! I did not expect the 12x12 panel to fill the room. This was a proof of concept build before making the larger panels. I have had a set Martin Logan CLSIIs for about 30 years and the panels are about shot, and wanted to try my hand at building a replacement. Once you have had electrostats it is hard to listen to anything else.

I do get a good spark from the supply. I had tried several voltages with increasing sparking with voltage level and the HV probe on the o-scope shows the voltage to be in the right area. Right now I have the board wired to about 3.2kV. I get a slightly diminished sound level when I drop the level to about 2.6kV. If I go up to 4.5kV or higher I get arcing somewhere in the panel.

When I unplug the supply mains, the sound keeps playing for probably 20-30 seconds, far shore of the couple of minutes, meaning I have some leakage somewhere. I will check the solder connection to the charge ring first.

I'll also check the polarity of the Antec windings. I didn't think of that when I first wired it up.

SyBorg 17th February 2013 08:20 PM

Thanks Gerald,

Definitely all the primaries are hooked in parallel. The real question is if they are the correct polarity...

geraldfryjr 17th February 2013 08:23 PM

You can simply just swap the polarity of one of the primary winding's rather than take the whole thing apart.
Unless the polarity's of the red and black wires on each core are screwed up.

Who knows with them!!

On the AS-1206 that I have been testing I found that the Green wire (LV) is the same polarity as the Black wire (HV).

I hope that helps but again you never know !!! :/

jer :)

SyBorg 17th February 2013 08:27 PM

I'll probably pull the whole thing apart anyway. That will let me check the primary and secondary windings for polarity against the lead color and clean up the wiring. This was a quick-and-dirty job - I was too anxious to hear the panels.

tyu 17th February 2013 09:26 PM

gOO jOB...I cant wate for your CLS panel work...I have a pr CLS here now.... i have reworket the ML panels for years ...10pr now there great if you can get them a part?? Just take your time....the largest panles i have got into are the 16"x 48"....but i have.. an you can get a EZe 3-4db more output than new panles from ML... buy recoting an runing new the right bias feeds..bias feeds are the Key...ML wonted to sale panels in the old days ...not so much now.......
thanks for posting have fun....good luck

SyBorg 17th February 2013 10:14 PM

Charlie, I filed down the solder joint on the charge ring. It was not as bad as it looked in the picture, but it did make a difference. After filing, I put a piece of polyester tape over it for extra insulation. After re-assembling the panel the volume output was not any better, but, it did keep playing for about 4 1/2 minutes instead of 30 seconds or so.

Charlie and Gerald, I ripped apart the transformer assemblies. The green wire of the low voltage side was in phase with the green low voltage wire and both red wires of the high voltage side on one transformer and in phase with the green low voltage wire and black high voltage wires on the other transformer. ( Go Antec! ) After rewiring so that the phases lined up, I powered up the panel and got lots of output.

Lesson learned - check the phases of all Antec transformers.

Thanks for everything! Now it is time to build the full size panels...

All times are GMT. The time now is 04:31 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2