|Planars & Exotics ESL's, planars, and alternative technologies|
Please consider donating to help us continue to serve you.
Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
||Thread Tools||Search this Thread|
|18th February 2013, 04:46 PM||#21|
Join Date: Feb 2008
Syborg, in your second picture I noticed a sharp point in one of the holes.
It is areas like this that can and will wreck havoc on your newly built panel.
Especially when you start to crank it up.
You should consider removing this sharp point and then add some more coats to the coating.
Especially along the edges .
Even though you may have passed an arc test with your bias supply you will be applying well over 10KV to more than 20KV p-p across the stators on music peaks.
The high frequency ones are the ones most prone to cause punch thru of the stator coatings.
I was pushing 25KV P-P when my panel lit up like an arc welder on the side and finally on the edge on the inside from the very same bad spot that had been repaired before.
Burning the stator very badly.
Once it burns it leaves channels of carbon left behind buried in the coating and this is like little rivers that are just like exposed wires at a high voltage.
Very difficult to real seal once burned and usually requires completely stripping it down to the bare metal to reseal it as Charlie had mention.
Last edited by geraldfryjr; 18th February 2013 at 04:56 PM.
|8th March 2013, 12:06 AM||#22|
Join Date: Jan 2010
Hi. Very interesting thread and I'm keen to experiment with ESLs, and as I am an electrical engineer by trade, I know of the dangers of high voltages and proper insulation. I have a few questions if anyone can answer.
1) If I were to build an ESL of about 4ft square surface area, could my LM4780 Gainclone amp drive it, It has +/- 25 volt rails?
2) At work we use perforated stainless sheet which has 6mm holes in it. It is roughly 1.5mm thick. It is actually used for sewage screening, but looks like it could be used for the stators, any thoughts?
3) What is the best way of insulating the stators? I heard that powder coating isn't usually that successful as the holes in the sheet can get pretty small. I have used Plasti-kote before for various DIY stuff, would this be any good for stator covering?
Plasti-kote, Welcome to the home of spray paint
4) For the DC bias, I have seen mains supply versions using a 6v to 230v transformer and then a trippler circuit and I have also seen 15V DC supply circuits which use a 555 timer to create a SMPS to get the voltage. Which is the best method? The mains supply one seems simpler and I'm not a fan of SMPS in audio.
5) My current set-up is a DIY LM4780 amp, DIY DAC and DIY full range speakers (CSS EL70) in the Lake District Series, Derwent cabinets (Designs from frugal-phile website) along with some DIY super tweeters. I am very happy with the sound produced from these, so is it worth the effort to build some ESLs?
6) If I build some ESLs, I will probably use a 8 inch or 10 inch woofer for the low-end. Would a crossover point of about 250Hz be OK and would a 2nd order crossover be OK?
7) Could passive crossovers be used before the step-up transformer? As I don't really want to build active crossovers, as I am happy with my amp and don't really want to bi-amp the hybrid ESLs/woofer.
Thanks in advance for any help
Another project of mine: (a very big one at that!)
|8th March 2013, 01:51 AM||#23|
Join Date: Feb 2008
I have had good results from using spray paint and those tests can be found here,
High strength Dielectric Coatings, fact or fiction
A Segmented Stator Desktop ESL
A Segmented Stator Desktop ESL
Do use a primer and stay away from any pigmented paints as a top coat.
A base coat is fine for color then just use clear after that.
I have gotten between 1700v to 2100v per mil before breakdown occurs using typical off of the shelf types.
I think Stainless would be a nice material to work with as long as the paint will stick to it well and as long as it is flat with no wraps or ripples.
I think the biggest issue you will have is getting rid of all of the sharp edge as it is some pretty hard stuff.
6mm holes are large and I wouldn't use anything larger.
Many have had good results using that size holes, But personally I have not tried it yet.
The Bias supply can be made from any method you chose fit and are willing to go through to get it working should you choose to do a switching type of some sorts.
All that is required that you have a high voltage at a low current, Typically from about 2.5Kv to 6.5Kv.
Here are a couple of links for Bias supply's.
One is the one I designed and is variable although it is a bit of an undertaking to build.
And the other one found on Jazzman's ESL (Charlie) site.
Does anyone have schematics of a varible HV power supply
Jazzman's DIY Electrostatic Loudspeaker Page: The Electronics Package
Another method that works well to make a variable supply is to use a small chipamp fed from a opamp sine wave oscillator with a volume control feeding a low voltage power transformer in reverse into the voltage multiplier stage.
Anything above that 7kv and you start to run in to dielectric breakdown issues especially if you are using a smaller D/S of about .0625".
Your chipamp may power it okay but don't expect a lot out of it.
The impedance of these speaker systems drop to and/well below 2ohms at the higher frequency's depending on the transformer step up ratio and the transformers themselves.
If you follow the guidelines shown in some of these threads you will have a decent working system.
I would consider using I paralleled chipamp setup.
I have been planning on building a Pair of PA200 or PA250's out of the 5 Lm4780's I do have.
I was thinking of two PA200's and use the 5th one to power a pair of my 8" subs for my next bigger model.
But I do have some very big amps at my disposal so it is last on my list of things to do.
You might want to consider going with a 12" woofer for the lower end as even a 10" can run out of steam quickly depending on your listening style.
It all depends on what you have already available.
I have a bunch of 8's laying around and in some sim's that I have done just 2 or 3 of per side can put out some very decent SPL's.
I used just one of them when I first powered up my little panel for the first time in years.
It was loud enough to make it difficult to hold a conversation and it was being powered off of a cheap CD player type all one one box.
There is a picture of that setup in one of these threads.
A 1' X 4' panel can really get up there in SPL's if it is driven properly.
A passive crossover can be designed, But it will take a lot of testing to get it right and it is much cheaper to just get an active crossover or build one out of a few opamps.
Not to mention the cost of the components to build a quality passive design.
Do consider going the active route!!
If you want to get down to 250Hz then you will need to go with a large D/S of say at least .075" to .110" or so.
Using a large D/S means higher voltages and step-up ratio's to make up the difference in loss of efficiency.
Also the transformer is an issue here as well as the core saturation points of your typical power transformer will be about 350hz to 330hz for a 40Vrms signal into the 6v winding.
A 100 watt core may fair a little better in this range but even the one I have been testing can only get to 300hz max for a 40Vrms input.
If you don't go above 20Vrms to 30Vrms it may be possible to get good results down to 250Hz.
The cost of getting to a lower crossover frequency is efficiency, due to core saturation.
The step-up ratio sets the maximum SPL at this frequency and if you use a higher step-up ratio then you lower your impedance that your amplifier sees.
So therefore you have to add more cores.
Unless you wind your own transformer.
Some have used as many as 8 per speaker to get a lower frequency response.
I think 4 would do it and is what I use for my little panels.
I will be going over this soon in another thread as I am going over some measuring technique issues this very moment.
So, there are a few trade off's to consider and it just takes a little planning is all.
Most the that trial and error stuff has already been done in these threads.
Last edited by geraldfryjr; 8th March 2013 at 02:14 AM.
|8th March 2013, 07:21 AM||#24|
I do not know Plasti-kote but I can say that you should prepare for a difficult job if you want to make perforated sheets insulated well(>>8 kV or so) over it's entire area. You will need to round sharp edges of holes and spray many, many layers of paint. If you want to use canned stuff then prepare to buy loads of them. The problem with perf. stators are edges of each hole.
a) Difficult to tune in and do any equalization
b) Increased distortion of step-up transformer because source impedance is increased at lower frequencies
c) Eats up efficiency and so on
My personal view is that wire stators are by far superior to perf. metal for DIY, unless you are targeting curved designs like MLs. I suggest you read the following thread for introduction(which is quite old):
Wire stator gluing method
|8th March 2013, 08:11 AM||#25|
Join Date: Feb 2008
Yes, I will have to agree with your disagreement!! He,he,he
I should have stated 250hz and below!
But typically the next size in thicknesses for materials that are found in the .050" to .0625" range is .080" to .093".
Depending on your method of construction it can be made adjustable using what is available in your area.
Powder coating has a few issues.
I have used it but you will spend more time getting your powdercoater to do them correctly than you will with just painting them.
Powder coating is more brittle than paint and it can develop micro cracks in the coating and this will lead to failure of the panel.
I had to use two coats to make it work on some wire mesh and the coater refused to do a third coat for whatever reason, I never got one either.
It could have been the type of material that was used.
Seven years later I had to re-coat them again with some clear acrylic due to the cracks causing weak spots and they eventually failed and I burned them up.
But a lot of this was my fault from seeing how high I could push the voltages too.
It can be done yourself but it will take some time to get it right, with lots of experimenting for sure.
I must say I do prefer the wire method as well and it can be a lot cheaper on materials if you are resourceful.
Have you thought about segmenting the panels into several smaller ones?
This is where the wire method is very nice because you can electrically segment the stator's
This will help to widen the horizontal dispersion of a wider panel of 12" should you not physically segment them in smaller sections.
And not to mention it will make for an easier load on the amplifier if it is electrically segmented as well.
This is were a larger D/S will play for a major importance as I did not have good luck with diaphragm stability on my 8" wide panels with a D/S of .0625 especially below 300Hz to 200Hz.
I ended up with 2mm to 2.5mm and just increased the bias voltage to make up loss in efficiency.
Diaphragm tension had a lot to do with this as well since I didn't have a mechanical stretcher and I am using .25 mil (6um) mylar.
Just I few more thoughts.
Last edited by geraldfryjr; 8th March 2013 at 08:13 AM.
|8th March 2013, 03:23 PM||#26|
ESL manufacturers I know of used bias voltages in the order of about 2kV/1mm. There are good reasons not to exceed this by much :
a) Air starts to ionize
b) Diaphragm sticking to stators
It's not easy to compensate significant increase of d/s spacing by only putting a lot more bias voltage because of limitations above. Remember doubling d/s spacing requires four times the voltage to get efficiency back.
To get 6um mylar stable on 8" panel with 2mm spacing requires either strong mechanical tensioning or adding more supports. One ESL manufacturer who is now out of here(Capaciti) has mentioned multiple times against using strong tension due to partial resonances of membrane.
I have found a quite interesting applet for simulating modes of a rectangular membrane :
Rectangular Membrane Applet
It is difficult to tell how many of these will actually show up in measured performance.
In terms of D/S spacing I think 2.5-3 mm is already enough for full range operation. I have been testing panels with only 2mm full range and it didn't strike stators with music. Test signals could do this if targeted at resonance with high levels.
Last edited by Bazukaz; 8th March 2013 at 03:26 PM.
|8th March 2013, 04:39 PM||#27|
Join Date: Feb 2008
Yes, I agree.
I have found that about 2mm is good for getting down to that range on both my 3.25" and 8" wide panels.
If one where to use the sticky tape method of construction then there are only A few choices of material to choose from and the .0625" thick stuff would be marginal for use at the lower frequency's.
I will be exploring this in greater detail soon as my new panels D/S is adjustable as was my last one and is set at around 1.85mm.
As little as a 15% to 30% increase of the D/S spacing can make your break your low frequency performance at such low frequency's due to not having enough room for excursion.
And the increase of bias voltage would still not be more than about 50% or so to make up the difference.
I have measured close to about a 6db increase of efficiency for every doubling of the bias voltage.
That is 12db more going from 3kv to 12kv but as I stated before, anything much above 7kv is impractical due insulation issues.
It was not at all easy to get to those high of voltages with out issues.
I did get DS clipping with my smaller 3.25" wide panel and had to increase the DS spacing to reach high SPL's at below 200HZ or so to .090" from about .072" total spacing.
And yes I took a hit in the efficiency and it was very noticeable.
Being that they have a naturally rising frequency response Diaphragm excursion is not an issue at higher frequency's.
I do have to admit that the voltages I have reported that I am using are true but are a bit on the extreme side of things.
A larger panel won't need such high voltages to perform well but when you want to get down below about 300Hz the transformer and the D/S start to become some things to consider in the design.
|8th March 2013, 06:47 PM||#29|
Join Date: Apr 2011
With the difficulty Antec has in producing phase identifiable transformers, one must wonder if their recent factory fire (as I have heard) was due to electrical problems in their factory?
|8th March 2013, 07:46 PM||#30|
Join Date: Feb 2008
HA,ha,ha,ha, Interesting....I should call them and give them hell and say I told you so !!! He,he,he,he,he
For not listening to me when I called them on it!!!!
They don't use very thick insulation in between the winding's on their transformers either!!!
|Thread Tools||Search this Thread|
|Thread||Thread Starter||Forum||Replies||Last Post|
|high efficiency easy to drive esl||pforeman||Planars & Exotics||3||29th March 2010 08:51 AM|
|Efficiency of esl goes down withing 10 minutes||WrineX||Planars & Exotics||3||2nd November 2009 09:33 PM|
|Split an esl panel to two way panel||pwgtang||Multi-Way||1||13th August 2008 07:57 PM|
|ESL tweeter efficiency||blu_line||Planars & Exotics||10||23rd May 2007 01:45 PM|
|DIY super high efficiency/frequency ESL line array||thadman||Planars & Exotics||17||20th January 2007 09:42 AM|
|New To Site?||Need Help?|