Fresh ESL Research

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
For Bear

I don't think you can make a transformer that actually works full range and results in a flat response without other stuff that reduces the sensitivity and/or flattens the response.

OK you have piqued my curiosity what is the patent number so I can find it?

About the transformer I'm currently reading the RDH and am learning a lot. Still foggy on some of the concepts but when my noodle gets aound it enough to ask some good questions I'll be a askin.

Mark
 
Jim C. Strickland Patent #4,323,736 (U.S.)

If you are referring to the Acoustat biformer patent this is it, Jim C. Strickland Patent #4,323,736 (U.S.). you might just catch a copy at the ESL circuit. I think that Bear is right about building a true wide band transformer. I also think that it is a wasted project as dynamic driver are easier to use and cheaper to build for bass purposes. I do believe that hybrids are the way to go. Good luck and have fun. Hope this is what you wanted. Regards Moray James.
 
Multiple cell sandwich

TO Moray:

I think you mentioned an idea in the can you get high SPL ESL's. It is roughly the same as noted in Roger Sanders ESL book about stacked cells to increase output. It would be intriging if someone has a spreadsheet program that could spit out the max SPL for a given cell area and stator spacing plus the stator polarising voltage. Wagner has most of the math down. But I must admit that my math skills are dimmly limited to the basics. Calculus is a distant foggy memory. If such a sandwich panel could be fabricated the cell could be connected in series to lower the total capacitance. And maybe even horn coupled! Now there is the SPL answer!:devilr: :devilr: :devilr: :devilr:

THe actual design could maybe have common stators for every second cell to simplify things a bit. Such a sandwich could get a bit of air moving if there were enough layers.

Speaking about fresh research. Anybody have any idea of the approximate positive/negative movement of a real esl assembly? I'm guessing that it could be derived from the SPL generated from a known power input and understanding the amount of driven surface area. That would allow you to get a rough volume displacement. From that point knowing how much pressurized air makes how much sound volume could lead to a plus minus travel measurement on the cell membrane.

MArk

P.S. Moray thanks for the tip on the patent number.
 
Calling for Moray

I just picked up some 30 guage wire on enormous rolls. They have 20,000 feet on each. The ESL I'm planning on will need tons of it if I go with the .03" spacing. So my question is what great method did you come up with to evenly space the wires and fix them on to the grid. I understand the idea of using a threaded rod to create an even space between the wires. The fixing in place part leads me to think of some epoxy or something to set the ends of the loops into.

As a side question did you experience any rattling of the wires against the grids. Or did you set them a discreet distance from the grids so that they never really touched them?

I next have to get the acrylic grids. Will find a source that is economical sooner or later. Last will be the membrane.

Mark
 
Moray,

The idea that the mylar backing for recording tape is heat shrinkable seems entirely counter intuitive since you want the most dimensionally stable material you can find for that application - are you 100% certain about this??

Mark,

The wires must be held 100% solidly firm and unable to move whatsoever. That usually requires both tensioning and some sort of glue or embedding.

Look at the JanZen articles/patents for one way to handle wires.
There was a nice article in TAA some years back on a way to "clone" Acoustat panels, called "Amber", iirc. That's another way to go.

_-_-bear
 
Oh,

and imho, the goal is full range operation or as nearly so as you can manage.

...I don't like xovers in the 400-2kHz range if they can be avoided, and especially if they involve moving from one type of driver to another or between drivers made from different materials.

If possible it would be nice to go from ~80Hz up... should be do-able using an Acoustat type of approach and using a slightly lighter membrane/coating to get slightly better HF response, at the expense of making it all the way to 35-40Hz. - which the Acoustats do just fine.

just so everyone is on the same page - you can make a transformer with "wideband" frequency response just fine, but the ESL cells don't care! They won't and don't follow the response of the xfmr, they're going to be governed by the impedance match vs. the capacitive reactance curve no matter what you do.

_-_-bear
 
Been off linr a while

Mark: I usen a threaded rod to set the spaceing of the wire. The panels I have now are 20 wires per inch and I plan to move up to 32 wires per inch. The wires are bonded down to the grid frame with adhesive made from weld on #4 solvent mixed with some base plastic. I plan on trying to make the bond useing only the #4 solvent.

Bear: the idea behind the shrink film (must be biaxially oriented) is so the film can be coated and then shrunk to incerase the density of the coating. Best regards Moray James.
 
Great Minds Think Alike ?

Thanks for the tip Moray. Missed your posts!

The 32nd of an inch spacing is about the maximum that can be pulled off. My great master plan is to make a apeaker with three 24" x 36" panels each. It should give me a decent amount of surface area. About 18'^2 each side. I'm going to punch this into the Linkwitz dipole calculator and see what and where the low end roll off will start. Below the low frequency suckout I will be transfering the real work to horns. Still a clean sound.

So in the fastening of the wires to the eggcrate three or four adhesion points seems to make sense. Less rattling.

The reading up on the transformer design stuff is getting along to the point that I am actually understanding the trade offs involved. One thing that kind of haunts me is if I could use a better insulator inbetween the layers. I have tons of teflon tape. If I understand the idea behind the lowering of capacitance is that a better thinner insulator would work out as a gain in lowering the overall capacitance of the transformer. Open to correction on this one for sure.

The Strickland patent seems to imply that there is really to transformers that are crossed over passively. It really is a good idea. Seems to kill all the stones with one bird. Keep the lowend as a big ol X-former that can't pass the high end. And the mid to high end with a properly tailored and tweeked X-former. Anybody have any real life experience with this transformer? Or even any ideas. This will probably be the route that I try to go with. I have some immense I E cores sitting around. And to go the high frequency route I may have to find something better. I read with interest the article about winding up old microwave X-formers. That may be another avenue of adventure to pursue.

Mark
 
I puncha Dah numbers and this is what I got

Her is the big number crunch. Dipole driver A is the 18'^2 single panel. And Dipole driver B is the two together. There will be room gain as the frequency goes below about 60hz of around 12db/octave. So all in all this is fairly respectable. That's if I used the correct info :)

Mark
 

Attachments

  • spl calc.gif
    spl calc.gif
    52.2 KB · Views: 315
SY said:
Amber. Yes, that's it. Fine article.

Acoustat used a machined "comb" to space the wires. They were glued to the eggcrate using polystyrene dissolved in a thinner (probably MEK or something like that). The polystyrene doesn't really stick to the wire so much as encapsulate it.

If I have it right, the "comb" was on either end of the plastic grids... not machined into the grids themselves, as was JanZen's molded parts...

Along the grids the Acoustats' wires are glued with what is likely polystyrene & solvent dope... since the grids are polystyrene.

_-_-bear :Pawprint:
 
Re: Great Minds Think Alike ?

mwmkravchenko said:
Thanks for the tip Moray. Missed your posts!

<snip>

The Strickland patent seems to imply that there is really to transformers that are crossed over passively. It really is a good idea. Seems to kill all the stones with one bird. Keep the lowend as a big ol X-former that can't pass the high end. And the mid to high end with a properly tailored and tweeked X-former. Anybody have any real life experience with this transformer? Or even any ideas. This will probably be the route that I try to go with. I have some immense I E cores sitting around. And to go the high frequency route I may have to find something better. I read with interest the article about winding up old microwave X-formers. That may be another avenue of adventure to pursue.

Mark

Yeah!

I have Acoustats, they do work.

If there is any point to question it is that there does seem to be a slight delay between the smaller and larger transformer which can be seen if you look very closely, but isn't likely to be particularly audible. Make sure ur caps are high quality on the input side especially... hmmm... never tried biamping it!

The HF transformer actually covers fairly low, like down to 200 Hz or so - it doesn't look like a classical xover at all. And the HF xfrmr is about a <3.5" EI core unit - not all that big at all. The LF xfmr is beefier and from recollection it's got a core about 6" EI...

_-_-bear :Pawprint:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.