Give Oskar a Wedgie and Heil Smooth Out and Won't Be Such a Dip!

[toddalin]

The dip in the Heil that lies in the 5-6K range has now been well documented. I use the Heil in a 3-way system crossed over at ~2,500 Hz and the dip is readily obvious as shown in the included picture. The mic is about 3 feet from the tweeter in direct alighnment.
http://www.audioheritage.org/photopost/data//500/medium/DSC_0013.JPG
Certainly this could be handled using eq if one desired. But I run various sets of speakers and the eq for one set would differ from another, and who could remember all of these or want to deal with them when simpling selecting a different pair of speakers? The obvious answer is that we want every set of speakers to be as flat and smooth in the room as possible without the need for electronic modification.

Then perhaps the dip could be handled with the crossovers. I've spent countless hours in various configurations, but anything of value has a volume penalty and when you are trying to match a JBL 2241H 18" woofer, you need as much volume as you can muster.

But what of a "mechanical" solution. I've done extensive work with the JBL 075/2402 placing a slant plate assembly in front of the tweeter. This not only increased the horizontal dispersion, it took the peak (and "in your face") away routing this energy to the sides. It also increased the low end of the tweeter, undoubtedly through the effect of "loading" it.

So why couldn't something like this work with the Heil?

Well, it just so happens that it can! I initially held an engineering scale up to the tweeter and there was a flattening of the frequency response, in a very positive way!

I decided to make a wedge to the size and shape of the ruler (triangular and 1" on each side) by laying two strips of 1" wide brass strip next to each other on a piece of masking tape and folded it to 60 degrees by lying it between two scales that were taped together. I put superglue along the seam and added a piece of lucite to the inside to hold the shape. This was glued to a piece of 3/8" thick acrylic as a stand and to let the wedge protrude ~0.3" into the "void" to test various distances from the Heil diaphragm. The "void" (plastic ledge between the very front and the diaphragm) is 5/16". A piece of foam was glued to the bottom of the acrylic to take up any slop (not shown).

http://www.audioheritage.org/photopost/data//500/medium/DSC_000312.JPG

http://www.audioheritage.org/photopost/data//500/medium/DSC_00065.JPG

This piece sits in front of the opening. My testing has shown that I am getting the smoothest response with the point of the wedge right at the front edge of the plastic housing.
http://www.audioheritage.org/photopost/data//500/medium/DSC_0010.JPG

http://www.audioheritage.org/photopost/data//500/medium/DSC_00092.JPG

And of course the proof of the pudding is in the eating.

Without moving the mic from the past photos, this shows the modified frequency response with the wedge in place. Not only does it remove most of the dip, it flattens the response over the entire range.
http://www.audioheritage.org/photopost/data//500/medium/DSC_00112.JPG

So, right after I finished testing/photograghing, this arrived. It is pretty clear to me that Oskar is in for a wedgie.

http://www.audioheritage.org/photopost/data//500/medium/DSC_0003_2_.JPG

[SEAWOLF97]

I use the Heil in a 3-way system

Every ESS AMT system that I've seen is a 2 way. I wonder if that has a bearing
on things. :dont-know:

[toddalin]

I got over to the workshop today and used the laser to cut the bases for the wedgies. I was suprised that I got it dead on the first time. I also used the chop saw to cut the lucite triangle and it came out better than I could have hoped for.

The JBL 077 won't have anything on these bad boys!

http://www.audioheritage.org/photopost/data//500/medium/DSC_00048.JPG

While I was there I also used the laser to cut the grill cloth frames for the 2251Js. The cloth/screen will be sandwiched between the two pieces and the four holes in the "lower" piece locate magnets to attach to the screws that hold the woofer to the baffleboard.

http://www.audioheritage.org/photopost/data//500/medium/DSC_00072.JPG
http://www.audioheritage.org/photopost/data//500/medium/DSC_00072.JPG

[toddalin]

OOOH that's a tight squeeze!

http://www.audioheritage.org/photopost/data//500/medium/DSC_0004_2_.JPG
http://www.audioheritage.org/photopost/data//500/medium/DSC_0006_2_.JPG

[10 Watt Street]

How about a wedge on the back side?

[toddalin]

Well that would make sense as that's where wedgies are typically given. ;)

But I have my terminals back there and they would be in the way of the base. Also, when I held the brass wedge back there, I didn't get the flattening of the curve.

http://www.audioheritage.org/photopost/data//500/medium/DSC_00593.JPG

[10 Watt Street]

I would think you would lower distortion with wedge front and back.

[toddalin]

Just spent the last hour and a half listening to Joni Mitchell and Diana Krall. The wedges make all the difference! Inteligability went up and the ladies no longer have any cloud in their voices.

The wedges do cost about 1.5 dB overall, mostly from the range shown in the pics. I had padded the Heil by an additional 5 ohms beyond the 5 ohms to achieve ~8 ohms, which works out to be ~1.5 dB. Removal of this additional resistor brought the overall balance back except that now there is a lot more where the dip was and into the crossover region where is was needed.

Very happy with the results for a $12 investment in parts. :applaud:

[badman]

Hi Todd:

I strongly suggest trying the wings- they add a lot of support to the lower frequency range of the driver, and smooth response. I'd also suggest you figure out a way to round the back end of the triangle- you're adding a diffraction slot, which often smooths/spreads response, but also has artifacts associated with it. A teardrop profile would suffer from this much less than the triangle.

Continue playing sir! If you get REALLY into them, you can try the various diaphragm types, they do respond differently. Some old ones are polyethylene and aluminum- low durability, but high self-damping, vs. the later kapton (IIRC) phragms.

[toddalin]

Hi Todd:

I strongly suggest trying the wings- they add a lot of support to the lower frequency range of the driver, and smooth response. I'd also suggest you figure out a way to round the back end of the triangle- you're adding a diffraction slot, which often smooths/spreads response, but also has artifacts associated with it. A teardrop profile would suffer from this much less than the triangle.

Continue playing sir! If you get REALLY into them, you can try the various diaphragm types, they do respond differently. Some old ones are polyethylene and aluminum- low durability, but high self-damping, vs. the later kapton (IIRC) phragms.

Thanks Jeff.

How are the new "digs" working out?

I have designed a base that will allow me to use a pair of laser-cut panels rather than the wedge. This way I can try different lengths, angles, and even leaving the center open. I can also round the edge where the two come together (rather than a sharp point) with a piece of tape/bead of glue at the junction. I'll try to get to the workshop next week to make the cuts.

Take care.

[toddalin]

A video of the "Ethyl Mermans" (JBL 2241H/JBL 2251J/Heil AMT with Clear Solutions Diffuser Wedge/"Home Brew" Crossover) doing their thing with a male voice. The Clear Solutions Heil Wedge really allows the Heil to intergrate with the JBL2251J presenting a nice clear midrange (and vocals) without undue high end fatigue.

This is in "Pure Direct" with no subs, eq, or digital processing involved. Enjoy.


https://youtu.be/cO8T6R2WhTE

[SEAWOLF97]

A video of the "Ethyl Mermans" (JBL 2241H/JBL 2251J/Heil AMT with Clear Solutions Diffuser Wedge/"Home Brew" Crossover) doing their thing with a male voice. The Clear Solutions Heil Wedge really allows the Heil to intergrate with the JBL2251J presenting a nice clear midrange (and vocals) without undue high end fatigue.

This is in "Pure Direct" with no subs, eq, or digital processing involved. Enjoy.


https://youtu.be/cO8T6R2WhTE

on my PC speakers (120Ti's), your vid sounded very good. thx.

For some years I've been posting here about the Heil AMT , but largely ignored (except the forum elder who belittled them). They are truly fine drivers.

I didn't know about the FET connection.

https://en.wikipedia.org/wiki/Oskar_Heil

https://www.precide.ch/eng/eheil/eheil.htm


Dr. Oskar Heil, noted physisist and inventor of the Field Effect Transistor, began his research into loudspeaker design, not with abstract theory of how a loudspeaker should work, but with a study of the peculiarities of the human listening aparatus. The result of this intensive year-long research program led to his discovery of the principle on which the Oskar Air Velocity Transformer is based. By applying this principle to the design of a loudspeaker diaphragm, he was able to achieve a revolutionary breakthrough in solving the fundamental problems of diaphragm mass, inertia and self resonance.

[badman]

Thanks Jeff.

How are the new "digs" working out?

I have designed a base that will allow me to use a pair of laser-cut panels rather than the wedge. This way I can try different lengths, angles, and even leaving the center open. I can also round the edge where the two come together (rather than a sharp point) with a piece of tape/bead of glue at the junction. I'll try to get to the workshop next week to make the cuts.

Take care.

Cool, I look forward to your experimentation. Regarding measurement- I use and like the Dayton Omnimic setup (it's quite straightforward, requires minimal equipment, and is powerful enough for almost anything). RTA has some pretty significant limitations, it's great for balancing tonality when a few measurement locations are used but not a great optimization/problem solving tool. Horns/acoustic lenses/diffraction have a high degree of specificity and need better resolution to truly optimize.

It's good here, just got back from a family cruise with a couple ports in Mexico which really drove home how goddamned cold it is. I'm hoping to grow a better motivator unit this week, get the basement set up for guests better, and invite some of the mass audio crew to come visit in the next few weeks.

[toddalin]

Cool, I look forward to your experimentation.

I got to the workshop and cut a pair of bases and vanes to go with them. I made provisions to test the 60 degree angle as well as a 30 degree angle and even a center vane for if the outer vanes don't come all the way together. I also did vanes in 1, 1-1/4, 1-1/2, and 1-3/4" lengths. The slot that holds the vane at the base is 3 times longer than the tab on the vane, so the vanes can be adjusted forward/rearward to adjust the opening between them. This will allow for a lot of variation.

I did up two sets. Typically I will test with one speaker with the RTA, then apply that modification to both and do lots of listening.

http://www.audioheritage.org/photopost/data//500/medium/DSC_00074.JPG

These are 30 and 60 degree angles at 1",1-1/4", 1-1/2", and 1-3/4". Doing up two sets allows for a much quicker comparison of angles and lengths and the tape is applied front and rear so they pretty much hold their shapes and just plug in. We'll see how this shakes out before going to individual vanes.

PS, Wife hates the smell of the burnt acrylic so these need to air out a couple days.

http://www.audioheritage.org/photopost/data//500/medium/DSC_0006_2_1.JPG

[badman]

I got to the workshop and cut a pair of bases and vanes to go with them. I made provisions to test the 60 degree angle as well as a 30 degree angle and even a center vane for if the outer vanes don't come all the way together. I also did vanes in 1, 1-1/4, 1-1/2, and 1-3/4" lengths. The slot that holds the vane at the base is 3 times longer than the tab on the vane, so the vanes can be adjusted forward/rearward to adjust the opening between them. This will allow for a lot of variation.

I did up two sets. Typically I will test with one speaker with the RTA, then apply that modification to both and do lots of listening.

http://www.audioheritage.org/photopost/data//500/medium/DSC_00074.JPG

These are 30 and 60 degree angles at 1",1-1/4", 1-1/2", and 1-3/4". Doing up two sets allows for a much quicker comparison of angles and lengths and the tape is applied front and rear so they pretty much hold their shapes and just plug in. We'll see how this shakes out before going to individual vanes.

PS, Wife hates the smell of the burnt acrylic so these need to air out a couple days.

http://www.audioheritage.org/photopost/data//500/medium/DSC_0006_2_1.JPG

Nice! Cool experimentation! Looking forward to when you get a chance to play. I suspect one of the narrower angles will sound better, but that's to be seen. With all that variance available to you all the more reason to get a higher res measurement system. One of my earliest experiments with drivers was actually one of my more successful- using a grill to locate absorbent material in front of the center of a 12", as part of the low pass filter action. Worked quite well, but could have surely been optimized further. Some years later armed with high res measurements I went about making a constant directivity supertweeter using a modified soft dome, horn, and absorbent materials. I've never seen better top-octave polars. If I remember to logon I'll grab the end graph. They wound up being a little too squirrely about the passive crossover to fit in my system (they wanted to push much lower than I wanted to XO), but they measured spectacularly from ~3kHz up, below that the horn loading fell off smoothly but it wasn't very usable response. The resolution of measurement made a big difference as the base horn loaded dome was very, very good, but the absorption took it to a whole 'nother level of smoothness and helped a lot with LF rolloff. I would have seen very similar results on an RTA but with unsmoothed high-res polars, it was easy to see some artifacts coming from the horn lip interaction.

[toddalin]

Test video. Heil is fed pink noise from the Behringer through a 13.3 mfd cap and in series with a 4-ohm Mills 12.5 watt resistor. This creates a high pass Butterworth filter at ~1,500 Hz.
Next step is to make a 30 degree from acrylic rod. http://www.audiokarma.org/forums/styles/default/xenforo/smilies/banana.gif The wider wedges didn't seem to work as well as the smaller, but I only tried a couple at this point.

http://www.audioheritage.org/photopost/data//500/medium/DSC_2496_1280x854_.jpg


Mic is on-axis at 2 meters.
http://www.audioheritage.org/photopost/data//500/medium/DSC_2493_1280x854_.jpg


Voila:
https://youtu.be/o8Ftyl303Gk

Edit, I guess they don't make acrylic rod in that angle. I guess I could try to bend some metal, but it won't have the same finesse.

[toddalin]

Because I can't find what I want. I've designed my own wedges from acrylic plates that will be stacked to give the desired height. I have black, smoked, and clear acrylic and will do some form of pattern up the wedge.


The pieces will be bolted together as the Heil is. A screw will come from the top and pass through the wedges:
http://www.audioheritage.org/photopost/data//500/ScreenHunter_193_Feb_18_18_27.jpg
The bottom wedge will be cut to hold a 8/32" machine nut so that the ~27-32 pieces can be bolted together. These are actually 40.6 degrees to leave enough meat to fit the nut and keep the overall length to about 1.2". Longer pieces showed less effect.


http://www.audioheritage.org/photopost/data//500/ScreenHunter_194_Feb_18_18_28.jpg
The screw/stack will proceed further through the upper base that is slotted to allow the wedge to be adjusted. Note the cutout for the "Heil glue" so it can snug up to the Heil base.
http://www.audioheritage.org/photopost/data//500/medium/ScreenHunter_195_Feb_18_18_28.jpg
This is glued to a lower base that will allow for a nut to be tightened with a small pliers.


http://www.audioheritage.org/photopost/data//500/medium/ScreenHunter_196_Feb_18_18_29.jpg
I've also designed an "assembly bed" that has the same angle as the wedges that will hold them to assist in putting the pieces together and deciding on a color scheme. The bed is marked for the correct length making this much simpler.

[toddalin]

Made these today. Tomorrow will cut ~70-80 wedgies with a round hole.


http://www.audioheritage.org/photopost/data//500/medium/DSC_00093.JPG http://www.audioheritage.org/photopost/data//500/medium/DSC_00101.JPG
http://www.audioheritage.org/photopost/data//500/medium/DSC_00121.JPG

[toddalin]

Done. Now for some listening. http://www.audiokarma.org/forums/styles/default/xenforo/smilies/music.gif


http://www.audioheritage.org/photopost/data//500/medium/DSC_00122.JPG http://www.audioheritage.org/photopost/data//500/medium/DSC_00151.JPG http://www.audioheritage.org/photopost/data//500/medium/DSC_00171.JPG