[matevana]

Inspired by the feel of the new Linkwitz LX-521 design, I began to work on a progression of the original Hestia project. Dubbed the Hestia-SL, in recognition of Siegfried Linkwitz's LX-521 main baffle shape.





Initial Project design goals:

Continued use of the MCM Audio Select 10" lower midrange driver
Continued use of the Vifa D-19, or optional drop-in replacement of the XT-19 tweeter.
Introduction of a 5" coupling driver for enhanced mids
Incorporation of a H-Fame carrier for the LF driver with anti-vibration mount (or optional bridge)
All active or active/passive hybrid with a choice of 2, 3, or 4 way operation.
Floorstanding height with 5 degree incline: (42" H x 11.5" W x 15" D at widest point)
Alignment of all driver's acoustic centers
SL inspired main baffle shape for enhanced off-axis performance
Pleasing form-meets-function artistic design
Best bang-for-the-buck approach
 

At the present, the proposed driver selection includes:

Dayton Pro Sound 10" mid woofer
Vifa XT-19 dual ring radiator 3/4" driver (D-19 option)
Celestion 5" Neo Midrange (PE Buyout driver)
Usher 8955 8" bass driver in H-Frame   

UPDATE: Here's a pic as of 12/21/12 with the 3rd baffle shape iteration.   



UPDATE: Fourth and final baffle revision 1/5/13:





Regards,

Ed

[AJinFLA]

MCM Audio Seelect 10" mid woofer
Vifa XT-19 dual ring radiator 3/4" driver (D-19 option)
Celestion 5" Neo Midrange (PE Buyout driver)
Usher 8955 8" bass driver in H-Frame   

Regards,

Ed

Hi Ed, unless you already have that driver, you could probably do a lot better for the $$. Not going to get much bass out of those in a dipole

cheers,

AJ

[Trismos]

Looking forward to watching the project progress!

[matevana]

Hey AJ,

All of my designs use a single, smallish, articulate LF driver. It's that Allison Model One approach where sacrificing the lowest possible octave in hopes of achieving great sound in the upper bass, lower mid section can sometimes pay big dividends. I currently have a design that uses an Usher 8137a 8" woofer and the bass is solid down to the mid 50's in OB. More importantly to me, the 100-300 Hz range sounds right, which from a musician's standpoint gives the instruments their proper weight. This philosophy seems to be lacking in many of the OB designs I've heard, that try to makeup for this critical range by hitting the lowest register extra hard. The Ushers are also amazingly quiet in OB, with almost no motor noise or resonances.

Right now I'm testing a pair of the 8955's in stereo in their H-Frame carriers. Listening to them by themselves, I am convinced the bottom end will surprise many.     

[lowtech]

Looks nice.  No rear tweeter, though?  One of the main design goals of SL's new speaker is to maintain dipole directivity up to 7kHz.  That obviously won't happen without two tweeters.  Obviously you aren't trying to clone his design, so my comment is just an observation.

[matevana]

Looks nice.  No rear tweeter, though?  One of the main design goals of SL's new speaker is to maintain dipole directivity up to 7kHz.  That obviously won't happen without two tweeters.  Obviously you aren't trying to clone his design, so my comment is just an observation.

Yes, that's exactly right. The 5" midrange driver will cross to the tweeter in the area of 5-7k which is still to be determined, with mostly cardioid directivity above that. I personally experience more harm than good by running dipole beyond that, due to the less than ideal distance to the wall rear wall (about 30") in my listening room. Even in a more ideal space, I'm not sure I prefer dipole fullrange, but acknowledge that many do. 

It's interesting when looking at the new designs from both SL and John K. that they are perhaps more similar than they are different. After decades of experimentation and refinement, both designers use similar driver and crossover topologies, baffle shape, width, etc. I find it reassuring when two of the neo-classical dipole guys have converged on common ideas. If you are going to borrow, why not borrow from some of the best.   

[sjhomey]

I see that the Usher 8955 is no longer available at PE. Looks like it has been discontinued.

[scorpion]

But of course there will soon be another Usher driver !

/Erling

[JohnR]

Looks interesting. I may have misread but you would be running the H-frame up to 300 Hz?

[matevana]

I see that the Usher 8955 is no longer available at PE. Looks like it has been discontinued.

Hey SJ,

The 8955 is still current but unfortunately PE no longer carries them. Not to worry, the 8137a is still available and is just as quiet regarding mechanicals. A variety of 8" drivers with decent excursion will work. They should be able to be crossed as high as 150Hz and go down to around 40. The 8" servo sub from GR looks very interesting but I have not heard them. 

[matevana]

Looks interesting. I may have misread but you would be running the H-frame up to 300 Hz?

The plan is to cross them no higher than 150Hz, and probably closer to 120. The 8" woofers seem to lend themselves to higher xo points if desirable in the overall design. Note that the 10" lower mid will partially overlap this driver and run from about 50Hz to the point where it meets the coupling midrange driver. 

My comment earlier related to the critical region of 100-300 Hz, where you can feel the music in your gut. It is this area where I believe many of the OB designs are lacking. I spend a lot of time making sure this sounds the way it should. I do not use any active eq in my designs, instead relying on varying driver levels, crossover points and the occasional passive notch filters to compensate for dipole roll off. 

[matevana]

Here's a few shots of the H-Frame Carriers I am testing. They are universal in design and will allow a variety of top baffles that attach with a unique shock mount. They are very compact measuring only 11.5" wide by about 15" deep at the widest point. They are built with a baffle incline of 5 degrees, and will hold the main baffle approximately 30 mm behind the woofer baffle to better align the driver's acoustic center and provide some time alignment as well.   





I am happy to say there is no baffle noise at near concert levels. The Ushers produce copious amounts of bass for their size and are extremely quiet, quieter than any OB bass driver I have used.  This is true for both the 8955 shown in the picture as well as the 8137a kevlar bass driver. They are also very fast and articulate (return to zero). I have spent the entire day listening to the stereo pair of H-Frames, at varying crossover frequencies. They can be crossed much higher than typical, but the actual crossover point has yet to be determined.





Materials used are full 1 inch thick poplar and 3/4" MDF. Panels are bolted with flex seal. 

[richardcooper2k]

I am thinking about building the hestias and maybe trying the SLs as well in the new year

I live in the UK

What have folk found is the best way to buy the MCM and other drivers from the UK ?

I investigated on Amazon.com last night but it stopped me at the checkout

[matevana]

I am thinking about building the hestias and maybe trying the SLs as well in the new year

I live in the UK

What have folk found is the best way to buy the MCM and other drivers from the UK ?

I investigated on Amazon.com last night but it stopped me at the checkout

Hi Richard.

You should be able to purchase directly from MCM. Their site is www.mcmelectronics.com  Not sure of their policy on international shipping but right now they are offering 15% off on most items. You have to use promo code AFC909 after the items are in your cart.  I hope that helps!  Let me know if you have any questions if you decide to build either Hestia project. 

[sjhomey]

Hi Matevana

Looks nice. I like the circular sides. How is it all put together? The lower horizontal brace fastened to the sides and the bass baffle fastened to it and the upper panel fastened to the bass baffle? Are all three fastened to the sides? I think there are some pieces hiding between the bottom brace and the bass baffle and also between the bass baffle and the upper panel. No? Are they part of the vibration isolation design? Is flex seal that fabled audio vibration isolation product that I can also seal my gutters with as seen on tv?

You mentioned an active/passive setup. In the interest of estimating costs what sort of active parts are you considering? I assume as with the Hestia one option will be mono bass with a plate amp. Will the top end have an all passive setup?

[matevana]

Hi Matevana

Looks nice. I like the circular sides. How is it all put together? The lower horizontal brace fastened to the sides and the bass baffle fastened to it and the upper panel fastened to the bass baffle? Are all three fastened to the sides? I think there are some pieces hiding between the bottom brace and the bass baffle and also between the bass baffle and the upper panel. No? Are they part of the vibration isolation design? Is flex seal that fabled audio vibration isolation product that I can also seal my gutters with as seen on tv?

You mentioned an active/passive setup. In the interest of estimating costs what sort of active parts are you considering? I assume as with the Hestia one option will be mono bass with a plate amp. Will the top end have an all passive setup?

The H-carrier in the photo is assembled as a rigid unit.  The shock mount isn't in the picture at this time, but will help to decouple the H-carrier from the main baffle which hasn't been built yet. The theory behind the shock mount is similar to the design of theater walls that are decoupled from the floor and ceiling. The hardware connecting the two will "float" so as not to transfer vibration from the H-Frame to the main baffle. I will post a more detailed drawing of this system when I can. The flex seal is an acoustic sealant that remains viscous and minimizes panel flex. It is used in conjunction with mechanical fasteners.   

Most of my testing to date has been with an all active, tri-amped setup plus mono plate amp for the subs. I am learning a lot about what works and what doesn't. I'm not sure how easily this can be duplicated passively, but perhaps a hybrid might be the way to go, such as combining the lower mid with the mid, or the mid and tweeter. The other thing that's interesting, is that an all active system doesn't necessarily have to cost more, with high quality amplifier watts pricing out at about a dollar per watt (in the form of tripath amps used to drive the mids and highs). This can compare favorably to the rising cost of premium quality inductors, caps and resistors. Still lots of things to test, but it's coming along very nicely... if only I had more time to spend on the hobby  

[matevana]

Here's a very rough sketch of the shock mount concept.



Basically the main baffle is sandwiched to the H-frame carrier at three fulcrum points. The mechanical fasteners are only rigid where they come into contact with the main baffle. The pass through holes in the carrier are at least twice the diameter of the fasteners, so as not to come in direct contact with the hardware. The main baffle is suspended between two foam rubber donuts, about 1/4" thick. Vibrations traveling to the main baffle should be diminished.




Here's a view from the side. The order from front to back would be lag bolt, main baffle, rubber donut, H-frame carrier, rubber donut, fender washer, lock nut. The baffle will be torqued just to the point where the rubber compresses and enough force is exerted to sandwich the baffle in place, but where the rubber rings still maintain their resiliency.  Nylon inserts in the lock nut will prevent the nut from backing out over time. The rubber donuts or large o-rings will be made from foam rubber pipe insulation (tubes), sliced down to an appropriate thickness (approx 1/4" to 1/2").

[sjhomey]

Would this page be helpful?

http://www.amazon.com/Isolate-It-Sorbothane-Vibration-Isolation/dp/B0055SJEDG/ref=sr_1_fed0_14?ie=UTF8&qid=1354629061&sr=8-14&keywords=audio+isolator

[matevana]

Great catch!  I currently use sorbathane pads as dampening feet, but have not seen isolators made from the same material. I may just order a set while I'm testing baffle designs.

On another note, I am beginning to appreciate the intricacies of incorporating a midrange driver into a design. It's not as easy as it seems and truly a balancing act. At this point I'm not too happy with the results, especially when I step back and compare it to a properly implemented 2.5 way design. And I'm prototyping with a 4th order active analog system where adjustments can be made in real time.

On the plus side, both tweeters (D19 and XT19) sound really nice when crossed low with a 4th order slope.

[matevana]

After 3 weeks of crossover development, the following is being presented as version 1.0.  Most critical to the design is the mid level attenuation. Not being able to decide between more detailed/slightly forward and slightly less analytic, I decided to use a variable resistor (P1) to allow the listener to dial-in his/her preference, which may also depend on the source material.

The current driver/component combinations for version 1.0 are as follows:

Low/Mid: 
Dayton 10" Pro Sound Driver 8 ohms
L1 = 1.80 mh
C1 = 30 uf
R1 = 8 ohms

Mid Coupler:
Celestion 5" Neo 8 ohms
L2 = .33 mh (18 ga)
C2 =  50 uf
P1 =  8 ohm potentiometer, 15 watts

High Frequency:
Vifa XT19TD00-04 4 ohms
C3 = 6.8 uf
R2 = 0.51 ohms

Note that the approximate crossover points are 700 Hz for the Low/Mid driver and 5800 Hz for the tweeter. The mid coupler crosses at approximately 400 Hz and 5700 Hz. The overlap/deficit is intentional and is used to help EQ the design. The drivers were chosen in part due to their low slope tolerance over the selected range.

The schematic is labeled version 1.0 for an important reason. The drivers are being tested on a rectangular temporary baffle and are not at their finished height. The actual completed baffle will also contour to help enhance off axis response. It is therefore anticipated that further adjustments to the component values will be necessary.

Also note that the Usher low bass driver is powered and crossed separately with a Yung plate amp, similar to the original Hestia design. The approximate x/o is 100 Hz.