[matevana]

The Hestia-V Dome:

This time around I got the itch to develop two kits at the same time after coming to the realization that the AMT Tweeter in the "V" isn't everyone's cup of tea.  So it was back to woodshed to design the Hestia V-Dome, a smaller more conventional Hestia V.

For those looking to recreate the design, Melby Audio will offer both a finished and unfinished version of the baffle/stand which is milled from void free 19mm ply with a maple face. The kit is shipped as a flat-pack and arrives looking like this:




At the present time, the mock-ups are shown here, still awaiting final selection and testing of the critical midrange driver:





From what I have heard so far, I really like the new Markaudio Alpair 6P (paper) driver. Unlike the original V which is designed to feature the Dayton AMTPro Tweeter, this version will use a dome tweeter and lower midrange to simply compliment the fullrange driver at both ends. The plan is to run the Alpair either fullrange or rolled-off slightly on the low end for greater power handling.



 

[danvprod]

Wow -- this one is interesting too. I'd be very curious to hear your impressions should you try the Alpair 6p in here.

[matevana]

Wow -- this one is interesting too. I'd be very curious to hear your impressions should you try the Alpair 6p in here.

I only have one Alpair 6p that I got for testing. The North American shipment is apparently on the slow boat from China (literally) and should arrive at Madisound by late August. It's a bit frustrating that the one driver has promise, but I really need to listen to a stereo pair to fully evaluate them in the Hestia V-Dome. In the meanwhile I will probably test a few other drivers, if for no other reason than to kill some time.

[Don_S]

Are the pointed shoulders on the baffle around the woofer critical to performance?  I think rounded to follow the contour of the woofer would be better aesthetically. Give the speakers a pleasing, rounded jug shape. But that is just my preference.  I find the sharp corners distracting and potentially dangerous.

[matevana]

The overall baffle shape was done in homage to Siegfried Linkwitz and his LX-521 design. He states that the baffle shape was derived after many acoustic free field measurements. His design evolved from minimal width baffles around each driver which might yield a more uniform radiation pattern.

My designs are primarily focused on the combination of driver selection and passive equalization that allows for single amp use with no additional active circuitry. In my last two design (The Hestia V and Hestia V Dome) I borrowed the overall baffle presentation in hopes of benefiting from some of Mr Linkwitz's design goals and research. I also happen to really like the shape of the more angular LX-521. Programs like the Edge often suggest that round baffles around woofers do not yield great results, particularly in OB, where baffle diffraction is concerned. Perhaps this is why we see so few designs like this.  Regarding safety and the protruding shoulders, I have not found this to be an issue, even in tight quarters. 

[Russell Dawkins]

When the distance from a circular driver's centre to the edges of the surface it's mounted on is similar in all or most directions, a circular baffle representing the extreme case, the deleterious effect of the baffle shape on the frequency response is maximized. Counter-intuitively, when the baffle is rectangular, the distance from the driver centre to the baffle board edges varies as you move around the driver, smearing the edge effect over a range of frequencies with no particular emphasis at any one frequency, especially if the driver is not centre mounted. Some of Linkwitz's baffle board designs go further than a rectangular baffle and spread the effect over a still wider range of frequencies.

Lyngdorf does something similar in their DP1:
http://lyngdorf.com/products/loudspeakers/dp-1

[matevana]

Thanks Russell, that actually makes perfect sense. In some designs I offset the tweeters in hopes of spreading the defraction properties so they are irregular and somewhat less apparent.

[matevana]

This weekend I will do some listening tests with the Alpair 6p as well as the Seas MU10RB-SL in the design. The MU10 is the same upper midrange used in the Linkwitz LX-521.


;

 

[Don_S]

Thanks for the informative replies.  I get it now.  Size (and shape) matters. 

[matevana]

Sadly, I'm not sure that the Alpair 6P is the best choice in OB. The "built-in" baffle step compensation produces a somewhat hollow midrange. I can see where this voicing might work well in a more traditional boxed enclosure, requiring fewer components to counter baffle step. I really wanted to like this driver, considering its build quality and good looks, but I find myself missing the rich mids from my Seas FU10's in the other room (Hestia V). I look forward to trying the newer Seas MU10 on the Hestia V-Dome this weekend. It is very similar to the FU10, but has a slightly lower Qts and a modified rubber surround. It is also 4 ohms nominal.

[JohnR]

Interesting - thanks for the update 

[matevana]

Completed build with the Seas MU10's mounted, front and back views. Now for a weekend of testing and tweaking! I'm getting to appreciate crossover design at least as much as driver sourcing, baffle design, and fabrication.


 

[matevana]

Minimalism

Despite their similar appearance, the Hestia V-Dome differs from the Hestia V in several ways. The ultimate goal of the Hestia V project was to exploit the strengths of the Dayton AMTPRO driver, namely its speed and clarity. The Hestia V-Dome, on the other hand, strives to be liked by the full- range enthusiast (despite the appearance of multi-drivers) for its design philosophy and execution.  The V-Dome exploits the merits of the Seas MU10 driver and attempts to make up for the few weaknesses it has.

In the typical full-range, single driver system, the appealing sound characteristic comes from the lack of a conventional crossover. In a two-way speaker, for example, the acoustic phase of each driver rotates in opposite directions around the crossover point. While the combined phase may be correct at the crossover point, away from that point, the phase will be constantly changing. The human brain is designed to locate sounds primarily by phase. We are particularly sensitive to phase in the 300-3000 Hz range. If the phase relationship is less than correct, the sound stage often lacks precision and depth. By eliminating the crossover and the attendant phase problems, a single driver speaker sounds much more natural and coherent.

But single driver systems in an open-back alignment have their own challenges. Without the support of a box to cancel the driver’s rear wave, most medium sized drivers will not offer enough bass support in OB. And while many capable full range drivers remain clear and articulate up through the higher octaves, many can also benefit from the reinforcement and wider dispersion that a small dome can offer.


 
The crossover design in the V-Dome should be considered minimalist, allowing the Seas MU10 to play throughout its entire range, with only a single high quality resistor used for level matching. There are no inductors or capacitors which would effectively alter this driver’s phase relationship. The Eminence bass driver and the Vifa ¾” dual ring radiator are only used to support the MU10 at its extremes, in areas where the inherent OB properties might override the benefits of a single driver approach. Even their integration is done with a first order methodology to minimize phase abnormalities.  With few components in the chain, the extra detail that is maintained by better quality caps and coils is well worth the expense and the benefits are extremely audible. Lastly, the upper and lower x/o points are positioned well beyond the frequencies that we are most sensitive to.

Driver:                            Component:

Eminence B102...............9.00 mH 15 AWG Sledge Hammer Inductor (series)
Seas MU10RB-SL.............Mundorf 15 ohm MOX metal oxide film resistor (series)
Vifa XT19.........................Mundorf  1.5 uF EVO Oil capacitor (series)

All drivers are wired in-phase.  There is also a 10 ohm shunt resistor placed across the two auxiliary terminals (in parallel) of the Eminence B102. Here too I am using the Mundorf MOX brand. This is a critical part of the woofer integration. Without it there is too much high frequency overlap from the B102.  The resistance across the terminals helps the large inductor track better and chokes at least one additional octave of high frequency information. In this alignment the Eminence B102 reinforces the lower octaves of the MU10's range and it sounds like one coherent driver. 




 

[tonytopshed]

I really like what you are doing here, there is an elegance to the shape that's appealing! Having played around with my quick OB set-up, I'd now like to follow your ideas of the V-Dome. As I will be making it (or similar) myself (in UK), is it possible that you could provide a drawing or basic dimensions? I'm attempting to replicate the design on CAD, but it's a lot of guesswork!

Is there any advantage in selecting the Seas MU10RB 4 ohm speaker rather than the 8 ohm FU10RB? Their specs look very similar, you have a larger resistor in the crossover compared to the V; I'm wondering if it is better to have 8 ohm driver that doesn't need restricted? Hope this isn't a silly question!

[matevana]

Hi Tony,

One of the reasons the Hestia Dome has a larger resistor value is that in it's design the resistor is the only component affecting the midrange driver. The Hestia V for example includes a bandpass circuit. The inductor in that circuit adds it's own resistance which gets compensated by using a resistor with a smaller value.

It's a bit odd, but I think I prefer the overall presentation of the MU10 to the FU10, even when used full range. That being said they are VERY close in character and I'd be hard pressed to tell them apart if I didn't already know which was which. Ironically the FU10 was designed by Seas as a fullranger, while the MU10 provides the midrange coupler function in Linkwitz's LX-521. I believe the surround material is supposed to be more transparent in the MU series, at least where midrange performance is concerned.

I'm on the road for a few days, but will post the critical dimensions of the Hestia Dome when I return.     

[matevana]

Tony,

I have the Autocad files (.DXF/.DXT) for the Hestia Dome baffle and base. If you PM me, I will email them to you. It is the same file used to cut the plywood on the CNC machine.

[tonytopshed]

Just wondering - what difference is there if an H-frame has or hasn't a top or bottom? I'm asking because I can get 8 1 foot squares from a standard 4' x 2' mdf sheet to make two H-frames. Each with two sides, top and the fourth square to mount the woofer. No bottom.

Is the woofer's mdf square positioned halfway into the frame, or about 1/3 in?

[matevana]

Hi Tony,

The baffle is positioned halfway between the front and rear of the cube. Theoretically you can omit the bottom, if the cube is placed on a well damped, solid surface. If you omit the top, there will be too much cancellation. Note that omitting either the top or bottom may compromise the integrity and result in additional cabinet resonances, even with 1" MDF. I continue to recommend the pre-fab "paper" cubes since they are surprisingly well damped and easy to assemble.

[Dimakiir]

How is this development going?
I really like the clean design and passive crossover. Possible to share more info of how to build it?

[matevana]

They are complete and I like them very much. I am going to be experimenting with an active crossover (bottom section only) over the holiday break. The passive crossover used with the Peerless SLS drivers sounds nice but is limiting the gain to the Adcom GFA-565 amplifier that drives them. I plan use a DBX 223xs crossover which will provide proper gain to the Adcom and offer a 24dB/oct rolloff. That being said I would not want to change the top section at all. In fact, I often run them w/o the Peerless and they sound very nice. I will post my observations over the break.