[matevana]

The Hestia OB Project





Background:

I’ve spent the past few months actively auditioning drivers that might be well suited for the Hestia OB project. An emphasis was placed on auditioning several mid bass/HF driver combinations that would become the cornerstone of this project. All were reviewed in the same test baffle and listening environment. So as not to overburden anyone with a gigantic post, I will be breaking this project into several installments over the next few weeks, complete with pictures, links and other information. Please feel free to mark this thread and follow along!

Hestia OB Design Goals:

Well integrated driver selection
Shallow slope crossover w/low component count
High efficiency
Neutral vocal presentation
Small 3-way all inclusive design
Refined “no shout” voicing characteristics
Extended horizontal dispersion
Low cost (no single driver > $50 US)
Consistency; ability to replicate the design for the DIY community

At least 75% of the success rate in OB can be attributed to driver selection. In addition to choosing drivers that perform well in OB, they must also integrate properly amongst themselves. It seems that many amateur OB projects consider the former but not the later point. Selecting drivers with well mannered roll-off characteristics will minimize the component count and preserve efficiency. Drivers with relatively smooth curves will allow for less complex crossovers. And perhaps most important in OB is the actual voicing characteristics of the driver itself. This can be a factor of the cone’s material and thickness, VC former construction, the baskets ability to suppress additional reflection, and a number of generally accepted TS parameters that are conducive to OB. The fact is that many acclaimed (and often expensive) drivers typically used in conventional box designs make poor choices in OB. Conversely there are a myriad of lower priced drivers that make excellent candidates.

One of the primary design goals of the Hestia was to address vocal shout that seems to be present in many amateur and professional designs alike. For the purpose of this project, shout will be addressed primarily at the driver selection stage and secondarily in the crossover design. Many hours have been spent listening to raw drivers and driver combinations in the 1k to 5k Hz frequency range. I truly believe this one factor can make or break a design.     

As if the above mentioned design goals aren’t lofty enough, cost should also be a big factor in the DIY arena. For this reason I decided that no single driver should cost more than $50 US, and where possible, even less.

From previous OB projects I have decided that I really enjoy the presentation that a well suited 10” pro sound driver provides in OB for the critical midrange segment. Unlike traditional 10” woofers, they are often voiced to provide stronger midrange up through 4-5k Hz. They also provide a full lower midrange on an open panel, where otherwise two smaller drivers might be needed. Since it will be crossed relatively low, beaming is not much of a consideration here.

I spent a few months listening to eight 10” driver candidates made up of pro sound, guitar and bass guitar drivers  One had an aluminum cone, another two were made of hemp and fiber and the rest were paper or paper blends. One driver had an alnico magnet, two had smallish neo magnets and the rest were ceramic. One driver even had an all paper former; the rest were poly. Most were in the .5 to .9 Qts range and all had an efficiency of at least 92dB at 1w/1m.

Based on a variety of listening tests, two drivers emerged at the top of the list, but one was cut due to cost (considering the mantra of no single driver > $50). The winner in this impromptu shoot-out was the MCM audio select 10” pro sound cast frame driver, # 55-2981. Every time I considered this driver I could not believe how ridiculously low priced this driver is positioned. It is basically a very well made cast frame driver with a 2”voice coil, a treated paper cone and a very useable xmax of 6.5mm and Qts of .54.with an average sensitivity of 93dB. It is often seen on sale by the manufacturer for around $25 US. Truly unbelievable. More important, it has a very nice low mid presentation with a rapidly descending roll off above 2k and no shout when crossed properly. A perfect choice for the Hestia project.

http://www.mcmelectronics.com/product/MCM-AUDIO-SELECT-55-2981-/55-2981

Next installment, conquering the high-end.   

 

[matevana]

Baffle dimensions showing Right Panel (Left is mirror image)





MCM Audio Select, Cast Frame 10" Pro Sound Driver, #55-2981 < $30 US!





MCM Audio Select, Manufacturer's Stated Frequency Response & Impedance Rise

[Poultrygeist]

Can't wait to read more.

[Brad]

Can't wait to read more.

Exactly - this looks like a fun project.  Eager to hear about x-over points, total sensitivity, etc.
Might be a great match for my Scott 222c
Thanks for sharing

[this_is_vv]

what a timing...about to create a 3 way OB and here comes another contender....cant wait....

V

[lowtech]

I spent a few months listening to eight 10” driver candidates made up of pro sound, guitar and bass guitar drivers...

Based on a variety of listening tests, two drivers emerged at the top of the list, but one was cut due to cost (considering the mantra of no single driver > $50). The winner in this impromptu shoot-out was the MCM audio select 10” pro sound cast frame driver, # 55-2981.

If you don't mind sharing, what where the other seven?  Also, were you listening to all of them full-range?

[Poultrygeist]

Wow, for the price of one Beta 12LTA I can buy two MCM's. I have the Betas and love them but are the MCMs as good?

[matevana]

If you don't mind sharing, what where the other seven?  Also, were you listening to all of them full-range?

Sure thing. I have access to a bunch of Eminence drivers, many originally intended for guitar or bass applications. Remember that in the context of this project, they are being used as mid/upper bass drivers. Many of those tested had limited excursion. I was mostly interested in experimenting with different voicings, cone materials, alnico magnets, etc. They were all fairly efficient but many were far from flat in their response.

MCM Audio Select model 55-2981 (ultimately selected for this project)
MCM Audio Select model 55-2961 (not bad but for the small difference in price, the above was a better performer).
Eminence Lil' Buddy (very nice sounding driver, hemp cone, extremely flat, very limited excursion, a little too pricey for this project)
Eminence Copperhead
Eminence Legend 1028k (alnico)
Eminence Basslite CA2010 (aluminum cone)
Eminence Basslite CH2010
Eminence Legend BP102

I previously tested a 10" Tone Tubby hemp cone driver, which initially peaked my interest in using these types of drivers as the "heart" of a 3-way OB. The Tone Tubby was a little more lively than the Eminence Lil Buddy and not nearly as flat. One day I plan to look at a project with the Eminence Lil' Buddy; I was particularly impressed with its flat response in my area of interest and it's nice, laid back appearance. It's Qts is perfect for OB but it's xmax is extremely limited so it would need to be filtered on it's low end as well.   

[matevana]

Wow, for the price of one Beta 12LTA I can buy two MCM's. I have the Betas and love them but are the MCMs as good?

I think it depends on where you plan to cross them. The Beta 12LTA (with it's whizzer cone) is intended to play fairly high. Some people prefer to cross larger diameter drivers low. In my project, the MCM starts it's roll-off at around 1400Hz. With the right combination of drivers and crossover points, it has a very elegant lower mid presentation. Last night I was listening to a Strunz & Farah CD where they were trading high speed acoustic guitar arpeggios. It just sounded right, with the proper weight, attack and decay. I am very happy with it. 

[mcgsxr]

Perhaps my aging eyes missed it, but are you going passive Xover or active?  One amp, or multi?

Looks to be an interesting project, thanks for sharing!

[matevana]

Perhaps my aging eyes missed it, but are you going passive Xover or active?  One amp, or multi?

Looks to be an interesting project, thanks for sharing!

Nope, you didn't miss it.     In a nut shell the Hestia is a 3-way dipole hybrid with two different 10" drivers and a 3/4" dome.  It is crossed passively between the tweeter and mid/upper bass driver with a 1400 Hz LR-2 low pass and an 1800 Hz LR-2 high pass (I will post the schematic shortly). Bass duties are handled by the 2nd 10" woofer and crossed at 90 Hz LR-4 with a Bash style class-D plate amp. Based on a quick mic session, the average efficiency of the mid/high side is around 91dB @ 1w/1m. The system operates in mono from 90Hz down and is true dipole up through around 1800 Hz where it starts it's transition to a cardoid pattern.

One of the design goals was easy and consistent replication, so the passive crossover can be built on a readily available Dayton pre etched PC board. It's just a matter of dropping in two coils, two caps and 8 solder points. The driver integration process included level assessment so there is no need for an additional l-pad on the tweeter side.

[matevana]

Tweeter selection was almost as daunting.  I know from previous use that I really enjoy the neutral sound of the Usher 9950-20.  But in keeping with the design goal of < $50 per driver, I set out to test a bunch of lower cost drivers that exhibited similar neutrality with a nice airy top end. In a two week period I borrowed as many tweeters as I could get my hands on. One aspect that became apparent was the enhanced dispersion characteristics of a ¾ dome vs. a one inch or larger diameter. But in a world of trade-offs, I wondered if it would be possible to meet the mid driver at a relatively low frequency w/o high distortion. In short, the answer was maybe… at least in theory.

I was unable to get my hands on the newest series of neo magnet Vifa/Peerless ¾ dome tweeters that have a surprisingly low Fs for their size. I did have an XT19 ring modulator style driver, along with a classic (and still available) D19TD-05 poly dome. I used the Usher as my benchmark and compared it’s neutral sound to a few Morel models, the above mentioned Vifa’s and several small Dayton neo domes and one Hi-Vi model as well. I started by crossing each of the drivers at 2k Hz LR-2 (not too taxing) and just listened to them. It’s amazing what you can hear when you isolate a driver before attempting integration. I listened to a bunch of cymbal and click tracks from a drumworks test CD, paying particular attention to the air the tweeter produces. Right off the bat about half of the test samples seemed to produce a “tizzy” quality to their top end. The Morel drivers were nice and a little on the warm side but had the poorest horizontal dispersion of the group, possibly due to their 1 1/8” diameter dome. The Dayton drivers were nice and small but seemed to suffer from consistency among the sampled pairs. In two cases, the overall clarity was questionable; one was a 5/8” dome and integration would likely have been problematic anyway.

The Vifa XT19 was an interesting bird. It didn’t measure too well with a simple LR-2 alignment, but I understood from reading about this driver that it would likely need more attention. The best results came with a 3rd order alignment using a cap on either side of the shunted coil. It did sound nice, but the more complex x/o might present phase integration problems with the intended lower order x/o of the mid woofer. The additional parts would also add cost to the project. On the plus side, it did have excellent horizontal dispersion at 30 and even 45 degrees off axis.

Interestingly enough, the driver that I preferred the most was the Vifa D19TD-05. It’s a classic style poly dome that Vifa OEM’d for use in many boutique shop branded mini monitors. It has excellent dispersion properties and next to the Usher, shared many of the same characteristics at a much lower cost. The pair that I have are made in Denmark. Vifa/Peerless currently produces this driver in Asia and many of the reviews have been favorable regarding the current facility’s ability to reproduce a faithful copy of the original. This driver will likely be produced for the long haul since it’s a drop-in replacement for so many mini-monitor style speakers. It also plays loud and clear and seems to handle a fair amount of power. Best of all, it currently retails for around $16 US. 

On the down side it has a fairly high Fs, which might otherwise be problematic for this particular project. Based on my listening impression and the bargain basement cost of this nice tweeter, I wanted to see if there was a way to make it work.

Next up, passive crossover design and combating driver “shout”         


The Tried & True $16 Vifa D19TD-05-08

[matevana]

A few pics of the prototype:

[matevana]

The inductors are Jantzen 18 AWG air core and the Caps are Dayton Audio Polypropylene. I use the Dayton pre etched 2-way PC board number 260-130 which makes a nice finished product. I typically mount them on the baffle back with rubber washers.

L1 =  1.80 mh 18 AWG
C1 =  7.5 uf
L2 =  1.40 mh 18 AWG
C2 =  5.6 uf

[MarvinTheMartian]

You will probably want to change the orientation of one of your air coils to reduce inductive coupling.
google speaker crossover inductor orientation the first few hits will explain the issues.

Danny also has a good crossover primer here
http://www.audiocircle.com/index.php?topic=83325.0

Shawn

[matevana]

You will probably want to change the orientation of one of your air coils to reduce inductive coupling.
Shawn

Yes, the coils are just attached with a strip of velcro for now to test the prototype. The finished board will have the smaller inductor flipped on its side and zip tied to the board. 

[JohnR]

I'm just curious if anyone has ever either measured or heard adverse effects from inductors not all positioned at right angles in four dimensions.

[matevana]

The bass driver:

The new Dayton line of pro sound drivers were selected for this project to handle bass duties, in particular the model PA255-8 10” Pro Sound woofer. Essentially there were three factors that went into this decision, namely (1) a high sensitivity of 95dB; (2) an Fs in the mid 40’s which is about 10 Hz or more lower than most 10” pro sound drivers, coupled with favorable Qts, xmax and MMS measurements; and (3) low cost ($43 US).

During development I routinely alternated the positions/duties of the Dayton with the MCM driver and came to this conclusion. I definitely prefer the MCM driver for mid duties, in fact it is the cornerstone of this project with its accurate and unassuming presentation, beating out a myriad of other tested drivers. That being said, the Dayton driver only offered a slight advantage in bass duties over the MCM in its ability to play slightly lower, slightly louder and handle slightly more power. Could the project be built using two MCM drivers at a lower cost? Absolutely… just remember that the cut-out for the MCM is around 1/8” larger. Still though the Dayton offers slight advantages in the bass position and stays within my goal of sub $50. It also comes with rear gaskets that help dampen unwanted resonances down low. 

Next up: Pulling it all together

[Poultrygeist]

Did you consider an H-frame design for the bass drivers?

[matevana]

Did you consider an H-frame design for the bass drivers?

Fair question. Here's my philosophy on H and U Frames as it relates to this project, although some may disagree 

Martin King clearly demonstrates that the same driver will go lower in both an H and U alignment, at the possible expense of some SPLs. While I won't suggest (as others have) that the additional structure starts to take on sonic characteristics of a "box", I will say that I really enjoy the low profile, small footprint appearance of a flat panel system. In a much more comprehensive system like Orion, the speakers can be used at very high SPLs and produce ample bass. If anyone tried to build the Hestia project (with it's one 10" bass driver, panel mounted) I suspect they would be satisfied with the low output up to moderate levels, and perhaps employ a separate sub beyond that. I believe that would STILL be the case if these same drivers were used in an H or U. Hope that makes some sense.