[matevana]

The Djinn (pronounced ‘Gin’) is my fourth Open Baffle project documented on Audio Circle. I wanted to try several design concepts that were new to me and report on my experiences.


Project Goals:

•   Compact flat baffle with near full range performance; driven by a single amplifier without DSP/ASP. 

•   Investigate the supposed benefits of slot loading a low frequency (LF) driver.

•   Employ a minimalist crossover for the full range (FR) driver, combined with radical driver/baffle orientation designed to reduce component count and address diffraction issues.

•   Use of a modern, class leading, metal cone FR driver without the need for a separate tweeter.

•   Purpose built passive corrective filters on the low frequency driver to address baffle step, equalization, and roll-off.

•   Use of low Q drivers to improve motor control in a boxless environment. (attack & decay).

•   Layered baffle materials designed to suppress resonances and maintain rigidity through Young modulus analysis.


Over the holiday break I will try to document each the above goal outcomes, for better or for worse… 

Please stay tuned and feel free to share your own experiences and opinions.

[matevana]

•   GOAL 1: Compact flat baffle with near full range performance; driven by a single amplifier without DSP/ASP. 

For many people, size, aesthetics and complexity represent barriers to owning open baffle loudspeakers. Full range designs are often augmented with low frequency drivers, (and sometimes high frequency drivers) that may require a separate power source (bi-amping). These systems generally include some type of variable signal processing. Frustration can occur when the listener is not privy to the original design standard and may over/under compensate for perceived shortcomings of individual recordings. The Djinn attempts to address these issues with its compact, understated aesthetics and passive, purpose built crossover. The latter allows the listener to use his/her favorite amplifier without introducing obstacles such as signal splitting, level and phase abnormalities, and becoming overwhelmed by a variety of user settings. 

The baffle is constructed of an inner and outer shell made of different materials and densities, separated by a damping sheet, and laminated with green glue and Kreg mechanical fasteners. The dimensions are 21” wide, by 20” tall, and only 8” at its deepest point. The baffle is angled back 10 degrees to better align each driver’s mechanical center. They are mirror images of one another.




Inner & Outer Baffle Orientation, Mirror Imaged:

[matevana]

•   GOAL 2: Investigate the supposed benefits of slot loading a low frequency (LF) driver.

Slot-loading appears to have both speculative and verifiable outcomes. Engineers & Enthusiasts such as Nelson Pass have designed successful slot loaded open baffle systems (SLOBs); still many others are skeptical of the merits. While some of the reported benefits are speculative, I have tried to concentrate on the more verifiable outcomes in this design.

Generally speaking, when slot loading a driver the front wave exits through an opening whose area is considerably smaller than that of the driver’s radiating surface. At the same time the rear wave remains unaffected. As the frequency rises, the polar pattern is altered so its forward radiation is no longer symmetrical with that of the rear wave. This can assist the designer by way of a less complex electrical crossover, when the altered cancellation properties are considered as part of the overall system. By its very nature, the slot also serves to physically protect the large LF driver and change the overall aesthetic as well.

Other more speculative claims, which I have not observed, include increasing the forward wave’s velocity which results in greater air mass movement that surpasses the cone mass of the driver. The supposed outcome here would be higher efficiency and increased transient response. Others have claimed a lowering of the driver’s Fs as well as force cancellation. 

It should be noted that there are many variations on slot loading, including orientations that attempt to address pressure on both sides of the driver (Ripole).

Passing the low frequencies through a smaller aperture, and attempting to benefit from a change to the radiation pattern as frequencies rise (due to the slot resonance), is a basic tenet of this project.

The finished slot is approximately one third of the driver’s radiating surface area for this project. For testing purposes, I simply started with the LF driver mounted to its inner shell (a traditional baffle with a 14” circular cut out) and placed two boards, roughly half the size of the finished baffle in front, meeting at the center. I then took measurements as I pulled the boards apart, and calculated the area of the simulated “slot” at different increments. While in room measurements on open baffle speakers can be challenging, the slot creates an additional problem with larger than normal variances between the near and far field, possibly due to velocity changes as the wave exits the slot. In the end I relied mostly on in room listening tests with the prototype crossover in place. By playing with the area of the opening, I was able to eliminate an additional peak in the LF response that otherwise would have required attention. Most observable changes took place when the area of the slot was between one third and one fourth of Sd.

Perhaps of lesser importance to me, is the protective qualities and improved aesthetic that a slotted baffle can bring to the table. It’s funny how many guests have complimented the design aesthetic, since the large pro-sound woofer is barely visible from the front. I guess this speaks to the acceptance factor, when placed in more traditional rooms.     



 

[matevana]

•   GOAL 3: Employ a minimalist crossover for the full range (FR) driver, combined with radical driver/baffle orientation designed to reduce component count and address diffraction issues.

Throughout the design process, I had to resist the temptation of making the FR driver's crossover more complex then it ended up being. I am pleased with this design choice, due in part to the Eikona's generous excursion and superior distortion characteristics. The driver is attenuated with a simple voltage divider, making use of high quality metal oxide resistors. There is no high pass (HP) filter in the crossover per se, and adverse distortion remains in-check at all but the highest playback levels. Instead, a HP function is created by placing the driver in close proximity to the baffle edge.

The driver was measured at varying distances from the edge, until the desired low frequency cancellation approximated the targeted roll-off. This design choice negated an additional 80 mfd of in-line capacitance, as originally dictated by the crossover design. I am not a fan of introducing significant amounts of capacitance in series with any full range driver, but most will require this or be subject to excessive distortion at higher volume on an open baffle. The Eikona seems to be an exception to the rule, as it stays well mannered at fairly significant levels. 

The driver is offset and mirror imaged with it's mate to reduce baffle diffraction. It's a bit of an illusion from the front, but the FR driver is actually much closer to the bass driver than it appears. You can see this from the butt shot in the 2nd post.

[Chicken Man]

That was a great read 'matevana', and very descriptive too, I am looking forward to your conclusions with 'slot loading' and any benefits it may afford bass drivers on smaller baffles.

Visually, very large bass drivers (12 "and above) tend to have a rather low WAF and any technique that can improve upon that would enhance OB's appeal with the other half of the species.   

Merry Christmas to all.

C.M

[Bemopti123]

Really looking forward your discoveries and conclusion.  How does the slot design of the woofer impact upon its efficiency?  That is a key question.  I guess the balancing between different sensitivities of the components have something that will have to be addressed via crossover component choice.  Merry Christmas and hope the New Year will bring great things in this study. 

 

[RoadTripper]

For a music enthusiast, there can be no better website/forum than Audiocircle. This post exemplifies that belief to the hilt.

[matevana]

•   GOAL 4: Use of a modern, class leading, metal cone FR driver without the need for a separate tweeter.

For some time now I have wanted to experiment with the newer breed of FR drivers. I am specifically referring to some of the exotic metal cone drivers as designed by Mark Audio, EJ Jordan, and the like. Many of these drivers are reported to exhibit smooth, non-fatiguing performance with extended high frequency response. For this project I actually purchased two sets of drivers to play with; the EJ Jordan Eikona 2 and the Mark Audio Alpair 10M. Both share numerous enthusiastic reviews from the DIY community, with little to no negative feedback. It should be noted that these drivers are more similar than they are different.

Both drivers have had prior iterations and are in their 3rd design generation or later. The Eikona 2 employs a cast aluminum basket based on a Scanspeak frame that I have worked with previously. It is very well vented and unobscured. The 10M’s frame is a type of plastic composite and a bit more bulky, but does the job nonetheless. Both have their camp of devoted followers, although it seems that few people have heard both. 

For the Djinns, the Eikona 2 ultimately beat out the Alpair in my design for a few reasons, despite the Jordan’s higher retail cost. The Eikona 2 offered similar performance, but occupied less real estate on a very compact baffle, which helped achieve one of the primary design goals. The Eikona 2 also seemed to do everything the Alpair 10M did, but in a more refined way. This included, in my opinion, a slightly less grainy high frequency characteristic (although perhaps a bit less extended) and a more realistic reach into the lower mid/upper bass registers. Their cleanliness is amazing and their low frequency extension is surprising for a driver this small. That being said, the Alpair 10M is a fine driver, and could also be considered for this design. It would require a taller baffle and some additional attenuation as it is a few dB’s more efficient than the Eikona.

Neither driver is particularly suited for stand-alone OB duty, but would work well within the confines of this project, or another design employing a helper woofer. It should be noted that both drivers were broken-in on the same 12” x 36” baffle for about 100 hours before critical evaluation.

[matevana]

That was a great read 'matevana', and very descriptive too, I am looking forward to your conclusions with 'slot loading' and any benefits it may afford bass drivers on smaller baffles.

Visually, very large bass drivers (12 "and above) tend to have a rather low WAF and any technique that can improve upon that would enhance OB's appeal with the other half of the species.   

Merry Christmas to all.

C.M

Thank you. I'll try to address what worked and what didn't at the end. Completely agree! The world would look a lot different if everything was built strictly for performance and didn't compromise elsewhere.

[matevana]

How does the slot design of the woofer impact upon its efficiency?  That is a key question.  I guess the balancing between different sensitivities of the components have something that will have to be addressed via crossover component choice.  Merry Christmas and hope the New Year will bring great things in this study. 

 

I wish I could demonstrate that the slot added efficiency, but I can't. Others have suggested this, but I was only able to measure a difference in some of the cancellation properties, none of which occurred in the lowest two octaves.  Perhaps the baffle size itself is a limiting factor in this design. 

[matevana]

•   GOAL 5: Purpose built passive corrective filters on the low frequency (LF) driver to address baffle step, equalization, and roll-off

Another design goal was to allow the FR driver to play on its own, with minimal electrical intervention. With this in mind, it became necessary to tailor the response of the LF driver to meet the specific roll-off of the Eikona in its intended baffle, and to address baffle step loss and low frequency equalization. This was accomplished by splitting the inductor of a typical 2nd order electrical circuit into two, shunting across the first half with a resistor, and continuing on with the remaining inductance necessary to complete the circuit. The value of the first leg of inductance, the resistor’s value, and the overall circuit inductance in series with the woofer (along with additional capacitance and resistance in parallel) are all variables that were used to tune the low frequency behavior of the woofer. A single LCR notch filter made up of a capacitor and resistor, (borrowing inductance from the existing circuit) was used to tame a peak around 500 Hz. A second more audible peak which occurred at 1600 Hz was completely eliminated by the band pass action of the baffle slot. The woofer begins its roll-off around 180 Hz. 

Since passive equalization sacrifices efficiency, it was necessary to select a LF driver that was approximately 10dB more efficient than the Eikona. Fortunately there are many good pro sound candidates that meet this requirement and are surprisingly linear as well. 

I ultimately selected the Dayton 15” Pro Sound driver, model number PA380-8 based on its TS parameters, dozens of positive reviews, and cost. The FaitalPro 15PR400 (even more linear) and Eminence Delta Lite II 2515 could also be used in this design with little to no changes.

The topology of the LF crossover is shown below. The left leg is the (+) from the amp, the right leg is (+) to the driver's plus terminal. If anyone is interested in trying this design with the selected drivers, PM me and I will gladly provide all of the component values.

[matevana]

•   GOAL 6: Use of low Q (Qts) drivers to improve motor control in a boxless environment. (attack & decay).

While there's something to be said for the traditional targeted total Q of .7 - 1.2 in OB designs, there are still reasons to consider a lower Q driver as a valid design choice. High Q drivers generally perform best in sealed enclosures as they rely on the additional compliance provided by an air tight environment. The air inside a sealed enclosure acts as a spring, which counteracts the driver's movement. In an open baffle environment we have an absence of this effect and the driver (with its weaker motor) may be prone to unwanted movement. This can affect attack and decay properties, or the driver's ability to start and stop instantaneously. This effect may be compounded when two or more drivers are used in a full range design, and the bass driver has different compliance properties then the FR driver. Some believe this can lead to smearing of the image, ringing, and confusion in complex musical passages.

Bass guitar players often refer to speaker choices as “fast or slow” based on their attack and decay properties. It is not surprising that “fast” drivers are often those with a Qts of .35 or less. Many people define fast as a driver’s reaction to an electrical signal. More properly, a fast driver is one that comes to rest instantaneously upon the absence of a signal. 

Lower Q drivers are often found in vented and reflex enclosures, and are built to stop without the benefit of the additional compliance provided by air tight enclosures. One might suggest that for this property, low Q drivers are more akin to OB than their higher Q counterparts.

On the down side, lower Q drivers will require a larger baffle (and sometimes lower crossover point) and/or some type of compensation to equalize their low-end response, when compared to their higher Q relatives. Passive EQ is employed in this design, at the expense of about 8 dB of relative loudness. Fortunately many of the low Q 15” drivers are in the 95-100dB sensitivity range.     

Other benefits of low Q drivers may include: (1) Stronger motors reduce lateral cone movement which may reduce distortion; (2) Lower Q drivers typically provide a lower Fs; (3) Lower Q drivers often have better excursion capabilities; (4) Stronger magnets found in low Q drivers tend to increase efficiency; (5) A wider selection of project choices. 

I have built open baffle speakers with both high and low Q bass drivers, the latter requiring some type of compensation be it active or passive. While it’s difficult to generalize, the notion of a properly equalized low Q driver in OB is very appealing. 

[matevana]

•   GOAL 7: Layered baffle materials designed to suppress resonances and maintain rigidity.

With prior projects I have had success in laminating two or more baffle materials to address mass, stiffness, and anti-resonant properties, which are sometimes opposing in nature. The goal was to produce a dense, acoustically dead baffle that would not contribute to the sonic characteristics of either driver. I also wanted an end product that would not warp over time, when subjected to the relative humidity in South Florida.

Finished Outer Baffle:





The outer baffle is made of marine grade, void free ¾” (23/32-inch) birch plywood. The overall dimensions of 21” wide by 20” tall are actually comprised of 4 pieces, attached from the back using Kreg joints. The two side pieces are 8” x 20” while the middle pieces are each 4” x 5”.  When assembled, this yields a slot opening that is approximately one third of the bass driver’s surface area. I was careful to maintain the grain orientation of the plywood when making cuts so it would look like one large piece when re-assembled. My hope is that the multi-layered plys will add dimensional stability to the baffle.

Outer Baffle showing sections:





The inner baffle is made of ¾” MDF which was prefinished in white melamine so there are no exposed edges. This layer is designed to add mass to the baffle.

Inner Baffle being test fit:





In between the two layers is a sound dampening sheet (visco elastic polymer) borrowed from the after-market auto sound industry where it is used to reduce interior cabin noise. Products like these claim to convert vibrational energy into heat energy. Green glue was spread with a 1/8” tile mastic trowel on each inner surface of the baffle, before the two sides were sandwiched together and fastened with Kreg screws. The combined baffle thickness is just shy of 1 ¾”.

Inner and Outer Baffles Combined:





Working at a university and having access to some pretty cool gear, I was hoping to test the baffle’s rigidity and resiliency on a Taber tester. I would like to be able to demonstrate that the benefit of the combined materials is greater than the sum of its parts, but timing over the holiday break did not permit. I have had conversations with some of our engineers about construction techniques, who were more than happy to provide their insight. 
 

[Bemopti123]

Keep it coming.  Your ideas and insights are very clear to me.   

PS:  Hopefully you will have a list of low end pro woofers and different crossover components choices for the final version. 

[Chicken Man]

'Matevana' did mention the drivers listed below .......

I ultimately selected the Dayton 15” Pro Sound driver, model number PA380-8 based on its TS parameters, dozens of positive reviews, and cost. The FaitalPro 15PR400 (even more linear) and Eminence Delta Lite II 2515 could also be used in this design with little to no changes.

I'll be using the FaitalPro 15PR400, but not on this project, at least for the moment.

I'm already into the construction of an OB using the FaitalPro, Satori MW16P-8 and the SB29RDC-C000-4 Tweeter.

I am presently marking out the baffle now, this time I wanted something a little more stylish than the usual slab of wood I often engineer OB designs to be.
Maybe I'll have to wait for 3D printing to become a household norm in order to become more creative with front baffle design.   

I really am enjoying this thread, and I do agree with 'matevana' on low QTS drivers, some may not.

C.M

[matevana]

Recap:

I tried to remain objective when describing the outcome of the project goals, stating more than one POV if applicable. I would like to comment on what I believe worked and what did not, within the context of this project, and what I might have done differently.

What worked:

The Jordan Eikona 2.  I'm a fan. There is something very coherent about them. The driver is said to have evolved with each generation. The construction is first rate and the clarity is amazing. They remain polite at fairly high levels, even in OB. And this without an electrical high pass function. Experiments with adding a separate tweeter did not improve the sound, and in my opinion, are unnecessary.

Low Qts woofers.  People are quick to discard the notion of low Q woofers in open baffle.  I surmise that an equalized low Q driver can be superior. The equalization can come in the form of passive components within the crossover, at the expense of sensitivity. The properties of attack and decay should be considered as part of a design choice.

Compact baffle. With the proper driver selection and crossover design, it is not necessary to have Goliath sized baffles to enjoy deep, plentiful bass. I am pleased with the quantity and quality of bass offered by the Djinns.

Radical edge placement. I got lucky within the context of this project in that the Jordan performed well without an electrical high pass. This allowed me to make use of cancellation properties produced by mounting the driver close to the baffle edge. The alternative would have been a significant amount of capacitance placed in series with the driver. Even stacked poly caps detracted from the Jordan's sound, and inexpensive electrolytics were unlistenable.

Multi layered baffle. I am pleased with the outcome of the baffle construction.  I can detect no sonic contribution from the baffle itself. The footprint is smaller than most floor standers, and the height is not overwhelming.

I’d also like to put in a plug for these cool Wago lever nuts, which I discovered during this project. I’m a fan of point-to-point wiring and usually solder the crossover components to one another. The Wagos make a really tight connection on a copper buss bar, with either solid or stranded wire. They are reusable by lifting a lever and leave the component leads in pristine condition. Perfect for bread boarding or even the final product. They come in 2, 3 and 5 terminal options.






What didn't work:

While it wasn't a failure, I did expect more from the slot loading function of the baffle design. In fairness to the proponents of this technique, the diminutive baffle size may have been a limiting factor in this particular design.

At low frequencies, I did not observe a change to the figure eight radiation pattern that impacted cancellation. I did not measure additional amplitude as a result of the slot in the lowest octaves. There were no apparent benefits in protecting the driver from over excursion due to pressurization. Any perceived benefit in the near field, fell off rapidly at the listening position. While slot loading may lower the driver's Fs, the baffle size did not allow me to benefit from this. This may however make some higher Fs woofers more useful in OB.

Within the context of this project, the slot did provide some lesser benefits as mentioned in the outcomes above. The slots cosmetic contribution was not originally considered, but proved to increase the acceptance factor greatly. 

What I might do differently:

I am still intrigued by the potential merits of a properly designed slotted baffle.  In the future I would like to apply similar techniques to a larger baffle (or a woofer with a higher Fs) in an attempt to more accurately measure its contribution.

 

[fdandrews]

Interesting project. I appreciate that you shared your design concepts, and the actual results. I have decided to try the Nelson Pass SLOB, with the slot next to the floor. Has anyone tried slot loading both sides of the woofer? Is this what is known as a Ripole? Please keep posting as you get more experience with this speaker.

[matevana]

Visiting Relatives:  GONE.  Visiting Friends: GONE.  Finally an opportunity for some serious listening and tweaking.   

[matevana]

I'm already into the construction of an OB using the FaitalPro, Satori MW16P-8 and the SB29RDC-C000-4 Tweeter.

C.M

I really like the motors on the Satori woofers. They post some of the lowest distortion numbers. I've had one in my hands but haven't actually heard them. Interested to see what you've come up with for the three drivers mentioned. 

[EJJordanDesigns]

Although we know of other OB designs with our drivers, this is the first we're aware of that uses the Eikona 2. If anyone has any questions specific to the Eikona, I'm happy to try and answer them here or via the Sales email on our website.

Thanks to Matevana for choosing our latest drive unit for this unusual and thoughtful project.