Saturday, June 7, 2014

Speaker Enclosures - Part 4

As all things must come to an end, this is the last of a 4-part series discussing issues with speaker cabinet designs. By now you should understand a little bit about how cabinets influence the sound of a speaker by what is called internal resonances. These resonances are challenging to control but easy to predict in solid rectangular cabinets. Changing the shape of a speaker enclosure from a solid rectangle to one with fewer parallel sides eliminates some of the internal resonance issues and eliminating all of the parallel surfaces eliminates one of the modes that an enclosure resonates (axial mode).

Truncated pyramids with their totally non-parallel wall structure are good choices for such designs as are variations on the basic wedge shape. But again there is more about cabinets that can be visually discerned. Scrutinize the way the baffle board is designed (the board on which the drivers are mounted) in the two cabinet designs below and note the differences. Take your time and see how many you can find just by looking at them.


Baffle Boards 1 and 2

The first thing you should notice is that the RH baffle board flush mounts the drivers and the LH does not. Why do you suppose this is? Could there be a reason? You bet! It's called the First Reflection. As the sound moves away from the driver, it strikes something - anything - from the edges of nearby drivers to screws on the baffle board to - well you get the idea. Anything that sound can be reflected from will bounce back to the driver and disrupt the sound coming from it. The simplest thing to do is to make the front of the baffle board appear as flat and smooth as possible thereby eliminating as much of the driver-induced first reflections as possible. You can do this by recessing the drivers into the baffle board (countersink them) and to use flat-head screws instead of round-head, etc.

Good job! Now, look at how the drivers themselves. Why did the RH designer use three drivers and the LH designer two? Good question. Air is moved by the driver thereby producing sound but just like bicycle racers ride lightweight bikes to go fast, two smaller drivers are typically lighter than one bigger driver and they will move faster (have better transient response) than using a single driver. However, you never get anything for nothing. You double the cost by adding a second driver and the free-air resonance of the cone will be higher (less deep bass, all things being equal).

Good. So far you have noticed how drivers are attached, recessed, and arranged on the baffle board. BTW, why did the RH designer put the tweeter in between the two woofers? Doing so causes the relative sound wave created by the woofers to appear to emanate from the same point as that of the tweeter (called a point-source). Unfortunately, there is a compromise with such an arrangement that disperses the sound by the two drivers well horizontally but not so much vertically. So to get good listening from a wider range of listening positions, you should orient the speakers vertically as shown rather than laying them horizontally on their side.

OK, let's look at some more examples and see what differences can be observed. Again take your time and see what you notice just from the appearance of the drivers on the baffle board.


Baffle Boards 3 and 4

The most obvious difference is the way the woofers are mounted. These are both KEF speakers and it shows how the evolution of design takes place. The model on the left shows the typical convention for mounting a driver: put a screw in every hole in the driver basket. The model on the right shows a marked departure from this philosophy as a result of some serious research: 3 screws only in a 120-degree pattern. KEF found back in the early 1980s when trying to minimize basket resonances that by using three screws to mount a driver created the minimum amount of mechanical resonance. This is something to note that seems to elude speaker manufacturers today. Most still plug all of the holes with screws.

Here is a free hint: if your speakers have 8 screws in them, back off the 5 screws that are not in a triangular pattern and see if you hear a difference in the way they sound. If so, remove the driver, plug the holes with RTV, and replace the driver using only 3 screws (one on top and two at the bottom).

One last thing that you need to understand about the baffle board and that is shown in the next picture.


Baffle Boards 5 and 6

These are two versions of the B&W model 801 as they have evolved over time. The LH older model shows an early attempt to round the edges of the enclosure at the baffle board. The RH newer version shows the implementation of a no parallel sides enclosure with much more radically rounded midrange and tweeter surfaces. What B&W and many others are doing here is to remove the effects of what is called edge diffraction where the mere presence of a square edge influences how the sound radiates into free space.

Rounding the corners of the baffle board is pretty common practice today and something you can easily see when making your next loudspeaker purchase. Those with well designed boxes using flush-mounted drivers, non-parallel surfaces, and rounded baffle boards at least give the loudspeaker a chance at sounding better by minimizing the physical interactions of first reflections and edge diffractions. It does amaze me that manufacturers still insist on using more than three screws to install drivers but hey, some folks never do learn from history.

Anyway, I hope you have enjoyed this brief explanation of speaker enclosures. At least now some of the mystery behind the curtain has been exposed and you can better understand why designers make the choices they do. Most is cost driven but some are aesthetic compromises. Regardless, all are just that: compromises. And BTW, compromise is not a bad word as some may lead you to believe.

Related Articles
See all entries about speaker enclosures in Part 1, Part 2, Part 3, and Part 4. Also, a related article on the effects of crossover network components on driver performance.

Yours for higher fidelity,
Philip Rastocny

I do not use ads in this blog to help support my efforts. If you like what you are reading, please remember to reciprocate, My newest title is called Where, oh Where did the Star of Bethlehem Go? It’s an astronomer’s look at what this celestial object may have been, who the "Wise Men" were, and where they came from. Written in an investigative journalism style, it targets one star that has never been considered before and builds a solid case for its candidacy.

http://www.amazon.com/dp/B00QFIAC3G

My other titles include:

Copyright © 2015 by Philip Rastocny. All rights reserved.

Speaker Enclosures - Part 3

In Part 1, we saw how all rectangular solids resonate at frequencies determined by their physical dimensions (dimensional resonances). And we saw how online calculators help determine where these resonances occur in your speakers and  if these dimensional resonances occur (group together) around notes on the musical scale. We also saw how to correct for these dimensional resonances by building new rectangular boxes of different dimensions but with the same internal volume and then stuffing it with the existing hardware.

In Part 2, we found that solid rectangular cabinets resonate in three planes: against two parallel walls (axial), against four walls (tangential), and against all six walls (oblique). We explored an alternative physical shape (the wedge) and learned why such a shape was even better (had fewer clustered resonances) than that of even a well-designed solid rectangle (fewer parallel walls meaning fewer axial mode resonances).

Here in Part 3 we will take this concept of removing parallel walls another step by exploring yet another cabinet shape to eliminate axial mode resonances. If you have anticipated this, you get a gold star and are beginning to understand audio, specifically acoustic resonances. Making a cabinet with zero parallel walls eliminates all axial resonant modes. But does this assure you that the remaining tangential and oblique resonant modes are random enough not to cluster? Maybe...we'll see.

First, a completely non-parallel surfaced speaker looks odd and few people are attracted to the radical departure from the solid rectangular reference although most artists and right-brained people readily embrace this departure (myself included). Such cabinets can be very expensive to build and beyond the means of the average aspiring garage-housed woodworking shop.




A Speaker Cabinet with Non-parallel Walls

But variations of the simple wedge shape can create an enclosure that has no parallel walls and is much easier to build. Sloping either the left, right, or both otherwise parallel sides of the basic wedge shape results in a cabinet resembling a truncated pyramid and a shape that is still within the abilities of the average DIYer to create in even the most humble garage.



Another Speaker Cabinet with Non-parallel Walls


Although you could also slope the remaining 90-degree faces on the basic wedge, this complicates construction and may not be required. Such alternative designs completely eliminate axial mode resonances and this is why they are used. However, this does not eliminate all major resonances.

No matter what shape a cabinet is, another major resonance creeps in: that of the fundamental resonance of the entire enclosure. If you tap anything - a wine glass, a fender on a car, a rubber ball - everything resonates at a frequency inherent to that shape. Even the earth has a natural resonant frequency and Nikola Tesla leveraged this resonance to transmit electricity from Colorado to Australia without wires. Such is the power of understanding resonances!

If this natural resonance occurs within the operating band of the driver, it will be excited by the driver at some time causing an emphasis at that resonant frequency. Several products are available to help reduce these resonances typically applied to the large surfaces of automobiles but are also useful in cabinet design.


A Typical Self-adhesive Sound Damping Material

So after careful construction of your truncated pyramid enclosure you abruptly rap its side and it rings like a bell, there is hope to recovering from an otherwise surmised disaster. Covering at least 1/3 of the internal surfaces with self-adhesive sound absorbing material will help lower the frequency of or totally eliminate cabinet natural resonances and salvage your back-breaking brow-sweating project. Making the shape of the damping material random rather than uniform may also help in distributing these peak of a resonance. Here trial and error will tell you if you have succeeded and your cabinet is as "dead" as possible.

Understanding the extent of a cabinet's natural resonance is as simple as rapping it with your knuckles while holding up your smart-phone's RTA application. If a natural resonance occurs within the operating range of the driver, this resonance will be excited at some time during a listening session (when the "note" matches the resonant frequency of the cabinet). Shifting the natural cabinet resonance off of the same frequency of a note on the musical scale will help minimize the audible effects of such resonances.

SUMMARY

Speaker cabinets do more than create a known volume in which a driver will optimally operate; it also introduces resonances to the speaker that can easily influence its sound. Changing the shape of a cabinet can help control the internal resonances especially by eliminating parallel walls in its design. Adding internal sound damping can reduce effects of remaining natural resonances by moving them off of a musical note or totally eliminating them entirely.

There is one more issue regarding cabinet design we will explore in Part 4 of this series.

Related ArticlesSee all entries about speaker enclosures in Part 1Part 2Part 3, and Part 4. Also, a related article on the effects of crossover network components on driver performance.

Yours for higher fidelity,
Philip Rastocny

I do not use ads in this blog to help support my efforts. If you like what you are reading, please remember to reciprocate, My newest title is called Where, oh Where did the Star of Bethlehem Go? It’s an astronomer’s look at what this celestial object may have been, who the "Wise Men" were, and where they came from. Written in an investigative journalism style, it targets one star that has never been considered before and builds a solid case for its candidacy.

http://www.amazon.com/dp/B00QFIAC3G

My other titles include:

Copyright © 2015 by Philip Rastocny. All rights reserved.

Friday, June 6, 2014

Speaker Enclosures - Part 2

In Part 1, we saw how all rectangular solids resonate at frequencies determined by their physical dimensions (dimensional resonances). And we saw how online calculators help determine where these resonances occur in your speakers and if these dimensional resonances occur (group together) around notes on the musical scale. We also saw how to correct for these dimensional resonances by building new rectangular boxes of different dimensions but with the same internal volume and then stuffing it with the existing hardware. In part 2, we will explore alternative physical shapes and learn why such shapes can be even better than that (have fewer clustered resonances) of evan a well-designed solid rectangle.

Before I begin, I understand that most of you do not want to redesign your speaker cabinets and nor do I recommend you do so. However, armed with knowledge of the compromises designers make in creating a production loudspeaker, you will better understand why choices are made and what these choices can sound like. If you see a cabinet of one design, you can listen for the resonances that always occur as a compromise of choosing that design. Remember, with knowledge comes wisdom.

Everything in the world is a compromise of some sort to gain something else and anyone who tells you different is not being honest with you. Saying it another way, to get this you have to give up that. When someone makes rational excuses for a compromise they are, as the old saying goes, “trying to sell you something.” (I find it interesting that “sales” is generally perceived as being a dishonest profession, one tolerated and even encouraged behavior from which the root of many of this world’s problems arise.) The joke “How can you tell when a salesperson (or politician, or...) is lying? When his/her lips are moving.” simplistically sums up the impression people have about honesty in general. OK, I’m off the soap box now and let’s get on with the show and get honest and talk about desired gains and resulting compromises.

The first alternative shape is the triangular solid or "wedge" enclosure. Here, there are only two parallel walls (gain: one major "side-to-side" resonance and several minor resonances) and such a cabinet is exactly half the volume of an equivalent-dimension solid rectangle (compromise: a box that is literally twice the physical size to maintain the equivalent internal volume). Wedges are relatively simple to build and you can place the drivers on one of the right-angle faces with the slope at the back preserving the current solid-rectangle appeal. Waste is also kept to a minimum with such a design.


A Wedge-Shaped Enclosure

Wedges are therefore a really good choice as an alternative-shape enclosure. Drivers can be placed on any of the three non-parallel surfaces so its unconventional looks can be somewhat disguised. It does, however, cost a little more to build this style enclosure. You can use an online calculator such as the one found for "Wedge 1" at http://www.the12volt.com/caraudio/boxcalcs.asp#wed to determine the equivalent volume of a wedge cabinet for your specific needs.

All speaker cabinets resonate in three "modes" just as rooms do: axial (simple back-and-forth from two walls), tangential (simple bounce off of all three or four walls), and oblique (complex bounce off of all five or six walls). Making a wedge-shaped cabinet uses fewer parallel surfaces (and fewer surfaces) and minimizes the number of resonant axial modes. Wedge cabinets do little to reduce the number of tangential or oblique modes. In the next part of this series, we will see if there are other ways to deal with the remaining resonant modes. Until then, listen to your stereo with your ears and not your eyes.

But there is more about basic cabinet design and you can read about it in Part 3.

Related ArticlesSee all entries about speaker enclosures in Part 1Part 2Part 3, and Part 4. Also, a related article on the effects of crossover network components on driver performance.

Yours for higher fidelity,
Philip Rastocny

I do not use ads in this blog to help support my efforts. If you like what you are reading, please remember to reciprocate, My newest title is called Where, oh Where did the Star of Bethlehem Go? It’s an astronomer’s look at what this celestial object may have been, who the "Wise Men" were, and where they came from. Written in an investigative journalism style, it targets one star that has never been considered before and builds a solid case for its candidacy.

http://www.amazon.com/dp/B00QFIAC3G

My other titles include:

Copyright © 2015 by Philip Rastocny. All rights reserved.



Thursday, May 22, 2014

Speaker Enclosures - Part 1

One of the challenges speaker designers routinely face is overcoming the laws of physics. How they choose to compromise their designs to address physical laws is an individual choice, some resorting to baffling, others to exotic shapes, and still others to exotic materials. All are good choices and none are free; there is always a downside to any choice in any design.

The biggest motivation a designer chooses something over something else is usually cost where if the best way to implement a design is this way but a good way is another less-expensive way, the latter will likely be chosen to permit higher profits and lower consumer costs. Is this bad? No, not at all. It's just a choice and you choose exactly the same exercise in everything you do when considering something to purchase.

For example, my wife loves purses and shoes and all of her life she bought those inexpensive accessories from mass-marketing stores - and she was happy. For her birthday, I sprung for an esoteric purse and that changed everything. When you step up in quality, you get more than a better-made product, you see how others choose compromises that such a purchase does not. Mercedes Benz automobiles, Brietling watches, and Gucci shoes are all examples of how everyday purchases can choose fewer or no compromises when implementing a design and owning such a product exposes the compromises made by inferior designs.

The single-most frustrating law of physics that plagues speaker cabinet makers is the internal resonances (aka dimensional resonances). Much like a tiny room, speaker cabinets are prone to vibrate at certain frequencies based mainly on parallel walls. The best thing one can do is to spread these dimensional resonances out so that they do not all occur at the same frequency, at the same frequency as a musical note, or with multiple resonances at the same frequency. You can measure the internal dimensions of your solid-rectangular speaker (subtract the thickess from the external dimensions) and calculate their existing resonances with the WINISD speaker design software available here. Such software automatically creates recommendations for cabet dimensions based on the volume required.

Such software (and designers) often use recommended ratios for height, width, and depth to build cabinets such as 1:1.14:1.39, 1:1.26:1.59, 1:1.28:1.54, 1:1.44:2, 1:1.60:2.33, 1:1.62:2.63, and 2:3:5. But such ratios are not really all that useful in predicting how many resonant modes will occur near each other (sum together creating a louder resonance), and what musical notes these resonances will exaggerate (over emphasize a note on the musical scale despite the anechoic response of the driver).

For example, plugging internal box dimensions of 20cm, 30cm, and 50cm into a simple resonance mode calculator yields harmonic resonances at 343Hz (near F4), 571Hz (near D5), 667Hz (near E5), and 685Hz (near F5). What this means is that such a box will artificially create louder notes at those points (especially D5, E5, and F5) independant of the driver's anechoic response.Changing the box dimensions to 19x30x48 better distributes the resonances in between the fundamental frequencies of notes on the musical scale. This means that all notes on a cello run will appear more uniform with the 19x30x48 cabinet as opposed to the D5, E5, F5 resonances with the 20x30x50 cabinet.

Other resonances can occur within a cabinet creating other issues above the operating range of the driver. For example, when harmonics of a resonance occur within a cabinet, the harmonic content of that fundamental note will also be exaggerated. So if a cabinet resonates at multiples of the fundamental frequency of D5 (667Hz), multiples of this frequency (2x, 3x, 4x, etc.) are its harmonics and resonances at these multiples change the way the character of that note is percieved (colors the sound and makes that saxaphone or piano note sound less real).

Aside from juggling the cabinet dimensions, what else can be done to minimize these internal resonances?
Logically, the simplest thing to do is to avoid using parallel surfaces (sides). Next to a solid-square box (one major resonance and the most resonant shape you can choose), the poorly-designed solid-rectangle vies for the second-worst shape since there are three parallel sides creating three characteristic resonances between them. But yet the vast majority of speakers built today use solid-rectangular enclosures. Why? In a word: cost. Straight cuts, square corners, and uncomplicated angles are mass-production friendly with low material waste. In other words, it is incredibly easy and highly efficient to build a solid-rectangular enclosure as opposed to a more esoteric shape.

Question: What can you do to improve your solid-rectangular speakers?

Answer: Build a new, properly-designed enclosure with the same volume and put the existing parts into it. For those of you with moderate wood-working skills, you can build another solid rectangle with minimal grouped dimensional resonances using the above calculator or hire someone to build you another cabinet. Tweak the dimensions to retain the required volume while adjusting the characteristic resonances to occur at fundamental frequencies that are not the same as those of musical notes.

For those of you with advanced woodworking skills, you can take this opportunity to build an alternative shape with fewer parallel surfaces. This approach will be covered in Part 2 of this series.

Related ArticlesSee all entries about speaker enclosures in Part 1Part 2Part 3, and Part 4. Also, a related article on the effects of crossover network components on driver performance.

Yours for higher fidelity,
Philip Rastocny

I do not use ads in this blog to help support my efforts. If you like what you are reading, please remember to reciprocate, My newest title is called Where, oh Where did the Star of Bethlehem Go? It’s an astronomer’s look at what this celestial object may have been, who the "Wise Men" were, and where they came from. Written in an investigative journalism style, it targets one star that has never been considered before and builds a solid case for its candidacy.

http://www.amazon.com/dp/B00QFIAC3G

My other titles include:

Copyright © 2015 by Philip Rastocny. All rights reserved.


Saturday, May 10, 2014

Onkyo P-3000R Preamp Review - Part 2

In Part 1 of this review, I evaluated the single-ended (RCA) inputs. The performance was stellar and I praised the circuitry for doing what I recalled it did best: sonic accuracy, imaging stability, and front-to-back depth.  However, also included in this preamp are some pretty sophisticated digital components worthy of mention.

In comparing the quality of any two sources, switching back and forth between two sources – preferably within two seconds – is the best way to evaluate such differences. Very few golden-eared folks have adequate long-term sonic memory to recall all of the nuances and subtleties when taking more than two seconds or so to switch between sources but devising a good method that itself does not influence the test results has been an equal challenge.

Some people believe that a Y-splitter is one way to achieve quick results and such a method will allow you to quickly switch between two RCA interconnect cables.  For example, two RCA cables can be connected to two inputs on a preamp (e.g., CD and AUX) and the other ends to the same output on a CD player via the Y-connector.  But I suspect that such a test method introduces errors in the form of ground loops between both cables.  So if the introduction of a Y-splitter is of questionable value when evaluating two sources, how can one do so without such degradation?

Y-Splitter RCA Cable Test

The answer is pretty simple: think light.  The Onkyo P3000-R preamp has four RCA and two TOSLINK optical-digital inputs allowing you to perform simple source A/B switching without introducing ground loops into the test.  So for Part 2 of this preamp’s evaluation, I will use my reference 1-meter RCA interconnects and a 1-meter Better Cables TOSLINK cable both connected to the same source: my stock OPPO BDP-105.

With such a configuration, I am able to play the same track on the OPPO and instantly switch the preamp between the CD and AUX inputs. What this configuration allows me to do is instantly hear the difference between the DACs in the OPPO (ESS Sabre32 Reference) and the DACs in the Onkyo (Burr-Brown PCM1795 192/32). This should be interesting.

P-3000R Rear Panel Inputs

Upon first hearing, I could distinguish very little difference between the two sources, something I did not anticipate. I listened to MP3s all the way up to DSDs and although there WERE differences, they were so slight that I could not consistently identify them from one format to another. Even my wife said that the differences were so small that she would have a hard time deciding which sounded better, other than the TOSLINK cable giving a slightly deeper bass output. From the get go, I knew this was going to take the right source material and more time than originally anticipated.

As my wife heard, the deep bass was a tiny better with the TOSLINK cable than the RCA so I was at an impasse. However, I had planned to make another modification to the woofer's crossover network in my speakers and now seemed like a good time to do so. With a 4-ohm woofer, the insertion loss of a typical large value wire or foil inductor - even with 12AWG wire size - approached a DC resistance of 0.5 ohms. This means that 10% of the power to the woofer is dissipated by the inductor.

I have always wanted to try a high-power toroid coil since the DC resistance in these inductors is about 1/10th that of any other coil type and now seemed like a good time to give it a go. I also wanted to lower the crossover point from 850Hz to 500Hz , a much better upper-frequency limit for a 12" driver. With a  quick check to Parts Express, I found a 3.3mH toroid that with a bit of unwinding would work in my crossover network.

Now unwinding this type of inductor is not as simple as measuring it with an inexpensive RLC meter. Because of the low DCR, higher currents are required to accurately determine the value and so I drug out my oscilloscope, sine wave generator, and a known value capacitor and resistor. It took about 11 turns off of this inductor to get it to the value I needed as determined by the LC tank circuit it created (carefully measure the minimum "null" and read the frequency, then simple math tells you the value).


A Toroid Inductor

I rolled this toroid inductor into the woofer crossover network and immediately I noticed a much cleaner, tighter bass. Considering that the moving mass on these woofers is an incredibly low 40g, this was not expected. But the surprise came when watching Star Trek: Into Darkness.  WOW did that extend the extreme LF response and remove muddiness! So the weak link in my chain was the woofer inductor and now I have never been happier - as is my wife who loves bass, especially explosions and deep thuds.

Back to the TOSLINK/RCA (aka Burr-Brown/ESS Sabre32 Reference) dilemma. Now that the LF extension and clarity is better, I hoped to hear more differences and be able to easily discern them. The large organ recorded by 2-L on the  Kare Nordstoga improvisation proved to be the right choice.


All throughout this 24/96 piece (also available in SACD), wall-rattling bass can be heard that captures the true character of a very large brass pipe. Through the Burr-Brown/TOSLINK version, the bass was present but not full nor did it reveal the true timbre of the largest ranks. Sounding muted rather than forward, this configuration - although pleasant - just did not compare to the OPPO/RCA version. The OPPO extracted more information from the source and left you feeling more like "you were there" listening in a pew somewhere forward of the pipes in the balcony. More depth, more width, and more height all combined to giving a well deserved and decisive advantage to the OPPO/RCA version. While other parts of the frequency spectrum were also different, these differences were not as noticeable nor as obvious as the very deep bass and hall ambience.

As I have mentioned in the past, if you are searching for ways to improve your system, look at the size and shape of the sound stage and then choose what sounds more realistic. For now, I will stick with the OPPO/RCA combo for decoding.

Yours for higher fidelity,
Philip Rastocny

I do not use ads in this blog to help support my efforts. If you like what you are reading, please remember to reciprocate, My newest title is called Where, oh Where did the Star of Bethlehem Go? It’s an astronomer’s look at what this celestial object may have been, who the "Wise Men" were, and where they came from. Written in an investigative journalism style, it targets one star that has never been considered before and builds a solid case for its candidacy.

http://www.amazon.com/dp/B00QFIAC3G

My other titles include: