Victorian Boombox Mk III. A brand new start.

[J. Wilhelm]

Dear ladies and gents, I have decided to give the pink slip to the Mk. II Boombox because I was not satisfied with the quality of the electronics, and while trying to compromise and address it's shortcomings, I basically engineered myself into a cost overrun of what was supposed to be a cheap, easy to build system. So for the moment I will abandoning further development of the Mk. II, and instead I will revisit the Mk. I concept.  


PROJECT OUTLINE (sorry for the wall of text - I'm just laying down the rules of the project).

The original Mk. I was a "big one-off" project built around a salvaged Altec Lansing 2.1 computer subwoofer speaker system from circa 2000, back when Altec Lansing was a more reputable and expensive sound house. The sub-woofer box provided a body onto which you could screw or bolt all sorts of steamy paraphernalia, including two cast aluminum lamp shades that served as housing for the relatively small satellite speakers and doubled as decorative Victrola horns. Enormously popular and successful, the Mk. I was paraded for years at Steampunk meetings, like Jeff Vandermeer's Steampunk Bible release party in Austin in 2011, and the streets of downtown Austin every single South by Southwest Festival since about the same time, but it was a monster. Too heavy to haul around and too delicate and expensive to sell. To make matters worse the heavy triple-transformer power supply was external and you needed to haul that too if you wanted to demo the unit (it's not battery powered).

Over the years the old box has developed issues related to multiple loose connections and a maze of wires that have not aged gracefully, and are rather difficult to replace. Nothing worse than an amplifier built around separate modules connected by a maze of thin wires that can rust and break. Unfortunately most large sub-woofer systems are built in a similar way. They were simply not designed to me moved around a lot or to have a long life. Not that I can't fix or replace the connections, but it will require open heart surgery and I'm, still left with a monster that is increasingly becoming more difficult for me to carry.

What I'm proposing now is a new architecture, a second "one-off" built around relatively expensive systems, but using different technologies that will let me change any boombox I build from now on. The Mk. III, as I call it will revolve around 3 separate technologies:

1. A central Hi-Fi receiver/amp unit, with a built-in power supply. I will try a relatively high quality (10% THD at 30 W) digital amp to save on weight and size.

Sony CMT-MX 500i mini shelf system

2. A set of two full range speakers instead of the central sub-woofer with smaller satellite speakers. The driver units will be mounted at the extreme ends of a structural beam that will double as a Transmission Line speaker enclosure (Inverted Horn Wave Guide Speaker).

Sony 1-826-587-12 100W full range speaker (plus 1-826-589-11 horn tweeter not shown)

3. A computer sound card that is able to partially operate (mixing analog and digital sources, microphone with sensitivity control, D/A conversion, Dolby AC-3 surround sound decoding, Creative CMSS surround synthesis) without being connected to a computer, and when it is connected to a computer by way of USB, is able to provide much higher functionality as you all know from using computer sound software. The sound card serves as both a switch/mix hub to allow multiple input and output connections as well as a quality headphone monitor output, also with a separate volume control.

Creative Extigy SB0130

I don't intend to use the 5.1 surround output for the boombox itself for obvious reasons. There are no sub-woofer, center and rear amplified speakers. Instead the front R/L channels decoded from 5.1 AC-3 could be routed to the boombox speakers and a separate set of rear channels and sub-woofer/center channel would use an external amplifier. This is a limitation of the digital amplifier. Each R/L channel is a closed loop with no common electrical ground. I can't "borrow the signal" to power a center channel, say, without needing to build a pre-amp circuit and a separate amplifier.

For other 2-channel surround sound synthesis methods (known as "Virtual Surround," I will probably post on that later), the boombox' built-in speakers alone are enough - I will use the sound card connected to a computer through the desired surround software -or- if the surround system is simple enough, like SRS surround -which is analog- then any stereo input from a decoding source, wired analog, digital wired/optical or Bluetooth, will be able to pass along the surround with no problem to be reproduced by the speakers - in theory. It all depends on the quality of the amplification - but should be no problem for the Sony amp.

In the Bluetooth case, the surround sound decoder is in the form of an Android or iOS based app - if it exists. I don't think that a wired connection directly from a tablet or phone would be able to transmit a good surround signal on account that most Android phones use really crappy internal power amplifiers to save the battery - but you can try. I don't know about present day Apple tablets/phones (I'm too poor to own one). But I'm willing to bet my lunch money that my ancient 2nd Generation iPod Touch beats the crap out of most present day Android phone amps - quality wise... Android needs Bluetooth.

The advantages of the new system architecture are multiple. The use of two large speaker drivers is meant to eliminate the need for the sub-woofer.  It will be a less delicate system, it won't require a separate power supply, it will be ready to accept next-generation sound sources, such as a Bluetooth connected smartphone or tablet, as well as more traditional Line Level RCA-based  connections, MIDI DIN based connection, RCA digital and laser optical connections (common things in computer sound cards), etc.

Instead of a sub-woofer I will place two high-power speakers into what is known as a Transmission Line enclosure. The "Transmission Line, really is a wave-guide, shaped as an inverted horn (the sound source is placed at the mouth of the horn instead of the neck of the horn). The theory is that the speaker acoustic loading (impedance) is provided by the mass of the air trapped in the wave-guide which dissipates the undesired sound next to the driver and channels the remainder of that sound far enough to a small open port (the neck) where the sound emerges at the same phase as the sound emanating from the front of the driver in order to avoid destructive sound interference.

An inverted horn transmission line speaker

The end result is similar to a bass reflex system, and not entirely unlike a regular "non-inverted" horn, like you see in Victrola horns. The energy in the undesired sound waves is re-routed to favor certain low frequencies that make up for a poor performance of speakers at the low range. There's a number of terms used to express this acoustic phenomenon, from "low pass filter," to "impedance matching," etc.  It's all the same, just seen from different spring-mass-damper perspectives. It's an acoustic transformer, basically. You want to discard undesired sound and pump what you can as usable low frequencies in the right phase if possible. That is how I can get rid of the sub-woofer altogether.

In fact I'm certain you've heard this system before, albeit in a much smaller scale. Does the name Bose ring a bell? Well. I'm going to use bigger speakers and use the "folded horn" as a structural beam for the Mk. III system. I will have to compromise and use low weight materials. I had considered using PVC pipes, but they're very heavy and can be expensive (5 to 7-inch diameter pipes). I had considered using tin chimney flue and A/C climate contrul ducts as well. They're too thin and will not dampen undesired sounds, plus they will resonate when your supposed to dampen any sound instead (this is NOT like a violin or guitar where you want the enclosure to resonate).

I need high density and rigidity. Particle board or MDF would be ideal. But I may consider strong light wood like poplar if its well braced structurally (more on that subject later). I can make due with rigidity and dampen the rest with batting if need be...

I'll see what I can come up with in the next few days... Until then,

Cheers!

J. Wilhelm

[Synistor 303]

Strewth!

[Banfili]

Strewth!

Indeed!
Looking on with interest to see how this develops!

[J. Wilhelm]

The build started yesterday with me picking up the Extigy I bought at eBay from the post office. I got off to a rocky start. The sound card works - mostly - and it's in very good shape externally, except that during its life someone dropped the unit to the floor, breaking the optical in /out jacks by way of the tiny plastic plugs that cover the led and photosensor. When I opened the enclosure, the jack modules were sheared off clean at their solder points.

I knew I had to either remove the units or try to solder them back before applying power to the unit (avoiding an electrical short circuit). I did the latter, but clearly the LED transmitter (which is visible red and on at all times) is not working. So I know I don't have Optical pulse modulation out. The photosensor also was sheared off and I re soldered it. It remains to be seen whether that is still working. I need to dig up an optical cable and try to hook two sound cards together.

In any case this will not stop me. I have to confess that I have another Extigy of mine that I have tucked in a box inside a storage unit somewhere in the city. Alas, the storage unit is far from where I live and I don't have a car, so it's a major expedition just to go get it (or an expensive über ride to go get it).That's why I bought one from eBay, the Post office is closer!

Anyhow, if anybody has a transmitter and receiver pair of Toshiba modules ( TX179 /RX 179) let me know! Otherwise I'll try to get them or it may be even easier to get another sound card. I will continue work with this one as is until further notice.

The card otherwise works well. I discovered that Debian Linux offshoots like Ubuntu have next to no support for the Extigy. The sound is too low when connected to the computer, and the card functions much better in stand-alone mode. All analog input /output types seem to be working. I haven't tried digital coaxial yet or MIDI in/out.

In the days when I was more savvy and compiled my own software in Slackware Linux, I remember extracting more functionality out of the same sound card. I probably need to compile another sound server and get some Source Forge drivers from some geek out there who has managed to run the Extigy in Linux.

[J. Wilhelm]

So à lot has happened in the last 24 hours, and I had an inspiration after visiting the local hardware store on my way home last night.

I had already drawn on paper possible arrangements for the sound wave horns in the SoundBeam™. The maze of folds in the transmission line enclosure could be made from the smoothest, straightest boards of 1cm pine craft boards that you've ever seen. Now, pine is not known for its acoustic qualities, but I'm not making a violin. The wood is not supposed to resonate but rather form a horn, or rather two folded horns. The wood just needs to be dense and rigid enough, and the maze of passages in the SoundBeam™ ensure that the enclosure will be very rigid. I can't put something heavy and gummy like particle board on account o need to be able to lift the boombox with my hands!

The arrangement of the horns is extremely compact. The theory is that the wavelength of the fundamental frequency of the speaker alone will determine the length of the horn. The speaker will go "on the wrong end" of the horn, acting as a vibrating closed end of a quarter wave resonator. I measured the speaker's fundamental frequency to be 84 Hz or so, which corresponds to à wavelength of about 4 m. Hence the wave guide should be about 50 cm folded in half...

From the outside the transmission line will look very unassuming just a rectangular beam

The idea is to finish the SoundBeam™ first and test it with different sources and perhaps stuff fiberglass batting in it as needed. A large area on each end of the speaker enclosure will connect to the speaker's back in an unspecified manner.

I have actually made more progress than I've shown, and I apologize for not disclosing that now, but I hardly have any time tonight to post anything, and thus I'll leave you in a cliffhanger... What I can tell you now is, "wow" this project is showing strong signs that it wants to live! You know how sometimes too many things just align together to be some sort of coincidence? This is one of those times.

Now, I must go back to my SundayChores™, because I only have one hour before midnight and I still have to eat and do the laundry!

[J. Wilhelm]

Do you know how sometimes things just fit together for no reason? I couldn't have found a perfect match for the speaker drivers if I was looking for it. One of the things I hated the most about the boombox was looking for suitable lamp shades or bells to use as Victrola horns. This time, I found this pair of crown molding quoins or corners, which are relatively expensive at $15 per piece, but they were far too convenient to overlook. They simply save me a lot of time and work.

Even though the pictures suggest the method of attachment is trivial, the edge of the speaker magnet and the rim of the driver cone actually interfere with an easy installation to the waveguide box. One solution is to add a "neck" to the crown molding quoin, and I will probably use a smaller wooden box serving as a tube connection.

In my next post I'll explore methods of attachment for the units above...

[J. Wilhelm]

When I was at the hardware store, I was inspired by the way the molding crowns fit over the boards I was using to build the waveguide. It just seemed like a perfect match and the geometry of the boombox was emerging naturally from the materials.

The only problem is that installation is not as easy as shown in the pictures. The magnet and rim of the Sony driver actually interfere with the install as shown. My first thought on that is that a small rectangular box could serve as a connection between the crown molding and the waveguide beam. This box is critical, because it adds length to the waveguide, so I will have to adjust the waveguide, probably by blocking the neck and using a perforation on the side of the box - or not, we shall see. But what I know is that I have to match the 1 meter quarter wavelength. I think that blocking the exit of the horn and making a hole of the right diameter on the right location (like a flute) might be enough.

The geometry of the crown molding box doesn't bother me much. It's sufficiently "spherical" inside, so it makes a good bend in the waveguide, like the bend in a saxophone. It's mostly keeping the length of the waveguide that matters, so there will be some tuning of the system while it's being built

[Banfili]

Looking very good so far, J.Wilhelm - no doubt the technical problems will sort themselves out as you go - all part of the fun!

[Gregor]

Wow! - very nice.  I will be watching this as well. I always like a good, steamy, audio build.

Lowes crown moldings are always a go-to for me when I need a decorative box-base for something. 
They come in a few different sizes and they save me time by not using the router table.
2 or 4 put together upside down, with the pointy part and side(s) cut to fit, a panel over the hole left by the pointy parts, and you have a chassis
for a music box, seismograph, or Flistershin dimensional door calibration device.

Good work!-gregor

[J. Wilhelm]

Wow! - very nice.  I will be watching this as well. I always like a good, steamy, audio build.

Lowes crown moldings are always a go-to for me when I need a decorative box-base for something.  
They come in a few different sizes and they save me time by not using the router table.
2 or 4 put together upside down, with the pointy part and side(s) cut to fit, a panel over the hole left by the pointy parts, and you have a chassis
for a music box, seismograph, or Flistershin dimensional door calibration device.

Good work!-gregor

Thank you for your comments! Ha ha! I practically live at Lowe's and Home Depot!   I have also used old picture frames to mount a PC motherboard.  The first Boombox has very thin crown molding and brown paisley-printed felt over the subwoofer.  The horns were aluminium wall lamps (discontinued) also from Lowe's. Easiest build ever, with a very high satisfaction return. For the second build, I used pvc for acoustic suspension enclosures and cloth lamp shades for horns! Also from Lowe's. This time, however the horns will be real, but hidden! I'm getting more cavalier and involved in the actual math of acoustics this time.

Cheers!

J Wilhelm

[J. Wilhelm]

Looking very good so far, J.Wilhelm - no doubt the technical problems will sort themselves out as you go - all part of the fun!

They certainly will. Today, I tried the digital amplifier with the larger set of speakers (the blue-coned ones, mounted in their original bass reflex cabinet). There is no doubt that the system sounds better with the larger speakers than the small ones (shown in the picture at the top of the page) . The problem with the smaller Sony system is that you only have 30 Watts to work with and the tiny bass reflex speakers sold with the receiver are too small to say, give you a good bass when listening to the music score of Ghost in the Shell (1995), for example, which has a lot of Japanese drum music (a great test for stereo systems). The blue speakers go considerably further down the registers, to actually give you a really nice bass, but it's just shy of subwoofer territory. I'm hoping the horns will drive the system a couple of octaves lower over the bass reflex methods (OK, I'm exaggerating   one would be enough).

Thinking that I might be short of power, today I made the mistake of buying from Goodwill another Sony receiver, this time identical to a wonderful system I actually have in storage too (but which would be a pain to dig out of storage). The system is a 100 W receiver with digital in/out AC-3 Dolby surround 6 channels, Virtual 3D surround, etc. Very nice, and I know the device like the back of my hand... I'm sooo partial to Sony. But for a good reason.

A 2000s Era Sony STR DE545 w/digital 5.1ch Dolby and 2 ch virtual surround

One problem though... Half of the power amplifier chips were blown. I took it back for a refund ($20), and they'd only give me store credit   So I have $20 in a card for whenever I find something nice at Goodwill, but I'm starting to think that maybe that expensive über ride to my storage would be worthwhile *sigh*. I carried the silly amplifier in my hand for blocks to/from the bus stop, and I was reminded why I never used a desktop amp for the Boombox. It rivals the subwoofer in weight. It's that power transformer. It accounts for 90% of the weight of the unit.

I'm afraid the 30 Watt digital amplifier will have to do. There's just no way a regular human being will walk around for blocks (even miles at South by Southwest) lugging such weight in one, even two hands. I did it with the Mk I for a decade, but then again, I'm not a normal guy  

Those second-hand places are definitely a gamble, but I learned a valuable lesson and also that I took care of my electronics much better than most people....

[Deimos]

Do you know how sometimes things just fit together for no reason? I couldn't have found a perfect match for the speaker drivers if I was looking for it. One of the things I hated the most about the boombox was looking for suitable lamp shades or bells to use as Victrola horns. This time, I found this pair of crown molding quoins or corners, which are relatively expensive at $15 per piece, but they were far too convenient to overlook. They simply save me a lot of time and work. ..

Very clever, my good sir. Veeeeeerrrrry cleeeeeverrrrrrr.

[J. Wilhelm]

SNIP
Very clever, my good sir. Veeeeeerrrrry cleeeeeverrrrrrr.

Thank you! That was holy serendipity. One of those once in a blue moon coincidences.

I've started assembly of the SoundBeam™, and I've fumbled a bit on my woodworking skills. The pine is very "splitty" (is that a word?  ), so I had to glue a split or two on the edges.

This is what the transmission line looks like now. It looks like I'm tuning some bizarre medieval instrument. It's very tricky to keep the planks aligned precisely while screwing them together.

I had to use screws because at this stage I want to be able to open the transmission line enclosure, to add and remove stuff like reflectors, batting, etc. So I used those self drilling "pocket hole screws I used last time. The only problem is they do make the thin wood split. My solution was to copy the assembly method of a cheap Scandinavian furniture company and drill holes in the wood so large that the screws are barely holding on to anything. There are so many screws that the minor misalignment and bending of the wood is enough to create enough friction. Once I'm satisfied with the operation of the enclosure, I will glue the planks together.

I may take advantage of those excess plank lengths protruding from the waveguide to form a "neck" to connect with the speaker boxes. I will need to estimate the additional distance traveled by sound and subtract the length inside the transmission line. I realize that's easy to adjust due to the sharp angle on the horn's bend. I can simply fill that triangular corner with a "reflector" that basically is the bend of the horn itself. Just move it back and forth a couple of inches to adjust the horn length. The rest of the triangle will be filled with a sponge material to absorb and deaden sound.

I'm crossing my fingers and hoping this horn design will work for the transmission line. Hopefully I won't just end up with a giant wooden stereo Vuvuzela.

[J. Wilhelm]

I think I found how to make end caps for the transmission line. It will most likely involve blocking the narrow part of the horn and open ports some other way. But in the meantime I found this wooden Kleenex box that seems to fit the bill for a waveguide "elbow" connection

If you find something that saves you a couple of steps, then it's worthwhile to do. In this case the wooden box is made of pine and the oval opening is made on a sheet of plywood. I'll investigate whether I can leave it like is or whether I have to make the hole larger, but I like it because the area around the elliptical hole gives me something to drill holes and use screws to attach the kleenex box to the waveguide. The bottom of the Kleenex box has a very thin sliding door which I can replace with a thicker piece of wood but it makes it easy for me to cut an opening large enough to make space for the Sony driver unit.

The goal is to try to make it as strong as possible without adding too much weight and acoustical problems

Theory has it that I will have to fill the transmission line with fiberglass or foam (better) and adjust the position and level of full inside each horn. For that purpose one of the sides of the SoundBeam™ remains attached only with screws so I may have access to the inside of the waveguide.

[J. Wilhelm]

The show must go on. If I'm going to die, I'm going to do what I like until the last day. Now I have plenty of time on my hands.

If you recall (just look up  ) I was getting together the parts to make the "elbow" of the speaker enclosure. The "false bottom" of the Kleenex boxes was made of ridiculously thin plywood, so I knew o needed to reinforce it with some solid board. I had some remnants from the planks I used to make the SoundBeam™, so I just roughly cut two semi triangular pieces to glue over the thin plywood.

There is a bit of complexity in cutting the corner of the Kleenex box with the tools I have in my room, now that we're in isolation. So I hacked my way with a "shark teeth" saw. I figure that it doesn't have to be pretty, it's all hidden anyway. The elbow will be filled with acoustic foam or batting. A British site that shows some plans for building TL speakers recommends that 80 percent of the length be filled with batting or acoustic foam. It all depends on which frequencies you want to dampen. I figure that together with the elbow, the batting in the wide portion of the horn will probably cover about 60% of the 1/4 wavelength. I'll just have to experiment. I will also place "reflectors," that is tiny solid corners on the inside corners of the elbow and the bend of the horn, to smooth the profile. The overall shape of the transmission line in my mind resembles a saxophone with the speaker mounted on the mouth of the saxophone.

[J. Wilhelm]

A little more progress tonight. The speaker enclosure will need to be finished for me to even start tuning the horns. I find it exasperating that I can't test the performance of the waveguide yet because without a sturdy baffle surrounding the speaker cone, out of phase sound from the rear of the cone is reflected forward and the speaker sounds really tinny.

So I started cutting pieces to finish the perimeter of the "elbow" and to make the front baffle.  I used a coping saw, because without a proper work bench I have to make due with what I have in my room.

This "elbow" extends the length of the horn. I need to be careful to add a reflector in the fold of the horn so the sound may travel the whole length of the horn, and not be extinguished in the very sharp corner of the fold in the horn, that should set the length of the horn to the right ¼ wavelength  for the speaker fundamental frequency of 84 Hz. The length should be one meter. But I don't know how sensitive the system is if I get the length wrong when I'm building the speaker. That requires some involved calculations, which are better done with an acoustics simulator.

[Prof Marvel]

....
This "elbow" extends the length of the horn. I need to be careful to add a reflector in the fold of the horn so the sound may travel the whole length of the horn, and not be extinguished in the very sharp corner of the fold in the horn, that should set the length of the horn to the right ¼ wavelength  for the speaker fundamental frequency of 84 Hz. The length should be one meter. But I don't know how sensitive the system is if I get the length wrong when I'm building the speaker. That requires some involved calculations, which are better done with an acoustics simulator.

Back in the bad old days, I was lusting after a set of the famous "Klipschorn" loaded speakers:




Since I could never afford them and did nit have the space anyway, I found a great DIY article in an audio magazine, and built a pair
of "loaded" speakers with a 12 inch rubber edge woofer, a 5 inch midrange and a separate horn tweeter. One "tuned" the box by sliding
a cardboard tube in and out of the box while playing specific frequencies, and watching a candleflame start to flutter in front of the tube.

point being, could you "tune" this horns with an adjustable tube? or is that an exercise in futility since your design is an open (loaded) horn?

yhs
prf mvl

[J. Wilhelm]

....
This "elbow" extends the length of the horn. I need to be careful to add a reflector in the fold of the horn so the sound may travel the whole length of the horn, and not be extinguished in the very sharp corner of the fold in the horn, that should set the length of the horn to the right ¼ wavelength  for the speaker fundamental frequency of 84 Hz. The length should be one meter. But I don't know how sensitive the system is if I get the length wrong when I'm building the speaker. That requires some involved calculations, which are better done with an acoustics simulator.

Back in the bad old days, I was lusting after a set of the famous "Klipschorn" loaded speakers:




Since I could never afford them and did nit have the space anyway, I found a great DIY article in an audio magazine, and built a pair
of "loaded" speakers with a 12 inch rubber edge woofer, a 5 inch midrange and a separate horn tweeter. One "tuned" the box by sliding
a cardboard tube in and out of the box while playing specific frequencies, and watching a candleflame start to flutter in front of the tube.

point being, could you "tune" this horns with an adjustable tube? or is that an exercise in futility since your design is an open (loaded) horn?

yhs
prf mvl

Love the idea of the candle flame  

Well, this horn is a 1/4 wave resonator, so the speaker is considered the closed end of the pipe, and thus its trivial to tune it by extending the length of the waveguide. Doing that with the thin rectangular slot which is the "outlet" of the inverted horn will be tricky but not impossible. I need to imagine a mechanism.

Really I have excess length now, and a nasty corner at the "bend"  in the middle of the horn that I need to plug (see post below). So it's really trivial to tune even now, you just have to open the bottom. I guess I could get creative and figure a way to lengthen or shorten the 1/4 wave resonator from within, but there's other issues that are more pressing. The horn is more than a resonator, it's an acoustic transformer by virtue of "absorbing" all wavelengths shorter than the resonance frequency. The horn's shape is partly responsible for that (tapering waveguide), but the transmission line calls for batting to be used for greater than 50% of the length of the waveguide to slow down the speed of sound as much as possible. The idea is that higher frequencies get attenuated as much as possible, including slowing the speed of sound for a specific range of frequencies, and the 1/4 resonator amplifies the lower frequencies close to the fundamental frequency of the driver. So I'm just following their instructions. If it doesn't work, I'll move to the next step, including what you described.

You can get a look at the setup in my next post below:

[J. Wilhelm]

So today I got to finish the baffle of the speaker. You really can't have any perforations or opening around the driver, because the speaker will sound like crap. So I had to finish the enclosure so that at least I can start testing the sound after mating the speaker to the waveguide/horn. The baffle is made from thick 1/2 inch pine board. Acoustic theory states that the ideal enclosure should be as rigid and dense as possible, and no gaps should exist. Those are the #1 and #2 rules to maximizing acoustic energy containment.

What I'm fearful about is that the "elbow" enclosure is far too complex with too many surfaces that basically waste acoustic energy, instead of directing it to the horn for processing. The ideal waveguide should be as smooth as possible inside, to direct and not waste the pressure waves.

If this project fails, it will most likely be due to the "elbow design" or because I got the fundamental frequency of the Sony driver wrong, and the horn should be much longer or shorter... We'll see. There is acoustic modeling software out there, but if I go that way that will turn into a bigger project. I'm not sure I want to look up (or derive out from the Euler Equations) the acoustic wave equations, and start calculating acoustic impedance by hand, element by element after changing the "elbow design" to something I can calculate by hand, like a Helmholtz Resonator mated to a waveguide or such (theoretically I can do that. I'm rusty, I took that class 20 years ago). Who knows? I may want to do that anyhow. This project is more of an experiment than anything else .

That little piece of wood I glued below is the "reflector" and the bend of the horn

The bottom of the SoundBeam™ is removable, and the first step is to add two reflectors, one for each horn (left /right). You'll immediately spot why this must be done. You can't have a sharp acute angle corner at the bend of the horn, because sound gets trapped in the crevasse, and the pressure waves are absorbed by the dense walls. That is in fact a great way to kill sound. So if this will be at least a 1/4 wave resonator, the bend of the horn should be smoother.

That elliptic opening probably needs to be widened or I can make additional holes to minimize acoustic energy being reflected backwards.

Cheers,. JW

[J. Wilhelm]

So today I started some tests to see where the sound of the transmission line (aka inverted horn waveguide) speakers is. At first value it sounds a bit flat compared to the equivalent ported bass - reflex speaker... Until I delved a little deeper by playing the opening theme to Ghost in the Shell (1990s movie) which features an impressive Japanese drum symphony. And that's where I realized that while the low to mid-range performance was ho-hum, the horn substantially outperformed the ported speaker I used for comparison..

Which brings me to the next point

I have a little confession to make. I'm cheating a bit because I know already what the Sony full range driver is supposed to sound like under optimal conditions. The donor speakers for the mid-range and tweeters I'm using are none other than a free set of Sony bookshelf speakers I got from a neighbor who abandoned a stereo system next to the garbage bin. The amplifier was toast, but the speakers were perfect. I'm finishing one of the two horns first and doing a side by side comparison with the equivalent ported system

The idea is to slowly match the performance of the bass reflex system and then *surpass* the performance of said system.

At the moment, I am not using any batting inside the waveguide, and the system is untuned other than having a quarter wavelength in size. On top of that I need to make a smaller cavity or enclosure to house the tweeter you can see on top of the wooden SoundBeam™. I need to figure how to reproduce the horn for the tweeter. Use Sculpey again? Or perhaps make a horn out of thin wooden veneer? The size of it can be seen on the upper part of the silver bass reflex cabinet.

The only other "tuning" I did was to open 4 holes around the existing elliptical hole to try to minimize the sound reflexion back into the elbow. As time goes by, I'll keep refining the interior the "elbow" to make sure the interior is as smooth as possible. I think I need to add that tweeter now so I can make a fair comparison.

But other than that, this is what I found: both the bass reflex and the transmission line speakers actually have very similar frequency response profiles (using my spectral analysis app on my phone and white noise - I'll post diagrams later), so you really can't tell by looking at the spectrum analysis; however, to the ear, you can hear some notable differences.

On the whole, the bass reflex system was giving better mid-range to low frequencies (without a tweeter installed, I can't comment on high frequencies), but when I played Kenji Kawai's Ghost in the Shell theme, the drums sets really came alive. At the lowest range, there were features of the drums sounds that simply could not be heard with the bass reflex system. The drums sounded much "cleaner" and natural, with a good physical vibration felt in the chest, like if you were next to the drums, whereas the drums on the bass reflex system sounded "boxy" and you could tell where the lowest frequencies were chopped off. But to hear those details you really need to play the right music material.

Kenji Kawai - Cinema Symphony - Ghost In The Shell OST

Clearly, although I can't see it at face value from the spectrum analysis, the lowest frequencies are being amplified below the cutoff point of the bass reflex, and the horns are reproducing lower registers very well which is exactly what I wanted to hear. The mid-range, I don't know yet. Batting will need to be added, since the purpose of the batting is to affect mid-range frequencies.

Quantitatively, transmission lines are said to perform similar to, but reach one or two octaves lower than equivalent bass reflex systems, and qualitatively are said to have a very "wide soundstage," whatever that means. I swear hearing the difference between an engineer and an acoustics industry professional is akin to listening into a conversation between aeronautical engineers and pilots. The professions are closely related, but both speak very different languages

So next, I will rush an installation of the tweeter. The test waveguide speaker is the left channel during tests, and the right is the bass reflex, but I'll want to make a monaural test to directly compare performance. I may need a better app to reveal the frequency response profile, before tuning in earnest. Or maybe not. Who knows? I'm a pretty impatient guy.

[J. Wilhelm]

I went all batty on my speakers...

I was about to go buy some fiberglass insulation at the hardware shop when I realized I had an old pillow that had been in storage for ages. 100% flame retardant polyester fiber, and no health hazard or trip to the store to collect more Coronavirus particles. Sounds like a good idea to me.

There's enough material to do this about 3 times from just one pillow.

[J. Wilhelm]

I added the tiny tweeter to the assembled enclosure already filled all the way from the speaker to the bend of the horn. So, about 60% of the total waveguide length was filled with batting.

The "elbow" speaker enclosure mounted over the horn. A tweeter is connected provisionally to match the other speaker
Notice a thin rectangular slot between the horn and the "elbow." This is where bass emerges from the horn.
The horn should be tuned to 83 Hz. Because of the geometry one slot faces front and the other back.

The initial results with the batting did not impress me. The batting just muffled the sound even more in the mid-low range. However, when removing the batting from the horn altogether, while keeping the elbow completely filled with batting, gave me the best results so far. The mid range sound still sounds a bit muffled, but a wider range of low frequencies opened up, and the drums in the music video above sound deeper and cleaner than they did yesterday with no batting. So at this point the mid-range is flatter but close to the bass reflex and the lowest range is now decidedly better than the bass of the bass reflex enclosure. I wouldn't say the waveguide is "better" at this point, but somewhat equivalent with a bit different sound coloration.

White noise plus the spectrum analyzer in my smartphone are no help at the moment. The difference between the bass reflex and the transmission line/waveguide is not something I can discern with this app, they're already very close in performance with a fairly flat frequency response, but my ears can hear the difference. Maybe more sophisticated software measuring one frequency at a time would quantify what I'm hearing.

The waveguide sounds nice already, and I could leave it as is, but I'm not too satisfied yet. I know it can do better. I need to hear a cleaner mid range and higher low range, and I'm under the impression that sound energy is being wasted at the elbow (the enclosure in the picture above, not the bend of the horn).

I need to think about why the batting works best for the elbow, but I do know that the overall length of the waveguide, including the elbow is too long, by the distance from the center of the driver's cone to the edge of the horn's inlet. Also I know the elbow is "acoustically dirty" so I'm not surprised that adding batting to the elbow improved the sound. Either way, the elbow is probably damaging performance with all those corners inside. .

I could try to extend the batting to the narrow part of the horn up to 86% of the quarter wavelength and see what happens, but what I will try to do next is to add "reflectors" inside the elbow to try to smooth the interior a bit and avoid reflected sound, that could travel back to the speaker, and see if that helps performance.

[SeVeNeVeS]

Now I can see the photos (I had a few problems http://brassgoggles.co.uk/forum/index.php/topic,50930.0.html)

WOW!

Watching with interest !!!

There is a Denon amp and a couple of speakers in my Mums back bedroom, so I may do something to plug my samsung mobile into, for Spotify.

[J. Wilhelm]

Now I can see the photos (I had a few problems http://brassgoggles.co.uk/forum/index.php/topic,50930.0.html)

WOW!

Watching with interest !!!

There is a Denon amp and a couple of speakers in my Mums back bedroom, so I may do something to plug my samsung mobile into, for Spotify.

Thank you!

And by all means go hook up that Denon. Here's a tip; you can buy one of those $15 Bluetooth earplugs type devices (stereo), and with minimal soldering skill, you can solder a phono plug or RCA style plugs to hook up to the Denon inputs. The quality will be decent because there's practically no power amplification related distortion when the output power is so low (if you try a portable speaker, the distortion is much higher, plus the Bluetooth speaker is relatively expensive and usually monaural).

[SeVeNeVeS]

I have most of the house covered from the PC to a Technics amp, gragh and Mission /Tanoy speakers, but I thought maybe a portable-ish device for the garden and workshop.
I was thinking direct input using one of these 3.5 to phono.

Spoiler: ShowHide

Anyway, a looooong way off, so many other projects on the go.

Sorry for rambling, please do keep the updates coming as they happen.