Guitar Transmission-Line Cabinet

[Brenda]

I have built a few types of speaker cabinets. I’ve made sealed and ported cabinets. I helped a friend make some line-array speakers. I’ve even helped repair a real subwoofer that was part of Pipe-Dreams ($16,000/pair used.) Though, I never made a transmission-line cabinet before.

Jensen came out with this cool guitar speaker, which is almost like the beloved P12Q, but it has a stronger magnet for +2-3dB gain over the P12Q, and not so much basket to reflect, so I got a Jensen Blackbird 40. I like the green-blue of the the P12Q; I just have not to dwell on the the back color, somehow. Anyway…

Looking at the speaker, which is really 1950’s technology, with a more powerful magnet. These speakers were designed to run on a plain baffle plate, hoping for an infinite baffle so the backwave never meets the front, but that doesn’t exist. Well, let’s see, if it doesn’t want to run sealed or ported, which is really sealed up down to the tuning point–what can be done to get bass out of it without dampening the cone itself?

Well, I thought it would be fun to make a 12" transmission-line speaker, for playing at the coffee shop. The semi-transmission line has a bit of taper to prevent standing waves in the port. The inside of the cabinet will be lined with neoprene-backed indoor-outdoor carpet to attenuate the wave going through the port. The cabinet and port will have some fill, but how much, I am not sure. Oddly, this thing, being a guitar speaker is not meant to accurately reproduce sound, but be musical. My goal is also recover some bass that would otherwise be lost through wave cancelling. The port length is 900mm, which is why there are folds in it. It’s cross-section is related to the driver frontal area.

The cabinet will be powered by a Parts-express 25-watt digital amp, housed in a separate box. It might have wheels and a pull-handle, because I have health issues.

Anyway this is what I have been working on. I have most of the wood cut. Sadly. my CNC machine is in storage, again, so I have to make do with the tools I have.

I didn’t fuss with the renders. It’s funny, I don’t recall ever rendering cardboard tube before. The black thing on the back is a simplified Speakon connector. I have to order those. Rhino’s decal worked very well for the cone in the rendering. I was able to do a volume calculation, and multiplied it by about 44lbs a cubic foot, which gave me an idea of the weight before lining it. The speaker is only 4.5 lbs, there is so much Alnico magnet, that it has more magnetic strength than some smaller neodymium magnets. They call the magnet “fortified.” I suspect that they did the anisotropic grain organization on it to get it so strong.

Minimal rendering set up except for that decal that’s really making the cone look like something, whereas it’s just 2 lines and an arc for a dustcap, revolved. There’s not even a bumpmap. LOL!

1/2 Sheet Layout

Parts before sanding.

Cutting these by hand needed some truing up. Yes, it’s a 12" speaker!

[Brenda]

I wasn’t sure if I wanted to buy a router table just for this project, but gnawing all of these box joints out with a jigsaw was almost as much a chore–as fitting them.

The later ones, where the wood was cross-grain scored with a sharp knife before cutting helped accuracy–because the cut-line wasn’t being splintered out.

The files have too much surface area cutting at once to cut in a timely fashion, so I made some sanding sticks, which really cut through the plywood end-grain. The sandpaper is held on with a bit of the 'ol DuPont #77 spray contact adhesive. The sandpaper is around #80 grit. The corners were done with a safety-file (one side smooth) and the riffler file. To the right, under my glove, you can see the sandpaper on the endgrain for cutting the groove root?

I nearly chipped out this crap plywood’s twice, and had to stop and glue the chips back down.

The irony that I own a CNC machine, which I have somewhere to set it up, is not lost on me.

The fullsize box-joiny template was seamed together with glue-stick, whilst up against a bright window, so you can see the lines through. It’s holding up well.

While sanding and filing, I wear gloves because the sawdust dries my hands out crazy. Then when I play the guitar the skin on fingers will crack.

[Brenda]

I am starting to glue the cabinet cabinet together with Titebond “Ultimate” “Waterproof” glue. For stuff like that will have lacquer for a finish, I don’t want to resort to polyurethane adhesive such as Gorilla Glue.

It’s one thing to design something, but it took me a while to think of the order of precedence for assembly, as the box needs to stay open–until it’s mostly lined with dampening carpet and polyfill. Also, this may be a one-off, so I don’t want to build a lot of fixturing for it.

The Home Depot plywood is meh, has too much moisture, chips very very easily, and if you want a another opinion–the veneer is way too thin. It has a rotting lettuce odor, too, instead of sprucy goodness.

Compounding the plywood issues, I am making it with handheld power tools, so, the parts need fitting. Also, it’s been rainy here in Silicon Valley, and the glue has been taking a long time to set, as in, it’s been six hours, and the glue was still soft in the middle?

Anyway, there’s progress. Speaking of fitting, I need to sand the top edge of the upright port parts because they are a little proud.

[Brenda]

The main port tube or last run in the transmission-line are glued in place. I am using waterproof glue, and it’s unusually rainy here in Silicon Valley, so it’s going to take time to dry.

Let’s see, weather… Oh Humidity 76%, Chance Showers then Mostly Clear. Actually it rained whenever I got too close to the garage door.

Yes indeed, I should have more clamps. You can see the 8mm thick cardboard tube parts, which is at this point, is just being used as spacers. Notice the canted scrap used as a wedge. Way back when, blacksmithed metal things like nails were so high-tech, they made wooden wedges. (Pst. A screw is just a wedge wrapped around a mandrel.)

This is one strange speaker box. If it were all CNC’ed, I would have been at this point the first afternoon. Instead my CNC is taken apart and is in storage, and I have to hand fit parts. I had read that a chisel doesn’t work too well for plywood, but if I ever make box-joints by hand, I’d try one. I sanded these to fit, using 2 different homemade sanding sticks. I have a safety and rifler file, but they cut so slow. The joint quality is poor, but I have glue and sawdust; everything will be okay. I was going to buy a router table just for box joints, but I considered the larger picture.

So, that toppish board going horizontally, is actually tilted to keep the standing waves down. Actually, the whole 900mm of transmission line is tapered, making a bit less of a transmission line. It’s going to look cool when all the radius tubing is glued in place.

As Nurdrage says, crushing your expectations: This is not going to one of those audiophile speakers, that you have to put your tweed jacket on, hold a glass of cognac over a candle, and hold your breath to make your blood its bluest. It’s going to be a busking-style electric guitar cabinet, using a 1940s inspired vintage-style paper coned speaker with tar around its perimeter. As a design objective: it must make a good coffee cup holder. It will likely have wheels and a pull handle. It has a “fortified” AlNiCo magnet, which I take to be the anisotropic process to align the crystals. Anyway, it’s supposed to have 2-3 dB/1W gain over the P12Q, which means longer battery life. Shunning fancy gold-plated binding posts, the cabinet will have a lowly SpeakOn connector, 4 pole, perhaps wired as 2-pole for lower resistance.

Wow! look at the void in the plywood. I feel dirty for buying Home Depot plywood.

[Brenda]

Shown from the back…
Diverters, and carpet tube radii added. Clamps were added after photo was taken. You start to get a sense of the twin 1/2-sized transmission lines. The T-line is 900mm long; the cabinet is, I think 340mm high.

The 8mm carpet tube as pretty solid, and not as resonant as the plywood. It soaks up glue like a sponge. This is all “waterproof” glue. I am not fussed about making the inside of the cabinet look tidy, inasmuch as making sure it is sealed and strong.

At this point, the enclosure weighs 8.8lbs, or about 4kg, so I think I will be on track for 20 lbs for the enclosure. The driver weighs 4.4lbs. The wheels and handle will weigh some too, as will the indoor-outdoor carpet, that will line the enclosure. I am going to use it to aid in attenuating the back wave.

Shown from inside…
I did eventually clean up that glue hangar. The 2 side ports are all the coupling the driver will have. I wish it were better. The transmission line port is tilted just a bit, to minimize standing waves. I am tempted to place a curved diverter to better couple the sound wave through the transmission line, though it will eat cabinet volume. Hmmm.

Looks a bit better, now that the glue is drying. The box is getting stronger, yet neither the braces, nor the back and top are on.

Part of my purpose in posting this is go encourage people to get some hand-on making the things they design.

One can tap on this case, press on the parts. You can can feel what resonates, what’s weak, and what’s strong. This case was a challenge because, unlike an audiophile that can be made of 25mm MDF, this speaker cabinet needed to be light and portable.

I try to be objective about what I design in Rhino.

Looking at the last pic, looking through the speaker(driver) whole, that vertical part, and the 2 small parts, really should have been one, and the uprights on either side, should have been notched. I thought this was more miserly with the 1/2 sheet of plywood, but doing it the other way would have speeded the glue-up.

That rectangle part attached laying on the bottom, could have been lightened, because there will be a tube section over it.

[Brenda]

I got a like from the person, from whom I had bought a guitar. How could that not cheer me up? : )

[hanscad]

This is very cool. I’m interested in the acoustics beyond what you’ve described.

The extra-long port should push the fundamental way down below what a guitar would produce, is that the main idea? Maybe even far enough that it doesn’t produce too many overtones either?

More imortantly, how does it sound?

[Brenda]

Oddly, firstly, the guitar speaker’s resonance is just slightly above the lowest note on a guitar, by about 100Hz. Still things get strange in music because, for instance, when you tune a guitar, it produces beats that are less than 1hz, but that’s not playing, yet when you bend certain notes, it does the same thing.

(Generally, it is inefficient for a bass driver woofer to produce tones below its resonant frequency.)

Generally, guitar speakers are a throwback from the pre-hifi days of radio and amplification. The suffer from edge modes from their paper surrounds. They have smallish voice coils for higher end response. Guitar speakers are largely full-range drivers, originally designed for open-back cabinets, baffle plates (optimally Infinite baffles.) With thin paper cones, and small diameter voice coils, most guitar speakers weren’t really designed to be put in a sealed cabinet.

Though, if you leave the back open on a baffle-only type enclosure, then the back wave may either cancel the front wave, or reinforce it, making peaks, like “wolf notes.”

So the mantra of my transmission line is: to couple and conduct the back wave away from the driver, with as little discontinuities as practical in a small case, and then to attenuate that wave as it travels down the transmission line.

The slight angle I used somewhat mitigates the whole transmission-line mantra, but for that loss, it helps prevent standing waves from forming inside the transmission line.

There are formulas regarding the ratio of the driver diaphragm VS the cross section of the exit “port” of the enclosure. Also, the length of the transmission line is proportional to the wavelength of the speaker’s resonant frequency.

[Though a transmission line differs from an “ported” speaker, a bass-reflex enclosure in that, the bass-reflex enclosure is tuned to a certain frequency. Above that frequency, the cabinet is largely sealed. Below that frequency, the wad of air that had plugged up the port tube–is expelled, like a pea shooter. Bass-refrex offer free bass just above the tuning frequency–at the cost of deeper bass below it. Sealed enclosures can deliver deeper bass at lower frequencies than a bass-reflex enclosure.]

So, I have a speaker designed to be efficient and musical. It doesn’t want to be in an sealed/acoustic suspension box, but I want a bit more bass than a open backed cabinet, and I don’t want the non-linearatiy of a ported enclosure.

Still, this is has been a fun project, and I don’t know how it will sound yet. Once more, I’ve never heard of a guitar speaker in a transmission-line cabinet. Though, I have heard my friend’s JensenP12Q, which has the same cone as the Blackbird 40, but the P12Q has a less powerful magnet. I am not sure how it will sound. : )

Because of the driver’s high resonant frequency, I am not worried about the cabinet having too much bass. Tapping on the speaker cone reminded me that this indeed a full-range classed speaker, and not a sub or really even a woofer, in the modern sense. The sound that comes from the port will be delayed and more importantly (to me) attenuated.

[hanscad]

Ok that all makes sense. It’s almost the opposite of a tuned subwoofer, where the transmission line is more acting sort of like a soft helmholtz resonator that eats up the cabinet resonance if I understand correctly.

[Brenda]

I have seen Helmholtz resonators inside transmission line cabinets. : O

It’s very freaky how closely they match the choke couplings in a radar waveguide. It’s also freaky how close sound, light, end RF-EMF acts.

Waveguide flange | Choke connection

If a transmission cabinet was long enough, it could be sealed.

In a better transmission-line cabinet , there should be no backpressure to offer wave reinforcement/deconstruction.

I feel that part of what has made transmission-line cabinets difficult is: the consistency of the attenuating stuffing. Generally people fill 1/2 of the port with stuffing. I was trying to find materials that will attenuate the sound yet be consistent in density. My first line of attenuation will be rubber-backed indoor-outdoor carpet, which has lined 2 cabinets, and 2 computers.

[Brenda]

Laminated cardboard concrete tube was blended with the enclosure. I decided to add another layer, though 2 layers are well-strong enough to lift the cabinet by. Three layers are over 8mm thick.

There are 7 more tube sections to cut, laminate, and glue in. : O

[Brenda]

Made some more tubing segments for laminating. Added the 3rd layer for the two radii I started. The layups are pretty strong for cardboard and wood-glue. There are several more radii that need to be made, with 3 layers for each, as well as various bracing, the top and back. The transmission-line will be lined with rubber-backed carpet to attenuate the back wave. There will be the traditional 50% fill of polyfill, too.

It was fun to for the first time really compare the real cabinet and the design.

[Brenda]

Laminated radius segments and three braces were added, all written in the passive voice. Unnecessary articles omitted. This may be a better way to do the bracing. Still needs dowels to couple the center-section and the cabinet side-panels.

[Brenda]

When assembling the case, I tapped on the upper side panel, and found that upper dowel brace in the upper transmission-line wasn’t needed. Another brace was changed, creating the opportunity to change the side-brace shown, which appears like a dog-leg.

Because the driver/transmission line coupling in this lightweight portable case is not optimal, I don’t want to further mitigate it.

So I am going to try a side brace made in such a way that the support is moved forward into the case nearer the driver, and away from the already troublesome t-line/port transition/coupling. For the sound waves that are more or less in line with the brace, there would be a mixture of eddies and diffraction behind the brace, but from other angles, keeping the brace away from the transition should help the coupling. Although the brace will still be in the way, because it is angled and curved, that should spread out the reflection time, but because there is no free-lunch, the amount of reflection heading back to the driver will be increased because there is a surface closer to the driver.

Filleting/raidusing the brace should help with the reflection. Perhaps I will put sound-absorbing material on it. I’ve been using random-looped rubber backed indoor carpet for my projects, with favorable results.

This is fun, and was useful, for instance showing how much affect the bottom brace now has. It’s good that it’s rounded in the real box. To some degree it can show fall-off, but obviously does not take frequency into account. I use an emissive copy of the cone. The contrast in the rendering is quite high. There is a Grasshopper plug-in for sound, but I had issues with it. I’ve made rendering that the reflection is also turned up, so the light comes out the port (end of the transmission line)

[Brenda]

The side braces are in. There is only about 10mm square cross-section of the brace actually inside the port VS an entire dowel, as before. The center-box is the strongest part of the case, and a good place to brace the side panels from.

If this was 25mm MDF, these side-braces likely wouldn’t be necessary. Weight and size were figures of merit on this portable case–for a 12" driver.

No, actually, I’m not concerned with the glue appearance inside the box. I’ve been adding glue to additional glue to the joints to keep a slight fillet–while the glue dries and shrinks.

[ptcdesignau]

Very cool project. I had an old Leslie type amplifier which I converted to a guitar amp. Sadly lost it a few years ago but have always wanted to recreate it rotating speakers and all. This gives me a bit of inspiration.

[Brenda]

The last radius in made, fitted, and glued in place. Speaker wire holes drilled. Speak-On connector recess done. Small brace added(not pictured) Corrections started…

…I have discovered that I had thought that I made two of the same part, but instead–I accidentally used the cutoff part for a real part. The parts are only a few millimeters different, but this bit of difference meant that the two vertical splitters leaned forward more than they should, so I have had to add material and shape a few parts. These things happen.

Otherwise, I am pretty pleased with how the cabinet is turning out. I figured on 44 USCU pounds per cubic foot for the material, and this plywood is lighter than that. So, the weight is also less than calculated. I need to weight and measure a piece of this plywood, just in case I use it again. The cabinet without stuffing is around 13lbs or 5.9KG without the grill.

Even at this point the enclosure is much stronger than I thought it would be. You can see in the top render, that there were dowels to strengthen the upper rear of the cabinet, but the 3-layers of cardboard tube were so strong that I found I didn’t need those parts, so they are out of transmission line. I had convinced myself that this cabinet would be a compromise over the last for strength, but even with the back and top still not on, I have stood on the side of the cabinet, full weight. The cabinet feels almost like a composite structure, but it’s wood, glue…and cardboard tube : )

You may see a little radius in the upper right that connects to a thin long strip. This worked pretty well to strengthen the long strip, which will be weak–even though it’s to strengthen the rear panel. These ribs end up being like guitar or piano soundboard braces. I will probably put a bit of radius on it, too.

Sorry about the glue mess, but it’s not going to be seen after the cabinet is done. While the glue joints dry, I’ve been adding glue to make sure there are no voids, and there’s a bit of radius, albeit a millimeter.

You might also notice that the two vertical splitters have a thin shim strip, correcting the leaning issue.

I think that it’s important to get out there and make some stuff. It makes you a better designer. Except for the cardboard adding so much rigidity to the structure, I had a good sense of how strong the plywood was before I started drawing this in Rhino, because I had made a few of these. Still, parts of this were harder to make than I though. For instance, I thought the inside radii and the outside radii would be about the same trouble to make, but the outside radii went quick and accurate, and the inside radii took a lot of work. Asymptote or not, I didn’t think it would be this bad. What made it harder was that I had to laminate up the inner raddi; and that finishing the radius to size was hard because it stood up to abrasives and cutting tools alike.

[Brenda]

So, when are you going to start on the Leslie : )
How many mistakes do I need to make before you start?

[Brenda]

The speaker grill is raidused(filleted), so that it doesn’t print-through/(telegraph in composite work). After making sure there is another room for the grill material, I will treat the edges of the grill.

Radiusd the transmission line outlet/port on the back panel. There is a small chip I put in the veneer, and a much larger void, which doesn’t look like standard tear-out. I have had problems with veneer delamination with Home Depots “sanded” plywood, which needed regluing, but the small chip was my fault. The inside of the port will be black and/or covered in rubber-backed indoor/outdoor carpet, to dampen the transmission line.

Edit: the back is fitted but not yet glued.

It’s fun when it starts looking like the Rhino drawing. : )
I am contemplating covering it in Tweed, after testing.

[Brenda]

Enclosure lined with indoor-outdoor carpet for better attenuation. Poly-fill added to top area. Top panel glued on. Large radius routing done on cabinet and grill. Front of cabinet lined with carpet.

Deviating from the Rhino design, larger fillets were used. Bevel/chamfer changed to fillet on upper grill.