Victorian Boombox Mk III. A brand new start.

[Prof Marvel]

Wow. J !

this took a sudden left turn!

I think you are on the right track with your bandpass and cutoff filters.
I haven't played with bandpass filters since my amateur radio days....

Using springs as coils is intruiging, if you are having issues with streching them out, simply anneal them first,
taking the "springiness" out. This can be done with a propane torch or on a gas stove burner.
The trick is to achieve the stretch first, say on a steel rod, then heating them until red and allowing them to cool .
that will make the spring take the spacing you made while hot.

I hope that makes sense.

Of course if they are plated be aware of any gaseous ickiness whilst heating.

btw if your aren't careful you "might" get some sort of "reverb' affect lol ....

yhs
prof marvel

[J. Wilhelm]

Wow. J !

this took a sudden left turn!

I think you are on the right track with your bandpass and cutoff filters.
I haven't played with bandpass filters since my amateur radio days....

Using springs as coils is intruiging, if you are having issues with streching them out, simply anneal them first,
taking the "springiness" out. This can be done with a propane torch or on a gas stove burner.
The trick is to achieve the stretch first, say on a steel rod, then heating them until red and allowing them to cool .
that will make the spring take the spacing you made while hot.

I hope that makes sense.

Of course if they are plated be aware of any gaseous ickiness whilst heating.

btw if your aren't careful you "might" get some sort of "reverb' affect lol ....

yhs
prof marvel

This is what I'm building tonight with a scrap piece of wood. The weirdest crossover, which I'll finish tonight, just to see what it does. The high resistance is useful at the moment.

The ticket will be to make inductors in pairs to insure symmetry. You stretch a coil and distend it then you measure its length and calculate its inductance. Trim the length to give you twice the inductance you need. Divide it in two, that's your pair of solenoids for left and right channels.

As for capacitors, I scrounged up what you see. To make the second pair, either I wait for mail order, or start making my own Leyden jars!!

PS I like the idea of making my own components and displaying them on the device in a Steampunk way. Impossible to service except for the original maker. "Our system is so exclusive, we make our own parts. Only inexpensive systems use copper. " Is that Steampunk snobbery?   Maybe it's time to start looking at a custom made vaccum tube amplifier  

[J. Wilhelm]

I'm glad to say that the 4th order low pass filter is low passing as expected. It's expectedly muffled and without any unwanted oscillations like I get with the Mk. II cheapo subwoofer.

So this is the test setup. I ended up using screws with washers as connectors, since everything is difficult to get at the moment. Those type of connectors were common in 1970s component systems when I was a kid, and I had forgotten how ridiculously obnoxious those connectors are.

I'm not sure how well it's doing, because I really don't know the exact cutoff frequency yet. But it sounds an awful lot like my Vic Mk I boombox sounds when I forget to connect the side speakers. After calibrating the cutoff frequency properly, it should be a matter of determining whether or not to use a separate amplifier. In other words, it's down to how much power I need in that speaker to get the effect I want. I can't do high power tests tonight because it's past midnight, and everyone is asleep on the house 

[Prof Marvel]

As for capacitors, I scrounged up what you see. To make the second pair, either I wait for mail order, or start making my own Leyden jars!!

Aha! I have just the thing for you: layers of aluminum foil and paper.
repeat
roll up if you want a tube, or just leave flat and encase in "something non-conductive".

If you can scrounge old power supplies you may find the caps you need...

[Banfili]

And here I was thinking the set of music lights I made many years ago that clipped onto the speakers in the back of my radio/tape player were pretty good - sigh! It was my first electronics project, and I was so proud of them!!

[J. Wilhelm]

And here I was thinking the set of music lights I made many years ago that clipped onto the speakers in the back of my radio/tape player were pretty good - sigh! It was my first electronics project, and I was so proud of them!!

Yes, but small steps lead to great things! It's all about persistence!

As for capacitors, I scrounged up what you see. To make the second pair, either I wait for mail order, or start making my own Leyden jars!!

Aha! I have just the thing for you: layers of aluminum foil and paper.
repeat
roll up if you want a tube, or just leave flat and encase in "something non-conductive".

If you can scrounge old power supplies you may find the caps you need...

I'm kind of hoping I won't have to go that far. But there's a certain satisfaction in building them yourself.

I'm ruminating taking advantage of those springs and use them unsprung as resistors and sprung as variable inductors. Traditionally a roller is used over the coil to shorten or lengthen the inductor. But stretching the spring will have the effect of reducing the inductance. Perhaps make a tunable subwoofer?

[J. Wilhelm]

I'm getting very good results from this setup. It's still not Vic Mk. I Subwoofer level, but I'm hearing a very deep rich bass, - very clean too - including subwoofer range sounds (between 30 and 60 Hz, based on sound coming from each port), and that is simply head and shoulders above anything coming out from the Sony Bass Reflex speaker now. The horn is waking up.

The bass sounded somewhat "soft" in Kenji Kawai's GitS theme, but the bass was definitely stronger. It's hard to tell in that song how low the bass is tuned because you're hearing drums that activate the whole bass bandwidth. But I have other sound sources that better measure the performance of the bass, like Karsh Kale's Realize song (see previous posts for links), and that's when the lower bass came to me super strong and clear. The reason is "Realize" is a song I've heard many times in the Vic. Mk. I boombox, so I know how deep the bass goes or should go - to my personal taste (in fact most of us have heard that song if you ever purchased a computer with Windows Vista, or an iPad /iPod, because that song was used a lot as a demo in consumer electronics in the last 15 years (?) - the royalties from the song must be awesome).

This is another cool sounding song that I stumbled upon. It sounds "laser sharp" on the horn, with a much better low bass.

Midival Punditz - Rebirth

 It's a good idea to test the mid range as well, because I'm hearing very clean pleasing sound from 200-300, and this probably means the little subwoofer is filling those low performance points we talked about in the graphics on the last page. Me thinks having the subwoofer tuned a little high has extra benefits beyond 100 Hz mark up to 300 Hz. It'd be great to see the curve flattened where the bass reflex ports steal energy from, to pump as lower frequencies.

The other source I like to hear in order to judge bass is Bebel Gilberto's 2000's-Era Electrónica Bossa Nova album, "Tanto Tempo," because she has this deep wonderful voice and the jazzy electronic mastering of the album is nothing short of phenomenal (if you've never heard of it, it's included in the book "1001 albums you need to hear before you die" ).

Bebel Gilberto - Cem Contencao

Bebel Gilberto - Tanto Tempo

Bebel Gilberto - August Day Song

Bebel Gilberto ~ Samba Da Bencao

Depending on your type of hi fi system you will either hear a wonderful bass or not much at all, because in the mixing they used synthesized low frequency sounds that are very restricted to the lower bandwidths unlike physical instruments (these were the glory days when Electronica had just become mainstream and Moby was making videos with Madonna). Listening to the speaker right now I can tell you that the Sony Bass Reflex is hardly registering some lower pitched sounds that come loud and clear in the Vic. Mk. I boombox, and the horn can play them but not to "wow" levels the same way that the Mk. I can perform. The difference is especially audible in the songs Çem Contençao, and August Day Song, above

So at this point, I'm regarding the passive subwoofer to be a necessity. Listening to some of my favorite songs it's becoming apparent that the Sony Bass Reflex, as good as it is, it just doesn't make the grade for what I would like. The new challenger to beat or at least match to an acceptable degree is the Altec Lansing ACS 340 active Subwoofer in my Victorian Boombox Mk. I. Getting to beat the Altec Lansing subwoofer will be very difficult and I may not even want to go that far (though I have a secret weapon I'm trying not to use - the "nuclear option" which by all means is tantamount to cheating),  but I do think that I should switch to the Altec Lansing as my baseline reference from now on during measurements. That means I have to dust the boombox off, jiggle the connections (that's the illness vexing my old Vic Mk I - the connections are loose and it will be suddenly stop working for now reason - it's usually a jiggly connector) .

[J. Wilhelm]

Guess what I found for $7!  A walk to my local second hand shop produced this unit:

It's the wired version of the subwoofer I bought first. From that unit I extracted this speaker below, which is not exactly identical, but very similar to the one I'm using for tests now. The one I bought today is  simple bass reflex box with no electronics nor a crossover, same dimensions as the other one - probably meant to work with a stand alone receiver. It's older - made in 2012. The one I purchased before was a wireless version dated to 2018. Consequently, the drivers are not the same and have different labeling outside, although that tells me nothing about how different they are. Our=twardly they are very close dimensionally, and both have the same (resistive) impedance of 4 Ω. The difference seems to be a slightly larger magnet for the older unit, about 2 mm. larger, and with a smaller cone, about 4 mm smaller. The center cardboard disk is the same size for both. The frame of the driver is externally the same for both units. I figure these are two variations on the same design over the span of 6 years, so I'll assume that specs will not differ substantially.

The unit on the left is 6 years older has a larger magnet (+2mm) and a smaller cone (-4mm)
Samsung part number on left unit (2012) AH81-07058A
Samsung part number on right unit (2018) AH59-02735A

The center disk is the same. When finding this unit I realized there was minor damage to the cone. It had a soft spot and a crease, presumably where the cardboard had been kicked in by something. The way to solve that is to mark the spot and apply epoxy glue behind the crease/tear to stiffen the cone again. The increase in mass should be negligible, and I added another dab of glue on the opposite side to balance the load (not that it needs it).

I could also, apply a coat of lacquer in the front , or if I want to get real fancy paint the cardboard with colored enamel, for show, like flat or gloss white, for example. But at this point I'm willing to accept minor spec differences and repair the small damage, because it's fairly rare to get two identical speakers in second hand shops, and I'd have to fork over anywhere from $40 to $120 for a new pair of subwoofer drivers from some place like partsexpress.com. I guess I could do that, but I'm interested in getting results now and $20 can be used for something more urgent at the moment like purchasing capacitors online. I'm going to assume that for very low frequencies, a small variance from left to right will go unnoticed, because the wavelength of those frequencies is very long (i.e. 4 meters for 80 Hz) and so the difference in sound pressure from left to right will be imperceptible. It's certainly better than using only one subwoofer mounted on one horn asymmetrically and better than just having one center driver with its own enclosure. And if I can always find another Samsung subwoofer somewhere, as long as the driver frame is the same size..

[J. Wilhelm]

Aaaargh! To be or not to be; a minor setback (maybe)

Today I moved the subwoofer driver from the end (start of the line/"end cap") of the horn to one side. I was going to place the subwoofer driver facing down - like most subwoofers spread the bass, in anticipation of building the other right horn speaker enclosure and having the horn lay horizontally, ready for comparison tests with the Vic. Boombox MkI. I wanted to re-introduce the end cap with the ramp, which I know can add 1 to 3 decibels at certain frequencies. I really never intended to have the subwoofer at the end, for fear of sound leakage from the main driver. I was also thinking of adding a plate reflector at the bottom to spread the bass 180 degrees around the room.

Today I moved the speaker from the top top the side and re-installed the ramped cap
I thought it was a good idea.  

The problem is that immediately after I hooked the speaker back I could hear a difference from what I heard in the last few days with the subwwofer mounted at the end of the horn. And the difference is not good. Basically I lost all of the rich bass that I had gained in my informal listening tests in the last couple of days. That literally was just a few hours ago, and I'm trying to find out what happened. The rest of the horn doesn't sound very different but informal tests (using my ear) and switching the subwoofer on and off, show that the bass effect of the subwoofer has gone from quite noticeable and *very* clean, to something barely perceptible at all - not worth the effort! I'm trying to think what could have caused this loss, and naturally, I checked the connections and whether the subwoofer driver + crossover are working - they are.

Perhaps the enclosure has become too crowded, and the waveguide is now acoustically compromised

The volume of the subwoofer driver itself has gone down dramatically. There's two possibilities:

1. The volume taken out by the number of drivers in the box is now compromising the integrity of the waveguide . The passage is too narrow now.
2. The subwoofer in "passive mode" (ie no external amplifier just feeding from the same line) was in fact resonating with the horn.

It makes sense that if I put too many things in the enclosure that the acoustic pathway will be compromised. At some point you destroy the ability to form standing waves. It's just that when I had that situation it was all the way back when I had those "elbow" enclosures.

It also makes sense that the subwoofer, placed at the end of the horn, was feeding back from the standing waves of the horn. Not as a passive resonator, but a combination passive plus powered resonance.  I recall the horn first achieving resonance when I placed the Sony driver at the end, and i got an "uncontrolled" resonance, in which the cone displacement was so large that the speaker was clipping sound. So I know that the horn "liked" to be fed directly from the end. But the clipping coming from the Sony driver was unacceptable. That's when I made up my mind not to place speakers at the end of the horn ("start of the line" in waveguide terminology). But this Samsung subwoofer driver is substantially stiffer that the Sony driver, and also has a much smaller cone, and a huge magnet in the back that is practically a solid wall by itself. It could be that standing waves are being formed on the surface of the magnet (the "start of the line"). It's not a ramp but a flat, very solid wall

I have a choice now. To ramp or not to ramp. I can't have my cake and eat it too.

*Sigh* Double sigh and triple sigh. It seems that the horn is giving me a choice. I can have the subwoofer driver in passive mode but no ramp at the end, or use the ramp at the end and place the subwoofer in a separate enclosure whether powered by a separate amplifier or not. The horn will not accommodate the subwoofer driver on the side. And I thought this was a "shoe in."   Now it seems at least I will have to build yet *another box* just to resume testing - and this time it's a fairly complicated affair, except that I don't have to build a ramped end cap any longer - which I'm sort of glad about , because the ramp is a "young female dog" to build.

The sound that I was getting yesterday was quite acceptable, and I was getting very strong resonance at the ports, especially the 60+ 30 Hz port of the horn. Very strong and you could feel the vibrations on your fingertips in the vicinity of the port - not just air blown out of the port, but string vibrations, even at low volumes. There's no doubt about it, I think that the horn responded well to the subwoofer driver placed at the end, and I'm unwilling to lose that. I'll make another test tomorrow to see if I'm wrong - but just one listen to the same song and I knew what I was getting.The thing that bothers me the most is having to build a new box - actually two. Now that I have the second subwoofer driver, I'm going to want to build both boxes at the same time.  

But I have to BUILD ANOTHER BOX, GAAAAAH!

[J. Wilhelm]

The sudden realization that you've created a monster.
It's so big, I'm having a hard time photographing it all.

Hmmm. I'm staring to get ideas....

[SeVeNeVeS]

Still watching the trials and tests with much interest, Keep going J.

[Prof Marvel]

My Dear J

it appears to have become....
less than portable?

yhs
prof marvelling at the endevour

[SeVeNeVeS]

My Dear J

it appears to have become....
less than portable?

yhs
prof marvelling at the endevour

Fair point, you may need to Steampunk one of these as the next project to transport it 

Spoiler (click to show/hide)

Spoiler (click to show/hide)

As a side note, years ago I went to a Dub Reggae night, the sound system was, VERY bassy, felt like I was being thumped in the chest and my, umm, mid gentleman's area, rattled with every beat, strange experience, me and my mate the only white faces in there, thoroughly enjoyed every moment.

[J. Wilhelm]

My Dear J

it appears to have become....
less than portable?

yhs
prof marvelling at the endevour

Not any less portable than a large bazooka or surface to air missile launcher!  

My Dear J

it appears to have become....
less than portable?

yhs
prof marvelling at the endevour

Fair point, you may need to Steampunk one of these as the next project to transport it  

Spoiler (click to show/hide)

Spoiler (click to show/hide)

As a side note, years ago I went to a Dub Reggae night, the sound system was, VERY bassy, felt like I was being thumped in the chest and my, umm, mid gentleman's area, rattled with every beat, strange experience, me and my mate the only white faces in there, thoroughly enjoyed every moment.

Fear not, it can be made collapsible. The speakers can be hinged and fold on top, behind or below the horn. Equivalently, if it accepts the intrusion I could try a 90 degree elbow again at those junctions. I've come too far to go back 

I'm not thinking I want the subwoofer that strong. The idea is now to have a passive "audiophile" mode with the subwoofers feeding from the main line, and perhaps a switchable "sub mode" with a third powered amp.

[J. Wilhelm]

Pain in the rear end to cut these 4 reflectors for the new enclosures without slicing your fingers with the plane used to bevel the edges. They are used to shape the lower end of the driver enclosure to the mouth of the horn. These are indispensable if I want to keep the sound clean.

All triangles are inclined to one side or the other , camera distortion makes some of them look symmetrical

I am pondering making similar reflectors, this time two very slender isosceles triangles for the start of the line (end caps) in lieu of the ramp which was replaced by the subwoofer driver. The idea is to still direct some of the sound from the Sony driver to the horn and also the sound from the Samsung subwoofer driver. But it may not be necessary, as they're very slender, almost an afterthought just to erase the top corners. I really don't want to make any more mistakes, so if I try the slender wedges it'll be temporarily attached.

I can rebuild the left enclosure immediately, but my problem is that I need the solenoids for the crossovers. That means that tomorrow (actually today - it's the wee hours of the morn), I will buy more springs. And I'm still short of one pair of 100 uF capacitors. Otherwise I could just put together the whole horn. I'm afraid that'll have to wait a little longer.


PS. I just noticed the wood grain on the front of the whole horn. It almost looks like the whole front came from a single plank of wood. That deserves a good stained finish.

[J. Wilhelm]

Today I had to deal with an unexpected health problem. SIGH. Nevertheless in the few hours I had of activity i did manage to finish the two enclosures and I finally cracked open the second Sony Bass Reflex, to move the drivers into one of the two new enclosures.

Now, because I only have one viable horn, I have to attach both enclosures if I want to continue with the tests. Until I have a 100 uF capacitor(s), I won't be able to make both sides be the same. One side will remain with the ramp cap as a two way speaker, and the other side will be a 3-way speaker with the 4th order low pass filter..

I've come to the conclusion that using hinges to fold the enclosures under the horn is the best option. A strong hinge at the bottom and a very strong spring loaded latch at the top should hold it together and avoid vibrations. As you saw above that waveguide transmission line speaker is very long now, but overall it's no more than 122 cm long, and 62 cm folded.

It is at this point that I need to look for two types of things:

1. Capacitors (and solenoids) to complete the circuit and install the second subwoofer driver
2. Decoration ideas /items for the horn

Decoration needs to start now, because the method of attachment (hinges) will affect how the horn will look. I've been searching for Bavarian folk art and Swiss folk art for inspiration, since, well, this is a horn! And I was thinking that the alphorn is actually a good analogy for the device. To keep in character with my Luftschiffengel persona, I'm looking at a mid to late 19th century central European style. The background is the Engelfolk diaspora from Northern German countries through Bavaria and into Switzerland and Austria, seeking protection from Austria-Hungary.

The problem is that my Google-Fu is not returning much by the way of folkloric styles other than clothes. My health issues today really slowed me down today.

[Prof Marvel]

Ah My Dear J

for Bavarian ( or Black Forest ) folk art you need to look for


and you may or may not like what you find

woodwork:
https://www.google.com/search?source=univ&tbm=isch&q=bavarian+folk+woodwork&client=firefox-b-1-d&sa=X&ved=2ahUKEwiTnunmhqnqAhXDbc0KHRjvAykQsAR6BAgIEAE&biw=1366&bih=613

folk art
http://www.bavarianartist.com/bavarian-folk-art.htm

Bauernmalerei
https://www.google.com/search?source=univ&tbm=isch&q=bauernmalerei&sa=X&ved=2ahUKEwjn3OX1iKnqAhXHWc0KHVgSADAQsAR6BAgJEAE&biw=1366&bih=613

Bavarian Wood carvings
https://www.google.com/search?sa=X&source=univ&tbm=isch&q=bavarian+wood+carvings

and Holzchnitzereien
https://www.google.com/search?q=holzschnitzereien&source=lnms&tbm=isch&sa=X&ved=2ahUKEwi67pjCianqAhV_B50JHda9B4gQ_AUoAnoECAwQBA&biw=1366&bih=613

it ...sort of defines "kitsch" ...

hope this helps
prof marvel

[J. Wilhelm]

Ah My Dear J

for Bavarian ( or Black Forest ) folk art you need to look for


and you may or may not like what you find

woodwork:
https://www.google.com/search?source=univ&tbm=isch&q=bavarian+folk+woodwork&client=firefox-b-1-d&sa=X&ved=2ahUKEwiTnunmhqnqAhXDbc0KHRjvAykQsAR6BAgIEAE&biw=1366&bih=613

folk art
http://www.bavarianartist.com/bavarian-folk-art.htm

Bauernmalerei
https://www.google.com/search?source=univ&tbm=isch&q=bauernmalerei&sa=X&ved=2ahUKEwjn3OX1iKnqAhXHWc0KHVgSADAQsAR6BAgJEAE&biw=1366&bih=613

Bavarian Wood carvings
https://www.google.com/search?sa=X&source=univ&tbm=isch&q=bavarian+wood+carvings

and Holzchnitzereien
https://www.google.com/search?q=holzschnitzereien&source=lnms&tbm=isch&sa=X&ved=2ahUKEwi67pjCianqAhV_B50JHda9B4gQ_AUoAnoECAwQBA&biw=1366&bih=613

it ...sort of defines "kitsch" ...

hope this helps
prof marvel

Thank you for the links! Well, folk art is folk art. A lot of it won't suit. What I see as a common denominator is floral arrangements, almost universally, from Trachten and Dindrl clothing to pastoral festivals and such. Idelic grassland themes, basically. I see a preponderance of red on various dark backgrounds, but there are many other colors, depending on the type of flora depicted. Some of that could be incorporated, depending.

I could choose a higher European style instead, following the Austrian Imperial styles. I want the decoration to tell a story of sorts, even if in a very abstract way. As usual, I may opt for the black "wrought iron" look hardware, as that is common enough to find. Or let me put it this way: Gold and white, though very Austrian forthe  period (eg uniforms) is not something I'm looking at using here. The wood deserves to be showcased, and I need to relate to the background story of the characters.

Alternatively I can invent something. For all practical purposes, these are real-world elves with an ancient lineage, even among germanic people with a clear history back to the encounter between Romans and German folk. There's plenty of history to draw from. I could try ancient graphics.

[J. Wilhelm]

So I've implemented a folding form of the horn using a pair of large door hinges and two spring loaded latches

There's a minuscule amount of play to the hinge and a small a thin gap between the speaker enclosure and the horn proper, but that can easily be solved by attaching felt to the horn side of the joint. Red felt ought to be a great antique looking effect.  Theghorn is begging for decoration now, and o think it'd look great with some simple graphic pattern. If it's some sort of floral arrangement, let it be. Red colored flora would look fine and match the felt, wood stain or not.

Unfortunately, the transmission line speaker is hardly portable. It's I'd say, more "luggable." It's a piece of furniture, not very different from just gluing two bass reflex speakers back to back and calling it a "boombox." The weight of the whole horn is now approximating the weight of the Mk I Victorian Boombox. So given the size, I'm projecting that Mk. I still wins over Mk. III in the portable design category, no matter what, although Mk. III has an open potential for much better sound (I'm not done developing that yet, but the frequency response of Mk III above 1 KHz is orders of magnitude better than Mk. I, so the only thing I need to beat Mk I is to give Mk III the same bass frequency response - totally doable, and very close to doing precisely that in the near future)

The Mk III in its current iteration, does make for a very fancy sound bar for your living room, like those found under flat screen televisions, but bigger and with full hi-fi performance without need for a Subwoofer and satellite speakers.. Perhaps that will be the destiny of this horn, an I intend to bring it to fruition, probably along that path. I'm just doubting that I'd ever carry this thing for blocks like I did the Mk. I for events and such. About the only way you'll see me carry this, is if I put wheels on it - which by the way, sounds like a good idea!

While I will finish Mark III with no defined purpose yet, I think it's about time to start thinking what I can do to apply what I've learned from this system while I built it. Outside of the last problem with the side placement of the subwoofer driver, I think I have a pretty good understanding of the design steisi need to take to come up with with an acoustically successful design.

First and foremost, the work told me that in a slender configuration, a horn will need to be hibridized with a two pipe Helmholtz resonator. The port design is done by complete octaves, with the lowest frequency being two full octaves below the resonance frequency of the main set of drivers. In other words, overtones are an energy transport system you can use to extend the bass two octaves below to the fundamental frequency of a given driver.

The inverted horn is designed to overlap with the first overtone, one octave above the lowest frequency and one half of the driver resonance frequency. This design constraint can be used for any set of drivers I can find. Interior volume defines Helmholtz resonance, length defines ¼ wave resonance. You can overlap both methods of achieving resonance, and for a thin waveguide you may have to. If I use a subwoofer driver it's best to use a very stiff driver at the end of the horn, not on the side. Maximum resonance is always achieved by placing the speaker in line with the axis of the horn, as long as the driver is not too soft, otherwise the driver may "clip" the sound and possibly deform.

The "start of the line," where standing waves form is critical; its where you will notice energy loss and sound distortion the most. The waveguide must be clean, uncluttered, smooth. No sharp bends if possible. The end of the line, or neck must be wide enough to get some sound. Unlike a regular horn, the inverted horn's job is to kill sound at all frequencies, except for one frequency you wish to enhance, so at the end of the line you have very little pressure to work with, that's why the narrow end of the horn must be large enough.

No parallel surfaces anywhere if possible, try to approximate a conical cross section. Did I mention no corners? Avoid corners because they garble the sound and make the speaker sound "boxy." Use ramps and reflectors to smooth the interior of the speaker for clean sound, and so you may avoid the use of batting - the inverted horn shape should be enough if it's conical enough. Batting is for square boxes only, and basically can make a volume seen larger, because it slows the speed of sound depending on frequency; by the same token its that frequency dependence in the batting that causes sound to be muffled - so I discourage the use of batting.

*whew*

Next, while I finish this Black Forest sound contraption, I will run some numbers to see if I can translate this design into a circular cross section pipe system with more folds. It's probably the only way I could get to better portability for the same drivers.  Large diameter PVC pipe fittings can get expensive, though. And practically the full range speaker would definitely have to be smaller than 5 inches, because PVC pipes can get very heavy too. So I could design a system adapted to smaller drivers and with more folds.

[J. Wilhelm]

So I've ordered the capacitors from Parts express. These are non polarized electrolytic capacitors, which led me to discover a fatal mistake when putting together my capacitors: there's two ways to make non polarized capacitors. The one everyone knows is to connect two capacitors in series with the polarity reversed from one to the next. That gives you a 1/2 capacitance non polarized capacitor. My way is to connect them in parallel with diodes. The diodes serve as valves passing current only in the acceptable direction for each capacitor. Unfortunately, I forgot to add the diodes. That means that I'm getting twice the capacitance and the capacitors are at risk of failure, even bursting open or exploding. The voltages in this case are very low, so I haven't had any mishaps. And perhaps the extra capacitance could make up for the fact that I may have overestimated the magnetic permeability of the iron cores on the inductors...

But in any event, I prefer to be safe, and make sure that the components have the right value. I had resolved to make the filter cutoff be 120 Hz. It turns out that commercial 5 inch subwoofer drivers have a peak performance at about 120 Hz, so it makes sense to allow that maximum peak to operate unhindered.

I will have to figure out how to measure the inductance of the door spring coils, so I can approach my design values and take it from there. At $26 for the capacitors and today $22 for extra springs and steel bar, it's expensive. That's about $50 for the electronic components of the crossover alone. And buying inductors for speakers is much worse. $11 for one straight 12 mH coil with ferrite core, and $76 for a 16 mH toroid solenoid! This is not cheap!

I guess I could buy the 12 mH solenoids, but that's 2 out of 4 required solenoids, all ofwwhich would be (6.89+10.98)×2 = 35.74 + tax + shipping!

[J. Wilhelm]

So I'm making my own inductors as I write this post...Warning! Stay away from hard springs! They will jump back at your face and possibly cause a serious injury! I just lost a bit of skin on my index finger from a spring recoil. The spring went supersonic and hit the ceiling.

The method is to overstretch a spring, and noting it's very difficult to predict how much it will stretch initially. Once a spring deforms plastically, it will take much less effort to stretch a second time, so you have to be careful and practice to get a feel for how to stretch to get the length you want for a particular type of spring (avoid hard springs!). The cool thing about this method however is that if you go over the desired length you can always compress it back to the length you want with a drop of glue on the iron core or some other method like rubber bands. You can actually adjust the inductance of the coil to a desired value from a random coil length. Assuming a relative permeability of 300 at 28 cm the coil above has an inductance of 8 mH, and if compressed to 22 cm it goes to 11 mH.

I've refined my technique since the last time and this time I'm using drops of epoxy to hold the ends together and insulate the spring from the steel bar

[J. Wilhelm]

Aaaargh! Disaster struck. I've had issues with the spring solenoids. Compressing them back to desired size resulted in whole bunches of loops in the coil shorting into one another, which means I don't have the right number of turns, and I can't do anything about now that there's glue involved. $17 down the drain just like that. DID I MENTION TO NOT USE STIFF SPRINGS? Gah!

OK actually that's not the worst part. The worst part is that the inductance I've been working with may have been much lower than I thought. The problem is that steel varies greatly in the relative permeability value. The magnetic properties of steel vary a lot depending on the mechanical work done one the metal. Cold rolled steel, which is a safe assumption the bars are made from have a significantly lower relative permeability than annealed steel or hot rolled steel. That means mechanical hardening has lowered the magnetic efficiency. I could anneal the bars, but as Prof. Marvel has suggested, the zinc plating will make the annealing job a rather psychodelic affair.

And I'm not in the mood to spend yet another $20 or so on a new set of springs.

Looking at the inductance equations, and running a few numbers, it seems to me like the closest I got to getting a reasonable inductance value was that toroid inductor I made with a steel bearing and the last bit of good magnet wire I had. Toroidal inductors are much more efficient, because the iron core returns the magnetic field back to itself, completely trapping the magnetic lines within the torus. There's quite a bit of difference, about an order of magnitude in efficiency which let's you better use your material.

Looking at the materials I have at hand, the best I could achieve would be to somehow make 5 cm rings out of the zinc bars, and wrap it with the smaller 10 cm long rings, to get values of approximately 1.3 mH, which is not a lot but in the same order of magnitude. Otherwise if I increase the radius from 1cm for the steel bar to à core about 0.7 x 1.2 cm, and use magnet wire (which I don't have), I can start getting values over 6mH which is far more useful and practical. Gaah!

It's starting to look like if I want to avoid further complications, I'd better fork over for those inductors from parts express. And I thought I was being clever... It turns out that my low pass filter was probably working relativelywell  well, because I underestimated the capacitors by a factor of 2 and overestimated the inductance by a factor of 10. Somehow the thing worked, but the capacitors as hooked will blow up eventually. Sigh.

[Prof Marvel]

My Dear J

here's a thought from the King of the Scroungers  Scavangers Them Who Manage to Snatch Good Stuff From the Jaws of Defeat  -

the springs can be salvaged as well as the rods

method the first:
take the glued on spring & rod assembly  and a dull boy scout knife
treat the spring as a threaded rod and "chase the threads" with the dull knife.
this will chisel the glue out, eventually.
several passes will be required

methid the second
build a small bbq fire ( or boy scout camp fire) in your back yard
throw the spring & rod assembly in
walk away
this will burn off the glue, and if the small fire is just large enough
(and you throw some bacon fat in) no one will notice the buring glue smell.
it will also anneal both spring and rod

Perhaps Next Iteration, put the spring on wooden dowels?

yhs
prof mumbles
(hey, I am solving other's problems for free, wddaya expect?)

[J. Wilhelm]

My Dear J

here's a thought from the King of the Scroungers  Scavangers Them Who Manage to Snatch Good Stuff From the Jaws of Defeat  -

the springs can be salvaged as well as the rods

method the first:
take the glued on spring & rod assembly  and a dull boy scout knife
treat the spring as a threaded rod and "chase the threads" with the dull knife.
this will chisel the glue out, eventually.
several passes will be required

methid the second
build a small bbq fire ( or boy scout camp fire) in your back yard
throw the spring & rod assembly in
walk away
this will burn off the glue, and if the small fire is just large enough
(and you throw some bacon fat in) no one will notice the buring glue smell.
it will also anneal both spring and rod

Perhaps Next Iteration, put the spring on wooden dowels?

yhs
prof mumbles
(hey, I am solving other's problems for free, wddaya expect?)

Good advice! But I think the correct path will be to move on to torus cores. They take a similar number of turns for one order of magnitude larger inductance.

What I might do is make a core out of these, below. Five of these two-inch washers will give me the value I need with only 150 turns. The rods used 141 turns spread over a 22 cm length. But I still need magnet wire. It's going to be difficult to scrunch 150 wire loops in the inner perimeter (the hole is 2cm wide). The width of the washers helps me, because the greater the cross section of the torus, the less turns I have to do. Would love to keep the number of turns below 100,but these are the biggest washers I can find. I could anneal them by heating them red hot with a torch, and just let them cool on air

The toroid on the left, below is the solenoid I made from a ball bearing. With corrections to the relative permeability, it went down to about 7 mH for 130 turns. But I don't have more of those bearings, they're likely high carbon, and I would most likely have to anneal them anyway. The picture on the right shows the stack of washers which should give me 11½ mH.

On the left, I used 22 Gauge wire which is slightly thinner than typical speaker applications (Gauge 20). 132 turns required about 140% of the perimeter of the ring. Those 150 turns will require a stack of more layers, especially because of the inner diameter being small. If I could get bigger washers it'd be better. There are some washers that have a larger inner hole, but they're smaller. Maybe I can stack more washers together. And I still need magnet wire... It's either this or about $50 on mail orders  .

Another alternative is making toroids with the springs, but I I'm limited to a 1cm diameter steel core (that's the only spring diameter I can find) . I can only get about 1⅓ mH per 5cm diameter toroid... That's not good.

[J. Wilhelm]

My Dear J

here's a thought from the King of the Scroungers  Scavangers Them Who Manage to Snatch Good Stuff From the Jaws of Defeat  -

the springs can be salvaged as well as the rods

method the first:
take the glued on spring & rod assembly  and a dull boy scout knife
treat the spring as a threaded rod and "chase the threads" with the dull knife.
this will chisel the glue out, eventually.
several passes will be required

methid the second
build a small bbq fire ( or boy scout camp fire) in your back yard
throw the spring & rod assembly in
walk away
this will burn off the glue, and if the small fire is just large enough
(and you throw some bacon fat in) no one will notice the buring glue smell.
it will also anneal both spring and rod

Perhaps Next Iteration, put the spring on wooden dowels?

yhs
prof mumbles
(hey, I am solving other's problems for free, wddaya expect?)

Well, I managed to wrestle the two springs off the steel rods, but not without a spring chomping off another bit of skin. Those things bite! The springs are destroyed, but the rods are intact. That saves me some money.

I'm down to these two options, the large washers and the linear coils over the steel rods.

The inductance I can afford after annealing (assumes a corrected Relative Permeability of 180 for low carbon steel) is about 6 mH for each of the two bars above (143 turns) , and about 11 mH for the toroid made with washers (155 turns).

Both are expensive in the sense that I either need a total of 6 bars to make the stereo crossover, or I need two torii and two bars to get the same result. I think I have enough straight bar to make all 6 linear coils, and so instead of ordering magnet wire, it's easier to use springs again.

I may have to anneal everything, springs and all. Everything is plated in zinc. So I have to take precautions. Alternatively I can anneal the bars and get softer springs. The only thing is that I refuse to work with those hard springs again. They're very dangerous. The short ones I used for the previous crossover are much softer, andddon't need annealing, but I need a long one. I'll be hunting for springs I think. I might even have to accommodate two strings in one bar to get the length I need.

PS, I just realized that this picture above spells the Mayan numeral 11