Hmmm... I'll have to think about that. I had imagined tubes just hanging "still" and emitting sound. So, part of the presentation is the observer trying to figure out which tube is chiming, etc.
OTOH, if I *kicked* (tipped) the tube so that it struck the hammer, there would be more visible kinetics to the presentation. It's more animated.
But, now I have a much larger mass to control... making sure it quickly returns to "rest" (since you don't know if it will be "struck" again RSN).
Yes, that was my thinking. The wire(s) may need to be enameled or otherwise subtly insulated (in case the tubes are conductive)
Ah! That's something I should explore! I think I have a spool here, someplace. Let *it* be the actuator tugging on a pivot mounted hammer! I'll have to see how fast it reacts and then recovers...
Cool! Assuming those tubes are 1+ inch dia, that must be a couple feet wide and "high"?
Do you play it like a xylophone? I.e., "flat" with hand-held mallets?
Isn't that what the wheels are for? :>
*Is* the sound good? One of the problems I face is settling on the right material. Finding some material that is
*available* in the various dimensions needed as well as sounding good over a wide range of frequencies.
E.g., Five foot O2 tanks make *great* chimes -- but don't fare well in the upper registers! :-/
And, I'm not yet convinced that the *air* vs. *tube* vibration is the primary sound source.
Hmmm... also cheating. :>
The only piezoelectric transducers I have any experience with are small and not capable of moving much mass (in particular, they are used to raise "pins" in a Braille display... little mass, little displacement).
Do more capable transducers exist? Or, do they end up getting too large?
(also, what range of frequencies can they be operated over?)
This means putting a cap on each end of the tube. I'm not sure that will be "acoustically neutral". Esp since there will be tension from one end cap to the other.
This is starting to sound difficult to make. How do you attach the spring to the magnet(s)? And, how do you keep the "attachment sides" lined up?
But, couldn't you get a similar result with a piece of tempered spring steel?
[ASCII art disn't survive quoting. But, I think I get the intent.]
What about one (or more?) clappers dangling downward. The shaft of a solenoid (at rest) allows the clapper to remain dangling. When actuated, the solenoid has a "conical" taper (apex *up*) such that the widening diameter of this cone drives the clapper(s) outward from their rest position.
The "hinge" can be nothing more than an attachment point of the clapper to the solenoid's frame (or whatever). The orientation of the clapper(s) is not critical -- since the actuator is conical in shape (i.e., if the clapper(s) move around to some other "side" of the actuator, that just affects where they are driven
*away* from the actuator... but, they will still *be* driven away!)
XXXXX XXXXX XXXXX XXXXX XXXXX | . | . | O / \ / \
-----
XXX is solenoid coil/frame ... is "whatever" supports clapper O is clapper line art is actuator (shown at rest)
Of course, nothing is to scale :>
Note that, in practice, the clapper may want to extend *below* the end of the actuator. If the attachment point is more "inward" than the tip of the actuator, the clapper support will be "more driven" outward as the actuator pulls in.
I think I could make this with small steel balls on stiff wires (the "..." in the drawing). And, the solenoid might be replaceable with muscle wire (makes the diameter smaller).
No "precision parts" (like pivots, linkages, etc.).
I also wonder if one of those "vibrator" motors found in pagers/phones could be of use. I.e., restrict the armature so it can only rotate through a fraction of a revolution. Spring load it so it wants to restore to an idle "unwound" point. Then, energize it to force that fractional rotation to drive a clapper outward, momentarily.
Have you considered experimenting with longtitudinal excitation? Mounting the drives at one end may be easier to convert into a visually pleasing arrangement. By smart profiling of the tube ends, such as chamfering, you may be able to excite some transverse excitation as well, the sound could be 'interesting'.
Unfortunately it's *very* expensive, but an exciter based on a Terfenol-D rod may be an effective option.
I was thinking about finding a highly magnetostrictive alloy in a tube form.(Terfenol-D) Then he could use a fast pulse through a coil to induce it to ring. It would even need to be touched! Lots of power over a very short time though. Like xenon camera flash circuitry. Or like a sonar transmitter - PING!
No, I don't really want to "experiment". It seems like too much "art" is involved -- no easy way to "The Solution" nor any real criteria to determine if you're headed in the right direction or not :-/
Things like:
make me realize there's way too much magic involved in the actual sound production so I'll let smarter (or more patient!) minds than mine deal with those issues. I'll "add value" in the "presentation", instead.
Take a look at Rijke Tubes. Supporting a tube AND heating an element, to start resonance, might be possible with the same set of wires.
When your device is 'on'/standby - you could perhpas keep a sub-somic 'holding current' on the heating elemnts, then, as it plays, send more current through as desired. I would expect quick notes - but it still may work OK for some tunes. Obviously requires some experimentation.
Imagine two cylindrical magnets side by side with their like poles adjacent. They try to repel. Now join them at one end with a hinge - maybe just a flexible tube. Their natural tendency is to straighten out.
But they can't, they're inside a chime, so the best they can do is to form a V shape with the free ends touching the inside of the chime.
Now a third, larger magnet approaches the open end of the V with its opposite pole foremost. The V closes as the two ends are attracted to the larger magnet, overcoming their natural antipathy.
Now pull the larger magnet away, and 'DING', the V opens up and strikes the inside of the chime.
I like magnets used like this as the action is very snappy. A slow pull on the large magnet (or you could arrange the smaller pair to move) results in a sudden springing open of the V. I tried this with a length of silicone tubing as a hinge. It works surprisingly well.
The problem with that idea is that they will not really strike: they'll press themselves to the inside of the chime (after rebounding off it a couple times, maybe). That means they'll be _stopping_ the sound just as effectively as they created any. You would get a sound more like the "DRRRRRRR" portion of an old-tyle telephone's "DRRRRRRRiiiinggggg".
You've neglected one central aspect of how the hammers in a piano (and human percussionists, too, even if they're not necessarily aware of doing it) actually work. Pressing down a piano key doesn't press the hammer against the string. Instead the mechanic _throws_ the hammer against the string, lets it rebound freely, and then catches it on a braking pad so it doesn't hit more than once.
You want to _hit_ that chime, not press it. I.e. the key to striking a chime, bell or string to get a nice sound of of them is that both the chime and the striker need to be able to move freely at the point of contact, in the direction of contact, and least as long as the contact lasts. And afterwards, the striker needs to be held elsewhere, and the chime has to continue to be able to move freely.
And there will be rebound. I.e. a free hanging chime will start to move after that strike, unless you manage to balance two exactly simultaneous strikes to exactly balance their effects.
Yes, I accept all that. Like hitting a drum and keeping the stick pressed against the skin - not good. But it's simple, and maybe if you hit the skin nearer the edge, or the chime nearer one end... It's all about mechanical impedances and the like. Probably.
You could maybe arrange for the large magnet to catch the smaller ones after their first strike, but that sounds (ha!) tricky. But then if it were easy, any fule could do it.
Yes - the pipes are 43mm diameter, and the whole thing must weigh >30Kg.
Yes, I used to lay it on a table.
It isn't stable on only two wheels. I should use one wheel and one foot, with a handle so you can "wheelbarrow" it. Maybe with a folding leg to support it vertically.
Surprisingly good. The lengths are in ratios of 2^(1/24), for semi-tone (2^(1/12)) intervals, but the material is variable enough that I had to make adjustments up to 2mm to tune them accurately.
Yikes! I have several 10 foot "top rails" off a chain link fence (we use them to lift the bedsheets onto the citrus trees on cold nights). They don't strike me as that heavy (by contrast, the weight of the sheet is *significant*).
Hard to gauge from the angle/perspective of the photo but I'm guessing the longest tubes are ~2 ft long? And, looks like 25 of them? So, maybe a bit under 40 ft total length? (I'm trying to imagine what four of my "top rails" would feel like)
Could you also "end strike" the tubes (like the way tubular bells are played)? Or, does that alter the tonal quality of the sound?
I had assumed that was how you *were* carrying it. I.e., sort of like a "rickshaw" -- imagining the "handle" to be in the center of the edge closest to the camera (so you would sort of drag it behind you like those suitcases with wheels)
I guessed that from the arrangement of the tubes (white and black keys of a piano)
Because of differences/variations in the thickness of the pipe walls?
How are the pipes supported/suspended? It looks like there are mounts on the top sides of the top tubes -- i.e. so they "hang" in that orientation. But, what about the bottom tubes? And, it can only (?) be played "lying flat"? (i.e., if you stood it up so that the pipes were vertical, they wouldn't hang freely any longer?)
Please post a link if you do so! The biggest problem (bigger than finding a suitable striker mechanism) I see with this project is coming up with a material that will "sound good" -- since you have to commit to a material *before* you can determine what it will sound like. And, apparently, as you move up/down the spectrum, materials that sounded good in one part don't sound quite as "rich" in other frequency ranges.
(that's a lot of effort to put into something to discover it doesn't work well! esp since you might need to change pipe diameters/materials to keep the same quality tone...)
Sorry for the delay, it's stored in a place where it's hard to get to it to measure. The pipes range from 30" to 15" long, so about 50' overall.
It doesn't make much sound that way - it's suspended at 9/40's from each end. I figured that experimentally - though a friend later did the math and said I was pretty close. Someone else said 2/9, almost the same.
The suspension is on 3/16" diameter machine screws an inch long, with two nuts on each - one locking the head to the support and one locked against the threaded hole in the tube. That way each support has about
3/4" of screw to isolate the vibrations from the mounting.
Right, but strings need to be altered in length by 2^(1/12) to make a semitone, whereas bars need only half that length change. That surprised me at first - I cut the second tube 5.9% shorter than the first and the pitch went up a full tone. Then I figured out why...
I believe it's manufacturing variability in the walls, yes. Luckily I received good ear training as a child, so it's well in tune.
No, both supports on each tube are fixed (though the screws can flex) so it could be played in any position.
I don't think that's a problem. Any uniform material in the same tubular shape will have the same oscillation modes. Though the dimensions might vary the relative amplitudes, the basic harmonic agreement (or discord) is fixed.
Yeah, well, I got the material for nothing and just used a couple of abrasive blades in the circular saw, so it was mainly just fun making huge showers of sparks while trying not to set my clothes on fire.
Actually it works fine anyhow, it's just too damn heavy. Lightweight tubes of the same diameter could be shorter and still cover the same pitch range with substantially the same tone, I believe.
I should strengthen the frame and have it installed in a kids playground, they'd have a ball with it.
Personally, I became rather good at playing "Yesterday Once More" by the Carpenters in it. That was my show piece, which used the whole range.
So, the supports are *rigid*! Yet, don't interfere (noticeably) in the sound?
I've a small "guitar tuner" that I'm hoping would be applicable. But, from what I've read, it depends on how you design...
OK. (I'm still trying to get used to the idea that such a rigid support works!)
Some of the material I've read suggests different "character" to different materials. E.g., imagine making the same device out of sch 40 PVC...
Understood. I'd love to be able to do it with *glass* but the tuning would be too "anxious". (but, it would *look* wicked cool!)
I think wall thickness influences tone a lot. E.g., a woman down the road has a single chime made from a gas tank (O2, N2, etc.) with its bottom cut off. It has an incredibly rich tone (and loud!) I can't imagine getting the same sort of character in a, e.g., PVC tube (of appropriate dimension).
Might be interesting to try putting a "SuperBall (TmReg)" on a tether and using that to strike the chimes (as a "play piece").
This gets back to my suggestion of supporting the chime with, instead of a solid screw, a hollow tube (whether threaded on the outside or brazed in place w/ever) and have the striker actuated by a rod pushing through the inside of the support. On the inside, the rod could be bent downwards at a 90deg angle and the end could have a small clapper/ ball on it to strike the opposite inside chime wall.
Because this is a vibrational node, there may be little damping of the sound. Still, as pointed out in one of the links in an earlier post, tubes do not have exactly the same behavior as solid rods or plates as in marimbas/xylophones. Still, it seems it works well with screws so.......
It seems that the more I get exposed to, the less certain the outcome! I.e., more "art" than "science".
There seems to be an awful lot of mention about *where* (along the tube) the supports should be located so its hard to imagine that they aren't critical (or, at least, "significant") to the resulting sound quality.
I guess I should just get some pipe and start banging on it and see what (if any) effect these sorts of changes have...
It's a nodal point, so why not? The tube isn't going to bend significantly, perhaps a millimeter or two at the anti-nodes, so the actual deflection angle at the mounts is tiny, and that is taken up as flexion in the 3/4" free length of the mounting screws. The screws themselves are pretty thin relative to the moving mass of the tube.
The exact position of the mounting point is critical. Move it by more than a couple of millimeters and the tone and Q both drop off hugely.
PVC has very low stiffness and very low Q. It would be like trying to make a guitar soundboard out of cardboard.
The "figure of merit" for guitar top plates is the speed of sound propagation in a wave propagating out from a point on the plate. If you halve the thickness of the plate, you halve the mass per area, but the stiffness drops as the cube; so the speed of sound drops as the square. of the thickness. That's why very low density but very stiff woods are used, especially spruce, which is really rather exceptional compared to most other woods.
However, in a guitar, if you use an ultra-stiff synthetic composite, the losses at very high frequencies are too low, and so the guitar sounds brittle and harsh. Wood is a very complex nanostructured composite, and doesn't behave like a uniform material at all scales.
I think glass would work just fine, but you'd need to have an effective grinding setup to tune it. You would have the "brittle" character I mentioned, but in a chime, it wouldn't seem harsh.
What I suspect is happening there is that you have many circumferential modes of oscillation, because the diameter is so large compared with the width. It's not really about the wall thickness, it's just that the longitudinal and lateral modes are in the same frequency range as the circumferential ones. Whether they're conchordant or dischordant will depend on a huge number of things, so its pot luck. But that's the case even with giant orchestral gongs. Despite being so expensive to make, two gongs that were made as close as possible to identical have very different character, and one in a hundred just stands out as being exceptional. Same with guitars from the world's top makers, by the way; they will only rarely have a complete failure, but can't explain or repeat their occasional exceptional instrument.
The basic issue with the suspension point is the damping it causes. the "2/9th" rule has to do with where the standing wave for the first overtone (where the chime is exactly one wavelength long) basically hits zero. IOW, there's no motion at the 22.42% point from the first overtone. And 2/9ths is within a percent of that point, and close enough for most purposes. If you suspend from another point, you'll damp the first overtone to varying degrees. Have you seen:
formatting link
But you're right, this is not cut-and-dried, especially given variations in materials.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.