From:  "bmh1944" <bmh1944@yahoo.com>
Subject:  Capping Long-Tube Chimes

I thought I might share some findings with anyone interested about the concept of capping one end of a long chime tube. While I'm a wireless communications engineer by profession, I hold a black belt in "tinkering".

I've used the cumulative advice from the many links and extremely valuable information provided through this website to quickly graduate from total ignorance to a level of educated novice. Many good presentations brush on the fact that orchestra chimes are capped at one end (with only a small hole in the cap), but have never really explored, expounded upon, or gone past the "why" factor. I decided to cut two long tubes of equal length (as per Lee Hite's Excel program) in the
C2 range at about 153cm. Due to advice I got from a close friend (who's a metallurgical engineer with a primary contracting supplier for Boeing), I decided to experiment with 1" OD, type-L copper tubing because it was easier to find/attain, easier to cap (by soldering on a copper pipe cap that is readily available from any plumbing store), had a hardness factor equal to most common grade aluminum-alloy tubing, and was actually a bit better in sustaining a resonant tone than the softer brass tubing used for orchestra chimes.

I horizontally hung two 153cm lengths of tubing (one open tube, and one capped with a 1/8" hole in the center of the cap) from two, long monofilament loops (on each tube) that were positioned at the "node" point near each end of the respective tube. As a "control" method, both tubes were hung in the same manner, side-by-side, from identical supports; a omni-directional microphone was placed exactly between both tubes and hooked to my laptop pc; and I ran the free "b-flat"
tuning program that's listed in one of the links provided on this website. Also, I applied the "Fred Flintstone" factor of simply using the human ear's concept of "perceived tone" and a stopwatch to clock the duration of such. The results were very interesting.

First, I'll expound a bit on something I DO know about - "resonance". Whether mechanical or electrical, any energy force having the property of continued excitation at a given frequency will have a measured length (wavelength) at which that particular frequency sees nearly zero impedance to sustaining it's action. It's actually much simpler to define this concept in an electronic environment because electron movement at "resonance" produces harmonic frequencies that are EXACT multiples of the fundamental frequency (just as an "octave" difference is defined with mechanical vibration). However, when mechanical vibration is studied, one is suddenly faced with the "overtone" concept which doesn't apply to exact mathematical multiples of the fundamental frequency. Why? Because each exact harmonic of the fundamental frequency is not riding on a "smooth plane" (ok, a level playing field) like the fundamental frequency enjoys. Instead (like modulation riding on an RF carrier) each mathematical multiple (harmonic) of the fundamental frequency is "riding" on the vibration pattern of the fundamental frequency.

Thus (without going into the complicated math), the first "harmonic" (exact double of the fundamental frequency) sees the logarithmic addition of the fundamental frequency's disruption of an, otherwise smooth, "playing field"; and the resulting addition of that factor to the exact multiple of the fundamental produces an "overtone" which is higher in frequency than the exact "harmonic". Each successive harmonic of the fundamental goes through the same mathematical addition of the phase influences of the fundamental frequency plus any lower-frequency overtones present; therefore, each higher overtone is an ever increasing value of it's particular "exact harmonic multiple".

The factor of "free air resonance" creates a major difference in "perceived tone" (cumulative audible sound of the fundamental frequency plus any and all overtones) with a solid rod compared to a hollow tube. A solid rod only creates air movement (audible sound) from it's surface vibrations; and, as Chuck has already pointed out, the hardness (or temper) of the metal only serves to increase the ability to continue vibrating and sustain the length of the sound produced when it is struck - not the frequencies produced by a given resonant length. But, since a hollow tube is also influenced by air movement within the tube itself (during vibration from being struck), the influence of that internal air movement can either help or impede the sustaining of any tube wall vibrations.

Looking at what I've seen with the pc program's spectrum analyzer and clocked with plain old human ear perception (and a stopwatch), I've noticed the following differences between an open tube and an equal length tube capped at one end:

OPEN TUBE:
1. Striking the tube in the middle produced the best and longest sustained "musical note" that could easily be defined by the pc program's little musical note display (and a "single note" perception by my ear as well). Noting the pc program indicated that striking the tube in the middle produced a major excitation of the tube's first overtone (above the fundamental frequency it was cut at) made a lot of sense because, just as a rock dropped in calm water, the waves
(vibrations) produced moved outward from the point of impact. In the dynamics of a resonant tube, the vibrations from a strike in the middle would naturally start in the middle and spread toward each end of the tube - which would naturally excite the "happy wavelength" characteristics of half the tube's fundamental resonant frequency.
2. Striking the tube at the end (according to the spectrum analyzer) initially produced a very equal array of fundamental and overtone levels up to around 1.5KHz. As suspected, the pc program could not readily identify any particular "musical note" and kept moving the little note display to different notes (about one second after striking the tube) because the fundamental frequency and many overtones were being equally produced by the strike at the maximum impedance node for all resonant frequencies (determined by the tube's length).
3. Quickly after either a strike in the middle or at the end, the spectrum of fundamental frequency and overtones quickly decayed to only the fundamental frequency (not a major audible "player" to the ear) and (in order of db strength) the 1st, 2nd, and 5th overtones with little or no 4th.

CAPPED TUBE:
1. With the pc program's indications, striking the tube in the middle caused a similar reading to that of the open tube being struck at the end; a defined "musical note" only lasted for a second before changing back and forth to different displayed "notes". I made the same conclusions from observing the fact that the fundamental frequency and many overtones were being generated at very similar db levels - thus causing confusion in the program's ability to "define" any exact note.
2. Striking the capped tube at the end created total chaos with the pc program's note identification process. Looking at the spectrum analyzer, I noted that the fundamental frequency and ALL overtones (from 1st thru 6th) were not only present in very similar db levels, but also decaying at about the same rate after the tube was stuck.

EAR/STOPWATCH/REAL-WORLD RESULTS:
1. Other than the capped tube sounding just a bit flat (lower perceived frequency) than the open tube (probably because the cap made the tube about 1/8" longer), there was NO difference in the octave or "perceived musical note" I heard.
2. Striking either tube in the middle gave more of a "clean and simple" perceived musical note; BUT striking either tube at the end gave a much richer and fuller sound to the ear because ALL frequencies were being excited to a more equal level - thus producing a "richer" sound to the human ear.
3. BIG DIFFERENCE - While standing very close to both tubes allowed me to hear the very low fundamental frequency (around 60+Hz) long after the "perceived" tone had diminished, the capped tube sustained the "perceived tone" about 20% longer than the open tube. While the pc program's "musical note identification" process was in total chaos, its spectrum analyzer confirmed the differences noted by my ear and the stopwatch.

Why the difference? I'll stick my neck out with a WAG (wild-assed guess) that it has something to do with the SWR (standing wave ratio) of the air column within the chime tube. Whether mechanical or electronic, an analog, sine-wave energy force at resonance produces the best "flow" of constantly moving energy waves along the medium involved as a conduit for such; the "decay rate" of excitation (initiated by a one-time "shot" of force) depends not only on the medium's inherent impedance, but also by direct influence from the medium's surrounding environment and/or ambient temperature. While resonance (a particular frequency at perfect wavelength conditions) provides the optimum, free-flowing energy movement of the energy
waveform, any impedance or length mismatch to that optimum resonant environment causes some of the "moving energy" to be reflected back toward the source because the "out of resonance" factor impedes the maximum rate of movement and waveform propagation (OK, like open floodgates of a dam still impede the "perfect" natural water flow rate of a river if the dam wasn't there in the first place). Any reflected energy waves (from being out of perfect resonance) is inversely "algebraically added" to the optimum flow of positive energy waves being generated; the algebraic "sum" of forward and reflected wave movement causes an effect that (mathematically/physically) slows down the apparent and measured "positive energy" wave movement speed from that which is possible at optimum resonance. The "true" end result of actual perceived, measured, and/or radiated aspect of any generated energy waveform is the quantitative amount that results from the ratio between forward and reflected power in a non-resonant environment (standing wave ratio). Creating a total barrier (maximum
impedance) to resonance at one end of any "normally resonant" medium causes a 90% phase shift in the total "algebraic sum" of the positive energy waveform vs. reflected energy; the result tends to make the normally free-flowing waveform movement look "stationary" and completely non-moving in nature.

Looking at the results, I've noted that capping the end of a very long (C1 to C3), open chime tube has little effect on the tube's mechanical resonance properties or "perceived musical note" heard by the human ear. However, since the cap creates a maximum, full-wave impedance to the "free air resonance" of the air within the tube, the SWR (standing wave ratio) of internal air movement tends to create a "stationary" waveform of internal air pressure - thus, helping to
sustain the mechanical wall vibration of the chime tube at the fundamental and ALL major overtone frequencies (translation: longer sustained "perceived tone" to the human ear). I qualify this thought with the results from my next experiment at capping BOTH ends of the same tube. When a 90-degree air movement phase-shift (SWR) was "added to" by a second 90-degree air-movement phase shift at the other end of the same tube, the resulting 180-degree total phase shift (of internal air movement) algebraically added to a total "phase cancellation" (equal, but opposite combination of two different waveforms) - which, equals ZERO, nada, nothing. Striking the tube (closed at both ends) resulted in nothing more than a "thunk". Yet, striking a long tube (closed at one end) created a 20% longer-sustained "perceived" tone than that produced from the
same strike on an equal-length tube that was open at both ends.

Tuning a very long chime tube with pc programs or spectrum analyzers? - FORGET IT!! As those more experienced than I am have already noted, it's the "perceived tone" (combination of all frequencies) to the human ear that really counts in the world of audible reality. Use all the formulas, charts, and Excel programs to get yourself in the "ballpark" for any musical notes you desire to have played at random by the wind's unpredictable moves. Using the longest chime as a "constant", you'll be better served in tuning the others (pre-cut to recommended Excel, or chart lengths) simply "by
ear" - just like grandpa tuned his violin or guitar's strings with musical respect to the "base/bass string" he decided to use as a reference point for the rest (regardless of whether it coincided with the proper, defined, textbook frequency for that particular musical note - or not). Using a pitch-pipe or pc tone generator will help you get started and provide somewhat of a base for comparison; but, like using a slide-rule, "interpolation" will be the best tool in allowing
your "ear" to do any fine tuning of each individual chime tube's sound with respect to the others in your desired configuration. Being a musician of 54 years, I can let everyone know right now that "orchestra chimes" ARE NOT used as a "standard" for tuning anything because their emitted sound is "multi-overtone" in nature; each "note" simply falls in the "arena of acceptability" to the human ear's musical perception of such. A precise, exact frequency (in Hertz) "musical note" is NOT the reason for their wonderful addition to an orchestra; instead, their bell-like, rich, ringing tone (sorta-kinda on the money) is the attribute that makes them so appreciated, desired, and admired. Personally speaking, using all the obsessive/compulsive desires for exact, particular notes or "musical scales" is a predestined waste of one's time because the wind will neither play your "song", nor punch out a beautiful pentatonic scale (or any other) that will suit even the lamest of your ideal expectations. I find the simple aesthetics of deep, totally random, long-sustained, bell-like sounds more relaxing than anything else.


Apologies for such a long dissertation that's mostly unfounded by any co-authored proof or supporting mathematical equations; but, I think I've found the reason that orchestra chimes are capped at one end (with a small, pressure-relief hole in the cap). I feel that nothing more than a longer-sustained "perceived tone" (to the human ear) after a strike is the overall goal and objective for this "anomaly" of design. If the same property is desired in long-tube, bell-sounding windchimes, then go for the concept. If I am in total or partial error, then please correct my assumptions. Brent Hamilton

Because I'm also a bit of a novice as well, I'll share a few things I've found that work very well. My experience comes from building a few sets of beautiful sounding long-tube chimes (longest tube around 66") from 1" diameter hard copper (type "L" and "K") which has about the same hardness as most aluminum tubing. The long, 8-chime sets take a little more thought and tinkering to build a striker (clacker) and sail combination that works efficiently enough to provide a good
sound to such a long set of chimes in low-wind conditions. Even though you may not be building long chimes, I've found the following works very well:

STRIKER - I played around with wooden wheels, plastic disks laminated between layers of wood, and a few other things that worked OK; but my best results came from a simple PVC sewer pipe plug that's available for about a buck and a half (in 2", 3", and 4" diameter) from just about any hardware store, plumbing supply, or building center chain.
This stuff is hard (but "kind" in nature like hardwood), about 1/4" thick, and built to resist the sun's UV ray degradation. Yes, it's white in color; but, with a little fine sanding, you can paint it with most exterior enamels if that's what trips your trigger (just don't paint the outer striking edge). The PVC plug comes with exterior threads and a raised, square protrusion so one can use a wrench to tighten it into a sewer pipe collar. Its round outer height is about 1" tall; but, using a piece of masking tape around its circumference, it can be hack sawed down to about 1/4" to 3/16" in
thickness. Leaving the threads in place provides a nice "knife edge" to strike the chimes; or, grinding them off (much work) and giving the whole thing a rounded outer edge is a little more aesthetic in appearance and works just as well. Drilling a hole through the raised square portion of the plug is easy, and you can use whatever method you choose to hang it from the striker's pendulum, cord, or chain.

SAIL - I've played around with everything from one of my wife's discarded bras to an AOL CD I got in the mail for use as an efficient wind-catcher; my conclusion from all experimentation is that flat sails suck! Why? Because they're only two-dimensional in nature! Here in the Texas hill country, the wind is an anomaly that plays back and forth from ever-changing directions; thus, a flat sail either starts spinning - or just lays there when the wind is wafting toward it's edge. I scampered off to the local Home Depot and came back with a 2-foot square piece of thin (26 ga.) sheet aluminum that (instead of being "plain") had an embossed pattern on it - to keep the wife happy. I cut a couple of rectangular pieces to
a size that would be adequate to let a moderate wind move my striker; placed one piece on top of the other, then vertically riveted them down the middle using hammered aluminum rivets (wife didn't like the looks of pop-rivets - grrrrrrrr). OK, so now I gave the whole tthing to my wife (along with a pair of heavy-duty scissors) and told her to just let the creative juices flow. On each side of the vertical rivet line, she cut out some really pretty butterflies (through both layers
of metal). I simply bent the layers apart (from the vertical rivet line) to form a large, 3-D, "X" shape with four butterflies at 90-degree angles from each other. This sail now catches wind from any direction and moves the striker against the chimes even better than the beer cans did (but don't tell my wife I've admitted to this).

Hi Chuck,
Yes, I also tried equal lengths (starting at about 66") of 1"-OD galvanized steel (electrical conduit), plain aluminum, anodized aluminum, and hard (type L) copper tubing. I tried my best method of comparing the different tubing by horizontally clamping a six-foot length of cast iron pipe across the top of a stepladder. From that, I used two long loops of thin monofilament fishing line to horizontally suspend each test "tube" (at each end's node point) from the iron pipe. I relied only upon my ear's perception, a stopwatch, my laptop pc, a stationary microphone, and the simple spectrum analyzer program that's listed in the information provided here in the group's message board. I struck each tube many times in the middle and at the end to get an average of the results; the "strike" was produced by a 1" diameter wooden ball suspended by a string from the same point (and dropped from the same height to swing against the tube). As you already know, there was a considerable difference in the perceived sound produced by the steel, both types of aluminum, and copper tubes; but, since beauty is in the eye of the beholder, I'll leave the desirability of each particular sound to the choice of each particular individual - because each one is better in some respects than the others.

Because it was easy to do so, I also experimented with the difference noted between an open hard copper tube and the same length (from the same stock) tube that I'd tin-soldered a copper plumbing cap over one end. If I had the knowledge and equipment to weld a simple plug over the end of either steel or aluminum tubing, I'd have tried that as well - because the results were not only much better in the perceived tone produced, but also provided an extremely efficient, simple,
longer lasting, and very aesthetic way of mounting any long tube as a windchime component. Brent

From:  cllsj
Subject:  Re: Long-Tube Experiments With Steel, Aluminum, And C

I did solder fender washers on the end of steel conduit to suspend some long tubes. So I know that it does work. I used a propane torch with acid core solder. Nothing special. I did put a piece of metal (angle?) on the floor, the washer next and then the tube holding it all down while I soldered. I tried putting the washer on top but with nothing clamping the washer down I kept knocking it out of position.
In addition it kept the solder from running down the tube and looking ugly.

I assume you want to flare the copper tube to put a fitting on the tube to cap it. I don't see why this wouldn't work. Chuck

From:  "bmh1944" <bmh1944@yahoo.com>
Subject:  Capped Tube Answers

I've done a lot of tinkering and experimenting with capped tubes for two major reasons: (1) that's the way most expensive orchestra chimes are constructed, and (2) single-line, center/top suspension of a tube allows for closer tube spacing and less tube-to-tube contact from rocking (as compared to upper-node, single-line, internal suspension).

As Fred Flintstone here, my job is to be terribly wordy while trying to put terribly complex mathematical and physical concepts into stone-age terminology that people of my ignorance on the subject can better understand. When my limited knowledge exceeds its bounds and any "logical assumption" (wild-assed guess) deviates from a course of science to science-fiction, then I totally depend on Chuck and/or Jack (our resident experts) to break their lengthy silence and slap me back in the right direction - to everyone's benefit here.

Anyway, capping a brass or copper tube is done very simply by soldering on a copper plumbing cap which can be obtained from any hardware or plumbing store; unfortunately, steel and aluminum tubing take a lot more effort for such. If you have a wire-welder, you can weld a steel fender washer (either externally or internally fitting) to steel or EMT tubing, grind/file/polish everything, then suspend the tube from the small hole in the washer. With aluminum tubing, you'd have to have a MIG attachment for the wire-welder and weld in an aluminum disk you had previously cut and pre-fitted.

There is one other option (for both steel and aluminum) that also works very well when one doesn't have a welder - good 'ole "JB Weld". Yes, most people put JB Weld in the same arena with duct-tape, bailing wire, and super glue, but it has a tremendous strength and very good UV resistance. Additionally, you can use a preformed cap (that fits) or capping disk of any metal because the insulating factor of the epoxy-based compound will prevent any dissimilar metal corrosion from electrolysis. I wouldn't recommend using it on extremely heavy (over 40-pound) tubes, but it works great on lighter rigs.

When you cap any tube, you create a lot of dynamic changes in the tube's many vibration modes and internal air-column movement; some of those changes are noticeable - and some are not. When using very long tubes that far exceed Chuck's "ideal length" (look for the calculator for such on his website here in the "links" section), I've found that capping a tube (many times) restores a "missing overtone" and makes the tube produce a much richer sound. But, if you read all the
information on Chuck's website, tuning a very long tube can be very frustrating if one has little patience - and a short supply of both tubing and whiskey - LOL.

I've not experimented with capping a tube cut to "ideal length", so it is impossible for me to say what affect it may have on the precise tuning of such. Capping a tube on one end creates two changes in the tube's normal vibration dynamics: (1) it creates a virtual "short circuit" at a major antinode of all vibration modes where, at that point, an open tube provides an infinite impedance (open circuit) to the resonant fundamental frequency, and (2) the capped tube creates an infinite impedance to the internal air-column movement where no impedance exists in an open tube.

Whatever tuning you desire or manage to achieve, the top/center (antinode) suspended, capped tube will usually produce just as long-sustained tone as a conventional, node-mounted tube providing you:
(a) use a strong, thin, very flexible suspension line like braided nylon or stainless steel cable (deep sea fishing "leader line") and
(b) leave at least 6" of line between the top of the tube and the suspension disk or "halo" it is hung from. Brent

From:  "bmh1944" <bmh1944@yahoo.com>
Subject:  Re: Adjustable Tuning

Art;

I've only drilled a hole in the center of a cap to end-suspend it like many orchestra chimes are done. Center/cap-suspension is done at a major vibration antinode; so, it must be done with a relatively thin and very flexible suspension line like braided nylon or braided stainless steel cable used for deep-sea fishing leader. To prevent dampening at the antinode, it seems that at least 5" of tube to support line must be used - anything shorter begins to dampen the sound from that of the same capped tube which is conventionally node suspended.

Whatever degree of influence (good or bad) created by a single end-cap is probably due in part to a virtual, cross-sectional "short circuit" of the axial and circular modes at that point; and in thin walled tubes, it may also be caused by a maximum impedance to the internal air-column movement at one end. Making the hole larger would do little to change the tube's cross-sectional difference at the end, and would lessen the impedance to the internal air-column. It seems like the cap alters the pitch slightly but I think that's more from changing the tube's shape and consistency at the capped end.
Regardless of the slight frequency shift caused by a cap, it seems the major effect is on the way it changes the relative amplitudes of the resident frequencies (their individual decay/sustain times) more than anything else. So, using a hole in the cap for tuning is probably nothing to explore. Brent

From:  "bmh1944" <bmh1944@yahoo.com>
Subject:  Re: Capped Tube Answers

I've not done any experimenting with anything other than a completely capped tube (with only a very small suspension hole in the center of the cap) because that's the way many orchestra chimes are constructed and hung. Capping one end of a tube changes many inherent dynamics of both tube mode-vibration and internal air-column movement as compared to a tube with both ends open. E.g.: most vibration modes see infinite impedance at each end (antinode) in a fully open tube, but see a major "short circuit" at the capped end of a tube; additionally, the normally-free air-column movement in an open tube, suddenly sees maximum impedance at the capped end.

I don't have nearly the depth of physics or mathematical knowledge to either quantify or qualify the results; but I have noticed that a capped long tube (exceeding Chuck's "ideal-length" criteria for the most precise musical note/tone produced) does produce a little longer duration of the total sustained sound produced by the tube – and seems to restore or greatly enhance the amplitude of many "missing overtone" anomalies that frequently happen in such long tubes.

I've tried suspending a long capped tube by top-suspension (at the antinode) by a single line which has a knot or crimped-on device which keeps the suspension line from pulling through the very small hole in the center of the cap; I've also tried the conventional node-mounting technique with the upper capped end totally free. I've not been able to detect any difference in overall sustained tone duration or change in decay rates of existing fundamental of overtone frequencies in either method of suspension. HOWEVER, since top-center suspension is done at one of the antinodes (major vibration point),
the suspension line must be of thin braided nylon, monofilament, or stainless steel cable, AND one must have at least 5" (maybe more as line stiffness dictates) of line between the cap and the main chime support disk/halo - otherwise, significant dampening of the tube's tone will result as the line length is decreased.

Note that my capping experiments have only been with very long tubes which far exceed Chuck's "ideal length" criteria; and I've not tried it with any ideal length tubes to see what effect it may produce. BUT, since Chuck's very excellent criteria (for predicted results) are based on many ratios of dimensions and air-column influence in an open tube, I would take his advice and not expect a capped tube to make anyone happy with the accuracy of produced results.

I understand Chuck added a little (not so terse) bit of ramblings about "perceived sound" to his website in an effort to warn those who experiment with very long tubes about the almost unpredictable results the brain's "fuzzy logic" will make one perceive - when the math and what is there in reality indicate what really exists. I will stand and argue forever that a tube capped at one end produces a much "richer" sound than an open tube; but, especially in a long tube, that "richness" is from many (non-musically harmonious) overtones being enhanced in amplitude and sustained duration. When the brain tries to put all this together, what you may hear or perceive as some musical note may very well NOT agree with what is
actually mathematically true and precisely indicated by a spectrum analyzer. Such being the case, using the Euler-based calculators to predict what you actually "hear" will quickly put you into the realm of total confusion because the brain doesn't care what the math and reality dictate when exceeding Chuck's "ideal length" criteria.

Yes, most high-dollar orchestra chimes are fully capped at one end, top/antinode suspended, and many times greatly exceed Chuck's proven "ideal length" concepts - BUT there is one major thing to consider. These folks have spent a few hundred years, thousands of whiskey gallons, and numerous brain meltdowns by trial-and-error experimentation with different length/OD/wall thickness ratios to individually "tune" each note of those very "rich sounding" chimes to something the brain's "fuzzy logic" perceives as the correct note and pitch - regardless of what is mathematically and physically present
in the overall sound package produced by the chime.

Chuck and Jack have gone to great lengths in coming up with an "ideal length" calculator that is excellent in predicted results; so, I always advise anyone to go there first instead of trying to "mathematically tune" something the brain's "fuzzy logic" completely disregards in many instances. Brent

From:  "bmh1944" <bmh1944@yahoo.com>
Subject:  Re: Capped Tube Answers

Hi Hock;

Glad to see someone else has verified the same results I've found and noted in the past about very long tube anomalies and the noticeable influence that capping produces in a relatively thin-walled tube. In general, I too have noticed capping drops the overall tone a bit in frequency, sustains the lower frequencies much longer, and tends to quickly cause extremely high frequencies to decay much quicker. I'm only guessing this is brought about more by the major change in the internal air-column's movement (and its influence on a thin-walled tube) than by changes in the tube's overall physical characteristics.
However, I suppose one would have to try capping very thick walled tubing as an experiment. If the same results were noted, then I'd suspect it was the change in tube properties which produce the difference (if any) because I begin to seriously doubt the internal air-column's ability to influence a very thick walled tube to any appreciable degree.

Your measured results on the very long tube (with respect to its length/OD ratio) points out another anomaly which has been noted many times in the past, but not really explained. If you look at your measured frequencies, you'll notice there is a problem - the 5^th natural frequency (4th Overtone) is missing. Looking at the tables of mathematical overtone progression, you should have a 5th natural frequency (4th Overtone) of about 518Hz present, but your measured results show a hole between the 4th natural frequency (342Hz) and the 6th natural frequency (712Hz). Whether it's a phenomenon of phase
cancellation between different modes of vibration or some other interaction (of which I haven't a clue), I've found the same results when using a relatively narrow OD tubing and pushing it deep into the C1 or C2 fundamental length range. Sometimes capping such a tube will partially or completely restore the missing overtone; but, likewise, I've seen capping create a missing overtone. On other occasions, I've seen the missing overtone return when the tube's length is slightly
changed - but keep making it a little shorter, and all of a sudden another overtone disappears - so go figure.

As far as the difficult tuning goes in very long tubes, I've used the "fuzzy logic" concept to describe how the brain mixes various non-musically harmonic frequencies (each representing a different musical note) to come up with some analog composite we hear as a "perceived" tone or musical note. Much of what we "perceive" as a result is almost totally dependent on the respective amplitudes of each frequency present (in relationship to each other) and the "net" amplitude of each frequency when the ear's extremely poor response curve changes those amplitude relationships when feeding information
to the brain. While the brain considers all frequencies (that's where one gets the "richness" of sound), the fuzzy logic process is going to "home in" on those frequencies which are predominant in amplitude as "factoring agents" in the cumulative "perception" of some musical note or tone.

I'm glad you posted your results because it may help me explain what can very easily happen with the brain's unexpected application of "fuzzy logic". Looking at your measured natural frequencies (with the 5th natural frequency missing) and roughly rounding them off to the musical note they correspond to:
38Hz = Eb1
105Hz = Ab2
207Hz = Ab3
342Hz = F4
712Hz = F5

Now we have a low E-flat, two-octaves of higher A-flat, and two higher octaves of F-major all being produced at the same time. Even if the lower three frequencies were being generated at the same amplitude as the higher two, they are much farther down on the ear's poor response curve and won't "appear" to be as high in amplitude when information is fed to the brain. So, at any distance, you won't hear the 38Hz at all; and the next two frequencies will be considerably lower in "heard" amplitude than the higher two (which are much higher up on the ear's response curve). The results in the above case will probably produce a predominant F-major note (which is a bit sour because the two "octaves" are not exact multiples) plus a little A-flat mixed in as well. So, you'll probably hear mostly an F-major sound - unless, for some reason the brain decides to "add" the 2nd and 3rd natural frequencies to a "perceived" 312Hz and toss a little Eb4 in there with the mix. You could go to the piano and get close to duplicating the mix by hitting two (octave-spaced) F-majors and a lower A-flat all at the same time. Now, hypothetically change the tube's length or OD just a bit (without changing the frequencies
produced very much) where the "missing overtone" suddenly comes back at a very high amplitude; POW - suddenly the old perception of F-Major quickly gets a strong C5 added for the brain to consider. Go back to the piano and press a strong C-major between the two F-majors (and a bit of lower A-flat) and see what "note" you perceive from that mix.

I suppose my point lies in the fact that very long tubes produce a multitude of frequencies, which, in most cases, are not producing the same musical note. Just about any OD/length ratio tubing can be cut to a fundamental resonance in the low C1 range; but in many instances, the particular OD/length ratio will most definitely change the respective amplitudes of each frequency component in relationship to each other. And it is that change (with respect to individual amplitude levels) that can suddenly trigger the brain's fuzzy logic "perception" of tone or musical note to radically shift from one perception to another. This is why tuning very long tubes can easily become frustrating because some minor length change can greatly alter the relative amplitudes of the natural frequencies (and sometimes suddenly restore or lose an overtone) - which, in turn, quickly causes the fuzzy logic process of perception to take an unexpected tack.

I've never taken Chuck's excellent calculator to it's upper limits because my own fooling around with lengths and ODs always seem to produce the best expected sound and octave range when kept within the limits his calculator figures when tuning for the first natural frequency. Experimentation seems to put the particular OD of chimes and bells in the same physical realm as big speakers vs. little speakers. While both sizes can be fed the same frequency information, the name of the real efficiency game is how much air can be moved by a device at a particular frequency. The little speaker is being fed the same amplitude bass frequency as the large speaker; but it's small diameter cone simply cannot move enough air at a very low frequency to overcome the ear's poor low-frequency response curve. The large speaker can really punch out a lot of air movement at low frequencies, but the sheer mass of it's large cone does not let it move at a very high frequency - thus high frequencies are quickly dampened to almost nothing by cone inertia. While many OD/length ratios can allow almost any tube to be cut to produce almost any fundamental resonance, the large diameter tube has considerably more surface area in vibration and can move more air at a low frequency than a smaller OD tube; but it's mass also quickly dampens out higher
frequencies produced in the same manner as the bass speaker will kill the tone of a violin. So, rather than trying to make any particular OD do everything one wishes to produce (in low to high frequency range) it seems better and more reliable to use shorter, narrow OD tubing for high octaves - and much longer, larger ODs for low octaves. Brent

From:  "bmh1944" <bmh1944@yahoo.com>
Subject:  Re: More capped tube madness

I think Chuck's theory has considerable merit because he's been able to push toward a logical quantification of the long-tube mystery that frequently defies the tuning process. I'm sure there's a deep, yet to be found, mathematical and physics explanation for the varying conditions which radically change as tube length/OD ratios go far beyond tuning for the first natural frequency in the "ideal" range. If and when we will be able to predict the measured results, we'll still have to contend with relative frequency component amplitudes which actually exist, then also figure in the "adjustment" to those
relative amplitudes when passed through the human ear's terribly poor frequency response curve - then feed those results to the unpredictable "fuzzy logic" process of the brain for a "perceived" end result.

I've tossed together a few "long tube" sets for friends who wanted only a pleasing, ringing, aesthetic sound of mixed low and high frequencies with no tuning other than a noticeable difference in the general pitch of each tube. Even doing something as simple as that has frequently resulted in a tube which (for no reason) sounded pathetically poor and almost void of any sustained ringing properties. I don't have a clue as to whether that was due to a physical anomaly of phase cancellation of air column or different tube vibration modes - or due to some relative component amplitude change that causes a problem with brain fuzzy logic perception - or a little of both. Anyway, simply cutting off another 1/2" or so brought back the nice full ringing sound. So, I just go with what I get in those instances and don't frustrate myself with any tuning which could very easily change things again for the unexpected worse.

When I get serious about a beautiful tuned note at any desired octave, I use Chuck's excellent calculator for ONLY the first natural frequency. If you keep plugging in different ODs until you find the "ideal window" for the fundamental (first natural frequency) desired, it's very hard to go wrong. I'm reasonably sure this always keeps the desired fundamental as predominant and also gives much attenuation to undesirable overtones. High notes are shorter lengths of small ODs; thus, eliminating any lower components even existing.

Very low notes are "ideal" with 5" to 6" ODs which move a lot of air at that low frequency and tend to quickly decay higher frequency components due to the tube's increased mass and inability to move at a very fast rate. It all seems to work very well with a great predictable result and ease of making very fine tuning adjustments.

I agree with Jack's thoughts in feeling there's probably more scientific and mathematical merit in this group than may very well exist in professional windchime or orchestra chime makers. It stands to logic that, if they had some miracle magical formula, they would be making every note in the C1 to C9 range out of 1"OD tubing - LOL. I think Chuck (and Jack) have come up with the formula for predictable tuning at the first natural frequency that many professional manufacturers have spent a lot of time filling the scrap yards with past experiments in finding the right OD/length combination for each particular note or range. All of them go with smaller OD for higher notes and much larger ODs as the note goes lower in octave.

Brent

From:  "abhunkin" <abhunkin@uncg.edu>
Subject:  Re: More capped tube madness

Having built a set of 1"OD aluminum windchimes with lengths from 10" to 33" at every half-inch interval, I've found the following:

In the "ideal length" range (see Chuck), the first harmonic (fundamental) is strong and resonates a long time; the fourth
harmonic is weak and only present at strike time (no ring).

As the tube gets longer, the first harmonic becomes softer (it seems to ring just as long, you just can't hear it as long!), the fourth harmonic gets louder, rings longer and eventually (with longer tubes) completely overpowers the fundamental. In addition, other frequencies begin to appear (at the strike point only - no sustain). I stopped at 33" because beyond this, the fundamental was effectively lost.

*Above* ideal length (shorter tubes), the fundamental tone is pure and the fourth harmonic gradually disappears even at the strike point, but the fundamental itself loses its sustaining ring. Shorter than 10" is more resembles the "clunk" of the fourth harmonic strike in the ideal length range.

As for me, I'll stick with Chuck's ideal length ranges, and – wishing to have as wide a pitch range as possible - work with mixed diameters ranging from 1/2" to 1 3/4". I'll have two octaves of pitches, all fairly pure and resonant. Art Hunkins

From:  "abhunkin" <abhunkin@uncg.edu>
Subject:  Re: More capped tube madness

Just bought some "Knock Out Seals" for electrical boxes, which work just as well as the "Metal Hole Fillers" - and cost a lot less. They come in a variety of sizes. The 1/2" size fits 1" aluminum tube. (The seals will fit tube up to 2" OD or so.) Also, their prongs make for a more secure fit inside a tube.

Found that drilling holes in them do not change the pitch any, so I'll be drilling 7/8" holes in the larger seals to insert PVC connectors sized variously, to see if this is a good way to micro tune pitch. Art Hunkins

From:  "abhunkin" <abhunkin@uncg.edu>
Subject:  Re: More capped tube madness

Thanks to Brent for the great info. Here are my latest "findings."

To clarify: all my chimes are hung from the node - either double-hung or single-hung from an axle. Shortly, I'll try Brent's critical idea of hanging by an end cap. My only concern is that the caps are friction fit, and I fear may well separate from the tube (disaster?) I'll surely need to experiment on my cheaper tubes rather than Gil's good ones.

I've directly compared the two methods (double vs. single hung at the node): with a 1" chime (ideal range) double-hung I get a 30-35" ring time; with a knock out seal in the end, 25-30" ring time. Same but single hung: 35"+ ring time, 25"+ ring time with a knock out seal. Single hung has a slight advantage, which seems neutralized by inserting the knock out seal. (Single-hung *tone* also seems just a bit clearer or more resonant. But I could be imagining this; the difference is *subtle*.)

The knock out seal lowers the pitch one-half step (like from C to B), and decreases the ring time modestly. (It also seems to cut down modestly on the upper harmonic content or its duration. With respect to upper harmonic content, it is also clear that striking the chime at the end is brighter, with more harmonic content, than striking it in the middle; and this effect is stronger than that of adding the knock out seal. Has anyone ever mentioned this?)

The main experiment: I used a single-hung (at the node) 1.5" chime to test further, adding CPVC .5" connectors. This chime (also cut by Gil - and within its ideal range) has a ring time of 40-45". Adding a knock out seal, it rings 30-35"; adding to the 13/16" hole in the seal (which has no effect on anything) the CPVC connector, reduces the ring time slightly to 25-30". The seal lowers pitch a half step; the full CPVC connector, an additional half step.

Note that *in total* we're only talking about lowering the pitch per tube a whole step (as in D to C). (This is what I mean by micro tuning, which is usually thought of as changes in pitch less than a half step.) Using a shorter CPVC connector (by cutting some of it off) allows for micro tuning the tube through the range of a half step. Ring time doesn't seem to change depending on the length of the connector.

At worst, the tone quality is slightly less bright when adding these "encumbrances"; but at least part of this seems to be less of the upper harmonics, as mentioned above.

My summary: this is quite a good way to do micro tuning with larger OD chimes (1.5" and up). Tone quality and ring time changes only minimally, and within quite acceptable limits. Of course, if such tuning is not desired, the knock out seals should be removed (an easy job; they are friction fit).

Gil is working with me on a comparable solution for smaller OD tubes (down to .5"). Knock out seals (or metal hole fillers) are available for smaller size tubes, but the CPVC connectors are too large, and I can't find anything comparable that is smaller. Gil, however, is making something! I can't wait.

Oh yes, one other thing: I tried adding one or more strong, ceramic (one-inch circular) *magnets* to the knock out seal. Forget it! Though the pitch did change (lower) as anticipated, I got mostly "clunk", rattle - and even the strong magnets quickly begann to slide off. Can't win them all! Art Hunkins

From:  "abhunkin" <abhunkin@uncg.edu>
Subject:  Two More Capped Tube Experiments

More capped tube madness - but unfortunately both experiments are duds.

I was really hopeful, thanks to Brent's words, about the idea of end (antinode) single-point suspension, from a cap.

I used my own 1" OD aluminum tube (within ideal range). Using a traditional axle-type two-point hanging, the tube vibrates for 20". Interestingly, putting a knock out seal on the bottom end (lowering the pitch a half-step) keeps the vibration at near 20" as well.

Suspending the same tube from a top end cap (in this case a metal hole filler) very much muffles the sound, giving only 5" of ring. I tried two ways: with a great big knot (the hole was 1/4" plus), then with the cord wedged between cap and tube but rising through the hole. Same result with either method. This is with a good friction fit between cap and tube; no danger of falling off, and no rattle-just really under whelming. Adding a knock out seal to the other end as well, was equally so.

Brent, any idea how this result might be notably improved? Don't know whether currently I'd be inclined to permanently affix a cap to one tube or not. And I don't imagine changing the size of the hole in the cap will change anything - right?

Second experiment: I tried two methods of screwing a 1/4 to 5/16" bolt through a lower end cap (metal hole filler). Both obtained snug fits, particularly with the addition of a snugging nut. And both pretty much resulted in "clunks." I'm convinced there is no future in pursuing this approach, even for tuning within a very restricted range.

From:  "bmh1944" <bmh1944@yahoo.com>
Subject:  Re: Two More Capped Tube Experiments

Art;

You are now a sworn member of the frustrated experimenters club which resides at this site - lol. Experimenting is fantastic because iit's a challenge that most of us "tinkers" really enjoy; many times the results are pretty under whelming, but there have also been many occasions where things have been discovered that let us all take a major step forward. We not only learn from everyone's success, but it's also very valuable to learn many things not to do or try.

I don't think anyone has really explored the differences between capping one end of a tube (compared to an open tube) of all metals and all ODs to arrive at any concrete, always predictable result. In a tube open at both ends, the fundamental transverse mode at resonance creates a "standing wave" property where forward and reflected waves of vibration blend into a wave that doesn't appear to be moving; so, antinodes of maximum vibration movement and nodes of less physical movement always lie at the same mathematical points along the length of the tube. However the internal air-column movement (and influence on a thin-walled tube's vibration modes) is freely changing flow direction and ambient air pressure both
internally and at each end of the tube. When one end of a tube is capped, there is suddenly a maximum impedance to internal air-column movement at that end - which may very well cause the reflected air-column movement to also take on a "standing wave" characteristic to some degree.

I've found that whether using a preformed cap (over the end of the tube) or an internal disk or washer that's welded or epoxyed in place always produces about a 1/2 step lowering of the fundamental and related overtone frequencies. I don't have a clue whether this is due to a change in the internal air-column's movement and influence, or because it creates a major null to other vibration modes which are also resident along with the predominant transverse mode, or perhaps a little of both; but I do know it happens with almost any OD of type of tubing. A few have said that capping both ends of a tube didn't create any problems; but I've never found that capping both ends of a tube has ever resulted in anything but another "windthud". The hole in the end of a cap should only be large enough to let a very thin suspension line pass through it and be held in place by an internal knot in the line. Making the hole larger will not cause any major changes in the impedance to the internal air-column movement, nor will it change the major null to other vibration modes to any degree.

Most all modern orchestra chimes are capped at the top and single-line, center suspended. There are a few orchestra chimes that use an internal node axle for tube suspension, and the hole in the cap is large enough so the sides of the hole do not touch the suspension line and dampen the tone. Yet, most orchestra chimes seem to prefer being center suspended from the cap itself. I've noticed that the cap seems to allow the fundamental frequency (first natural frequency) to sustain a little longer while also increasing the amplitude of some overtones.

While the unknown factors presented by a cap are yet to be fully documented or explained, it seems to help with some types of tubing and be a detriment to other types. Note that almost all capped orchestra chimes are brass - not aluminum or steel. My thoughts as a musician reason that brass (and copper) produce a considerably more mellow, less loud sound that blends well with other instruments; and perhaps the cap helps increase the tonal qualities of brass or copper. But, since windchimes are not used in an orchestra, the louder, brighter sound of aluminum or steel is more desirable to many people; and I've not done any comparative experiments with either capped aluminum or steel tubing vs. open tubing.

I use a lot of capped tubes when going for very long tubes where tuning is not a factor and only the aesthetics of different tones is desired. I've used capped tubes which are both cap suspended and node suspended with no real discernable difference between the two suspension types. You are correct in the fact that an end strike of a very long tube will produce a fuller sound than a center strike because the end strike excites all vibration modes and overtones to the maximum. In very long tubes (which greatly exceed Chuck's length/OD ratios when tuning for the first natural frequency), tuning becomes an unpredictable nightmare from the many things we've already tossed around about missing overtones, radically changing frequency component amplitudes with minor length adjustments, and the brain's fuzzy logic process that takes an unpredictable tack when amplitude ratios change.

So, when I'm out to make a precisely tuned chime with the greatest degree of mathematically predictable results, I NEVER exceed the "ideal length" window (to much degree) of calculated length/OD ratios (for a given note and octave/pitch) as derived from Chuck's calculator when tuning for the FIRST natural frequency. Following those guidelines will create a tube with a very predominant and audible fundamental frequency with little distracting influence from non-harmonic overtones. When using this method, precise tuning by minor length adjustments is very predictable and relatively easy to do. Thus, there is no real need to install any add-on or mechanical tuning device that usually is a bit distracting in eye appeal.

I was really hoping that suspending a capped tube by your threaded tuning rod would have been a success, because that would eliminate the need for all the little grinding and filing processes of tuning. Plus, being able to end suspend such from the antinode would allow tuning without worrying about changing the node-hole location of a new fundamental frequency brought about by tuning. Oh well, keep trying different methods and you still may find something that takes all another step further in knowledge - I could sure use some - LOL. Brent

From:  "abhunkin" <abhunkin@uncg.edu>
Subject:  Re: Two More Capped Tube Experiments

Just for fun (?) I did one last experiment suspending my 1" tube from a cap. This cap was a knock out seal with a very small (1/16") hole in it. The wind chime cord has a single knot inside the tube. (My previous attempt was with a 1/8" hole and a huge knot.)

The result was not as disastrous as before, but still not good. Whereas a double-hung node-point mount (with the same cord) gave a resonant 20" ring, the end cap mount ring lasted at most 15" and was not as full and resonant.

While I was at it, I tried *both* ends capped (with end cap suspension), inserting the bolt and nut into the lower cap. Same result as the bolt and nut previously - "thud."

Conclusion: give up the end cap suspension idea. It doesn't begin to compare (at its best) with either variety of node suspension.

The lower end cap with magnets and with bolt (and nut) are also disasters, and "history."

So I'll patiently wait for what Gil is going for me (the equivalent of CPVC connectors inserted in an end cap, but for smaller diameter tubes). Incidentally, he came up with a *beautiful* experimental "headpiece" (suspension header) for me, which I'm just now experimenting with and will report on. Art Hunkins

From:  "Brent" <bmh1944@yahoo.com>
Subject:  Re: Center Suspension Ideas

Howard;

I hope I'm being a help more than a confusion factor with the advice; Doug may be able to help you more with finding good (inexpensive) sources for stainless steel cable in small amounts; but most fishing supply places only sell solid wire leader - not cable (wire rope).

Doug is more experienced in using cable suspension for small/short chime tubes than I am because I've been notorious here for experimenting with very long tubes over 5 feet. I never even tried cable suspension until happening upon some extremely thick-walled, 5" diameter hard copper tubing in a metal scrap yard at a give-away price. Since the longest tube in my "chimezilla" was a bit over 8 feet and weighed 178 pounds, SS cable was the only solution because chain caused major dampening and braided nylon couldn't be trusted to hold the weight after a few months of UV exposure. So, for what they're worth, I'll give you my thoughts on everything you requested.

I only recommend center suspension from an inverted-V node axle when the tube's OD and length ratios fall within Chuck's "ideal length" criteria for tuning at the FIRST natural frequency. I say this because tuning at any higher natural frequency makes the tube much longer with respect to it's diameter; and a long, relatively narrow tube will tend to rock on it's suspension axle (considerably more than a shorter tube) which causes the upper end of the tube to frequently make contact with the internal suspension line and dampen the tone. I haven't tried cable suspension of relatively
lightweight/short chime tubes because I've always figured (perhaps erroneously) that it would cause more dampening of sustained tube vibrations because it has absolutely no "stretch factor" and is considerably more stiff than braided polymer or monofilament line.
For lighter, shorter tubes I prefer to use "Spiderwire" braided polymer-fishing line because it's extremely small in diameter, stronger than steel (at the same diameter), has a higher degree of flexibility than nylon, and is extremely good at UV resistance.

If you are still seeking stainless steel cable, I ended up using 1/16" diameter, 7X7 stainless steel wire rope (480 pound test) for "chimezilla". It's a bit pricey at around 45 cents per foot, but it worked out really well. You can go all the way down to 1/32" diameter SS cable (110 pound test) at around 25 cents per foot if you go to (www.wrca.com). From their home page, you select "wire rope" then "specialty small ropes"; you'll only find some of the prices from their "aircraft pricing" PDF sheet - so, it's best to call their toll free number to get a better quote. Doug may have a better source
for crimping sleeves, but Bass Pro Shops (www.bassproshops.com) has a number of different sizes available. I've used nothing more than the ends of plain old crimp-on electrical wire lugs that I had cut the business end off with a pair of wire cutters.

One warning about steel cable is that it is terribly hard and extremely abrasive on the chime tube's suspension axle. I had very little luck in finding a cable thimble (for very small diameter cable) that would protect from abrasion; so, I tried using very small diameter copper tubing for such, but I could hear a little chattering going on when the tube was struck. I ended up making a double, overlapping wrap of the cable around the inverted-V suspension axle (before pushing the axle up inside the tube), installing the crimping sleeve, then (using a pair of vice-grips on the short end) pulled the cable very tight around the axle before crimping the sleeve in place. After making sure everything was aligned properly, I finished it off with a drop of super-glue that not only held things in place with no slippage, but also would bleed into the joint and help prevent any corrosion problems from dissimilar metal contact.

Cutting tubing has always been labor intense - especially when tuning a tube. Grinding off small amounts during tuning always leaves a nasty mess that has to be shaped and smoothed (inside and outside) with a file after each grinding process. Likewise, cutting tubing with a carborundum-type blade in a chop-saw makes an equally nasty cut that takes a considerable amount of work to smooth; plus cutting very small amounts off the tube's end (during tuning) is very difficult with an abrasive blade. I started out by outfitting a small bench top band saw with a metal cutting blade, but (perhaps for safety
reasons) those narrow metal cutting blades are getting harder and harder to find. So, depending on how many six-packs it takes to get the wife to agree (and how many it takes for you to get the cahoonies to ask), Delta and Ryobi both make metal cutting band saws in a chop-saw fashion that sell for about $175 if you're going to do very much tubing or rod cutting in the future. A fine-tooth blade will let you cut almost any tubing with such a very slight degree of burr that a quick spin in steel wool and a fast internal ream with an old pocketknife will cure. Additionally, you can easily and quickly cut as
little as 1/16" off the end of a tube (for tuning) in the same clean and square manner.

 With steel prices going through the roof, there's not a lot of price difference between steel, aluminum, and hard
copper anymore:

Steel (galvanized EMT) is probably the best for loudness and favoring/sustaining the higher frequencies better, but there's the problem of rust at the cut ends and node holes; plus, any welding of an internal suspension axle (if desired) discolors and burns away the galvanized coating.

Hard copper (or brass) is easier to work with, but is considerably more mellow in sound produced, and seems to favor/sustain lower frequencies better than other metals. Internal axles can be quickly soldered or brazed in place with a little propane torch and smoothed to an almost unnoticeable presence on the tube's exterior. Copper or brass can either be highly polished and coated with a thin layer of polyurethane outdoor-grade varnish, or can be chemically treated to a
beautiful blue-green patina finish. Another factor to consider is that copper is much heavier than steel, and requires additional strength in suspension criteria.

Aluminum is the lightest of practical metals to use, doesn't rust, and can be preserved at a highly polished luster with a thin coat of polyurethane varnish if preferred. It has almost as good an ability at sustained high frequency tone as steel, but not as good at sustaining low frequencies as hard copper or brass. It is as easy to cut, smooth, and polish as copper and brass; but (other than using JB Weld), permanently installing an internal suspension axle requires a small MIG welder, which seldom comes at less than $300. Being considerably lighter in weight than other metals, aluminum doesn't require nearly as much strength in suspension line or support disk/halo; but those properties are not good in areas of a prevailing
high wind condition because the lightweight chime tubes fly around in a brisk wind almost as much as the striker/sail assembly does.
CAPPING ONE END OF A TUBE:
Not caring to re-invent the wheel, I wondered why orchestra chimes are capped at one end. Since I was already playing with hard copper, soldering a little plumbing cap over one end was simple. Using my same "Fred Flintstone" testing method, I noticed that the little spectrum analyzer program now showed an almost equal level of sustain and decay rate in the fundamental, first, second, third, fourth, and fifth overtones (the second and fourth hadn't been there much before). Not only was the "chime" sound even richer and fuller, but the tube also resonated it's deep sound for at least a second longer
than the open tube. My metal guru was of absolutely no help with this phenomenon because they fill all tubes with plastic foam to completely kill any possible resonating vibration at any frequency (metal fatigue equals the airplane "buying the field").

Most of us tinkers understand the "standing wave" concept of the vibration pattern in a resonating object with a finite length. However, the air inside a hollow tube is pumping in and out both ends of the tube (from the peristaltic rippling of the tube's walls) in an equal but opposite vector direction (180 degrees out of phase with each other). Making only a wild guess and unjustified assumption, I'm thinking that capping one end of the tube reflects any air movement in that vector direction and creates a similar "standing wave" property to the internal air column as well. If the air column's standing waves were somewhat in phase with the standing waves of the tube's walls, I would imagine that this would help sustain the
oscillating action of both entities - and result in a small decrease in the decay rate of any resonant frequency and/or it's overtones. Don't ask me to quantify this, but I've just noticed the difference - and nothing more.

One more very nice thing about capping the tube (a long one anyway) has come from the considerably more efficient and aesthetic ability to mount the tube to the chime assembly. I've tried all sorts of axles, types of line, and various techniques to mount the tube at its upper node point; but, for a very long tube (from 3 to 6 feet in length) the concept really sucks in practicality. When you suspend a 5-foot tube from the node point, you've created a long lever with the "fulcrum point" about one foot down from the top of the tube. Not only does a little wind start the whole thing in a rocking motion, but the suspension lines must be spread at a very wide angle to prevent the tube from contacting the suspending line and dampening the sound (thus, your mounting "halo" would have be larger than the diameter of a basketball hoop for an 8-tube chime set). A single center line (down the inside of an open tube) that's attached to an internal axle is even worse on a long tube because the rocking motion makes the tube contact the single suspension line even more often and to a greater degree. Soooooo, since long orchestra chimes are capped at one end and suspended by a single line attached to the cap – I tried out the concept. After drilling a 1/8" diameter hole in the center of the cap, I threaded a 1/8" braided nylon cord down through the cap until it came out the open end of the long tube. I used a cigarette lighter to heat the end of the line and fuse all the fuzz together; then I tied a "half granny" knot at the end of the cord and secured it with a drop of super-glue. Next, I pulled the line back through the hole in the cap until the knot seated against the cap and hung the tube with about 6" of cord between the tube and the point from which I hung it. Going all the way, I cut an identical length of open tube and suspended it in the usual manner (with fine monofilament line through holes drilled at the node point) right next to the capped tube. Striking both tubes in the middle and at the lower end did not give any discernable difference between the two in tone, sustain/decay rate, or anything else - whoopee!!

As a result, I was able to hang eight 1" OD tubes from an 8" diameter "halo". Since I'm after the "chime" sound, each tube was hung with different lengths of cord so their bottom ends were all at the same level; and my striker hits each one of them about 1" from the lower end. It was interesting to note that striking the tube on the relatively thick cap produced almost as good a sound as striking the tube's lower open end - but not really practical because the striiker needs a little length to get better leverage on it's sail- induced swing. The aesthetics are great because there's only a single cord suspending each chime tube; the durability is also great because there's very little (if any) chafing effect on the 1/8" diameter (300-pound test) nylon suspension cord because there's almost no movement at it's junction with the tube's cap while the tube is swinging. Why is there little or no difference from single-line, end-suspension (from the middle of a cap) compared to the usual double-line suspension from node holes in the sides of a tube? I don't have a single documented clue!!! I would suspect that being suspended by two opposing lines (in different vector directions) creates a dampening
effect to the small vibrations that DO exist at the node, while the cross-sectional friction of the line through both walls of the tube also has a bit of dampening effect as well. Even though top/center mounting of a capped tube is done at a major vibration point, I think the single line (with much elasticity) doesn't create nearly as much opposing bi-directional resistance like the traditional dual-point, node-mounting technique does. Additionally, the elasticity and extremely limp properties of the nylon cord "lets the fish flop at the end of the line" with as little dampening effect as practically
possible.


Have you noticed yet why this is such an experimental group? There are so many different tastes in produced sound, desired aesthetics, suspension techniques, and limitations of one's patience (and available alcohol quantities) that avenues of pursuit abound in untold numbers. Brent

Links:

Making Wind Chimes by Jim Haworth

Windchimeconstruction  Join Yahoo’s message board and get more information about wind  chime making.

Updated 3-24-05