steel bar tunings by Rick Sanford

Rick Sanford writes:

>>>
Question number one is - How are the bars suspended
thus where are the nodes?

( are they held only at one end
or are they drilled at two (or more) points
and suspended there)
>>>
Darren Replies:

Rather than approach this one question at a time, I have included below all
of the characteristics of the bars. The main problem I have encountered
with tuning them is that grinding between the nodes causes a phase or
beating effect. I want to tune the bars, but preserve their timbre. I
have tried a few things which are detailed below. Any ideas you have,
Rick, or any one else on the list, would be greatly appreciated.

These bars comprise 1961 bell instrument that attempts to mimic cast bell
carillons. Characteristics of the bars:

1. Each has two solenoid strikers. One primarily emphasizes the
fundamental (placed at the end of the bar), the other the second harmonic
(place just inside the vibrational node). This creates an overlap
ofoctaves between the adjacent octave racks of bars. (5 racks of bars)

2. The bars are varying lengths from about 20 cm to about 56 cm. 4 mm
square at the end (3/16th")

3. the bars each have 3 brass (or bronze?) collars firmly attached along
predictable but somewhat
inconsistent places between the vibrational nodes. They follow a general
pattern but are not in exactly the same relative postion from bar to bar.
It is my theory that these collars are used to tune and/or supress the
inharmonic partials. The Collars are larger on the longest bars 5/16th"
wide, 3/32" wide on the shorter bars
I have experimented a little with moving them and
have found doing so creates significant changes in the timbre of the bar.
I have also noted that moving one collar further from its original position
causes a "beating" effect -- the further from original, the faster the
beating. This is not a very predicable, consistent effect.

4. Tuning of the bar by grinding metal from between the
vibrational nodes lowers the fundamental frequency as expected. It has the
unfortunate effect of causing a slight phasing of about 2 cycles/second,
and significant loss of the original timbre of the bar.
****What is the cause of this effect?**** This effect is somewhat
corrected by moveing the brass collars, but not perfectly.
Placing weight on the end of bar also lowers the frequency, but that also
causes a phase effect.
I am experimenting with various tuning methods to determine how tune
without causing a phase effect.

5. There are five octaves of bars.

6. There is a notch cut away on either side of each bar at the vibrational
nodes (20% of the length of the bar) effectively making the bars thinner at
the nodes. About 1/16th" thick at the nodes.

7. The bars are suspended with string. On one end the suspension point is
close to the vibrational node. On the other end, it is at bit further from
the node, and closer to the end of the bar. The suspension points are
consistent, relative to the total length of each bar. The longest bar is
56.6 cm. Each of the nodes are at 11.6 cm (20%). The suspension point on
the "magnet" end is at 7.5 cm (13% of the total length). The suspension
point on the "solenoid" end is at 10.6 cm (18%)

8. Each bar has a small but strong magnet one end that pokes up thru a
coil of magnet wire.