 |
Description  |
|
|
BACKGROUND OF THE INVENTION
This invention relates to musical instruments and more specifically to a
unique method and apparatus for enabling operation of electronic music
synthesizers by musicians accustomed to playing non-keyboard instruments
such as guitars, banjos, etc.
For many years, musicians utilizing single instruments have been presented
with electronic and/or electromechanical methods for controlling other
instruments. For example, various methods for utilizing guitars to control
electronic organs have been disclosed. Other methods for utilizing trumpet
instrument fingerings to produce corresponding tones on electronic organs
have also been disclosed.
With the advent of the electronic musical synthesizers, a tremendous range
of new sounds has become available. The first musical controller utilized
for the synthesizer was the standard piano-type keyboard which allowed
pianists and organists access to the synthesizer. Several years later,
guitarists were able to enter the realm of synthesizers with the
introduction of guitar-to-synthesizer interfacing devices, sometimes
referred to as guitar synthesizers.
The prior art guitar synthesizers are of two general types. One type
provides the user with a specially constructed guitar with sensing
elements in the neck to indicate tones being played. A second type
provides a special electromagnetic pick-up for the musicians' own guitar
to allow sensing of the strings vibrations which in turn are interrogated
externally by additional circuitry to determine tones being played. In
both cases, after the tone(s) are determined, conversions are implemented
by electronic means to produce voltages and control signals that drive the
electronic synthesizer to cause the production of corresponding tones.
Both of these prior art methods have severe shortcomings. With the
specially constructed guitar, as provided in the prior art, musicians must
either swap their own guitar for the synthesizer guitar each time they
require the special synthesized sounds, or play only the special guitar.
Since guitarists spend much time and money in selecting their own guitar,
there is strong resistance to playing another guitar full time, even to
acquire special voicings. The total inconvenience of switching guitars in
mid-song should be apparent.
With the special electromagnetic pick-up method, the user can continue to
use his own guitar and switch the special effects in or out as required.
However, this method creates severe problems in the electronic extraction
of the tone. The vibrating string is a complex sound generator and creates
tremendous problems when trying to determine the one tone that is being
played. This gives rise to poor reliability, high cost and extraneous
sounds being produced unless the musician alters his playing technique
drastically. Additionally, feed-through from one vibrating string to
another does not allow polyphonic synthesizers (more than one note at a
time) nor usage with hollow-body guitars. A hollow-body guitar by design
has more resonance and thus more intense string vibrations than solid body
guitars. Vibrations caused by the plucking of one string causes
sympathetic vibrations of the remaining strings and may cause a situation
where the tone extraction circuit cannot effectively operate in this
highly interactive environment.
A third type of musician-to-synthesizer interface that has not found any
commercial implementation but has appeared in some periodicals is
mentioned here as a comparison. The interface is typically made in the
form of a flat typewriter-type keyboard with rows and columns of keys.
Playing the instrument is done by pressing one or more of the keys using
typewriter techniques. Even if the keys are capacitive touch keys instead
of mechanical action switches, this interface does not approximate a real
instrument of any kind and is useful only as an experimental device for
musicians not accustomed to either keyboards or guitars.
From the above description it can be seen that all existing
guitar-synthesizer interfaces suffer from one or more deficiencies; the
most prevalent being:
(1) causing the guitarist to modify his normal playing technique by
requiring extra care in plucking or fretting;
(2) eliminating several normal guitar characteristics such as sustain, open
notes, chord capabilities or hammer; and
(3) causing the guitarist to totally give up his own guitar for a specially
modified device.
SUMMARY OF THE INVENTION
It is in general an object of this invention to provide an improved
stringed instrument-synthesizer interface.
It is another object of this invention to provide a control mechanism for
electronic musical synthesizers and the like which can be manipulated in a
manner familiar to musicians of stringed instruments such as guitars,
banjos, etc. or other non-keyboard instruments.
It is a further object of this invention to provide a device allowing the
non-keyboard musician access to electronic music synthesizers while
allowing him full use of his own instrument during non-synthesized
passages.
It is another object of the invention to achieve highly reliable tone
conversion without any change in playing techniques.
It is yet another object of the invention to provide a non-keyboard
synthesizer interface apparatus which can be polyphonic and compatible
with guitar, banjo, and mandolin techniques.
These and other objects and advantages are achieved according to the
invention in conjunction with a fret-bearing surface having at least one
string stretched thereover. Electronic means associated with the
fret-bearing surface is designed to provide accurate generation of volume,
gate and frequency control signals for an electronic synthesizer. Novel
aspects include means for accurately detecting string depression, means
for detecting slide and/or hammer movements, means for detecting
open/close string status and means for properly responding to muting of
strings, as well as combination of these various means into a cooperating
whole. In a preferred embodiment, a micro-processor is used to time-share
a plurality of strings to provide any or all of the above novel means.
Novel means are also provided for minimizing and simplifying the
multiplexing of the electrical interface with the fret-bearing surface.
Another novel aspect of the invention is that the resulting synthesizer
interface device may be attached to the musician's own instrument, thus
allowing him access both to the synthesizer, as well as his own
instrument. The invention is more fully pointed out in the appended claims
.
BRIEF DESCRIPTION OF THE DRAWINGS
The preferred embodiments and best mode presently contemplated for
practicing the just summarized invention will now be described in detail
in conjunction with the drawings of which:
FIG. 1 is a perspective view illustrating a stringed, fret-bearing surface
according to the preferred embodiments of the invention.
FIGS. 2A and 2B illustrate mounting of the apparatus of FIG. 1 in
conjunction with a typical guitar or microphone stand.
FIG. 3 is a circuit schematic diagram illustrating an implementation of the
preferred embodiments.
FIG. 4 is a waveform diagram illustrating trigger pulse production from a
string pluck.
FIG. 5 is a waveform diagram illustrating peak detection of the output
resulting from a string pluck.
FIG. 6 is a schematic block diagram illustrating the frequency control
signal production technique of the preferred embodiments.
FIG. 7 is a schematic circuit diagram illustrating the fret pair detection
technique of the preferred embodiments.
FIG. 8 is a timing diagram further illustrating the fret pair detector
operation according to the preferred embodiments.
FIG. 9 is a schematic circuit diagram illustrating an open/close detector
according to the preferred embodiments.
FIG. 10 is a timing/waveform diagram illustrating the open/close detector
operation.
FIG. 11 is a timing/waveform diagram illustrating slide timer operation.
FIG. 12 is a schematic block diagram illustrating an embodiment of the
invention employing a micro-processor.
FIG. 12A illustrates a trigger pulse detector particularly useful in the
embodiment of FIG. 12.
FIG. 13 is a flow chart illustrating operation of the micro-processor of
FIG. 12.
FIG. 14 is a flow chart more particulary illustrating the frequency control
signal production portion of the micro-processor operation of FIG. 13.
FIG. 15 illustrates a multiplexing fret assignment scheme useful with the
preferred embodiments.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS OF THE INVENTION
FIG. 1 illustrates a guitar parallel fret board or "neck" 11 according to
the preferred embodiments. FIGS. 2A and 2B illustrate mounting techniques
which provide mounting to an ordinary guitar (FIG. 2A) or to a microphone
stand (FIG. 2B).
The fret board 11 has six strings 13 stretched across a neck-like structure
somewhat shorter than the neck of a normal electric guitar. The strings 13
are not used for vibrational tone production, but for touching and
therefore grounding certain switching elements embedded within or upon the
keyboard surface. There are fifteen frets 17. Each fret 17 is composed of
electrically conductive material divided into six segments, one fret
segment for each string. The six conductive segments of each fret are
electrically insulated from each other. Beneath the keyboard surface,
certain frets 17 are wired together for electronic minimization of certain
multiplexing hardware as discussed hereafter. While conductive fret
segments are utilized to detect string depression in the preferred
embodiment, many other switching means may be used without departing from
the scope of the invention.
The six strings are stretched for a short distance past the last fret 17
and pass over six electromagnetic sensors 19. These additional lengths of
the strings 13 are bridged by a rear bridge 21, and the strings are
heavily muted from beneath with firm foam material 23. The strings are
attached at a rear mounting bracket 25. The short string segments between
the rear bridge 21 and the rear mounting bracket 25, are used for
plucking. When a string is plucked, the electromagnetic pick-up 19
generates a short pulse train (20 milliseconds) with amplitude
proportional to the intensity of the pluck. The peak-detected version of
this pulse train is to be used to control the volume of the generated note
thereby giving control of dynamics to the player. The foam muting material
23 prevents interaction between fast plucks since even the fastest plucks
are typically separated by 50 milliseconds or more.
A sheathed cable 27 exits from one end of the neck, carrying various
control signals to external electronic circuitry as required. These
external circuits perform operations based on inputs from the parallel
keyboard 11 and in turn generate control signals for the synthesizer such
that the synthesizer will play the notes as played by the musician on the
parallel keyboard. These circuits will now be described in more detail in
conjunction with FIG. 3.
FIG. 3 shows a system for operation with one string. A description of the
structure and operation of the system will now be given to explain how the
system performs much like a normal instrument. To facilitate discussion,
explanation will be given initially for only one string, with the
extension of the technique to six strings illustrated thereafter.
The purpose of the overall system in FIG. 3 is to define to a synthesizer
(1) what note to play by producing a frequency control voltage (2) when to
play that note by producing a gate output, and (3) how loud to play the
note by producing a trigger amplitude signal.
The frequency control voltage is generated by a frequency control signal
generator 65. The digital frequency control signal is converted to analog
form by a digital to analog (D/A) converter 67, and supplied to the
synthesizer by a sample/hold circuit 69. The frequency control voltage
represents the note selected on the keyboard 11 by the musician.
The gate control signal is provided by a gate latch 59 which holds the gate
control information provided by a gate generator 57. The operation of the
gate generator 57 is controlled by two control signals, the trigger strobe
on a line 30 and a clear gate control signal on a line 63. Upon setting of
the gate generator 57 by the trigger strobe signal, the output of the gate
generator 57 on line 58 makes a transition from zero to a positive
voltage, causing an envelope volume increase in the synthesizer output.
When the clear-gate signal is sent over line 63, the gate control signal
drops to zero and decay of the envelope of the synthesizer output
commences.
The volume control signal is provided by the output of a peak detect and
hold circuit 31. This signal causes the synthesizer to modify its volume
in direct proportion to the amplitude of the peak stored by the peak
detect and hold circuit 31.
In FIG. 3, plucking of the string 13 initiates all tone production by
setting the gate generator 57. Each time the string 13 is plucked, a
trigger pulse signal is generated by the electromagnetic pick-up 19
directly beneath the string 13. This trigger pulse signal is amplified by
a pre-amp 28 and fed to a trigger detector 29. The trigger detector 29
converts the trigger pulse signal to a trigger strobe signal of voltage
level suitable for driving other elements of the control circuitry. This
is done according to the well-known technique of passing the plucked
output through a threshold comparator to generate a TTL logic level and
supplying this TTL logic level to a retriggerable one-shot to generate a
single pulse for each pluck. The resulting waveforms are shown in FIG. 4.
The trigger strobe appears on a line 30.
Additionally, the preamplified trigger pulse is peak detected and sample
held by the peak detect and hold circuit 31 to provide an amplitude signal
for driving the synthesizer's volume control. The resultant waveforms are
illustrated in FIG. 5. In this manner, the synthesizer output becomes a
direct function of the intensity of the plucking.
To faciliate detection of the particular note selected by the musician, the
string 13 is selectively subjected to zero volts (grounded). When
grounded, the string 13 is enabled to apply zero volts to each fret
segment it touches when the string 13 is pushed down at any of the one of
the fifteen fret positions.
The frets 17 are individually connected to the inputs, M.sub.1 through
M.sub.15, of a 16:1 multiplexer 35. The string 13 itself is connected to
the right most multiplexer input M.sub.0 such that with no depressed
frets, an "open string" indication will be given.
The frequency control signal generator is shown in more detail in FIG. 6.
In FIG. 6, generation of the proper frequency control signal is performed
by the cooperation of a fret scan generator 39, a preliminary note latch
41, a valid note latch 48, a fret pair detector 45, a slide timer 47 and
an open-close-open (OCO) detector 49. Two AND gates 51, 53 also
participate in this operation.
The fret scan generator 39 provides a select code to the multiplexer 35 and
to the fret pair detector 45. This code represents the fret being examined
at successive time intervals. Movement from fret to fret is caused by a
system clock called INCR (increment).
The fret pair detector 45 indicates a note has been selected by the
fingerboard player and provides a valid fret pair control signal to other
circuits. The valid note latch 48 latches the note actually to be provided
to the synthesizer and supplies it to the D/A converter 67.
The multiplexer circuit 35 (FIG. 2) is a slave to the control circuitry of
FIG. 6 and may be a Texas Instrument 74150 16:1 multiplexer. Four bits of
select code are then sent from the fret scan generator 39 in the
controller to the multiplexer 35. These four bits indicate in binary
format which of the sixteen multiplexer inputs M.sub.0 . . . M.sub.15 is
to be observed. The multiplexer inputs are successively scanned from left
to right M.sub.15, M.sub.14, M.sub.13, . . . M.sub.0. In response to the
select code, the multiplexer 35 generates a logic "1" output if the fret
segment identified by the select code has a string depressed against it or
a logic "0" output if it does not. These outputs are provided over a
connection 43 to the fret pair detector 45.
From this feed back on line 43, the fret-pair detector 45 ascertains
whether two adjacent frets are depressed or only one. The inventor has
found that detection of two adjacent frets being depressed reliably
indicates a note selection by the musician, whereas detection of only one
fret being depressed indicates a change in note selection such as release
of the current note or a slide to a new note.
In the circuit of FIG. 6, fret pair detection can be accomplished by
monitoring the multiplexer output 43. Upon detection of a first logic "1",
the select code of the fret scan generator 39 is latched by the fret pair
detector 45. If a second logic "1" is detected by the fret pair detect
logic 45 upon provision of the next successive fret scan generator output
to the multiplexer 35, the fret pair logic 45 produces the valid fret pair
signal on line 46. This signal causes the code latched by the fret pair
detector 45 to be transferred to the preliminary note latch 41. If, on the
other hand, the next successive fret scan generator output results in a
zero at the multiplexer output 43, no transfer is made to the preliminary
note latch and, preferably, no further scanning of the string is allowed.
If all sixteen frets 17 are scanned by the fret scan generator 39 without
detection of a valid note or a single closed fret, transfer of an all-zero
indication (an open string) to the valid note latch 48 occurs.
FIGS. 7 and 8 illustrate hardware and timing for the fret pair detector 45.
The hardware includes first and second fret flip-flops 121, 123, a
four-bit latch 125 and two AND gates 127, 129. The latch 125 and
flip-flops 121, 123 are supplied with the system clock INCR, illustrated
in the first line of FIG. 8. Initially, the first and second fret
flip-flops 121, 123 are cleared. Scanning begins at fifteen and progresses
to zero. Scanning is stopped at zero and re-started or will be stopped
when a valid fret pair is found.
When a fret is first found depressed, the first AND gate 127 is partially
enabled by the "fret depressed" signal on line 43. As shown by the "scan
number" in FIG. 7, the first fret found depressed in this illustration is
fret number nine. Since the first fret flip-flop 121 is clear, the Q.sub.1
input to the AND gate 127 is high causing production of the "load EN"
signal to the four-bit latch 125. The INCR signal going negative 128 then
loads the scan number representing fret nine into the four-bit latch 125.
This same edge 128 sets the first fret flip-flop 121, causing Q.sub.1 to
disable the first AND gate 127 from allowing any further loading of the
latch 125.
If the next fret, number eight in FIG. 8, is not depressed then the next
fall of INCR 130 will clear the first fret flip-flop 121. Scanning may be
terminated by production of a "cease scan" signal produced at the output
of an AND gate 126. The inputs to this AND gate 126 are the inverted form
of the signal on line 43 supplied by an inverter 124, the INCR signal and
the output Q.sub.1 of the first flip-flop 121.
If the next fret number eight is depressed, then the next fall 130 of the
clock INCR will set the second fret flip-flop 123 via the AND gate 129 and
clear the first fret flip-flop 121. The second fret flip-flop output
Q.sub.2 becomes the valid fret pair signal at the output of an OR gate
133. This signal allows loading of the preliminary note latch 41 with the
previously held scan number (number nine).
If other conditions discussed hereafter are valid, this number (nine in the
example) can be transferred to the valie note latch 48. The valid note
latch 48 is loaded when enabled by the "INCR" waveform but on the positive
going edge 131 of the clock INCR, which allows all decisions to settle
prior to loading the valid note latch 48.
If sixteen frets 17 are scanned without detecting an output on line 43, a
count sixteen circuit 132 produces the valid fret pair signal. This
counter circuit is incremented by the clock INCR and is reset by Q.sub.1
or upon reaching the count of sixteen. This signal produced by the count
circuit 132 enables loading of an all zero pattern indicating an open
string.
As noted above, the contents latched by the fret pair detector 45 are
supplied to the preliminary note latch 41 upon production of the valid
fret pair signal. Before the note can be latched by the valid note latch
48 as a valid note for presentation to the A/D converter, two further
conditions must be met at the AND gate 51.
The first condition is that the slide timer 47 is still running, as
indicated by the timer running signal supplied to the AND gate 51 by the
slide timer 47. The second condition is that a change from open to closed
or closed to open has not been detected, as indicated by the input to the
AND gate 51 from the OCO detector 49. Hence, the valid note latch 48 is
prevented from being updated if the slide timer 47 has timed-out or the
OCO detector 49 has detected a change in the open-closed condition of the
string 13. The operation of the slide timer 47 and OCO detector 49 will be
described in further detail below.
The OCO detector 49 is one of three method used to determine when to lower
the gate to the synthesizer. It stores the state of the string (open or
closed) when a pluck occurs and allows the gate to the synthesizer to
remain on until the opposite state occurs. If a player plucks an open
string, then the open note will sustain until the player depresses that
string or bridges it via his finger to an adjacent string. If a pluck
occurs with the string depressed onto a fret, the note will sustain until
the string is released.
A circuit embodiment illustrating the structure of an OCO detector 49 is
shown in FIG. 9. This circuit includes an open/closed (OC) flip-flop 80
and a gate flip-flop 81, as well as an exclusive-or (EX-OR) gate 82 and a
NAND-gate 83. The NAND-gate 83 is supplied with the signals on line 50
from the preliminary note latch 41. The output of the NAND-gate 83
indicates a "1" if the note represented by the contents of the preliminary
note latch 41 is an open note (all zeroes) and a "0" if that note is
closed. The output of the NAND-gate 83 is supplied to the "D" input of the
D-type OC flip-flop 80. The clock input 84 of the OC flip-flop 80 is
supplied with the trigger pulse on line 30. The Q output and the "D" input
of the OC flip-flop 80 form the inputs to the EX-OR gate 82. The trigger
signal, the output of the EX-OR gate 82, and the touch indicator signal on
line 38 (FIG. 3) provide the inputs to the gate flip-flop 81, which is
illustrated as a pair of cross-coupled NOR gates.
The operation of this circuit may be described as follows: Each time a
trigger occurs, the OC flip-flop 80 is updated--its Q output going to zero
if all zeroes are present at the inputs to the NAND gate 83 or going to
"1" if there is at least one "1" present at the input to the NAND gate 83.
The trigger also sets the gate flip-flop 81. The inputs of the EX-OR gate
82 match (11 or 00) and hence its output goes to zero, releasing any clear
to the gate flip-flop 81. If, following the trigger on line 30, the
preliminary note latch 41 supplied a second note code which represents a
change from open to closed or closed to open, then the inputs to EX-OR 82
will be different (10 or 01), thereby producing a "1" output which causes
the output 63 of the gate flip-flop 81 to go to a zero thereby dropping
the gate. The output of the EX-OR gate 82 is also inverted by an inverter
86 to provide a signal "NT" which will prevent the valid note latch 48
from latching the note representing a change in open/closed condition.
"NT" indicates that an open to close or close to open transition has not
occurred due to any valid note updates.
FIG. 10 presents waveforms illustrating the overall operation of the OCO
detector 49 for two cases. Case I shows an open note present when the
trigger occurs, setting the gate generator 57. Thereafter, the open note
going to a closed note clears the gate generator 57. Case II shows initial
selection of a closed note at the time the trigger on line 30 sets the
gate generator 57. Thereafter, the occurrence of an open note clears the
gate generator 57. The valid note latch 48 retains the original note.
The slide timer 47 is configured of conventional digital counter circuitry.
Each time a trigger pulse is generated on line 30, the slide timer 47 is
started. If the timer 47 reaches its maximum count of approximately 1/3
second before any new notes are made available to the valid note detector,
the timer will disable itself and disallow any further latching of new
notes. If, however, a new valid note is detected by the fret pair detector
45 and supplied by the preliminary note latch 46, for example, as a result
of a slide or a hammer, the timer 47 is started again, provided the new
note does not represent a change from open to closed or closed to open. As
long as new valid notes are supplied by the preliminary note latch 41
within each 1/3 second window, the repeat can continue indefinitely. As
soon as a pause of greater than 1/3 second is encountered, only a trigger
pulse will again start the slide timer 47 running. The slide timer 47 is
reset by the output of the AND gate 53 whose two inputs are the signal NT
from the OCO detector 49 and the valid fret pair signal from line 46.
The waveforms of FIG. 11, illustrate the operation of the slide timer 47.
When a trigger occurs on line 30, a closed note is present and the timer
running signal (line 52, FIG. 6) and gate signal (line 58, FIG. 3) start.
After a first interval of less than 1/3 second, a new valid note (closed
note #2) is found to have been selected and that note's number is
permitted to be transferred to the valid note latch 48 by coincidence of
the timer-running signal and NT signal. A second new note (closed note #3)
is detected within the next 1/3 second interval, and it is also latched by
the valid note latch 48. However, the third new note (closed note #4)
occurs after the timer 47 has run-out and the timer-running signal is not
present. Hence, the third new note cannot be latched by the valid note
latch 48.
The slide timer operation just-described allows the musician to move to a
new hand position on the fret board 11 following a note without having the
synthesizer play the note before a trigger, yet gives slide and hammer
capability to the instrument only if the movement occurs within 1/3 second
of a pluck. A typical "slide" is a movement of the finger along a
depressed string made in such a way as to cause progressive changes in
frets being touched. More musically it could be interpreted as glicendo
with discrete frequency points during movement. An example of a hammer-on
is the rapid movement of a finger down between two frets to cause a new
note to be sounded without a second pluck after a first note has been
initiated by plucking. An example of a hammer-off is the rapid release of
a fretted position to a new fretted position without a second pluck.
Hammer-ons and offs may be alternated rapidly after a single initiating
pluck for "trill" effects. Such movements are part of the technique and
artistry of the fret-board player and can be expected to vary.
The above discussion illustrates the basic techniques of applicant's
invention. These techniques may be readily extended to a keyboard having
six strings by use of several sets of circuitry or, more efficiently, by
use of a micro-processor As is well-known, a micro-processor contains
circuitry such as latches, timers, flip-flops and logic functions which
can be readily programmed to incorporate the basic techniques already
taught herein. That is, the latches used can be replaced by memory
locations, the flip-flops can be storage bits (so called flags) and
sequential operations can be replaced by coded instructions such as ADD,
SHIFT, COMPARE, etc. The conversion of such hardware is readily apparent
to those skilled in the art of micro-processor programming and
implementation aided by this disclosure.
FIG. 12 illustrates a six-string system utilizing a micro-processor 141 for
replacing or sharing most of the hardware of FIG. 2. The micro-processor
140 used can be one of several available single chip micro-processors such
as the INTEL 8048 with port expanders to increase the number of input and
output lines available. The processor 140 has internal storage locations
and a software controlled timer. Such a processor typically has available
"tri-state" ports which, in addition to providing logic "1" and "0"
outputs, also may provide a third high impedance state of several
megaohms. In the preferred micro-processor embodiment, the string enable
outputs, Port 1, are such tri-state ports.
FIG. 12 shows the multiplexer 35 and the D/A converter 67 being shared by
the processor 140, the remaining frequency determining hardware 65 of FIG.
3 having been incorporated into the processor. The other labeled blocks of
FIG. 12 are duplications times six of the corresponding circuitry from
FIG. 3 and operate as previously described. For micro-processor
implementation, it is advantageous to use the leading edge of the trigger
"pulse out" (FIG. 4) to set a "D" flip-flop which is then cleared by the
processor 140 when it reads the trigger status into the port. This avoids
erroneous readings on elongated pulses. This device is illustrated in FIG.
12A.
A clarifying description of the timing sequence is now provided in
connection with FIG. 12 to indicate generic operation of the
micro-processor interface. Each string is handled by the micro-processor
140 in sequential fashion from 1-6, then 1-6 etc. Each string is given
exactly 1/6 of the allocated sequence time (assume 2 ms) for its service
routine. That is, the longest series of coded instructions necessary to
perform the longest operation must be less than 1/6 of the total.
Otherwise, all service routines must be elongated and the total elongated.
If a service routine is shorter than this maximum, a wait state is entered
to fill the time before progressing to the next string. This allows time
measurements to be performed by each service routine since each entry to a
service routine is exactly 2 ms later than the previous entry. For
example, to do the slide timer operation, one can start with a storage
location set to zero when a trigger occurs for string "n" and increment
the location by 2 ms. each time that string is serviced. After 166 entries
we have accumulated 332 ms. which is 1/3 second as required. If a new
valid note is received during that 1/3 second (which is a condition
previously described to re-start the timer) the micro-processor 140 can
reset the timer location of string "n" to zero and continue counting.
The sharing of the D/A converter 67 by all strings is done by connecting
six sample/hold modules 143 to the single output of the D/A converter 67
and selecting the appropriate module based on the string selected at the
time. To avoid slow discharge of the sample/hold capacitors within the
modules, the sample/hold module 143 is refreshed at least once during each
2 ms. scan time.
The sharing of the 16:1 multiplexer 35 is straightforward. Port 2 of the
micro-processor 141 is sequenced to produce the scan inputs to the
multiplexer 35 as previously described, but only one of the six strings
has ground applied to it via Port 1 so the resultant fret depressed output
43 corresponds only to the string being observed at that time.
Referring to FIG. 13, upon power up ("start" FIG. 13), all gates are
inactive on Port 6, the Port 1 outputs are set to give string 1 a zero and
all other strings a tri-state condition, and all trigger flip-flops have
been cleared.
Assuming it has been more than 50 ms. since the last trigger, test 150 is
satisfied and the micro-processor then reads Port 10 to see if string 1
has produced a trigger. If no trigger is present, test 151 is negative,
and no action is taken on that particular string except to transfer the
previous valid note to the D/A converter 67. When, on subsequent scans, a
trigger is observed and test 151 is satisfied, a timer is started to
prevent looking at the trigger again for at least 50 ms. This prevents
"chatter" from occurring by multiple looks at the same trigger. The
program then issues a clear to the string 1 trigger flip-flop via a Port 9
bit.
In response to the trigger detected in test 151, the program also scans the
keyboard and determines what note is being played and sets a flag bit
indicating whether the note is open or closed. On subsequent keyboard
scans, the OC status is compared to allow lowering the gate if the
opposite state occurs.
Having found a valid note, the program outputs the note to the D/A
converter 67 and issues a load EN (load enable) the the Sample/Hold module
associated with string 1, causing storage of a voltage corresponding to
the note being played.
The trigger also starts the slide timer to allow 330 ms to look for a valid
change in the note. Each time the program finds a valid different note
that is not a status change note (open to close, or close to open), the
note is transferred and the timer restarted. The trigger also starts the
previously mentioned 50 ms. timer which disallows further triggers from
being observed.
If the test 150 is not satisfied, that is, if more than 50 ms. have not
elapsed since the last trigger, the program flow branches to the left to a
test 153 which examines the gate status. If the gate is not up, the flow
proceeds from test 153 to blocks 155 and 157. The operations performed are
to transfer the previous valid note from a processor storage location to
the sample hold for refresh, and to then appropriately decrement the
various timers, time-out and proceed to "start" for the next string.
If, at test 153, it is determined that the gate is up, another test 154 is
peformed to ascertain whether the slide timer has timed out. If it has,
the flow proceeds to block 155. If the slide timer has not timed out, a
scan of the string for new valid notes is performed.
If a valid note is found, test 158, and the open/close status has not
changed, test 159, the flow proceeds to test 160 where it is determined
whether the valid note is a new note. If test 160 is satisfied, the slide
timer is restarted, block 161, and the note is transferred to the D/A
converter, block 163. If test 160 determines there has been no change of
note, the previously valid note is transferred to the D/A converter, block
163.
If test 159 determines there has been a change in the open/close status of
the string, the gate is lowered, block 162, and the previously valid note
is transferred to the D/A converter, block 163.
The frequency control signal production portion of the string scan of FIG.
13 is illustrated further in the flow diagram of FIG. 14. The fret
segments 17 are scanned from left to right by the scan generator. When the
string 13 is found depressed upon a fret 17, test 70, the scan (fret)
number is saved, block 71. The scanner is then decremented by one, block
72, and one additional multiplexer input is examined, block 73, to
determine the status of the next fret. If the string 13 is also depressed
against the next fret 17, then the scan number saved in block 71 is
supplied to the OCO detector 49. Assuming tests 109, 110 are satisfied
(i.e. NT is true and the slide timer is still running), the saved scan
number is transferred to the valid note latch 48. If any of the tests 73,
109, 110 are not satisfied, no transfer to the valid note latch 48 is
permitted, the saved scan number is cleared, block 75, and the routine
goes to the next step, test 158 or "raise gate" in FIG. 13.
If, in block 70, FIG. 14, the string interrogated is not depressed against
the fret examined, a test 77 is made to ascertain whether all frets have
been checked. If the test 77 shows all frets have not been checked, the
scanner is decremented, and the flow returns to block 70 to examine the
next fret. If all frets have been checked, the mute status of the string
is examined, block 105. If the string is muted, no transfer of a new note
is performed. If the string is not muted, the fret proceeds to a test 106.
Assuming NT is true and that timer running signal is present, tests 106
and 107 are satisfied, and the open note indication (all zeroes) is
supplied to the valid note latch 51. If after block 105, either test 106,
107 is not satisfied, the routine proceeds to block 75.
When the string 13 is in the high impedance state and therefore not being
used for fret indications, a special touch detect circuit 37 may be used
to detect when the player's fingers are touching the string 13 without
depressing the string 13 to the extent necessary to contact a fret 17. The
string 13 is switched as required between the high impedance and ground
states by a string enable sequencer 40. This function will be described in
more detail below.
On a typical guitar, the musician sometimes plays chords where one or more
of the strings are required not to sound. To accomplish this, the musician
allows one of his fretting fingers to touch the appropriate string, thus
muting the sound. It is desirable in the multiple-string parallel keyboard
configuration to have a similar action available. Since no string
vibration occurs, however, an alternate method is required. The method
used is to measure the resistance of the user's fingers if they bridge two
strings 13. More specifically, the string enable sequencer 33 enables one
string (provides 0 volts) and the fret pair detector determines what frets
if any are depressed as already described. If no frets are depressed, then
the user is either playing an open note or has muted the string. In this
case, to detect muting, an additional operation is provided whereby the
string enable sequencer provides zero volts (ground) to the adjacent
strings 13 and a high impedence to the string 13 being interrogated. The
touch detector 37 then measures the resistance of the selected string to
ground. If 2-10 megaohms of resistance is detected by the touch detector,
the string 13 being interrogated is being bridged by a finger and the gate
is cleared such that no tone is allowed to be produced. If an open circuit
is seen, then an open tone is allowed to be produced.
Another feature useful for the multi-string mode is an improved method of
fully muting a chord by properly lowering (clearing) the gate signals for
each string when the musician has ceased playing that chord. The OCO
detector works well to control gate operation with single string melodies
but in chord playing there are certain situations where some of the notes
in the chord are closed and some of the notes are open. In these
situations, it is not convenient to release all closed notes and then
close all open notes in order to end the chord. Although it is possible to
use the mute strategy of measuring the bridged finger resistance as
previously described to mute the open notes of the chord, that strategy is
used most effectively for preventing unwanted open notes from sounding at
all when striking a chord initially.
Hence, an improved strategy may be provided to control the ending of a
chord. A mute controller provided by the micro-processor 140 accomplishes
this end-of-chord control. The mute controller monitors the gate condition
(set or clear) of all six strings and the notes associated with each
string as to whether they are open or closed. When the mute controller
detects a mixture of open and closed notes on the strings, it overrides
the OCO detector to clear necessary gates in two cases. First, when a
mixture of open and closed notes are present and then all closed notes go
open, the gates of all the open notes are also cleared to fully terminate
the chord. Second, when only open strings are present than detection of
any one of the opens going closed will clear all other opens. In either
case, the OCO detector still clears the gates of those strings which
actually experience a change in their open/close status.
Hence, after an initial trigger due to a pluck of the string, the touch
detector 37 is employed to detect whether the string should be allowed to
sound at all. After the correct strings in the chord have sounded, the
mute controller serves to control clearing of the sounding strings when
the chord is released. In the micro-processor embodiment, the string
enable sequencer operation and mute controller operation are incorporated
into the micro-processor 140, while the touch detector circuit is external
to the processor.
According to another feature useful in the multiple-string embodiment
certain fret segments are wired together in an advantageous way, saving
extensive hardware.
In a six by sixteen matrix of | | |
