GB2190495A - Ultrasonic transducer assembly - Google Patents

Abstract

A transducer assembly for use in the automatic filling of a container with a beverage comprises: (a) a housing (28); (b) a transmitter crystal (30) having a first lens (32) connected to its lower surface; (c) a separate receiver crystal (34) having a second lens (36) connected to its lower surface; (d) said transmitter crystal and first lens being located within but not touching a first non-ferrous metal tube (38); (e) said receiver crystal and second lens being located within but not touching a second non-ferrous metal tube (40); (f) said first and second metal tubes being located within said housing; and (g) polyurethane foam (50) substantially filling the remaining space within said housing and within said tubes and holding said tubes and said crystals in place therein. Crystals (30), (34), are thus electrically and acoustically separated. <IMAGE>

Description

GB 2 190 495 A 1

SPECIFICATION

Transducer assembly Field of the in ven tion 5

This invention relates to beverage dispensi ng a nd more particularly to a transducer assembly for use in a system for automatically controlling the filling of beverage containers, such as with post-mix carbonated soft drinks.

10 Descrip tion o f the prior art

Heretofore, attempts have been made to provide apparatus to automatically f ill beverage containers, such as cups, in response to the proper positioning of a cu p u nder a nozzle of a dispensing valve assembly of a beverage dispenser. Such apparatus used, f or exam pie, 1 iq uid level detectors such as either conductive or 15 capacitive electrical probes to measure the liquid level.

It is also known to measure various liquid levels within containers using ultrasonic energy and associated circuitry.

Summary of the invention 20

According to the invention there is provided a transducer assembly for use in the automatic filling of a container with a beverage comprising:

(a) a housing; (b) a transmitter crystal having a first lens connected to its lowersurface; (c) a separate receiver crystal having a second lens connected to its lowersurface; 25 (d) said transmitter crystal and first lens being located within but nottouching a first non-ferrous metal tube; (e) said receivercrystal and second lens being located within but nottouching a second non-ferrous metal tube; (f) said first and second metal tubes being located within said housing; and 30 (g) polyurethane foam substantially filling the remaining space within said housing and within saicitubes and holding said tubes and said crystals in placetherein.

Preferablythe invention further includesfirst and second electrical wires extending from the top of said assembly out of said foam, said firstwire including first and second leads each being connected to a metal plating on opposite surfaces of said transmitter crystal and a third lead connected to said first metal tube, said 35 second wire including first and second leads each being connected to a metal plating on opposite surfaces of said receiver crystals and a third wire connected to said second tube.

Further preferably said first lens is shaped to produce a beam having a footprint attwelve inches (30.48cm) that is about3/4 inch (1.91 em) wide and about 21/2 inch (6.35 em) long.

According to another preferred feature of the invention said crystals are ceramic crystals made of one of 40 M-4 or M-5a material and said lenses are made of one of ABS or acrylic plastic.

According to another preferred feature of the invention each of said lenses is glued to the respective one of said crystals.

According to another preferred feature of the invention said crystals and lenses are circular discs about 1/2 inch (1.27 em) in diameter. 45 According to another preferred feature of the invention the invention includes ultrasonic energyabsorbing walls extending downwardly below said transmitter crystal to absorb ultrasonic energy transmitted toward said walls.

According to another preferred feature of the invention each of said crystals is recessed into a cavity in a lowersurface of said foam. 50 Brief description of the drawings

Certain examples of the invention will now be described with reference to the accompanying drawings wherein like reference numerals referto like elements and wherein:

Figure 1 is a perspective view of a beverage dispenser having fourvalve assemblies each using the 55 automatic filling system of the present invention; Figure2 is a perspective view of one of the valve assemblies of Figure 1; Figure3 is a cross-sectional side view of thetransducer assembly shown in Figure 2; Figures4A and 48 are a master block diagram of the automatic control system of one embodiment ofthe present invention; 60 Figure 51s a microprocessor block diagram; Figure 6 is a schematic circuit diagram of part of the receiver subassembly including the receiver, the receiver front end, and the D/A gain reduction of Figure 4; Figure 7 is a schematic circuit diagram of another part of the receiver subassembly including the detector threshold comparator and the time varying detection generator of Figure 4; 65 2 GB 2 190 495 A 2 Figure 8 is a schematic circuit diagram of the power supply and output switch of Fig ure4; Figure 9 is a schematic circuit diagram of the transmitter subsystem of Figure 4; Figure 10 is a schematic circuit diagram of the frequency divider of Figure 4; Figure 11A is a schematic circuit diagram of the dip switch of Figure 4, and Figure 'I l B is a table showing how to set the switches fora desired ice level; 5 Figure 12 is a schematic circuit diagram of the front panel module with the manual fill switch and the over-ice and filling indicator lights; Figure 13 is an elevational view showing a nozzle, the transducer assembly, the grate, and a cup; Figures 14-26 are flow charts if lustrating the main routine and the sub- routines of the software for operating the microcomputer 66 in the block diagram of Figure 4; 10 Figure 27 is a perspective view of another embodiment of a valve assembly useful on the beverage dispenser of Figure 1; Figure 28 is a cross-sectional side view of the transducer assembly shown in Figure 27; Figure29 is a cross-sectional end view of the transducer assembly of Figure 28; Figure 30 is an exploded perspective view of the transducer assembly of Figure 28; 15 Figures 31A and 3 18 area master block diagram of the automatic control system of the preferred embodiment of the present invention; Figure 32 is a microprocessor block diagram; Figure 33 is a schematic circuit diagram of part of the receiver subassembly including the receiver, the receiver front end, and the D/A gain reduction of Fig u re 31. 20 Figure 34 is a schematic circuit diagram of another part of the receiver subassembly including the detector threshold comparator, the time varying detection generator, and the 60 Hz detector of Figure 31; Figure 35 is a schematic circuit diagram of the power supply and output switch of Figure 31; Figure 36 is a schematic circuit diagram of the transmitter subsystem of Figure 31; Figure 37 is a schematic circuit diagram of the frequency divider of Figure 31; 25 Figure 38A is a schematic circuit diagram of the dip switch of Figure 31 and Figure 38B is a table showing how to set the switches fora desired ice level; Figure 39 is a schematic circuit diagram of the front panel module with the manual fill switch and the over-ice and filling indicator lights; and Figures 40-46are flowcharts illustrating the main routine and the subroutines of the software for operating 30 the microcomputer in the block diagram of Figure 31.

Detaiffid description of the preferred embodiments

With reference nowto the drawings, a first embodiment of the present invention will first be described with referenceto Figures 1-26, and then a second embodimentwill be described with referenceto Figures 27-46. 35 Figure 1 shows a post-mix beverage dispenser 10 having four identical beverage dispensing valve assemblies 12, each of which has been modified to include the automatic filling apparatus of one embodiment of the present invention in place of the usual cup actuated mechanical leverthat normally extends beloweach valve assembly 12for operating thetwo solenoids of the valve assembly (one being for syrup and onefor carbonated water). Each of thevalve assemblies 12 is used for dispensing softdrink 40 beverages (usually a different beveragefrom each valve assembly) into various sizes of cups 14 and 16, supported on a cup supporting surface or grate 18. One particular beverage dispenser 10 and one particular valve assembly 12 is shown, however, anyvalve assemblies and any beverage dispenser can be used.

Beverage dispensers and beverage dispensing valve assemblies, such asthose shown at 10 and 12 respectively in Figure 1, arewell-known and thus no detailed description thereof is needed. 45

With referenceto Figures 1 and 2,the automatic filling apparatus of thefirst embodiment of the present invention includes a transducer assembly 20 located on the bottom surface 22 of the valve assembly 12 and 5 behind the nozzle 24, and a control module 26 attached to the front of the valve assembly 12.

The transducer assembly 20 is best shown in Figure 3 and includes a plastic housing 28 in which is contained a transmitter crystal 30 having a plastic lens 32, and a receivercrystal 34 having a plastic lens 36. 50 Thetransmitter and receiver crystals are located inside of brasstubes 38 and 40, respectively. A pairof shielded cables 42 and 44 are connected by a clamp 46to the housing 28. Each cable has a shieldwire connected to a respective one of the brasstubes and also a pair of wires connected to a respective one of the crystals at opposite locations thereon, as shown in Figure 3. Each of the crystals has a metal plating on each of its upper and lowersurfaces. Thewire connectionsto the crystals include a pair of 34 gauge wires soldered 55 one each to one of the metal platings on the crystal and then in turn soldered to two 22 gauge wires inthe cables 42 and 44. The cables 42 and 44are about six inches (1 5.24cm) long and terminate in a single IVITA connector 48,for connection tothe control module 26. All of the space within the housing 28 isfilledwith urethane foam 50.

The crystals 30 and 34 are preferably M-4 ceramic crystals (a generictrade designation for a particular 60 crystal material), which are a combination of lead titanate and lead zirconate. Each of the crystals 30 and 34is attached to its respective lens 32 and 36 preferably by using about 1/2 drop of glue such asthatsold underthe trademark Eastman 910. The plastic lens is preferably made of ABS, polycarbonate, acrylic or polystyrene plastic.

The plastic housing 28 has a pairof flanges (flange 52 is shown in Figure 2) each having a screw holefor 65 3 GB 2 190 495 A 3 attaching the transducer assembly 20 to the valve assembly 12.

The brass tubes 38 and 40 have the functions of electrically shielding or isolating the crystals, of sound isolating the crystals, and of mechanically holding the crystals (along with the urethane foam which is poured into a mold or fixture used to holdall of the elements of the assembly 20 in place, and which isthen allowed to harden). 5 The selection of the most desirable frequency to use was made as follows. Regarding the upper limit, the attenuation of ultrasonic sound in the air becomes too great to use for more than a few inches at above approximately 600 KHz. In addition, it was desired to avoid the 455 KHz broadcast band IF frequency, and the 550 KHz to 1.65 MHz AM broadcast band. By staying away from FCC assigned frequencies, and by using a transmission that is not too strong in relation to radio stations, interference on radio receivers that are 10 operated in close proximity to the automatic control system of the present invention is precluded.

Regarding the I owerlim it, because the beam pattern is constricted by the close proximity of other valve assemblies, a 2 inch (5.08cm) spread at 14 inches (35.56cm) atthe 3db pointwas assumed forthe beam pattern. This 2 inch (5.08cm) spread yields an angle of approximately8 degreestotal. Due to thespacing considerations, a 1/2 inch (1.27cm) diameter crystal was selected. A400 KHzfrequencywas selected asthe 15 preferred frequency. Other f req uencies in the range of 200 KHzto 450 KHz could alternatively be used.

Regarding the beam shape, at 14 inches (35.56cm) thetotal maximum beam pattern needsto be lessthan 3 inches (7.62cm) wide atthe limits of detectability (-40 db) in the sideto side direction, and approximately3 inches (7.62cm) frontto backatthe -3 db points. The gain atthese points would need to be asfiatas reasonable. The crystal pattern was chosen empirically asthe one giving the best level gain from frontto 20 backwith the crystals 30 and 34 aligned from frontto back between the nozzle 24 and the splash plate 25.The resulting overall gain pattern with a 3db gain at 12 inches (30.48cm) had a resulting spread sideways of 3.5 degrees, and a resulting spread frontto rear of 12 degrees.

To achievethe desired beam pattern, itwas necessaryto lensthe crystals. A 2 inch (5.08cm) concave radius produced the 8 degrees to 3.5 degrees narrowing from side to side, and a 4 inch (10.1 6cm) convex radius 25 produced the 8 degrees to 12 degrees spreading from frontto rearwhich formed a fan shaped beam pattern with an elongated footprint having a width of approximately 3/4 inch (1. 91 cm) and having a length of about 21/2 inches (6.35cm) at 3db gain and 12 inches (30.48cm) awayfrom thetransducer assembly 20. This beam shapefootprint has its long dimension extending frontto back relative to the dispenser.

Coupling from the crystals 30 and 34to the airwas calculated asfollows: 30 Characteristic impedance of PZT-4 is equal to 66 x 10E6 rayls (E= exponent throughout the following descciption).

Transmitted power is (Tp) 35 (N2/N1) Tp=(N2/N1)+1 XPc N2 =The characteristic impedance of air. 40 N1 =The characteristic impedance of M-4.

Pc= Power output of crystal.

Tp=12.6 X 1OE-6for 1 watt in, or.001 26% goes to the air.

If athird material isintroduced between the air and the material we getthe following equation: 45 4(N3/N1) (N2/N3) TP = ((N3/N1)+i_) x ((N2/N3W) x P' 50 4 (N2/N 1) - X Pc ((N2/Ni)+1)+(N2/N3)+(N3/N1) 55 Most materialsof interest for the third material have characteristic impedances between.1 X 10E6to 10 X 1 0E6 rayls.

For.1 X 1 0E6, Tp=25 X 1 OE-6 60 For 10 x 10E6,Tp= 22 x 1OE-6 4 GB 2 190 495 A 4 Thus, for any lossless material used as a coupling to airwith a characteristic impedance between. 1 x 10E6 to 10 x 10E6rayls, the resulting input power is at least doubled and the energy transmitted to the air varies by on iy 10%. The preferred lens material is one of the plastics such as acrylic or ABS. The lens should:

(1) be a plastic for production, (2) have a %" (1.27cm) diameter and be approximately.08"(0.20cm) thick, 5 (3) have a concave radius of 2" (5.08cm) in one axis and a convex radius of C (10.16cm) in the other axis, and (4) be cemented to the crystal face with about 1/2 drop of glue (preferably that sold under thetrademark Eastman 910, or equivalent).

Regarding the lens mount, to diminish acoustic coupling between the receiver and transmitter, the lenses 10 are mounted in polyurethane foam. A brass tube surrounds each crystal and its inner foam mountwhich provides electrical shielding and is soldered to the shield of the cable wiring to the crystals. The crystals are leftfloating, i.e., both electrodes are at a potential not referenced to ground. This gives greater electrical isolation in the receiver since it does not pick up ground referenced noise. The brass tubes are held in place bythe polyurethane foam to the desired package shape. The lenses protrude from the bottom surface foam 15 package.

Regarding the crystal shape and material, the transmitter crystal is preferably 1/2" (1.27cm) OD x.20C (0.51 em) fora series resonance of 400 KHz. M-4 material was chosen for the crystals 30 and 34 as the best compromise in strength, efficiency, and ease of workability. The receiver crystal is preferably l/2"(1.27cm) OD x.190"(0.48cm) fora parallel resonance of 400 KHz, and is also made of M- 4 material. 20 Regarding the electrical wiring, a twisted shielded pair of 22 gauge stranded wire is used. The wire shielding is soldered to the brass tubes 38 and 40. The brass tubes are isolated from each other electrically. A pair of 34 gauge solid wires is soldered across the metzi plated crystal faces and is then soldered to the 22 gauge lead wires. All wiring is foamed in place. The black wire of the twisted pair is attached to the outside crystal face which is marked with a small dot. 25 The control module 26 houses the control circuit board to which the crystals 30 and 34 are connected bythe cables 42 and 44 and the connector 48. Figures 4A and 413 together provided a master block diagram of the control circuit 60. The control circuit will now be described with reference to Figures 4-12.

The receiver transducer 62 (Figures 4,5, and 6) is a 400 Khz, 1/2 inch (1. 27cm) diameter para l lel resonant piezo-electric crystal 34 made of M-4 material. The crystal is coupled to the air by means of a plastic lens 36 30 which is shaped to receive the beam pattern. The transducer assembly 20 incorporates a brass tube 40 that is 518 inch (1.59cm) In diameter and is used for electrical isolation. The crystal 34 is mounted such that it is centered in the tube with the lens 36 exposed atone end of the tube. The tube assembly is foamed with polyurethane for acoustic isolation.

The receiver section 64(Figures4,5 and 6) has a total gain of 96db and is comprised of two protective 35 diodes 110 and 112 and two MC1350P IF amplifiers 114 and 116 that are interconnected through a tuned transformer 118 with another tuned transformer 120 to interconnect the second amplifier 116 to the detector 68. These amplifiers 114 and 116 have provisions for gain control from Pin 5 and are used in this application bythe microcomputer 66.

The detector circuit 68 (Figures4,5 and 7) changesthe400 Khzfrom the receiver64to a DCAnalog signal. 40 This detectoris special inthatitcan notonly detectthe envelope of the pulse butsince it is a DCcoupled detector, ithas no offset shift due to pulsewidth variations. By having a balanced detector system, the temperature drift is very low.

The receiver gain reduction 70 (Figures 4,5, and 6) is comprised of five resistors that form a binary weighted current sinking " D to X' converterthat is driven bythe microcomputer 66, which allowsfor 45 thirty-two stages of gain level control.

The threshold comparator72 (Figures 4,5, and 7) is comprised of an LM393N comparator 122 and is used in conjunction with the time varying detection to convertthe analog receiver signal to a digital signal which is then fed to the microcomputer 66. Within this circuit is a means for adjusting the slope of the time varying detector using a 1 OOK potentiometer and a means of adjusting the threshold detector using a 500 ohm 50 potentiometer 125.

The time varying detection generator 74 (Figures 4,5, and 7) uses the gate signal that the microcomputer66 sends to thetransmitter, and charges a 15 Nanofarad capacitor 124 to two volts, which sets the peak level of the time varying detectorwave form. This circuit is comprised of a 2N4126 switching transistor 126 and the power supplyto supportthat circuit. 55 The modulator 76 (Figures 4,5, and 9) is comprised of a 12 volt Zender diode 128 and two transistors 130 and 132 that perform an (Anding) function forthe transmitter gate signal (T) and the 400 KHz signal from the oscillator. This (Anded) signal is then level shifted through the 12 volt Zener diode 128 and the 2N4402 transistor 132 to the gate of the final amplifier78.

The final amplifier78 (Figures 4,5, and 9) is comprised of a BUZ-71A MOSFET 134, a resistor 136 and a 60 transformer 138. The resistor discharges the gate source capacitor of the MOS-FET 134. The MOS-FET 134 switches the output transformer 138 to the minus 20 volt supply in response to the gate drive signal. The transformer 138 steps the voltage up to the transmitting crystal 30 to approximately 2000 volts.

The transmit transducer 80 (Figures 4,5, and 9) is comprised of a 400 KHz 1/2 inch (1.27cm) diameterseries GB 2 190 495 A 5 resonant piezo-electric crystal 30 made of PZT-4 material much the same as the receiver crystal 34 with the exception of the thickness. The crystal 30 is coupled to the air by means of a plastic lens 32 which is also shaped to form the beam pattern. The assembly of the transmittransducer80 is exactly the same as forthe receiver as described above.

The microcomputer 66 (Figures 4 and 5) is a General Instruments Pic-1 654 and contains the intelligence 5 and control functions of the entire system. it communicates to the rest of the system through tweive 110 pins.

It also contains the oscillator circuit, the master clear circuit, and the real time clock counter input.

The crystal 82 (Figures 4 and 5) and components of the 4 MHz crystal comprises passive componentsthat form the feedback networkforthe oscillator in the Pic-1 654.

The power-on reset circuit 84 (Figures 4 and 5) forms a 10 millisecond reset pulse to the microcomputer 66 10 at POWER-ON that allows the 4 MHz oscillator crystal 82 to start and the microcomputer 66 to become initialized.

A divide by ten counter 86 (Figures 4,5, and 10) converts the 4 MHz computer clock to a 400 KHzsquare wave signal to operate the transmitter.

The divide bythree counter88 (Figures 4,5 and 10) converts the 400 KHz signal to a 133 KHz signal that is 15 applied to the microcomputer 66 as the real time clock counter input. Numberthirteen and numberfourteen are encompassed within the same IC (74HC390) divider chip which has a divide byten and a divide bythree circuit.

The front panel module 90 (Figures 4,5, and 12) consists of two LED indicators 92 and 94. One is an "Over-Ice/Cup Remove" (Figures 4,5, and 12) red indicator 92 and the other is a green " Fill " LED 94, 20 indicating thatthe cup can be filled or is being filled. This indicator 94 remains "on" steadywhen a cup is ok until filling starts. In the eventthatthere is too much ice in the cup, orthatthe cup is not recognized as a cup, the red indicator light 92 will flash on and off.

There is a programming dip switch 96 (Figures 4,5, and 11 A) comprising five individual switches accessible by removing a cover (not shown) on the lower rear surface of the control module 26. One switch is 25 used to select between a normal flow or a fast flow valve assembly, depending upon which type of valve assembly the automatic control system is being attached to. Another switch is used for selecting a foamy or flat product such as water. The otherthree switches are used for selecting ice level ortest position. Thetest position is used for alignment of the receiver during manufacturing and has no field use. The binary output of the three ice level switches allows for seven ice level selections from 118 cup to 7/8ths cup, as illustrated in 30 Figure 11 B. The multiplexer circuit 98 (Figures 4 and 5) allowsthe microcomputer 66 to read eitherthe dip switches or to setthe gain of the receiver as necessary. It is comprised of five signal diodes.

The powersupply 100 (Figures 4,5, and 8) uses 24volts ACfrom the 50 VACtransformer (not shown) inthe dispenser 10. The present control system consumes lessthan 2 volt-amps at 24volts AC. The 24voltsACis 35 rectified and filtered to form a minus 20 volt DC supply and a plus 25 volt DC supply. The minus 20voltsupply is regulated with a Zener diode and supplies powerto the transmitter. The plus 25volt supply is unregulated but has a 39 volt Zener diode used as surge protection. The 25volt DC supply is regulated down to 15 voltsfor the receiver subsystem by a 78L1 5threeterminal regulator 140. An MPS-A42 transistor 142 is used as a fly-back oscillatorto provide the plusfive volts needed to operatethe computer circuitry. The 4.3 voltZener 40 diode 144connected between the plusfive volt supply and the base of a 2N4124 transistor 146 serveto regulate the fly-back oscillator.

The output switch 104 (Figures 4,5, and 8) forthetwo solenoids of thevalve assembly 12 is operated from eitherthe microcomputer 66 orthe manual push button 102 on thefront of the control module 26. The resistor diode networkcouples the microcomputer66 and the manual switch 102 to the base of a 2N4124 45 transistor 148which then turns the outputtriac 149 on or off, which then turns the two solenoids in thevalve assembly 12 on or off.

The softwarewill now be described with reference to Figures 13through 26. Figure 13 is a side elevation viewshowing the transducer assembly 20, the lenses 32 and 36, the nozzle 24 of the beverage clispenservalve assembly 12,the control module 26,the splash plate 25,the grate 18 and a cup 16 having a cup lip 17, a cup 50 bottom 19, and a top level 21 of ice in the cup.

The software includesfour (4) major routines which are labeled Initialization Routine (INIT), Cup Detection (CUPDET), Fill Routine (FILL), and Cup Removal Routine (CUPREM).

The software also includes five (5) subroutinesthat are defined as Time Delay (WAIT), Absolute Value of the Difference of Two Numbers (DIFF), Grate/Overflow Detector (LGRATE), Transmit (T13DQ, T13DW, and TLD 55 as described below), and Receive (REC).

The Transmitter Subroutine sets the variablesforthe receiver routine and outputs a 25 microsecond pulse (10 cycles at400 KHzwhich occupies.1" (0.25cm) air space) during which time the transmitter is active. The selection of receiver variables is made through three different entry points (or surfaces off which the transmitted beam reflects): TB13Q (Transmit Bottom Detector), T13DW (Transmit Bottom Detectorwith 60 window), and TLD (Transmit Lip Detector).

The receiver has 32 steps of gain controlled bythe software. The gain is setto minimum from the startof transmitto approximately 1X (3.30cm) target distancetime (180 microseconds). Atthattime the gain isset equal to the gain variable set up in the entry point routines' ForTLD, the gain is always setto maximum. For T13DQ and TB13W, the gain is determined by the calling routine. In T13DQ and TLD, the distance of thefirst 65 6 GB 2 190 495 A 6 echo detected is captured for processing. In TBDW, a lip masking window is enabled which ignores any echoes closer than the lip distance +.2W (0.64cm). This allows a higher gain to be used to look at liquid level rising inside the cup. Under all entry points, 5 transmissions and receptions are made with the echo distances stored in RAM. The processing algorithm looks for two samples that correlate within. 1 " (0.25cm) for TLD, or 1 12.54cm) for T13DO, and TI3DW.The average of the two distances is used as the echo distance. A 2 5 millisecond delay is incorporated before each transmit to allow previous multiple reflections to decay.

WAIT is a programmable delay subroutine that returns to the calling routine immediately if the manual push button is pressed. It has a maximum delay of 1 second.

DIFF is a subroutine that calculates the absolute value of the difference of two numbers.

LGRATE is the Grate/Overf low detector subroutine and is used during the FILL routine. It uses TLD to detect 10 with maximum gain and no window. If the subroutine detects an echo distance less than the I ip distance minus. 1 " (0.25cm), the overflowflag is set before returning. If the subroutine detects an echo distance within.2W (0.64cm) of the grate distance, the cup removal flag is set before returning.

IN IT is used when the microcomputer is initialized by the "Master Clear" (hardware). During power up, the first instruction processed is at location 777 octal. This instruction "GOTO IN IT" commands the computerto 15 begin executing this routine, which comprises the following: (1) the RAM is cleared; (2) wait l second for power to stabilize; (3) run the diagnostic routine if enabled; (4) use TLD to look with maximum gain and no window for an echo distance between 7"(17.78cm) and 13"(33.02cm);(5) if it does not detect an echo within this range, the "Over Ice" indicator on the front panel flashes; (6) if it does detect an echo distance within 7117.78cm) to 13"(33.02cm), the distance is stored in RAM as the Grate distance and the program continues 20 at CUPDET.

CUPDET is the Cup Detection routine. This routine collects data using TLD and accepts a cup using the following procedure:

A.The manualfill switch onthefront panel is monitored continuouslyto assure proper operation. If the manual switch is pressed,the computer beginsthe Cup Removal routine immediately. 25 B.Astable lip distance mustbe established morethan X' (7.62cm)from the grate. Astable lip distanceis defined as 5consecutive echo distancesfrom TLD separated by6 milliseconds that correlate within X'(0.51 cm). This corresponds to the cup lip being stablefor 130 milliseconds.

C.Acup bottom orice level must be discernedthatis morethan.1 10.25cm)abovethe grate and morethan 30.25"(0.64cm) belowthe lip. This is accomplished by using T13DWand varying the gain asfollows: 30 With minimum gain, obtain an echo distance using TBDW. If the echo is not morethan.1 " (0.25cm) closer than the grate, then the gain is increased 1 step and anothersample istaken. Ifthe gain reachesthe maximunthe Over-ice indicator flashes andthe Cup Detection routine beginsagain.

D.The icelbottom height is calculated from the lastdistance obtained as outlined in (c) above and thegrate, andthen stored asthe actual ice height. Thecup height is calculatedfrom the lip distance andthe grate.The 35 cup heightis divided by8 andthe quotientis multiplied bythe 3 bit binary numberinputas selected ontheice level programming switches.This allowable ice height is compared withthe actual ice height. If the actual ice height isgreaterthan allowed bytheswitch selection,the Over-ice indicator f lashes and the Cup Detection routine begins again. If the actual ice height is lessthan the amountselected bythe switch,the FILLroutine begins. 40 The FILL routine controlsthe completefilling and top off operation. The routine limits the solenoid operationto a maximum of 3 On/Off cycles. Aftereach of thefirst2 cycles, the routine waits for the foam to settle beforestarting the nextcycle. Afterthefoam is settled and the cup iswithin 7/20"(0.89cm) of beingfull, the Cup Removal routine begins. If the manual switch is pressed atanytime during the FILL routine,theCup Removal routine begins immediately. Each of the cycles has a maximum solenoid on timewhich if exceeded 45 causesthe Cup Removal routineto being.

Adetailed description of the FILL routinefoliows:

A. Before the valve assembly 12 solenoids are actuated, several checks and corrections are made. The gain is initiallysetat 1 1116of maximum gain. If the Lip Distance is lessthan 4"(10.16cm),the gain is adjustedwith the empirically derived equation: 50 Gain= Gain- 118 W' (1 0.16cm)- lip distance).

If the Lip Distance is less than 4"0 0.1 6cm), the lip Distance is adjusted with the empirically derived equation:

Lip Distance= Lip Distance - 1/8 (4"(10.16cm)- Lip Distance) If the Lip Distance is less than. 1 " (0.25cm), the Lip Distance is set to. 1 " (0.25cm) to a 1 low the cup overflowto 55 function properly.

TheTime Constantforthis particularcup height is calculated with the equation:

Time Constant = cup height- 2" (5.08cm).

Thistime constant is used in each of thethree cyclesto provide a maximum "Solenoids On',time proportional to the cup height. 60 B. The gain must be adjusted such that the fluid level is detected and the lip is not during the period when the cup vibrates such as atthe beginning of a FILL. To accomplish this a period of time proportional to cup height is programmed to allowfilling to start and gain enough weightto minimize cup vibration and adjust gain as necessary. During this time period the routine uses T13DQ to check if the echo distance is within.75" (1.91cm) of the Lip Distance. If it is, the gain is reduced one step. If the gain reaches minimum, the cup 65 7 GB 2 190 495 A 7 removal routine begins. If the cup is removed during this period, the solenoids will not turn off because the G rate/Overf low detector subroutine is not called during the period due to trying to get as many sa m pies as possible to adjust the gain. At the end of the period, the solenoids stay on.

C. A second maximum time period begins that is also proportional to the cup height. During this time period, the routine uses T13DW to monitorthe liquid level and turns the solenoids off when the liquid level is 5 with in.5" (1.27cm) of the Lip Distance. The G rate/Overflowdetector!3ubroutine checks to see if the cup has been removed or if TI3DW has missed the I iquid level rising and an overflow is imminent. If the cup is missing, the cup remova I routine begins. If there is an overflow indicated, the solenoids are turned off.

D. A 5second pause begins at this time to allow the foam to settle.25"(0. 64cm) below the cup lip. The Grate/Overf low subroutine checks once each second too ascertain that a cup is sti I I in place. If the cup is 10 missing, the cup removal routine is started.

E. After the 5 second pause, a minimum number of seconds for the foam to subside is set at 16, and once a second, an echo distance is obtained with TBDQ. If 2 consecutive echo distances are within.1 "(0.25cm) of each other, or if the period times out, the top off cycle begins. The Grate/Overflow detector subroutine checks once a second for a missing cup. Once a cup is found missing, the cup removal routine begins. 15 F. The top off cycle uses TBDQ to determine if the liquid level is within 7/20"(0.89cm) of the lip. If this condition exists, the solenoids are notturned on. If the echo distance is notwithin 7/20"(0.89cm),the solenoids turn on until that condition is met.

G. A repeat of "D," "E," and "F" now occurs to implement the second top off cycle.

The cup removal routine (CUPREM) turns the fill indicator 92 off, the value assembly 12 solenoids off, and 20 the Over-lee indicator 94 on. It uses TLD and waits for an echo distance within.25"(0.64cm) of the grate. When this condition exists, a new grate distance is stored, the Over-ice indicator turns off, and the Cup Detection routine begins again.

As described above, the system of the present invention provides an ultrasonic method and apparatus for controlling the automatic f il ling of beverage cups. The system can be used with any beverage, such as coffee, 25 tea, milk, fruitjuice, and carbonated soft drinks. The beverages can produce foam during filling or not.

Different sized cups can be used and they can have icetherein.

The system can be used in conjunction with any known, standard beverage dispenser. In the case of carbonated soft drink dispensers, the transducer assembly and the control module of the present invention are located directly on the valve assembly. The cup actuated arm and the microswitch are removed from the 30 standard valve assembly; the triac 149 in Figure 8 takes the place of the microswitch and simultaneously turns the syrup solenoid and the carbonated water solenoid on and off.

The system of this invention is on and working wheneverthe powerto the dispenser is on. The power is often left on to the dispenser to maintain the refrigeration system on.

A brief overview will now be provided without reference to the details of the system, which have already 35 been described above.

The system first obtains a grate signal and stores it in the RAM. The way it doesthis is to transmitfive25 microsecond pulses (having a length in air of about 0.1 inch (0.25cm)), each spaced apart about2 milliseconds. If two signals are not received that arethe same within 0.1 inch (0.25cm), then this first setof pulses is discarded and a new set of five pulses is immediately (in abouttwo milliseconds) transmitted. If two 40 signals are received and arewithin 0.1 inch (0.25cm), and if they arefrom a distance of from about7(17.78cm) to 13 inches (33.02cm), then the system decides that it isthe grate distance and stores it in the RAM.

The system then goes to the cup detection routine. The same set of pulses istransmitted and is received at maximum sensitivity. To determine that a cup is present, the system has to see 5 consecutive echo distances, each separated by 6 milliseconds, that correlateto within 0.2inch(O.51 em). That is, 5 sets of pulses are 45 transmitted with 6 milliseconds between each set. If at leasttwo signals are received from thefirstset of 5 pulsesthat arewithin 0.1 inch (0.25cm), then thatwill be one value (or one echo distance). After receiving 5 of those in a rowwithin 0.2 inch (0.51cm), the system knowsthat a cup lip (or something otherthan the grate) is present.

The system then goes to the next routine. In this routine the system looks for something greater than 0.1 - 50 inch (0.25cm) above the grate and greater than 0.25 inch(O.64cm) below the lip, that is, either the cup bottom or ice. If it finds this, it concludes thatthe something present is a cup (ratherthan just a hand, forexample).

When the ice or bottom is obtained, it is stored temporarily. The cup height is then calculated and the ice height is then calculated. It is then calculated whether or notthe cup has too much ice. If it does not,the system goes to the FILL routine. This routine is somewhat complex. 55 In the FILL routine there are fourf illing periods. A first period or initial fill that is not monitored but which is set as a time function based on cup height. It will fill to about 1/3 cup under certain usual conditions. The system then switches automatically, without stopping the filling, to the second period in which the filling is monitored, and in which the filling is shut off when the liquid level rises to within 0.75 inch(l.91 em) of the stored lip distance. The FILL routine then waits 5 seconds to allow foam to subside (if the control module is 60 setfor a foamy beverage). The monitoring continues waiting forthe foam to quit moving and when two distances are received within 0.1 inch (0.25cm), then it calculates if the level is within 7/20inch(O.89cm) of the lip. If it is not, filling is resumed and monitored until the level is within 7/20 inch (0.89cm) of the lip. If itis within 7/20 inch (0.89cm), filling does not resume. The "top off" routine is then repeated after another 5 second pause. 65 8 GB 2 190 495 A 8 After the end of the FILL routine, the fill indicator light 92 is turned off, the solenoids are turned off, and the over-ice indicator light 94turns off.

A second (and preferred) embodiment of the present invention will now be described with reference to Figures 27-46. An important difference between this preferred embodiment and that described above with reference to Figures 1-26 is that the preferred embodiment is designed so that two or more beverage 5 dispensing valves having the ultrasonic control system of the preferred embodiment can be located in close proximityto each other, such as by being adjacent valves on a dispenser, without interference therebetween.

However, many other features of the two em bodiments are identical.

Figure 27 shows a valve assembly 212,similar to valve assembly 12, that can be used as one or more of the valve assemblies on the dispenser l 0 of Figure 1. The automatic filling apparatus of this embodiment of the 10 present invention includes a transducer assembly 220 located on the bottom surface 222 of the valve assembly 212 and behind the nozzle 224, and a control module 226 attached to the front of the valve assembly 212.

The transducer assembly 220 is best shown in Figures 28-30 and includes a plastic housing 228 in which is contained a transmitter crystal 230 having a plastic lens 232, and a separate receiver crystal 234 having a 15 plastic lens 236. The transmitter and receiver crystals are located inside of brass tubes 238 and 240, respectively.

A pair of shielded cables 242 and 244 each consist of a shield wire connected to a respective one of the brass tubes 238 and 240 and also a pair of wires connected to a respective one of the crystals at opposite locations thereon, as shown in Figure 28. Each of the crystals has a metal plating on each of its upper and 20 lower surfaces. The wire connections to the crystals are 28 gauge wire soldered directly to the crystal plating.

The cables 242 and 244 are about nine inches long and terminate in a single IVITA connection (such as 48 in Figure 3), for connection to the control module 226. Substantially all of the space within the housing 228 is filled with urethane foam 250.

The transmitter crystal 230 and the receiver crystal 234 are preferable M5a ceramic crystals (a generic 25 trade designation fora particular crystal material), which area combination of lead titanate and lead zirconate. Each of the crystals is attached to its respective lens preferably by using about 1/2 drop of glue such as that sold under the trademark Eastman 910. The plastic lens is preferably made of ABS or polycarbonate plastic.

The plastic housing 228 has a pair off langes on each side thereof, each flange having a screw holefor 30 attaching the transducer assembly 220 to the valve assembly 212.

The brass tubes 238 and 240 have the same function as described above with reference to brass tubes 38 and 40. As shown in Figures 28-30, the transducer assembly 220 includes the plastic housing 228, a urethane foam filler 250, a urethane foam lid 400, a plastic cover 402, and the transmitter and receiver subassemblies 420 and 422, respectively, slid into a pair of spaced-apart cylindrical cavities in the foam filler 250. 35 The transmitter subassembly420 includes the transmitter crystal 230, the lens 232, a urethanefoam thimble 424 and the brass tube 238. The receiver subassembly similarly includes the receiver crystal 234, the lens 236, a urethane foam thimble 426 and the brass tube 240.

The lenses 232 and 234 are shaped as shown in Figures 28 and 30 with a square flange and a circular lipto receive the crystal. The crystal is glued to the lens as described above. The crystal-lens unit is then pushed 40 inside the thimble and the tube is pushed over the thimble. The lens has a recess for the wire connection to the lower face of the crystal and the thimbles have two grooves as shown in Figure 28 for the two wires connected to the crystal. No groove is provided for the wire connected to the brass tube.

The housing 228 has a pair of thin flanges 408 and 410 and a pair of thick flanges 412 and 414 with screw holes for use in connecting the transducer assembly 220 to the dispensing valve 212. The thick flanges 412 45 and 414 are used to adjust the position of the housing 220 and thus, the location of the transmitted beam.

As shown in Fig ure28, the lenses 232 and 236 are recessed into the bottom of the foam filler 250 to provide a baffle 251 there between. Also, the lower sidewal Is 253 of the filler 250 extend downwardly below the lenses 232 and 234. The baffle 251 helps prevent ultrasonic energy passing directlyfrom the transmitter tothe receiver. The sidewalls 253 helps prevent ultrasonic energyfrom being transmitted sideways to an adjacent 50 valve. Thefoam absorbs the ultrasonic energy.

The selection of the most desirable frequency to use is also the same as described above with reference to the first embodiment.

Regarding the beam shape, at 14 inches (35.56cm), the total maximum beam pattern needs to be less than 3 inches (7.62cm) wide at the limits of detectabiJity (-40 d b) in the side to side direction and approximately 3 55 inches (7.62cm) frontto backwith -3 db point in the front followed closely by the 0 db point and then tapering off to -6 db atthe rear. The gain at a point nearthe front of the pattern (toward the nozzle) needs to be a maximum with the gain failing off smoothly by about 6 db as the pattern reaches the back point. The crystal pattern was chosen empirically as the one giving the best cup lip to ice (front) ratio with the crystals aligned from frontto back between the nozzle 224 and the splash plate 25. 60 The resulting overall gain pattern at 12 inches (30.48cm) had a spread sideways of 3.5 degrees, and a resulting spread frontto rearof 12 degrees.

To achievethe desired beam pattern, itwas necessaryto lensthe crystals. A 2 inch concave radius produced the 8 degreesto 3.5 degrees narrowing from sideto side for both transmit and receiving crystals, a 4 inch (10. 1 6cm) convex radius produced the Wto 12'spreading from front to rear for the receiver crystal and 65 9 GB 2 190 495 A 9 a flat lens toward the fromt for 1/2 the crystal followed by a 3 inch (7. 62cm) convex radius to the rearforthe transmitter crystal which formed a fa n-shaped beam pattern with an elongated footprint having a width of approximately 3/4 inch (1.91cm)at -3db and having a length of about 21/2 inches (6.35cm) with a bright spot aboutl inch (2.45cm) in from the front with -3 db at the front and -6 db atthe rear and 12 inches (30.48cm) away from the transducer assembly 220. This beam-shape footprint has its long dimension extending front 5 to back relative to the the dispenser.

Coupling from the crystals to the air was calculated in the same manner described above forthefirst embodiment.

One change in this second embodiment is that the lenses are inset into the bottom surface of thefoam package. 10 Regarding the crystal shape and material, the transmitter crystal is preferably 1/Y' (1.27cm)01D x.20W (0.51 cm) fora series resonance of 400 KHz. M-5a material was chosen for the crystals 230 and 234 as the best compromise in strength, efficiency, low mechanical Q, and ease of workabi I ity. The receiver crystal is preferably l/2"(1.27cm) OD X AW(O.48cm) fora parallel resonance of400 KHz, and is also made of M-5a material. 15 Regarding the electrical wiring, a twisted shielded pair of 28 gauge stranded wire is used and soldered directly to the plating on the crystal faces. The wire shielding is soldered to the brass tubes 238 and 240. The brass tubes are isolated from each other electrically. The black wire ofthe twisted pair is attached to the outside crystal face which is marked with a small dot.

The control module 226 houses the control circuit board to which the crysta Is are connected by the cables 20 242 and 244 and the connector 248. Fig ures31 A and 31 B together provide a master block diagram ofthe control circuit 260. The control circuit will now be described with reference to Figures 31-39.

The receiver transducer 262 (Fig ures31,32, and 33) is a 400 Khz, 1/2 inch (1.27cm) diameter parallel resonant piezo-electric crystal coupled to the air by means of a plastic lens shaped to receive the beam pattern. The transducer assembly 220 incorporates a brass tube 240 that is 5/8 inch (1.59cm) in diameter and 25 is used for electrical isolation. The crystal is mounted such that it is centered in the tube with the lens 236 exposed atone end of the tube. The polyurethane foam provides for acoustic isolation.

The receiver section 264 (Figures 31,32 and 33) has a total gain of 96db and is comprised oftwo protective diodes 310 and 312 and two MC1350P IF amplifiers 314 and 316 that are interconnected through a tuned transformer 318 with another tuned transformer 320 to interconnect the second am plifier316 to the detector 30 268. These amplifiers 314 and 316 have provisions for gain control from Pin 5 and are used in this application bythe microcomputer 266.

The detector circuit 268 (Figures31,32 and 33) changesthe400 Khzfrom the receiver 264to a DCAnalog signal. This detector is special in that itcan not only detectthe envelope ofthe pulse butsince it is a DC coupled detector, it has no offset shift due to pulsewidth variations. By having a balanced detectorsystem, 35 thetemperature drift isvery low.

The receivergain reduction circuit270 (Figures31,32, and 33) is comprised for five resistors that form a binaryweighted currentsinking "Dto X' converterthat is driven bythe microcomputer 266, which allowsfor thirty-two stages ofgain level control.

Thethreshold comparator272 (Figures 31,32, and 33) is comprised ofan LM393N comparator322 and is 40 used in conjunction with thetimevarying detection to convertthe analog receiversignal to a digital signal which isthenfedtothe microcomputer 266.

Thetimevarying detection generator274 (Figures 31,32 and 34) usesthe manual MID signal fromthe microcomputer 266 and charges a 15 Nanofarad capacitor 324to two volts, which setsthe peak level ofthe timevarying detector waveform. This circuit iscompromised ofa 2N412Gswitching transistor326 andthe 45 powersupplyto supportthat circuit. 60Hzdetection is accomplished in the 601-1z detector shown in Figures 31,32 and 34. The incoming 60Hz, 24VAC poweris sensed, afterfiltering, by 1/2 ofthe comparator LM393N 322 and the output shunts the WID signal to groundwhich, becausethe detector268 signal is biased above ground, forces the detector comparator output highfor 1/2 ofthe 60Hzwaveform. The microcomputer 266 sensesthis and usesthefalling edge ofthe60Hzsignal from the detector comparator to start itssequences 50 and isthereby phase locked tothe 60Hz -24VAC powersystem. Adjacentvalve assemblies are separated in time by reversing their 24VAC wires 450 and 452 (see Figure32) so that adjacent units synchronizeto different 1/2cycles ofthe 60Hz powersupply and thereby do not interferewith each other. Alternatively, a switch can be provided having two positions labeled "A" and "B" to designatethetwo possible orientations ofthewires 450 and 452. Thus, ifonevalve assembly has an 'W' position, each immediately adjacent valve assembly 55 must havethe switch on the "B" position. Units spaced morethan onevalve assembly apart arefarenough apart notto interferewith each other.

The modulator276 (Figures 31,32, and 36) is comprised ofa 12 volt Zener diode 328 and two transistors 330 and 332that perform an (Anding) function forthe transmitter gate signal (T) andthe 400 KHzsignalfrom the oscillator. This (Anded) signal isthen level shiftedthrough the 12 volt Zener diode 328 andthe2N4402 60 transistor 332 to the gate ofthefinal amplifier278.

Thefinal amplifier278 (Figures 31,32, and 36) is comprised ofa IRF-523 MOS-FET334, a resistor336and a transformer338. The resistor discharges the gate-source capacitorofthe MOS-FET334.The MOS-FET334 switchesthe output transformer 338 to the minus 20voltsupply in responsetothe gate drivesignal.The transformer 338 steps the voltage up to the transmitting crystal 230 to approximately 2000 volts. 65 GB 2 190 495 A 10 The transmit transducer 280 (Figures31,32, and 36) iscomprised of a 400 KHz 1/2 inch diameterseries resonant piezo-electric crystal 230 made of PZT-5Amaterial much the same asthe receivercrystal withthe exception of thethickness. Thecrystal 230 is coupledtothe air by means of a plastic lens 232which isalso shapedtoformthe beam pattern. The assembly of the transmit transducer 280 is exactlythe sameasforthe receiver as described above. 5 The microcomputer 266 (Figures 31 and 32) is a General Instruments Pic-1 654 and contains the intelligence and control functions of the entire system. It communicates to the rest of the system through twelve 1/0 pins.

It also contains the oscillator circuit, the master clear circuit, and the real time clock counter input.

The crystal 282 (Figures 31 and 32) and components of the 4 MHz crystal comprise passive components thatform the feedback networkforthe oscillator in the Pic-1 654. 10 The power-on reset circuit 284 (Figures 31 and 32) forms a 10 millisecond reset pulse to the microcomputer 266 at POWER-ON that allows the 4 MHz oscillator crystal 282 to start and the microcomputer 266 to become initialized.

A divide byten counter 286 (Figures 31,32, and 37) converts the 4 MHz computer clockto a 400 KHzsquare wave signal to operate thetransmitter. 15 The divde bythree counter 288 (Figures 31,32 and 37) converts the 400 KHz signal to a 13 3 KHz signal that is applied to the microcomputer 266 as the real time clock counter input. Numberthirteen and number fourteen are encompassed within the same [C (74HC390) divider chip which has a divide byten and a divide by th ree ci reu it.

The front panel module 290 (Figu res 31,32, and 39) consists of two LED indicators 292 and 294. One is an 20 "Over-lce/Cup Rernove" (Figures 31,32, and 39) red indicator 292 and the other is a green "FilV LED 294, indicating that the cu p can be filled or is being filled. This indicator 294 remains "on " steady when a eu p is ok until fil ling ends. In the event that there is too m uch ice in the cup, the Over-lce/Cu p Remove red indicatorwil 1 light and stay lit until the cup is removed. If the cup is not recognized as a cup, due to mispositioning the green indicator light 294 will flash on and off. 25 There is a programming dip switch 296 (Figures 31,32 and 38A) comprising five individual switches accessible by removing a cover (notshown) on the lower rearsurface of the control module 226. One switch is used to select between a normal flow or a fastflowvalve assembly, depending upon which type of valve assemblythe automatic control system is being attached to. Anotherswitch is used forselecting a foamyor flat productsuch as water. The otherthree switches are used for selecting ice level ortest position. Thetest 30 position is used for alignmentof the receiver during manufacturing and has no field use. The binary outputof thethree ice level switches allowsforseven ice level selectionsfrom 1/8 cup to 718ths cup, as illustrated in Figure38B.

The multiplexer circuit 298 (Figures 31 and 32) allowsthe microcomputer 266to read eitherthe dip switches orto setthe gain of the receiveras necessary. It is comprised of five signal diodes. 35 The powersupply 300 (Figures 31,32, and 35) uses 24volts ACfrom the 24 VACtransformer (not shown) in the dispenser 10. This 24VAC isfiltered to remove any high frequency noise that might interfere with the system operation. The present control system consumes lessthan 2 volt- amps at 24volts AC. The 24voltsAC is rectified andfiltered toform a minus 20 volt DC supply and a plus 25 volt DC supply. The minus 20volt supply is regulated with a Zener diode and supplies powerto thetransmitter. The plus 25 volt supplyis 40 unregulated but has a 39 volt Zener diode used as surge protection. The 25 volt DC supply is regulated down to 15voltsforthe receiver subsystem by a 78L1 5 threeterminal regulator340. An MPS-A06 transistor 342 is used as a fly-back oscillatorto providethe plusfive volts needed to operatethe computer circuitry. The43 voltZenerdiode 344 connected between the plus five volt supply and the base of a 2N4124 transistor 346 serveto regulatethe fly-back oscillator. 45 The outputswitch 304 (Figures 31,32, and 35) forthetwo solenoids of the valve assembly 212 isoperated from eitherthe microcomputer 266 orthe manual push button 302 on thefront of the control module 226.The resistor, opto-coupler network couplesthe microcomputer 266to the Tirac 349 which in turn energizesthe valve solenoids in the valve 212when eitherthe microprocessor 266 orthe manual push button 302 so requires. 50 The softwarewill now be described with referenceto Figures 40through 46.

The software includes4 major routines which are labeled Initialization Routine (INIT), Cup Detection (CUPDET), Fill Routine (FILL), and Cup Removal Routine (CUPREM).

The software also includes six subroutinesthat are defined as time delay (WAIT), absolutevalue of the difference of two numbers (DIFF), Grate/Overflow detector (LGRATE), Transmit/Receive, checkfortest mode 55 (TSTCHK), and checkfor maximum value on time (TIMOUT).

The Transmitter/Receiver subroutine obtains a distance data by allowing the transmitter to operate fora period of 25 microseconds (10 cycles at400 KHzwhich occupies.1 " (0.25cm) airspace) and then monitoring the receiver outputfor reflections. Two Transmit/Receive periods are contained in thetime period of a single half cycle of the sinusoldal line inputvoltage. The synchronization permits transmission only during the 60 positive half cyclewhich allowstwo valvesto operate side by side without interference by reversing the line inputwires on adjacentvalves. Three different entry pointsto the subroutine select receiver options: TBD (Transmit Bottom DetectorLT13DW (Transmit Bottom Detectorwith Window) and TLD (Transmit Lip Detector).

The receiver has 32 steps of gain controlled bythe software. The gain is setto minimum from the startof 65 GB 2 190 495 A 11 transmit to approximately.9"(2.29cm) target distance time (180 microseconds). At that time, the gain is set equal to the gain variable setup in the entry point routines. For TLD, the gain is always set to maxim u m. For TBD and T13DW, the gain is determined by the calling routine. In TBD and TLD, the distance of the first echo detected is captured for processing. In TI3DW, a lip masking window is enabled which ignores any echoes closer than the lip distance +35" (0.89cm).Th is allows a higher gain to be used to look at liquid level rising 5 inside the cup. Under a] I entry points, 2 transmissions, each separated by two milliseconds of receive time and twomil I iseconds of waiting fora I I reflections to cease, are made and the received distances stored in RAM. The processing algorithm accepts the distances if they correlate within a.4" (1.02cm) and returns with the mean value as the correct distance. If the two distances do not correlate, then the routine waits on the synchronization signal and takes two new samples to correlate. 10 WAIT is a programmable delay subroutine that returns to the calling routine immediately if the manual push button is pressed. It has a minim u m delay of 3.25 msec, and a maximum delay of.9 seconds.

DIFF is a subroutine that calculates the absolute va I ue of the difference of two numbers.

LGRATE is G rate/Overf low detector subroutine used during the FILL routine to determine whether a cup has been removed or if foam or I iquid has risen above the I ip. The subroutine uses TLD to detectwith 15 maximum gain and no window. If TLD returns with a distance of exactly 13X' (34.80cm), the distance is rejected and TLD is ca I led again. 13.7"(34.80cm) is the maximum distance a I lowed by the receiver software and indicates no reflection was detected. If TLD returns with a distance less than.25" (0.64cm), the overflow flag is immediately set. If TLD returns with a distance more than. 1 "(0. 25cm) closer than the stored Lip Distance forthree consecutive cal Is to TLD, the overflow flag is set. If TLD returns a distance fartherthan 20 25"(0.64cm) above the stored GRATE value for twelve consecutive calls to TLD, the cup removed flag is set. If TLD at any time returns a distance that does not meet any conditions above, the subroutine ends with no flags set.

TSTCH K (Figure 44A) is a subroutine that reads the five position DIP (Dual In I ine Package) switch. The switch positions are stored in the location in RAM labeled SWITCH. If the switches in positions 1,2 and 3 are 25 a] I off, the Test flag is set.

TIMOUT (Fig ure4413) is used whenever the solendid valve is turned on. The subroutine decrements the "Valve on Time" register and checks to see if the value of the Register is Zero. If it is greater than zero the subroutine ends. If the va I ue of the register is zero, the routine enters a trap loop from which there is no exit except through a hardware reset. The trap loop turns the solendid off and alternately f I ashes the red and 30 green indicators.

INIT(Figure 45M is used when the microcomputer is initialized bythe "MasterClear" (hardware). During power up,thefirst instruction processed is setat location 777 octal. This instruction "GOTO IN17' commands the computerto begin executingthis routine, which comprises the following:

a. All RAM are cleared. 35 b. Wait 1 second for powerto stabilize.

c. Call TSTCHK and run the diagnostic routine if testflag is set.

d. Use TLDto lookwith maximum gain and no windowfor an echo distance between 7"(17.78cm) and 13"(33.02cm).

e. If itdoes not detect an echo within this range,the "Over Ice" indicator on thefront panel flashes. 40 f. If it does detect an echo distance within 7"(1 7.78cm) to 13"(33.02cm), an average of 8 samples is stored in RAM asthe Grate distance and the program continues at CUPDET.

CUPDET is the Cup Detection routine. This routine collects data using TLD and accepts a cup usingthe following procedure:

a. The manual fill switch on thefront panel is monitored continuouslyto assure proper operation. Ifthe 45 manual switch is pressed,the computer beginsthe Cup Removal routine immediately. The DIP switch is read bycalling TSTCHKand if thetestflag is set,the CUPDET routine ends andthe]NIT routine begins.

b. Astable lip distance must be established morethan X' (7.62cm) abovethe GRATE. Astable lip distance is defined as 5consecutive echo distancesfrom TLD separated by 60 milliseconds that correlatewithin 50.1 "(0.25cm). This correspondsto the cup lip being stablefor330 milliseconds. If the stable lip distance istoo 50 closetothe crystals (.6"(1.52cm)),the Lip Distance is rejected,the FILL indicator f lashes and CUPDETbegins again.

c. Acup bottom or ice level must be discernedthat is morethan.1 "(0.25cm) abovethe Grate and morethan 5"(1.27cm) belowthe lip. This is accomplished using TBDW and varying the gain asfollows: With minimum gain, obtain an echo distance using TB1)W. If the echo distance is not morethan.1 "(0.25cm) closerthanthe 55 grate,then thegain is increased onestep and anothersample istaken. If the gain reachesthe maximum,the FILL indicator flashes and the Cup Detection routine begins again.

d. The Ice/Bottom height is calculatedfrom the last distance obtained as outlined in (C) above andthe GRATE andthen stored asthe actual ice height. The cup height is calculatedfrom the lip distance andthe GRATE. The cup height is divided by8 andthe quotient is multiplied bythe3 bit binary numberinputas 60 selected on the ice level programming switches. This allowable ice height is comparedtothe actual ice heightandthe Lip Distance. If the actual ice height is greaterthan allowed bythe switch selection, butless than.511.27cm) belowthe Lip Distance,the cup is rejected andthe Cup Removal routine begins. If theactual ice height is with in.5"(1.27cm) of the Lip Distance,thecup is not positioned correctlyand the FILLindicator flashes before beginning the Cup Detection routine again. If the actual ice height is less than the level selected 65 12 GB 2 190 495 A 12 bythe switch, the FILL routine begins.

The FILL routine controls the complete fi I I ing and top off operation. The routine I imits the solenoid operation to a maximum of 3 On/Off cycles. After each of the first 2 cycles, the routine waits for the foam to settle before starting the next cycle. When the cup is fu I I, the Cup Removal routine begins. If the manual switch is pressed at anytimedu ring the FILL routine, the Cup Removal routine begins immediately. The 5 timeout subroutine is called during the time the solendidvalve is turned on by the fil I program to monitorthe valve on time. if the maxim u m valve on time is exceeded, the timeoutsubroutine turns the valve off and does not return on the fil I routine. A detailed description of the FILL routine follows:

a. Before the solenoid is actuated, several checks and corrections are made. The gain is initially set at maximum. If the Lip Distance is less than 4", (10.16cm), the gain is adjusted with the empirically derived 10 equation:

Gain= Gain - 118 W'(1 0.1 6cm)-lip distance) If the Lip Distance is less than 4"(10.1 6cm), the Lip Distance is adjusted with the empirically derived equation:

Lip Distance= Lip Distance -1/8(4"(10.16cm)- Lip Distance) The Time Constant for this particular cup height is calculated with the equation: 15 Time Constant= cup height/4 for SEV and cup height /8 for Fast Flow.

This Time Constant is used in the first of the three cycles to provide an initial fil I time proportional to the cup height.

b. The gain must be adjusted such that the f I uid level is detected and the I ip is not during the period when the cup vibrates such as at the beginning of a FILL. Also if the cup was not positioned perfectly, the Lip 20 Distance maybe slig h ly farther than originally detected. To adjust the gain, an initial f if ling period proportional to cup height is allowed tom inimize cup vibration and adjust gain as necessary. During this time period the routine uses TBD to check if the echo distance is within. 75"(1.91cm) of the stored Lip Distance. If it is, the gain is reduced one step. If the gain reaches minimum, the Cup Removal routine begins.

This time is also used to adjust the Lip Distance as follows: LGRATE is ca I led and if an overflow is detected 25 then the Lip Distance is decremented (an overflow in LGRATE is defined as more than. 1 "(0.25cm) less than the stored Lip Distance). if LGRATE detects a missing cup then the Cup Removal routine begins immediately.

At the end of this period, the solenoid valve stays on.

c. During the next time period the routine uses TB13W to monitor the liquid level. If the Foamy/Flat switch is set to Foamy, the solenoid turns off when the I iquid level is within. 5"(1.27cm) for SEVorY'(1.78cm) for FFV. 30 If the Foamy/Flat switch is set to Flat, the solenoid is not turned off until the I iquid level reaches.2"(0.51cm) for SEVand.X'(0.76cm)forFFV, at which time the cup removal routine begins. This condition must be met in two consecutive checks for the solenoid to turn off. The G rate/Overf low detector subroutine checks to see if the cup has been removed or if TB13W has missed the I iquid level rising and an overflow is imminent. If the cup ism issing, the cup removal routine begins. If there is an overflow indicated, the solenoid is turned off. 35 d. A 4-second pause begins at this time to a I low the foam to settle. The G rate/Overflowsubroutine cheeks continuously forthecu p to be removed. If it is, the cup removal routine begins i m mediately.

e. After the pause, another4-secondti me period starts. Using TBD, the foam level is monitored. If the foam drops below.411.02cm) for 10 consecutive cheeks, this period ends and the first top-off period begins. If the foam does not drop below.4"(1.02cm) within 4 seconds, the first top-off period begins anyway. The 40 Grate/Overflow subroutine continuously checks fora missing cup. If a missing cup is detected, the cup removal routine begins.

f. Thefirsttop-off cycle uses TBDto determine if the liquid/foam level iswithin.1 "(0.25cm) for a normal 11/2 ounces (42.53 grams) per second valve assembly and.05"(0.1 27cm) forthefaster3 ounces (85.05grams) per second valve assemblies. If this condition exists, the solenoid is notturned on and this cycle ends. If not,then 45 the solenoid isturned on until the condition is met. Forstability, the solenoid has a minimum on time of.25 seconds.

g. A repeat of "D", "E", and "F" occurs nowto implementthe second top-off cycle with the exception that k in "F" thevalues are.210.51cm) forthe normal 11/2 ounces (42.53grams) persecond valve assemblyand 50.3"(0.762c,) forthe faster3 ounces (85.05grams) persecond valve assembly. The cup removal routine 50 (CUPREM) turnsthe fill indicator off, the solenoid off, and the Over-ice indicatoron. It uses TLD and waitsfor an echo distance within.25"(0.64cm) of the Grate. When this condition exists, a new Grate distance is stored, the Over-ice inclicatorturns off, and the Cup Detection routine begins again.

Whilethe preferred embodiments of this invention have been described above in detail, it is to be understood thatvariations and modifications can be madetherein without departing from the scope of the 55 present invention as setforth in the appended claims. Forexample, other materials can be used forthe crystals andthe lenses and other numbers of crystals can be used and other arrangements and locations can be used forthe two crystals of the transducer assembly. In addition, a different ultrasonictransmitterand receivercan be used in place of the crystals, if desired, such asvarious ultrasonicfoii devices. Whiletwo specific control circuits have been described in detail, other control circuits and other components thereof 60 can be used. While a microcomputer has been described and is preferred, the control circuit can alternatively use a microprocessor connected to a remote RAM and ROM, for example. While the transducer assembly and the control module are shown attached to the dispenservalve assembly, this is not essential; they can be attached to the dispenser and just connected electricallytothe valve assembly.

This patent application is a divisional patent application of British Patent Application No. 8517430 65 13 GB 2 190 495 A 13 (published as GB 2161604A).

Claims (1)

1. 5 1. A transducer assembly for use in the automatic filling of a container with a beverage, comprising:

   (a) a housing; (b) a transmitter crystal having a first lens connected to its lowersurface; (c) a separate receivercrystal having a second lens connected to its iowersurface; (d) said transmitter crystal and first lens being located within but nottouching a first non-ferrous metal 10 tube; (e) said receivercrystal and second lens being located within but nottouching a second non-ferrous metal tube; (f) said firstand second metal tubes being located within said housing; and (g) polyurethane foam substantially filling the remaining space within said housing and within said tubes 15 and holding said tubes and said crystals in placetherein.

   2. The assembly as recited in claim 1 including a first and second electrical wires extending from thetop of said assembly out of said foam, said firstwire including first and second leads each being connected to a metal plating on opposite surfaces of said transmitter crystal and a third lead connected to said first metal tube, said second wire including first and second leads each being connected to a metal plating on opposite 20 surfaces of said receiver crystals and a third wire connected to said second tube.

   3. The assembly as recited in claim 1 or 2 wherein said first lens is shaped to produce abeam having a footprint at twelve inches (30.48 cm) that is about 3/4 inch (1.91 cm) wide and about 21/2 inch (6.35 cm) long.

   4. The assembly as recited in claim 1, 2 or3 wherein said crystals are ceramic crystals made of one of PZT-4 or PZT-5a material and said lenses are made of one of ABS or acrylic plastic. 25 5. The assembly as recited in any preceding claim wherein each of said lenses is glued to the respective one of said crystals.

   6. The assembly as recited in any preceding claim wherein said crystals and lenses are circular discs about 1/2 inch (1.27 cm) in diameter.

   7. The assembly as recited in any preceding claim including ultrasonic energy absorbing walls extending 30 downwardly below said transmitter crystal to absorb ultrasonic energy transmitted toward said walls.

   8. The assembly as recited in any preceding claim wherein each of said crystals is recessed into a cavity in a lower surface of said foam.

   9. A transducer assembly for use in the automatic f ill ing of a container with a beverage, su bstantia lly as herein before described with reference to Figures 2, 3 and 27 to 30. 35 Printed for Her Majesty's Stationery Office by Croydon Printing Company (UK) Ltd,9187, D8991685. Published by The Patent Office, 25 Southampton Buildings, London WC2A l AY, from which copies maybe obtained.