US3364295A

US3364295A - Control system for vacuum arc furnace

Info

Publication number: US3364295A
Application number: US374785A
Authority: US
Inventors: Roland W Roberts
Original assignee: NORBATROL ELECTRONICS CORP
Current assignee: NORBATROL ELECTRONICS CORP
Priority date: 1964-06-12
Filing date: 1964-06-12
Publication date: 1968-01-16
Anticipated expiration: 1985-01-16

Description

Jan. 16, 1968 R. vv. ROBERTS 3,364,295

CONTROL'SYSTEM FOR VACUUM ARC FURACE Filed June l2, 1964 J INVENTOR Ralund w. Renens 124 uc. uc.

United States Patent O 3,364,295 CDNTRUL SYSTEM FOR VACUUM ARC FURNACE Roland W. Roberts, Fox Chapel, Pa., assignor to Norbatrol Electronics Corporation, a corporation of Pennsylvania Filed .lime 12, 1964, Ser. No. 374,785 13 Claims. (Cl. 13-13) 'This invention relates to a control system for controlling the posit-ion of a ram in a vacuum arc melting furnace. More particularly, the invention relates to an improved system for continuously `adjusting the ram feed rate of a consumable electrode in a vacuum arc furnace.

Certain types of furnaces are provided for the purification of metals. Generally, a ram holds the metal to be puriiied within the furnace. The metal comprises an electrode which is remelted under a high vacuum by means of a DC electric arc. Itis normal practice to position the ram by electro-mechanical means. The rate of linear movement of the ram is made proportional to the difference between the `average DC voltage across the arc which is formed between the base of the furnace and the electrode material being melted and a reference voltage. The operation of this type of system is entirely dependent upon the amount of voltage across the arc which is determined by the length of the arc. The increment of arc voltage change for a unit incremental change in arc length (delined generally as arc voltage gra-dient) is a function of the metal or alloy being melted. A metal which exhibits a very low value of arc voltage gradient approaching zero cannot be satisfactorily melted by the usual control system.

To solve this problem, one solution has been advanced.

The solution uses the randomly spaced pulsations present in the DC arc voltage. The rate of occurrence of these pulsations is an indication of the average arc length and melting rate. The pulse rate increases with an increasing melting rate and a -decreasing arc length. The pulses are counted during a fixed time interval by a digital electronic counting means. The total from this count is compared with a desired reference count and the ram is moved by an `amount dependent on the count error between the actual count and the reference count.

.The disadvantage of this solution is that it produces an undesirable step motion of the ram. This is due to the fact that the time interval in which the pulses are counted is substantial.

The 'present invention approaches a smooth uniformram feed rate using less complex equipment.

I provide a control system for controlling the position of a ram in an arc melting furnace comprising means detectin-g voltage pulses present across the arc, means quantizing and amplifying the detected voltage pulses and coupled to the means detecting, means integrating the difference between the amplified quantized voltage pulse and a fixed reference voltage, the means integrating coupled to the means quantizing, and means linearlyl moving the ram at a rate dependent upon the difference.A between the integral voltage froml the means integrating bistable amplifier having a control winding andan out` put. The control winding is electrically connected to the output of the means detecting. The bistable amplifier con- Verts pulses of random amplitude and duration from the means detecting into corresponding pulses of iixed amplitude and dur-ation.

I preferably provide that the integrating means comprises a capacitor coupled to the positive output of a bistable amplifier 4and a DC voltage coupled across the capacitor and having a negative polarity with respect to the positive output of the bistable amplier so that when an lelectrical charge coming from the bistable amplifier is built up on the capacitor, it is discharged through the negative side of the DC source and the voltage developed across the capacitor is the integral of the difference between the pulse rate of the bistable amplifier and the voltage from the 4DC source.

I preferably provide that the means linearly moving the ram at a rate dependent upon the difference between the integral voltage across the capacitor in the integrating circuit and a fixed reference voltage comprises a pulse oscillator coupled to the integrating circuit. The pulse oscillator drives an electric motor which controls a ram positioning servo system means which drives the ram at a rate which is dependent upon the rate of pulses present across the arc.

In the foregoing I have set out certain objects, purposes and advantages of this invention. Other objects, purposes and advantages will be apparent from the following description in which FIGURE 1 is a schematic of a half wave push-pull self-saturating magnetic amplifier;

FIGURE 2 is a schematic circuit diagram of a bistable amplifier providing .a controlled DC voltage at a load;

FIG-URE 2a is a graph of voltage versus time; and

yFIGURE 3 is a schematic of a control system for a consumable electrode vacuum arc furnace.

Referring to FIGURE 1 I have illustrated a pushpull half wave self-saturating magnetic amplifier generally shown as 10 which comprises a transformer core 12 around which primary winding 1'4 having terminals 16 and 18 is wound. A secondary winding 20 is wound around the core 12 and terminated at junction points 22 and 24.

Gate windings

26 and 28 are wound around cores 30` and 32, respectively.

Control windings

34 and 36 having terminals `38 and 40 are wound around

cores

30 and 32.

The

gate windings

26 and 28 are in series with rectifiers 42 and 44, respectively. The output from the magnetic amplifier 10 is taken across resistor load 46 at

terminals

48 and 50.

Gate resistor loads

52 and 54 are connected to junction point 24. Resistor 56 in series with rectifier 58 which is connected to Zener diode 60 are all connected between junction points 24 and 22.

The bias circuit comprises

resistors

62 and 64 tapping

gate windings

26 and 28, respectively.

The operation of the magnetic amplifier in the absence of a control current flowing through the

control windings

34 and 36 is as follows: An AC source is put across terminalsl 16 and 18 of the primary winding 14. When terminal 18 is positive, current IG1 and 162 will flow from junction point 22 through

gate windings

26 and 28 to produce a flux qsGl` and e562, respectively. pGI and ps2 are shown moving in the direction of the arrow on the figure. Assuming ideal conditions in which GI=G2, cores 30 `and 32 will fire (that is they will saturate) at the Sametime andproducea chopped since wave voltage across

resistors

52 and 54. The voltage across resistor 46 is the difference of the two voltages at

resistor

52 and 54. In this situationthe output voltage between

terminals

48 and 50 is zero. Now when the AC source goes negative and terminal 16 is positive with respect to terminal 18, current flows through rectifier 58, resistors 62 `and 64 whereby the fiux is driven back or reset to the high gain region. On the next half' cycle when terminal 18 is again positive the cores are fired again.

Now assume that a control current ows through

control windings

34 and 36 so as to make terminal 38 positive with respect to terminal 40. The direction of the control current shown in this situation is shown in FIGURE 1 by IC which produces a linx bcl and pc2 in the direction shown on the FIGURE l.

The control Winding 34 and the control current through it are such that ,fn-,I tends to oppose G1. The fiuX produced by the control current tend to opposed the flux produced by the gate current. The control winding 36 and the control current through it are such that C2 tends to add to G2. Therefore, whenever terminal 18 is positive with respect to terminal 16 and terminal 38 is positive with respect to terminal 40, core 32 will saturate prior to core because of the relationship of the fluxes in the respective cores. The opposing fluxes in core 30 cause it to take a longer time to saturate. The adding fiuXes in core 32 cause it to take less time to saturate. Therefore, between the time core 32 saturates until just before core 30 subsequently saturates, there is a voltage difference present across resistor 46 in which terminal 50 is positive with respect to terminal 48, This voltage is in the form of a pulse. Throughout the specification the term first polarity refers to the time when terminal 50 is positive with respect to terminal 48. In order to make terminal 48 positive with respect to terminal 50 (which is referred to as a second polarity) all that need be done is to reverse the polarity of IC and make terminal positive with respect toV terminal 38. This produces the opposite result in the cores by changing the direction of the flux produced by the control current IC. In this situation core 30 saturates prior to core 32 and terminal 48 has a positive pulse voltage with respect to terminal 50.

Referring to FIGURE 2 a rectifier circuit couples an AC source at

terminals

62 and 64 with a load 66. The rectifier circuit is controlled by a silicon controlled rectifier 68, hereinafter referred to as SCR. The SCR has an anode 70, a cathode 72, and a gate 74. The gate 74 and cathode 72 are connected directly to the output terminals and 48 of the magnetic amplifier 10. A full wave rectiication bridge

circuit comprising rectiiiers

76, 78, y80, and 82 are connected to the AC source at terminals 84 and 86. Rectifier 80 has an inductor 88 in series with it. The DC terminal 90 of the bridge circuit is connected directly to the load 66. The other DC terminal 92 of the bridge circuit is connected to the load 66 through SCR 68. The means coupling the inductor 88 to the gate 74 and cathode 72 of SCR 68 comprises lead 94 and rectifier 96. Rectifier 98 is used whenever inductive loads are applied at load 66.

The circuit described above operates as follows: Assume an output pulse from the magnetic amplifier 10 in which terminal 50 is positive with respect to terminal 48. This pulse triggers SCR 68. It must be remembered throughout the specification during the discussion of the SCRs that in order to trigger the SCR a positive voltage only must be applied at the gate. A negative voltage at the gate of the SCR will not fire or trigger it. Triggering or firing refers to the time when the anode conducts current to the cathode `as though a switch were thrown between the anode and the cathode. Until the SCR is triggered, it appears as an open switch and the only way to close the switch is to apply a positive voltage at the gates.

When AC terminal 62 is positive with respect to terminal 64 and at the vsame time assume terminal 50 produces a positive voltage with respect to terminal 48, the SCR is triggered. Current will flow from the AC terminal 62 to terminal 84, through rectifier 76, to terminal 90, through load 66, through SCR 68, to terminal 92, through inductor 88, through rectifier 80, to the AC terminal 86 of full wave rectifier bridge circuit, and return to the AC source terminal 64. In order to keep an SCR conducting anode voltage subsequently goes positive. After the first half cycle in the particular circuit described is ended, the current must go to zero. In order to retrigger the SCR 68 in the absence of a voltage pulse from the magnetic amplifier 10, an inductor 88 is used. When the current approaches zero across the load 66 and in the anode 70 of the SCR 68, the flux in inductor 88 collapses and produces a current through rectifier 96 thereby applying positive voltage on the gate 74 and retriggering the SCR 68. At that moment terminal 64 turns positive with respect to terminal 62 in the AC source `and sufficient current flows through the anode 70 from the full wave rectifier bridge circuit during the remainder of the second half cycle until the current -again .approaches zero when terminal 62 begins to turn positive with respect to terminal 64. When terminal 64 is positive with respect to terminal 62, the current flows to the AC terminal 86 of the bridge circuit, through rectifier 78, to the DC terminal 90 of the bridge circuit, through the load 66, through the SCR 68, through lead 94, to the AC terminel 92 at the bridge circuit, through rectifier 82, to terminal 84 of the bridge circuit, returning to AC source terminal 62. The voltage across the load 66 appears las shown in graph of FIGURE 2a. The dotted lines on the graph represent the voltage between the

AC terminals

64 and 62. A solid line represents the actual voltage as seen across load 66. The magnetic amplier has its power supplied from the same AC source through a pair of leads so that when terminal 62 is positive the AC voltage on the terminal 18 of the magnetic amplifier is positive.

I have shown fa bistable amplifier that switches power from an AC source to a load 66 depending upon the polarity and magnitude of the control current at

terminals

40 and 38 of the magnetic amplier 10. The AC power was switched in such a manner as to produce a full cycle of controlled DC voltage at the load 66 as shown in the graph.

Referring to FIGURE 3 a DC current regulated arc power supply 106 is applied to crucible housing 108 and ram 110 through leads 112 and 114. Ram 110 holds ingot 116. Ingot 116 is melted by an arc 122 between the bottom surface 118 of the ingot 116 and molten metal 120 resting on the bottom of crucible housing 108. Vacuum pump 124 creates a vacuum within the crucible housing 108 thereby facilitating the electric arc melting technique. As the ingot 116 is being melted, the ram 110 holding ingot 116 is lowered by .a chain 126 wound on pulley 128..

The length of the arc 122 is a function of the rate of melting. Random pulses of varying amplitude and duration are present in the voltage across the arc 122. The rate of occurrence of these random pulses is a function of the length of the arc 122. The rate of occurrence of the voltage pulses present in the arc 122 increases with an increasing melting rate and a decreasing arc length 122.

f Therefore, the rate of random pulses is used to control in a smooth feeding manner the lowering of the ram 110 on chain 126 from pulley 128.

Leads

130 and 132 couple the arc voltage to a standard L-C filter having a low-pass characteristic designed to remove any rectifier ripple and other high frequency noise. The L-C filter comprises inductor 134 and capacitor 136. The out-put of the L-C filter is coupled to

terminals

38 and 40 of the bistable amplifier through resistor 138 and coupling capacitor 140. The effect of the L-C filter coupled between

terminals

38 and 40 of the bistable amplifier 100 and

junction points

142 and 144, is to detect randomly occurring pulses of varying duration and amplitude between

terminals

142 and 144.

Bistable amplifier 100 acts as a pulse quantizer and amplifier which converts the random pulses into pulses of uniform amplitude and uniform duration. Therefore, one pulse delivered between

terminals

138 and 140 will pro-` duce one pulse of uniform amplitude and duration between

outputs

102 and 104 of bistable amplifier 100.*The

40 to the

control windings

34 and 36 of the bistable amplifier 100 referred to in FIGURE 1 and in FIGURE 2. The polarity of the

control windings

34 and 36 from

terminals

38 and 40 is such that an increase in magnitude in the arc voltage across arc 122 will tend to turn ON the output of the bistable amplifier 100. Control windings 146 and 148 are wound in the same displacement as

control windings

34 and 36 and likewise control

windings

150 and 152 are wound in the same displacement at

control windings

34 and 36. Control windings 146 and 148 lare coupled to a DC source 154 through resistor 156. The polarities of control windings 146 and 148 and voltage source 154 are such as to tend to turn OFF the output of the bistable amplifier 100 by creating an opposing magnetornotive force to the magnetomotive force produced by current flowing through

control windings

34 and 36.

Control windings

150 and 152 in series with resistor 158 have a polarity that also produces au opposing magnetomotive force to the magnetomotive force produced by current in

control windings

34 and 36, and therefore tends to turn OFF the output of bistable amplifier 100. Therefore, when the output from the bistable amplifier 100 between

terminals

102 and 104 is switched ON as a result of a pulse of sufficient magnitude being applied to the

input terminals

38 and 40 of the bistable amplifier 100, the output at

terminals

102 and 104 will be switched OFF immediately after one full cycle of output. The magnitude of the magnetomotive force produced by

windings

34 and 36 must be greater than the opposing magnetomotive force produced by windings 146 and 148 to turn ON bistable amplifier 100. DC source 154 and variable resistor 156 coupled to windings 146 and 148 act as a sensitivity control for the pulse input to bistable amplifier 100. The sensitivity control selects the minimum pulse input to bistable amplifier 100 which turns it ON.

The output of the vbistable amplifier 100 is coupled through rectifier 160 and resistor 162 to an integrator and reference network comprising a capacitor 164 and a DC source 166 in series with a resistor 168 all in parallel with the capacitor 164. The DC source 166 is the Reference and forms part of the integrator network. The capacitor 164 has a positive reference terminal 170. The DC source 166 has a negative polarity terminal 172 and a positive polarity terminal 174. For each output pulse from the bistable amplifier 100, an approximately fixed value of electrical charge is placed Ion the positive reference 170 of the capacitor 164.` This function is generally achieved whenever the voltage across the capacitor 164 at any one time is much smaller compared to the output voltage from the bistable amplifier 100. The amount of charge delivered to the capacitor 164 is determined by the valve -of the resistor 162. This value is selected to achieve a desired pulse rate. The rectifier 160 prevents the capacitor 164 from discharging back through the bistable ampliiier100. The battery 166 provides a constant current discharge from the positive reference 17 0 of the capacitor 164. To approach a constant current discharge characteristic curve, the voltage of DC source 166 should be much greater than the operating value `of the voltage across capacitor 164. When the desired average pulse rate is present in the voltage across the arc 122, the capactior 164 will receive equal average charging and discharging currents. Under this condition the capacitor 164 voltage will remain constant at whatever value may be present. If the voltage pulse rate detected in the arc 122 is greater than the preset desired value of voltage pulse rate, then the capacitor 164 will receive a greater average charging current than is being taken away by the discharge circuit and the capacitor 164 voltage will increase at `a rate directly dependent upon the magnitude of the difference (error) in average pulse rate. If the average input voltage pulse rate is less than preset value of voltage pulse rate, then the capacitor 164 voltage will decrease at a proportional rate. The value of the voltage across the capacitor 164 is represented by the following formula: N

The DC source 166 with

resistors

162 and 180 determines the unique pulse rate at which the capacitor voltage will remain constant, i.e., the value of the DC source 166 is calibrated against a desired pulse rate present in the DC arc voltage.

The output of the integrating capacitor is coupled at

terminals

176 and 178 through resistor 180 to a pulse oscillator circuit which comprises a bistable amplifier 182, rectifier 200 in series with resistor 202, an RC network having capacitor 204 in parallel with resistor 206, a resistor 208 in series with

feedback windings

192 and 194.

Terminals

176 and 178 are the input control winding terminals of bistable amplifier 182. Bistable amplifier 182 has the same internal circuitry as preceding bistable amplifier and

terminals

176 and 178 correspond to

terminals

38 and 40 of bistable amplifier 100.

Control windings

184, 186, 188, 190, 192 and 194 occupy the same displacement within bistable amplifier 182.

Control windings

188 and 190 are of such a polarity that whenever current is fiowing from DCsource 154, the magnetomotive force produced tends to turn ON bistable amplifier 182. Control windings 184 and 186 are of such a polarity that current flowing through them produces a magnetomotive force which tends to turn OFF bistable amplifer 182.

Feedback control windings

192 and 194 are of such a polarity that when current fiows through them the magnetomotive force produced tends to turn OFF bistable amplifier 182.

A net magnetomotive force is produced at the input of 'bistable amplifier 182 by

windings

184, 186, 188 and 190. Whenever the net magnetomotive force is positive, the bistable amplifier tends to turn ON.

When the bistable amplifier 182 is turned ON for one fullcycle of output, an appreciable charge is built up on capacitor 204 and a current passes through

feedback windings

192 and 194, turning bistable amplifier 182 OFF. The charge approaches a semi-peak of the output voltage from the bistable amplifier 182. In the absence of any output from bistable `amplifier 182, the current then discharges from capacitor 204 slowly through both the resistor 206 of the semi-peak charging RC network and the

negative feedback windings

192 and 194. When the voltage across capacitor 204 has dropped to a value in` which the magnetomotive force produced by the

feedback windings

192 and 194 is approximately equal to the net positive magnetomotive force produced by windings .184,' 186, 188, and 190, bistable yamplifier 182 will turn ON again for one cycle completing one full cycle of operation of the pulse oscillator. l p

The resistor 206 adjusts the proportionality between the pulse rate of bistable amplifier 182 and the input control current in windings 184 and 186. v

The exponential nature of the discharge of the capacitor 204 after being charged during one cycle of output from the bistable amplifier 182 causes a time interval between output pulses yof bistable amplifier 182 which is ya Afunction of the net positive magnetomotive force produced by windings 184, 1.86, 188 and 190. Therefore, the pulse rate of the pulse oscillator is controllable by the pulse rate present in the arc 122.

The purpose of the above-described pulse 1oscillator network is to drive a potentiometer gear motor 210 in a continuous series of incremental steps in which themotor 210 will move a step typically every few seconds dependent on the difference betweenra reference and the charge Ion the integrating capacitor.

The potentiometer gear motor 210 is mechanically coupled to a servo network system. The motor 210 is mechanically coupled to the command position arm 212; The servo network system comprises position command arm 212 and follower arm 218. The voltage taken from position command arm 212 and follower arm 218 is fed to amplifier 222. In the event there is an unbalance 'between the position command arm 212 and follower arm 218, there will be a voltage present (error voltage) at the amplifier 222. The output of the amplifier 222 is electrically coupled to mechanical drive means 226. The output voltage from amplifier 222 rotates the mechanical drive means 226. Mechanical drive means 226 is mechanically coupled to follower arm 218 and moves the follower arm 218 so as to restore the balance tending to create a zero input voltage to amplifier 222. The mechanical drive means 226 is mechanically coupled to pulley 128 through shaft 224. When the mechanical drive means 226 rotates, the ram 110 is lowered from the pulley 128 into the crucible 108. The pulse oscillator motor drives the position command arm which regulates the ram positioning servo system and produces a means for driving the ram downwardly at a very low speed which is proportional to the difference between the integrating capacitor voltage and the fixed reference 166.

Whenever ingot 116 is melting at a faster rate than desired, arc 122 Will become shorter in length. The shortened length of arc 122 causes an increased rate of voltage pulses which are detected Iby the low band pass filter and quantized and amplified into pulses of fixed durationand amplitude by bistable amplifier 100. The increased pulse rate is impressed across integrating'capacitor 164 causing a slow increase in capacitor voltage. This causes more current to flow in control windings 184 and 186 thereby reducing the net positive magnetomotive force acting to turn ON bistable amplifier 182. Bistable amplifier 182 therefore remains OFF and position command motor 210 remains stationary for a longer interval of time. Whenever this happens, there is a balance between the position command arm 212 and the follower arm 218 and the input and output voltages of amplifier 222 is zero whereby the mechanical drive means 226 ceases to rotate for a longer interval of time and the ram 110 is lowered at a slower rate. This condition wil-l remain until the rate of pulses present in arc 122 is decreased. This is brought about by increasing the arc length 122 which would be a natural result if the ram 110 is momentarily stopped from lowering the ingot 116 into the crucible housing 108. The average feed rate is then automatically adjusted to the value that gives the desired preset arc voltage pulse rate. p

While l have il-lustrated and described a preferred embodiment and practice of my invention, it will be understood that this invention may -be otherwise practiced within the scope of the following claims.

I claim:

1. A control system for controlling the position of a' ram in an arc melting furnace comprising:

(l) Means detecting voltage pulses present across the arc;

(2) Means quantizing and amplifying the detected voltage pulses and coupled to the means detecting;-

(3) Means integrating the difference between the amplifier quantized voltage pulses and a reference, the means integrating coupled to the means quantizing; and

(4) Means linearly moving the ram at a rate dependent I upon the integral voltage from the means integrating. 2. A control system for control-ling the position of a ram in an arc melting furnace comprising:

(l) Means detecting voltage pulses present across the arc; (2) Means converting input pulses of random amplitude and duration from the means detecting, into i corresponding pulses of fixed amplitude and duration;

(3) Means integrating the difference between the fixed amplitude pulses from the means converting and a reference, the means integrating coupled to the means converting; and

(4) Means linearly moving the ram at a rate dependent upon the integral voltage from the means integratmg.

3. A control system for controlling the position of a ram in an arc melting furnace comprising:

(l) Means detecting voltage pulses present across the arc and having an output;

(2) A bistable amplifier having a control winding and an output, the control winding electrically connected to the output of the means detecting, the bistable amplifier converting pulses of random amplitude and duration from the means detecting into corresponding pulses of fixed amplitude and duration;

(3) Means integrating the difference 'between the fixed amplitude pulses from the bistable amplifier and a reference forming part of the means integrating, the means integrating coupled to the output of the bistable amplifier; and

(4) Means linearly moving the ram at a rate dependent upon the integral voltage from the means integrating.

4. A control system for controlling the position of a ram in an arc melting furnace as recited in claim 3 wherein the means integrating comprises:

(l) A capacitor;

(2) Means coupling the DC pulses from the bistable amplifier to the capacitor; and

(3) Reference means discharging the capacitor.

5. A control system for controlling the position of a ram in an arc melting furnace as recited in claim 3 wherein the means linearly moving the ram at a rate dependent upon the integral voltage developed across the capacitor comprises:

(1) A pulse oscillator means;

(2) Means coupling the pulse oscillator means to the integral voltage developed across the capacitor; and

(3) A ram positioning servo means driving the ram at a rate dependent upon the rate of pulses from the pulse oscillator, the servo means coupled to the output of the pulse oscillator.

6. A control system for controlling the position of a ram in an arc melting furnace as recited in claim 4 wherein the means coupling the DC pulses from the bistable amplifier to the capacitor comprises a rectifier and a resistive -load in series connecting the Ibistable amplifier output to the capacitor.

7.` A control system for controlling the position of a ram in an arc melting furnace as recited in claim 4 wherein the reference means discharging the capacitor comprises a DC source.

8. A control system for controlling the position of a ram in an arc melting furnace as recited in claim 4 wherein the means linearly moving the ram at a rate dependent upon the integral voltage developed across the capacitor comprises:

(l) A pu-lse oscillator means;

(2) Means coupling the pulse oscillator means to the integral voltage developed across the capacitor;

(3) A motor electrically coupled to the pulse oscillator, the output voltage of the pu-lse oscillator controlling the rotation of the motor; and

(4) A ram' positioning servo means driving the ram at a rate dependent upon the rate of pulses from the pulse oscillator, the servo means coupled tothe motor.

9. A control system for controlling the position of a ram in an arc melting furnace as recited in claim 4 wherein the means detecting voltage pulses present across the arc omprises a band pass filter connected across the arc.

i0. A control system for controlling the position of a ram in an arc melting furnace as recited in claim 4 wherein the means detecting voltage pulses present across the arc comprises an inductor capacitor low frequency pass network connected across the arc.

11. A control system for controlling the position of a ram in an arc melting furnace as recited in claim 8 wherein the pulse oscillator means comprises:

(l) A 1bistable amplifier having an input coupled to the integral voltage across the capacitor; and

(2) Means applying delayed negative feedback to the bistable amplifier input.

12. A control system for controlling the position of a ram in an arc melting furnace as recited in claim 8 wherein the pulse oscillator means comprises:

(l) A bistable amplifier having an input coupled to the integral voltage;

(2) A feedback Winding providing an opposing magnetomotive force to the net input magnetomotive force to the bistable amplifier; and

(3) A semi-peak charging RC network coupled to the output of the bistable amplifier providing current through the feedback Winding whereby a delayed negative feedback is produced.

13. A control system for controlling the position of a 25 ram in an arc melting furnace as recited in claim 8 wherein the ram positioning servo means comprises:

(l) A compound potentiometer coupled to the motor;

(2) An amplifier having an input coupled to the command potentiometer;

(3) An electro-mechanical drive means coupled electrically to the output of the amplifier, the electromechanical drive means driving thc ram; and

(4) A fol-lower potentiometer mechanically coupled to the electromechanical drive means and electrically coupled to the amplifier input, when the voltage across the input of the amplifier is zero, the electromechanical drive means ceases its rotation until an unbalance exists between the command potentiometer and follower potentiometer.

References Cited UNITED STATES PATENTS 6/1965 Murtland et al 13--13 6/1965 Lyman 13-13 X BERNARD A. GILHEANY, Primary Examiner.

R. N. ENVALL, JR., Assistant Examiner.

UNITED STATES PATENT OFFICE CERTIFICATE OF CORRECTION Patent No. 3,364,295 January 16, 1968 Roland W. Roberts It is hereby certified that error appears in the above numbered patent requiring correction and that the said Letters Patent should read as corrected below.

Column l, lines 53, 57 and 6l, "quantizing", each Occurrence, should read quanttizng Column 2, line 62, "since" should read sine Column 3, line 8, "@Gl", first occurrence, should read @Cl line 9, "tend" should read tends same line 9, "opposed" should read oppose line 63, "gates" should read gate Column 4, line 20, "terminel" should read terminal line 69, "quantzer" should read quantitizer n. Column 5, line l0, "at" should read as line 62, "capactior""should read capacitor Column 7, line 3l, "quantzed" should read quantitized line 62, quantizing" should read quantitizng Column l0, line 3, "compound" should read command line l0, "electromechanical" should read electro-mechanical Signed and sealed this 7th day of October 1969.

(SEAL) Attest:

EDWARD M.PLETCHER,JR. WILLIAM E. SCHUYLER, JR. Attesting Officer Commissioner of Patents

Claims

1. A CONTROL SYSTEM FOR CONTROLLING THE POSITION OF A RAM IN AN ARC MELTING FURNACE COMPRISING: (1) MEANS DETECTING VOLTAGE PULSES PRESENT ACROSS THE ARC; (2) MEANS QUANTIZING AND AMPLIFYING THE DETECTED VOLTAGE PULSES AND COUPLED TO THE MEANS DETECTING; (3) MEANS INTEGRATING THE DIFFERENCE BETWEEN THE AMPLIFIER QUANTIZED VOLTAGE PULSES AND A REFERENCE, THE MEANS INTEGRATING COUPLED TO THE MEANS QUANTIZING; AND (4) MEANS LINEARLY MOVING THE RAM AT A RATE DEPENDENT UPON THE INTEGRAL VOLTAGE FROM THE MEANS INTEGRATING.