Bilateral digital data transmission system4519083Abstract A digital data transmission system for transmitting digital data from digital data handling hardware over a transmission line by analog signal representation and for receiving analog signal representation for reception of digital signals by the digital data handling hardware, incorporates an integrated circuit PSK modem. The modem has a transmitter which includes an encoder for encoding digital data into a first format for conversion to corresponding PSK signals and a transmitter modulator, implemented in switched capacitor circuits, for phase modulating a first carrier frequency according to the first format. The modem has a receiver that includes a demodulator, implemented in switched capacitor circuits, for demodulating PSK signals on a second carrier frequency to provide a second format for conversion to corresponding digital signals, and an encoder for encoding the second format into corresponding digital signals. In the preferred embodiment, the transmitter includes a buffer for the accommodation of a variety of modem protocols, a first filter for compensation of gain distortion in the transmission line, and a second filter for compensation of phase distortion in the transmission line. In this preferred embodiment, the receiver includes compensating filters for transmission line gain and phase distortion, automatic gain control and adaptive equalization filtering for reducing cross-talk noise between the transmission and receiving lines. Claims What is claimed is: Description BACKGROUND OF THE INVENTION
TABLE 1
______________________________________
Bell V22
______________________________________
Originate
(1200 Hz) Originate
(1200 Hz)
Answer (2400 Hz) Answer (2400 Hz)
______________________________________
Dibit Phase (degrees) Dibit
Phase (degrees)
______________________________________
00 +90 00 +270
01 0 01 +180
11 +270 11 +90
10 +180 10 0
______________________________________
RV
Originate (2250 Hz)
Answer (1150 Hz)
Dibit Phase (degrees) Dibit
Phase (degrees)
______________________________________
00 +270 00 +90
01 +90 01 +270
11 +180 11 +180
10 0 10 0
______________________________________
The current phase is added to the previous phase to cause a phase shift, or not. PLA198 receives inputs from logic circuits which in turn are input by the particular modem protocol employed. An answer/originate signal A/O is applied to AND gate 203 whose other input is BELL/RV-. Signal V22 is applied to one input of NOR gate 202 whose other input is supplied by the output of AND gate 203. Signal VAD is supplied directly to PLA198 and is inverted through inverter 199 and applied to PLA198. The output from NOR gate 202 is applied directly, and inverted through inverter 201 as well. Circuit 196 is made up of a pair of AND gates input to a NOR gate. Signal HLFSPD- is applied to one of the AND gates and signal HLFSPD is applied to the other AND gate. The Q output of flip flop FF71 is applied to the other input of the one AND gate and the Q output from flip flop FF70 is applied to the other input of the other AND gate. The output of the NOR gate of circuit 196 is signal D1- which is applied directly to the PLA198 and, inverted as signal D1, is also applied to the PLA198. The Q output of flip flop FF70 is signal D2 and the Q- output is signal D2-, each of which is applied to PLA198. The output signal of PLA198 is AB1, dependent upon the various selections. Signal VAD selects devices in PLA198 and provides one input to a selection circuit, specifically to NOR gate 211 and to AND gate 213. Signal D1 provides another input to NOR gate 211 and one input to AND gate 212. Signal D2- provides the third input to NOR gate 211 and the other input to AND gate 212. The output from NOR gate 211 and AND gates 212 and 213 provide three inputs to NOR gate 214, whose output signal AB2 provides an input to two bit adder 221. The sum is represented by bits DIN1 and DIN2, and their inversions from Q and Q- outputs of flip flops FF91 and FF92, respectively. In summary, the transmit buffer provides digital circuitry implemented in MOS for receiving asynchronous data and transmitting it in synchronous fashion. If the data received is synchronous, then the circuitry is bypassed. The user may select the particular mode of protocol desired. When selected, the dibit combinations are set, the data is scrambled or not, and the appropriate carrier frequencies are selected. TRANSMITTER FIG. 10 illustrates pseudo random shift register 230 which is provided with clocks 1 and 2, 180.degree. out of phase with each other, from clock generator circuit 232, which is identical to the clock generator circuit 33 of FIG. 3A, having an input from clock buffer 26. Flip flops FF101 through FF108, together with the PLA 231 and associated logic circuitry, form the pseudo random shift register 230. Inputs to PLA 231 are signals VAD and its inversion, and A/O and its inversion. Flip flops FF107 and FF108 have their outputs exclusively ORed through circuit 235. Flip flops FF102 and FF103 have their outputs exclusively ORed through exclusive OR circuit 234. Exclusive OR circuits 234 and 235 have their outputs exclusively ORed through circuit 233 which provides the J input to flip flop FF101. Clocks 1 and 2 clock the flip flops FF101-FF108. Output signals phi 1 and phi 1- are developed through selection circuit 236 and output signals phi 2 and phi 2- are developed through selection circuit 237. NOR gate 238 has inputs from the PLA 231 and its output, inverted, provides the preset inputs to the flip flops FF101- FF108. Signals phi 1 and phi 2, and their inversions, provide the clock signals for the switched capacitor circuits which are used in the transmitter and receiver circuits. They occur at a rate of 16 times the desired carrier frequency. FIG. 11 illustrates a circuit for developing sine and cosine gating signals for the In phase and Quadrature phase analog channels to be described. Signals phi 1 and phi 2 are input to NOR gates 242 and 243, respectively, which are cross-coupled, the output of NOR gate 243 providing an input to AND gate 245. Select signals VAD and A/O are applied to exclusive OR circuit 241 whose output, inverted, provides the other input to AND gate 245 and which directly provides an input to AND gate 244. The outputs of AND gates 244 and 245 provide the inputs to NOR gate 247 whose output provides the J input to flip flop FF111 and, inverted, to the K input of that flip flop. The Q output of flip flop FF111 provides the other input to AND gate 244 and clocks toggle flip flop FF112. The output of NOR gate 241 provides one input to OR gate 248 whose other input is provided by the signal phi 2-. The output of OR gate 248 provides one input to NAND gate 249 whose other input is provided by the signal phi 1-. Flip flop FF111 is clocked by the output of NAND gate 249. This clocking provides synchronization for flip flop FF111. The Q output of flip flop FF112 clocks toggle flip flop FF113 whose Q output clocks toggle flip flop FF114. The Q- output of flip flop FF112 clocks flip flop 116. The Q output from flip flop FF113 provides the J input to flip flop FF116 and the Q- output of flip flop 113 provides the K input to flip flop FF116 and clocks flip flop FF115. The Q output of flip flop FF114 provides the J input to flip flop FF115, and provides one input to circuit 251 and one input to circuit 252. Circuits 251, 252, 253 and 254 are identical to each other and are made up of a pair of two input AND gates whose outputs provide the inputs to a NOR gate. The Q- output of flip flop FF114 provides an input to circuits 251 and to 252. The Q output of flip flop FF115 provides one input to circuit 253 and one input to circuit 254. The Q- output provides one input to each of circuits 253 and 254. Signals phi 1- and phi 2- each provide an input to circuits 251, 252, 253 and 254. The output of circuit 251 is cosine signal phi 2I. The output of circuit 252 is cosine signal phi 1I. The output of circuit 253 is sine equal phi 1Q and the output of circuit 254 is sine signal phi 2Q. FIG. 12 illustrates a digital circuit for developing signals determinative of whether signals input to analog circuits should be inverted or not. Exclusive OR- type circuits 256-259 provide output signals phi 1LI, phi 2LI, phi 1LQ and phi 2LQ, respectively. circuit 256 receives signals phi 1- and DIN1, and phi 2- and DIN1-. Circuit 257 receives input signals phi 1- and DIN-, and phi 2- and DIN1. Circuit 258 receives signals phi 1- and DIN2, and phi 2- and DIN2-. Circuit 259 receives phi 1- and DIN2-, and phi 2- and DIN2. If phi 1LI equals phi 1, and if phi 2L1 equals phi 2, then signal VXMT will be inverted. If phi IL1 equals phi 2, and phi 2L1 equals phi 1, then signal VXMT will not be inverted. FIG. 13 illustrates the development of signals BBS and BBS- which are used in this transmitter circuit for putting a top notch in the square waves to cause square waves to look a little more like sine waves. Signal DBT is applied to the K input of flip flop FF118, and inverted, applied to the J input of that flip flop. The clock is provided by signal SCTA, inverted, and the Q- output is applied to the input of inverter 262 whose output is signal BBS. The Q output of flip flop FF118 is inverted through inverter 261 to provide the signal BBS-. Before proceeding with the discussion of the analog circuitry, FIGS. 14 and 15 should be referenced. Taken together with FIG. 11, the various digital signals developed are evident. Signals phi 1 and phi 2, applied in the circuitry of FIG. 11 are shown in FIG. 14 as non-overlapping clock signals. When transmitting at the high band frequency, signal A, the J input to flip flop FF111, is shown as the clock frequency divided by two. The clocking signal B to the flip flop F111 is shown as occurring at each of signals phi 1 and phi 2. The output of flip flop FF111, signal C, is shown at the same frequency as signal A, but displaced therefrom. In this portion of FIG. 14, the resultant waveforms come from the output of NOR gate 243. Following, the signals are developed with the NOR gate 243 output disabled to provide a low band frequency. In the "transmitting low band" section of FIG. 14, signal B is shown to be the same as signal phi 2 and signal C is at one half the frequency of signal C when it is transmitting high band. FIG. 15 continues following the signals through FIG. 11 with signal C, for the purpose of better illustration, being shown at evidently double its frequency of FIG. 14. In reality, it is being shown in expanded time scale. Signal D, the Q output of flip flop FF112 is shown at half the frequency of signal C; signal E, the Q output of flip flop FF113 is shown at half the frequency of signal D; signal F, the Q output of flip flop FF114 is shown at half the frequency of signal E; signal G, the Q output from flip flop FF115 is shown at the same frequency as signal F, but shifted with respect thereto; signal H, the Q output of flip flop FF116 is at the same frequency as signal E but shifted therefrom. The Q output of flip flop FF116, inverted, is signal CGT and the Q- output, inverted,, is signal SGT. Signal CGT is used to form a top notch in the cosine signal and signal SGT is used to form a top notch in the sine signal. These will be described below. As indicated above, four binary values (dibits) may be representative of four phases, those phases being dependent upon the modem protocol selected. Since four digital bits are represented by a single phase shift, the baud rate is one half of the bit rate. Typically, the baud rate is in the neighborhood of 600 while the bit rate is in the neighborhood of 1200 bps. Each pair of bits is used to select one of four possible phase shifts according to whether in answer or originate mode, or the particular modem protocol selected. Out of these selections comes four possible carrier signals, 1150 and 2250, and 1200 and 2400. Turning now to FIGS. 16A and 16B, the transmit analog circuit 13 is shown. Reference voltage Vxmt is a voltage at which the switched capacitor circuits operate. This voltage is somewhat elevated from that used in the conventional MOS circuits heretofore described. In this preferred embodiment, Vxmt is plus 51/2 volts and the MOS operating voltage is plus 5 volts. The initial circuitry to which reference voltage Vxmt, signal BBS and signal BBS- are applied, together with the various phase signals, provide level shifting, shaping and a first order filtering function. In referencing the field effect transistor components, the electrodes will means, interchangeably, the source and drain, with the third connection being the gate. Voltage level Vxmt is applied to one electrode of transistor T2 which is gated by the signal phi 1LQ. The other electrode of transistor T2 is connected to one electrode of transistor T8 and to one terminal of each of capacitors C3 and C4. The other electrode of transistor T8 is grounded and it is gated by the signal phi 2LQ. The other terminal of capacitor C4 is connected to one electrode of transistor T4 whose other electrode is grounded and whose gate has applied to it signal BBS-. The other terminal of capacitor C4 is also connected to one electrode of transistor T6 which is gated by signal BBS and whose other electrode is connected to one terminal of capacitor C8 and to one electrode of each of transistors T12 and T10. The other terminal of transistor T10 is grounded and it is gated by signal phi 2. The other terminal of transistor T12 is connected to the negative input to operational amplifier 266 and to one terminal of capacitor C7. The positive terminal of operational amplifier 266 is grounded. Its output is connected, through capacitor C7, to its negative input. The output of operational amplifier 266 is connected to one electrode of transistor T14 which is gated by signal phi 1 and whose other electrode is connected to one electrode of transistor T16 whose other electrode is grounded and which is gated by signal phi 2. The section just described in detail provides for shifting the voltage reference level to Vxmt and, for mixing, through the introduction of signals phi 1, phi 2, phi 2LQ and phi 1 LQ. Shaping is accomplished through the introduction of signals BBS and BBS- for notching the square wave input to cause that square wave to more closely resemble a sine wave. Filtering is accomplished through operational amplifier 266 and associated components, in a conventional manner, to produce the signal Qfout. This is the quadrature phase signal. The in phase signal development is shown above in a circuit that is identical to that described, except that signals phi 1 LI and phi 2 LI are applied in place of phi 1LQ and phi 2LQ. The output signal is Ifout. Referring again to signal Qfout, it is applied to two biquadratic filter sections, in series made up of operational amplifiers 267 and 268, and 269 and 272, respectively, and their associated circuits, as shown. The in phase section filter 270 is identical in structure to quadrature phase filter 271. The use of the in phase and quadrature phase technique is known in the prior art and is a comparatively straight-forward method for providing resultant output signals Iout, mixed with the cosine of the carrier and Q out, mixed with the sine of the carrier. Signal Iout is connected to one electrode of transistor T48 which is gated by signal phi 2I, with the other electrode connected to one terminal of capacitor C27 and one terminal of capacitor C28, as well as to one electrode of transistor T49. Transistor T49 is gated by signal phi 1I and its other electrode is grounded. The other terminal of capacitor C28 is connected to one electrode of each of transistors T46 and T47. Transistor T46 is gated by signal SGT and its other electrode is grounded. Transistor T47 is gated by signal CGT and its other electrode is connected to the other terminal of capacitor C27 where it is also connected to one terminal of capacitor C37 and to one terminal of transistor T51 which is gated by signal phi 2 and whose other electrode provides the negative input to operational amplifier 273. Signal Qout is connected to one electrode of transistor T41 which is gated by signal phi 2Q and whose other electrode is connected to one terminal of capacitor C29 and to one terminal of capacitor C31. It is also connected to one electrode of transistor T42 which is gated by signal phi 1Q and whose other electrode is connected to ground. The other terminal of capacitor C29 is connected to one electrode of transistor T45 which is gated by signal CGT and whose other electrode is grounded. The one terminal of capacitor C29 is also connected to one electrode of transistor T44 which is gated by signal SGT and whose other electrode is connected to one electrode of transistor T43 which is gated by signal phi 1 and whose other electrode is grounded. Transistor T42 has one electrode grounded, is gated by signal phi 1Q and has its other electrode connected to the other electrode of transistor T41. Signal PSKOUT is developed as an output of operational amplifier 272 and its associated circuitry. That is, capacitors C32-C36 are connected together at one end to one electrode of transistor T47 which is gated by signal phi 2 and whose other electrode is connected to the other terminal of capacitor C37 and to one electrode of transistor T48. The other electrode of transistor T48 is grounded and it is gated by signal phi 1. The other terminals of capacitors C32-C36 are connected to switches SW1-SW4, respectively. The other terminals of switches SW1-SW4 are connected to the negative terminal of operational amplifier 273. They are also connected to one terminal of capacitor C36 whose other terminal is connected to the output of operational amplifier 273. Switches SW1-SW4 are selectively activated by signals VLT, VHT, BLT and BHT, respectively. Signal VLT is the NOR function of signals VAD- and A/O-; signal VHT is the NOR function of signals VAD- and A/O; signal BLT is the NOR function of signals VAD and A/O; signal BHT is the NOR function of signals VAD and A/O-. By appropriate selections, the proper amount of capacitance is switched in and out of the circuit to adjust for the modem protocol selected and to thereby provide an appropriate PSKOUT signal. It should also be noted that the shape of PSKOUT is dependent upon signals VGT being applied to the cosine and signal SGT being applied to the sine of the carrier. Simply switching capacitors in and out of a multiple mode circuit, such as that represented by operational amplifier 273 and its associated capacitors and switches to obtain desired filter characteristics may cause parasitic capacitance problems. Referring to FIG. 17, a circuit is illustrated which eliminates the parasitic capacitance problem. FIG. 17 illustrates an operational amplifier providing signal Vout with its positive terminal grounded and capacitor C placed across its output and its negative input. Its negative input is in series with transistor T51 which is gated by signal phi 2 and whose opposite electrode is connected to one terminal of capacitor K1C, one electrode of transistor T52 and one electrode of each of transistors T59, T60 and T61. The signal Vin is applied to one electrode of transistor T55 which is gated by signal phi 1 and whose other electrode is connected to one electrode of transistor T54, to the other terminal of capacitor K1C and to one terminal of each of capacitors K2C, K3C and K4C. The other terminals of capacitors K2C, K3C and K4C are connected to the other electrode of transistors T61, T60 and T59, respectively. Transistors T59, T60 and T61 are gated by signals M4, M3 and M2, respectively. Connected to a voltage reference between each of capacitors K4C and transistors T59, K3C and transistor T60, capacitor K2C and transistor T61 are transistors T56, T57 and T58. They are gated by signals M4-, M3- and M2-, respectively. Note the similarity of the circuit of FIG. 17 with that of FIG. 16A at the output. It is evident that, for example, switch SW1 is equivalent to the combination of transistors T56 and T59 with signals M4- and M4 being applied, respectively, as gate signals. Without transistors T56-T58, the problem occurs when one or more of the capacitors are switched out of the circuit. A switching transistor used may have a very high off resistance, but still have a small capacitance between its terminals. This capacitance is in series with the capacitor that is switched out. The combination of these two capacitors is seen in parallel with the other capacitors that are switched into the circuit. This results in an effective capacitance that is larger than desired. Resulting filter response will be skewed from the desired response. This problem is magnified by the number of capacitors that are switched out of the circuit. When a capacitor is switched out of the circuit as shown in FIG. 17, then the appropriate transistors T56-T58 are gated and the switched out capacitors are connected to the reference voltage. Now the switched out capacitors appear as stray capacitance between the reference voltage and the stray insensitive nodes of the stray insensitive switched capacitor filter. Therefore, the switched out capacitors no longer skew the filter response. This switching arrangement is used throughout the switched capacitor circuitry of this invention. Signal PSKout may now be sent directly to the transmission line. However, it has been determined that both magnitude and phase are distored by transmission through a telephone line which is commonly used as the transmission line. To compensate for this distortion, further filtering is performed. In this preferred embodiment, the signal PSKout is applied to a biquadratic filter made up of operational amplifiers 274 and 275 and associated circuitry as shown in FIG. 18. All of the switches SW6-SW23 are of the type described with reference to FIG. 17. Input capacitors C38 and C39 are switched into the circuit by signal ALOT which is the NOR combination of an analog loop back signal and a test signal. These capacitors provide for switching in the proper amount of gain for testing purposes. The capacitor between the output of the filter and switches SW6 and SW7 that receive the ALOT signal is selected from capacitors C59-C62 by selection of switches SW8-SW11, activated by signals VLT, VHT, BLT, and BHT, respectively. In a similar manner, capacitors C47-C50 are selected by switches SW16-SW19, respectively, activated by signals VLT-BHT, as above, and are placed in the circuit between the input and output of operational amplifier 274. Still another bank of parallel capacitors C43-C46, switched by switches SW12-SW15, respectively, are connected between the output of operational amplifier 274 and the input of operational amplifier 275. Further, another bank of capacitors C51-C54, switched by switches SW20-SW23 is across the output of operational amplifier 275 and its negative input. Capacitors C55-C58 are connected in series with capacitors C47-C50, respectively, to the output of operational amplifier 275 whose output signal is VO2T. This filter described above is for magnitude equalization of the telephone line signal magnitude deterioration. At this point, signal VO2T could be transmitted directly down a telephone line. However, it has been found that the phase is distorted as well as the magnitude and in this preferred embodiment, signal VO2T is input to another biquadratic filter having banks of capacitors and switches as described above, for compensation of phase distortion. FIG. 19 illustrates this particular arrangement with signal VO2T providing an input to a quadratic filter made up of operational amplifiers 276 at the input and 277 at the output, and associated circuitry as shown. Signal VO2T is the input signal from the output of the magnitude equalization filter of FIG. 18. Capacitors C67-C69 are selectively switched into the circuit through switches SW25-SW27 through selection signals VHT, BHT and LBANDT. Signal LBANDT is the negation of the NOR function of signals BLT and VLT-. One terminal of each of capacitors C64, C65 and C66 is connected in series with capacitors C67, C68 and C69, respectively. The other terminals of capacitors C64-C66 are connected together to one electrode of each transistors T78 and T79. Transistor T79, gated by signal phi 2 has its other electrode grounded. Transistor T78 is gated by signal phi 1 and is connected to one terminal of each of capacitors C82-C89 and to the output of operational amplifier 277. The other terminal of capacitor C82 is connected to switch SW36 which is activated by signal L band T-. The other terminal of switch 36 is connected to one terminal of each of switches 28-31 which are activated by signals VLT, VHT, BLT, and BHT, respectively. Capacitors C71-C74 are selected by switches SW28-SW31, respectively. Capacitor C75 is connected in parallel with the combination of switches and capacitors, from the output of operational amplifier 276 to the negative input. The positive input is grounded. The other terminal of capacitor C83 is connected between switch SW28 and capacitor C71. The other terminal of capacitor C84 is connected between switch 30 and capacitor C73. Another bank of parallel capacitors, C77-C81 is positioned between the output of operational amplifier 276 and the input of operational amplifier 277. Switches SW32-SW35, activated by signals VLT, VHT, BLT, and BHT, respectively, are connected to switch capacitors C77-C80, respectively. Still another bank of parallel capacitors C85-C89 are located across the output of operational amplifier 277 and the negative input. The positive input is grounded. Capacitors C85-C89 are selected by switches SW37-SW40 with signals VLT, VHT, BLT and BHT being the respective activating signals. Capacitors C90-C93 are connected in parallel, with one terminal of each being tied together back to the circuit input, the other terminals being connected between switches SW37-SW40 and capacitors C85-C88, respectively. The output terminal TXALB is now ready for transmission down the telephone line, having been adjusted for both magnitude and phase distortion. The waveforms shown in FIG. 20 generally describe the transformation from digital to analog signals. Signal DIN1 is shown as a series of digital bits. The bit pattern shown is 1, 0, 1, 1, 0, 0. The shaper input illustrates the desired waveform mentioned above where the top of the digital waveform is notched. Note that in the series of two 1's and two 0's, that there is a notch between the two 1's and between the two 0's rather than an uninterrupted level. Signal PSK out is the filtered, analog equivalent of digital signal DIN1 before being compensated for amplitude and phase deterioration. RECEIVER Signal RCVA is received at the other end of the transmission line and represents signal TXPA after transmission and attendant distortion. Signal RCVA is received in the receive filters Section 18 shown in FIG. 1. Certain control signals are utilized in the receive filters. For an understanding of the origin of those signals, please refer first to FIG. 21. Pseudo random shift register 290 is shown being made up of PLA 300, flip flops FF121-FF127 and associated logic. A 4 MHz clock is applied to clock generator circuit 293 which is identical to clock generator circuit 33 described earlier. Clocks 1 and 2 from clock generator circuit 293 are used to clock dynamic flip flops FF121-FF127. The Q output of flip flop FF121 provides the D input to flip flop FF122 and so on. All of the Q and Q- outputs of the flip flops provide inputs to PLA 300. The preset terminal of flip flop FF121 is grounded and signal R20 applies the preset input to the remaining flip flops FF122-FF127. Clock signal CK1 from clock generator circuit 293, in addition to providing the input clock to the flip flops also provides an output signal FRCK1. The Q outputs of flip flops FF126 and FF127 provide the two inputs to AND gate 285 and the Q- outputs from those flip flops provide the two inputs to AND gate 286. AND gates 285 and 286 provide the inputs to NOR gate 287 whose output provides one input to NOR gate 289. Signals VAD-, L band- and ALT provide the inputs to NOR gate 288 whose output provides one input to AND gate 291. Signal R20 provides the other input to NOR gate 289 and the other input to AND gate 291. The output of NOR gate 289 and the output of AND gate 291 provide the inputs to NOR gate 292 whose output provides the D input to flip flop FF121. The selected lines of PLA 300 provide signals S11-S13, R11-R13, S21-S23, R21-R24. FIG. 22 illustrates a simple circuit wherein signals R21-R24 provide the inputs to NOR gate 302 whose output is inverted by inverter 303, providing output signal R20. R20 then is a 1 whenever any of signals R21-R24 is one. Signal R20, as indicated above, provides the preset to six of the flip flops and gating to two gates. FIG. 23 illustrates the development of signal theta 7. Signals S11-S13 and S21-S23 provide the inputs to OR gate 294. Signals R11-R13 and R20 provide the inputs to OR gate 295. The outputs of OR gates 294 and 295, together with signal FRCK1 from FIG. 21 provide the two inputs to each of AND gates 296 and 297, respectively. These two AND gates provide inputs to cross coupled NOR gates 298 and 299 whose output is inverted through inverter 301, providing output signal theta 7. FIG. 24 illustrates the development of signals ALT and phi 3 and phi 4. Signal theta 7 clocks toggle flip flop FF129 whose Q output clocks toggle flip flop FF128. The Q output of flip flop FF 128 is signal ALT. The Q-outputs of flip flops FF128 and FF129 provide the inputs to NOR gate 306 whose output is applied to NOR gate 309, and through inverter 307, to NOR gate 308. NOR gates 308 and 309 are cross coupled with the output of NOR gate 308 providing a gate signal for transistor T91. The output from NOR gate 309 provides the gate signal for transistor T92. The one electrode of transistor T91 is connected to voltage VDD and the other electrode of T91 is connected to one electrode of T92, the other electrode of T92 being connected to ground. A connection between the electrodes of transistors T91 and T92 provides signal phi 3. The circuitry developing signal phi 4 is identical to that for developing phi 3 except that the input signals are signal ALT and Q- from flip flop FF129. FIG. 25 illustrates the development of signals phi 7 and phi 8. Signal theta 7 is inverted through inverter 311 and provides one input to NOR gate 314. The output of inverter 311 is inverted through inverter 312, providing one input to NOR gate 313. NOR gates 313 and 314 are cross coupled, with NOR gates 314 providing output signal phi 7 and NOR gate 313 providing output signal phi 8. FIG. 26 illustrates the input to the receiver section of the modem. Input signal RCVA is impressed on continuous anti-alias filter 18A. Filter 18A is made up of series resistor R1 which is in series with the parallel combination of resistor R2 and capacitor C101. The other end of resistor R2 provides the positive input to operational amplifier 281. The other terminal of capacitor is connected to the negative input to comparator 281. The positive input of comparator 281 is tied to AC ground, through capacitor C102. The output of comparator 281 is tied back to its negative input and provides the output signal CAAF. Signal CAAF provides the input to samples anti-alias filter 18B. Signal CAAF is applied to one electrode of transistor T81 which is gated by signal phi 7 and whose other electrode is connected to one terminal of capacitor C103 and to one electrode of transistor T82. Transistor T82 is gated by signal phi 8 and its other electrode is grounded. The other terminal of capacitor C103 is connected to one electrode of transistor T83 and one electrode of transistor T80. Transistor T80 is gated by signal phi 8 and its other electrode is grounded. Transistor T83 is gated by signal phi 7 and its other electrode is connected to one negative input of operational amplifier 282. The positive input of operational amplifier 282 is grounded and its output is connected through capacitor C104 to its negative input. Its output is also connected to one electrode of transistor T84 which is gated by signal phi 7 and whose other electrode is connected to one terminal of capacitor C105 and to one electrode of transistor T85. Transistor T85 is gated by signal phi 8 and its other electrode is connected to ground. The other terminal of capacitor C105 is connected to one electrode of transistor T86 and one electrode of transistor T87. Transistor T86 is gated by signal phi 7 and its other electrode is grounded. Transistor T87 is gated by signal phi 8 and its other electrode is connected to the negative terminal of operational amplifier 283 and also is connected, through capacitor C109 to the input of sampled anti-aliasing filter 18B. The positive input of operational amplifier 283 is grounded and its output provides signal RCVIN. Its output is also connected through capacitor C108 to its negative input and through capacitor C107 to the negative input of operational amplifier 282. The output of operational amplifier 283 is also connected to one electrode of transistor T89 which is gated by signal phi 7 and the other electrode is connected to one electrode of transistor T88 and through capacitor C106 to the one electrode of transistor T80. Transistor T88 is gated by signal phi 8 and its other electrode is grounded. In summary, the development of signal RCVIN comes about by filtering through a continuous and a sampled low band pass filter combination. These filters act as anti-aliasing filters to prevent high frequency signals from entering. Signal RCVIN provides an input to PSQ receiver filter 18C. Also, signals LBANDR and LBANDR- provide input signals. Also input are signals BLR, BHR, VLR and VHR, as shown, as are the output signals from NOR gates 316-319. Signals VAD and EFA/0 provide the inputs to NOR gate 316. Signals VAD and EFA/0- provide the inputs to NOR gate 317. Signals VAD- and EFA/0- provide the inputs to NOR gate 318. Signals VAD- and EFA/0 provide the inputs to NOR gate 319. Control signals SM0-5 and LG1-5 also provide inputs to PSK receive filter 18C. Referring now to FIG. 27A and FIG. 27B, the development of signals SM0-5 and LG1-5 is shown. Signals SM0 and 6/1 are input to OR gate 321 whose output provides one input to NOR gate 322. Signal LOW6 provides the other input to AND gate 322, whose output provides one input to AND gate 325. The other input to NOR gate 325 is signal DOWN1-, from the automatic gain control circuit. The output from NOR gate 325 is signal M DOWN1. An identical circuit made up of OR gate 324, AND gate 326 and NOR gate 327 having signals SM5, 6/1 for OR gate 324, signal LG5 and the output of OR gate 324 as input for AND gate 326 and finally signal UP1 and the output of AND gate 326 for NOR Gate 327 to develop signal M UP 1. Circuits 328-333 are identical, each being made up of two 2-input AND gates, the outputs of which provide the inputs to a NOR gate. Signal M UP 1 provides one input to each of circuits 328-333. Signal M DOWN 1 provides one input to each of circuits 328-333. The output of circuit 328 provides the K input to flip flop FF131 and, inverted, the J input to that flip flop. The Q output of flip-flop FF131 provides another input to circuit 329. The Q- output provides another input to circuit 330. The output of circuit 329 provides the K input to flip flop FF132. NOR gate 336 provides the J input to flip flop FF132. One input to NOR gate 336 is provided by the output from circuit 329. The other input to NOR gate 336 is provided by NOR gate 334 which has three inputs. One input is provided by the Q- output of flip flop FF131, another input is provided by the Qoutput of flip flop FF132, the third input is provided by the Q- output from flip flop FF133. The Q output of flip flop FF132 provides another input to circuit 328. It also provides another input to circuit 330. Circuit 330 provides the K input to flip flop FF133 and, inverted, provides the J input to that flip flop. The Q output of flip flop FF133 provides the another input to circuit 329. The Q output of flip flop FF133 provides another input to circuit 328. Signal STEP and signal 6/1 provide inputs to NOR gate 337, the output of which clocks flip flops FF131-FF133. Signal 6/1 provides the preset for those three flip flops. Signals DOWN and SM0 provide inputs to AND gate 361. Signals SM5 and UP provide the inputs to AND gate 362. Signal 6/1 provides an input to OR gate 363 whose other two inputs are provided by the outputs of AND gates 361 and 362. The output of NOR gate 363 is NORed with the STEP signal, NOR gate 338 providing the clock signals for flip flops FF134-FF136. The output of circuit 331 is connected to the J input of flip flop FF134 and, through an inverter, to the K input. The Q output of flip-flop FF134 is connected to the input of circuit 333 and the 2-output is connected to the first AND gate of circuit 332. The output from circuits 332 is connected to the J input of flip flop FF135. It also provides one input to NOR gate 341, whose other input is provided by NOR gate 339. The inputs to NOR gate 339 are provided by the Q output from flip flop FF134, the Q- output from flip flop FF135, and the Q output from flip flop FF136. The Q- output of flip flop FF135 is also connected to the input of circuit 331 and to the input of circuit 333. The output of circuit 333 is connected to the J input of flip flop FF136 and, inverted, to its K input. The Q output of flip flop FF136 is connected to the input of circuit 331 and the Q- output is connected to the input of circuit 332. Flip flops FF131-FF133 and associated logic described form part of a counter for counting either up or down in 1 DB increments, depending upon the output of the automatic gain control circuit to be described later. The output of this counter is realized through NOR gates 342-347 which produce output signals SM0-SM5. The inputs to NOR gate 342 are the Q- outputs of flip flops FF131, FF132 and FF133. The inputs to NOR gate 343 are the Q output of flip flop FF131 and the Q- outputs of flip flops FF132 and FF133. The inputs to nor gate 344 are the Q outputs of flip flops FF131 and FF132 and the Q- output of flip flop 133. The inputs to NOR gate 345 are the Q outputs of each of flip flops FF131-FF133. The inputs to NOR gate 346 are the Q output from flip flop FF131 and the Q outputs from each of flip flops FF132 and FF133. The inputs to NOR gate 347 are the Q- outputs from flip flops FF131 and FF132 and the Q output from flip flop FF133. Flip flops FF134-FF136 and the associated logic circuitry form part of a second counter which counts in 6 DB increments, either up or down, depending upon the inputs from the automatic gain control circuit. The output of the 6 DB counter is represented by signals LG1-LG5 and LOW6. Signals LG1 and LG2 come from NAND gates 352 and 353, respectively; signal LG3 comes from inverter 354; signals LG4 and LG5 come from NOR gates 355 and 356, respectively; signal LOW6 comes from NOR gate 357. In addition, the STEP signal mentioned above is generated from NAND gate 376 whose inputs are signals Phi UP from the AGC and the output of OR gate 375. Inputs to OR gate 375 are signals UP and DOWN from AGC circuit 19, and the output from NOR gate 373, inverted. The inputs to NOR gate 373 are the output from NOR gates 339 and 348. The inputs to NOR gate 348 are the Q- outputs of flip flops FF134 and FF136 and the Q output of flip flop FF135. NAND gate 352 has inputs Q- from each of flip flops FF-134 and 136. NAND gate 353 has inputs from the Q- outputs from flip flops FF135 and 136. Inverter 354 inverts the output from Q- of flip flop FF136. NOR gate 355 has inputs from the Q output of flip flop FF134 and the Q- outputs of flip flop 136. NOR gate 356 has inputs from the Q output of flip flop FF135 and the Q- output of flip flop FF136. NOR gate 357, which provides output signal LOW6, has inputs from the Q output of flip flops FF134, FF135, and FF136. This second counter counts up and down in 6 DB increments, depending upon the signaling from AGC circuit 19. Therefore, depending upon the adjustment required, the increments of change will be up or down in either 1 or 6 DB steps. FIG. 28 illustrates the development of signal 6/1 and signal EDT. Signal XORAGC, from the AGC circuit, is inverted and provides one input to AND gate 366. The other input to AND gate 366 is signal LG5, inverted. AND gate 366 provides one input to cross coupled NOR gate 368 whose output is signal 6/1 and which is an input to NOR gate 369. When signal 6/1 is high, the 6 DB counter is activated in exclusion of the 1 DB counter. Another input to NOR gate 368 is the output from NOR gate 369. Signal SM5 is inverted and applied as one input to NOR gate 367. Signal LG5 is inverted and applied as another input to NOR gate 367 and finally, signal UP1- (from AGC circuit) is applied as an input. The output of NOR gate 367 provides one input to NOR gate 369. Another input to NOR gate 369 is provided by signal BSRST from the 6 DB increment counter. Finally, a power up circuit provides the fourth input to NOR gate 369. Signal MKEDT- is applied as one input to NOR gate 371 and the output of NOR gate 368, signal 6/1 is applied as the other input. The output of NOR gate 371 is signal EDT. FIGS. 29A-29E, when placed together as indicated, form a schematic of the receive filter. The telephone line, which is the ordinary transmission line over which a phase shifted signal is transmitted, has a pair of lines, one low band and one high band. For example, a low band frequency may be in the order of 1200 bps with a high band being in the order of 2400 bps. If the low band is received, there will be high band pickup. In addition, the telephone lines distort the signal in amplitude and phase. The purpose of the receive filter is to remove the unwanted high band or low band interference and to compensate for distortion in the telephone line. It will be remembered that on the transmitter end, an equalization circuit is used to minimize distortion. On this receive end, the distortion is further minimized by the receive filter. The receive filter of FIGS. 29A-29E is made up of seven sections, each section being a biquadratic filter similar to those discussed earlier. The difference in one biquadratic filter from another is in its transfer function. Also, in these particular filter sections, it will be noted that capacitance may be switched in and out of the circuit to adjust frequency response for the modem protocol and the gain determined by the AGC control 19. Turning first to FIG. 29A, signals BHR and VHR are shown as inputs to NOR gate 389 with the resultant signal LBANDR, which is used throughout this filter. Switch SW40, which is identical to the switch of FIG. 17, is controlled by signal LBANDR to isolate the filter when desired. Signal RCVIN is switched into the circuit through switch 40. Section one provides for selective introduction of capacitors C117-C119 through switches 42 and 43 to the input of operational amplifier 386. The introduction of capacitors C125 and C126 by switches SW46 and SW47, activated by signals VLR and BLR, respectively, to the input of operational amplifier 387 vary the capacitance depending upon protocol. The output signal from operational amplifier 387 is VO1 which proceeds, through transistor switch T111 into the second section of the filter which is made up of operational amplifiers 388 and 390 and the associated circuitry. Capacitor switch network 370 includes four capacitor switch circuits in parallel: switch SW53, activated by signal VHR, and capacitor C111; switch SW54, activated by signal BHR, and capacitor C112; switch SW55, activated by signal VLR, and capacitor C113; switch SW56, activated by signal BLR, and capacitor C114. This grouping will be used as an example for further discussion of the filter and provides a variable capacitance input to operational amplifier 388, dependent upon the protocol. Signals VHR, BHR, LBANDR, switch in capacitors C133, C134, and C135 by switches SW48, SW49 and SW50, respectively. Capacitors C141-C143 are switched in by signals LBANDR-, VLR, and BLR, through switches SW57, SW58, and SW59, respectively. These capacitor banks provide additional, selectable, capacitance to the operational amplifier 390, whose output is signal VO2. Signal VO2 is switched into the third section of the filter as shown in FIG. 29D. Signal VO2 enters through transistor T115 with a capacitor-switch arrangement (identical to capacitor switch circuit 395 of FIG. 29A) being positioned across the two operational amplifiers 391 and 392 of the biquadratic filter of this section three. Positioned between amplifiers 391 and 392 is another capacitor switch network 398 identical to capacitor switch network 395 except for values of capacitance. Finally, still another similar capacitor switch network 397 provides capacitance to be added to fixed capacitance at the input to operational amplifer 391. The output from operational amplifier 392 is signal VO3 which, as shown, is introduced into the fourth section of this filter. Signal VO3 is switched into section four through transistor T116. This section has capacitor switch network 377, similar to capacitor switch network 370, except for the capacitor values, connected to the input of operational amplifier 394. Another such network 376 is connected from the input of operational amplifier 393 to the output of differential amplifier 394. Still another similar network 375 is connected between the output of operational amplifier 393 and the input of operational amplifier 394. Additional capacitance to the input of operational amplifier 393 may be switched in through signals BHR, VHR, VLR and BLR in a network similar to that previously described for network 370. The output of this fourth section is signal VO4 which is shown entering the fifth section in FIG. 29E. Signal VO4 is shown entering through transistor T117 to provide an input to the section made up of operational amplifiers 401 and 402, and the associated switches and capacitors as shown. Connected to the input of operational amplifier 402 is a capacitor switch network 382 similar to 370, except for the capacitor values. Likewise, capacitor switch network 383, similar to network 370, is connected from the input of operational amplifier 401 to the output of operational amplifier 402. Connected between operational amplifiers 401 and 402 is another similar network 385. Finally, similar network 384 is connected to the input of operational amplifier 401. The output from operational amplifier 402 is signal VO5 shown entering the sixth stage of the filter in FIG. 29B. Signal VO5 enters the sixth stage through transistor T118, the sixth stage being made up of operational amplifiers 403 and 404, and associated switches and capacitors. It should be noted that in this section, capacitors are switched in and out of the circuit in similar capacitor switch networks as discussed earlier, but with the switches being activated by signals SM0-SM5 from the 1 DB step counter of FIG. 27A. Capacitor switch network 406 has six capacitors, each switched in by switches activated by signals SM0-SM5, respectively. This network serves as an input to operational amplifier 403. Also connected to that input is capacitor switch network 381, again similar to capacitor switch network 370. Connected across operational amplifier 403 is capacitor switch network 380, similar to network 370. Similar network 379 is connected between operational amplifiers 403 and 404. Switch capacitor network 378 is connected across operational amplifier 804. Switch capacitor network 407, controlled by signals SM0-SM5, similar to network 406, is connected to the input of operational amplifier 404. The output of operational amplifier 404 is signal VO6. It can be seen that the response of this filter is dependent not only on the protocol, but also on the automatic gain control circuit 19 which controls the counter providing signals SMO0-SM5. FIG. 29C illustrates the seventh and final section of this receive filter with signal VO6 being introduced into the filter by way of capacitor switch network 410. Network 410 is made up of five capacitors switched in by signals LG1-LG5, respectively, to the input of operational amplifier 408. A similar network 411 is connected between the input to network 410 and the output of operational amplifier 409. Switch capacitor network 373, similar to network 370, is across operational amplifier 408. Network 374, similar to network 370, is connected between operational amplifiers 408 and 409. Capacitor switch network 372, also similar to network 370, selectively adds to the capacitance of network 373. Network 371, similar to network 370, is connected, through transistor switches, to one side of each of networks 373 and 372. This section gain is controlled by signals Lg1-LG5, from the counter controlled by the AGC circuit 19, in addition to other gain components. The output signal FRCV has now been filtered and is now ready for phase shifting. FIG. 30, in block form, illustrates the automatic gain control and associated circuits. RCV filter 18A, 1 DB steps 18B and 6 DB steps 18C represent the seven step receive filter described above wherein receive filter 18A represents the first five sections, 1 DB steps 18B represents the sixth section, and 6 DB steps 18C represents the seventh section. The 1 DB steps and the 6 DB steps are controlled by logic 19E, the logic shown in FIGS. 27A and 27B that provides signals SM0-SM5 and LG1-LG5. Count 19D represents the counters for the 1 DB and 6 DB increments, also described in detail with respect to FIGS. 27A and 27B. The remainder of the automatic gain control system to be described is the rectifier 19A, filter 19B and comparator 19C. In this preferred embodiment, rectifier 19C is a negative rectifier so that if the incoming signal is above the voltage reference level, it is inverted. If the incoming signal is below that reference, it is not inverted. The rectified signal is filtered and then a comparison is made to determine whether the signal is too high or too low in magnitude. Depending upon that determination, an UP or DOWN signal is generated for use in the counters of FIGS. 27A and 28B. FIG. 31A illustrates the rectifier circuit made up of comparator 405 and operational amplifier 406, and associated circuitry as shown. In fact, the capacitors and transistor switches shown in connection with operational amplifier 406 provide a first order filtering function so that the output signal RECTD from operational amplifier 406 is rectified and somewhat filtered. From there it goes into section 19B shown in FIG. 31A being made up of operational amplifiers 407 and 408, forming a biquadratic section, together with the supporting circuitry shown. The output from operational amplifier 408 is signal BIFIL. The comparison circuitry 19C, for comparing the input signals with a standard and for causing the adjustment of those signals is shown in FIGS. 31A, 31B, and 31C. FIG. 31A illustrates clock phi 4 as one input to NOR gate 423. The other input is signal 6/1. NOR gate 422 has 6/1- and the Q output from flip flop FF143 applied. The output of NOR gates 422 and 423 are applied as inputs to NOR gate 424. This logic arrangement determines whether clock phi 4 or the output of flip flop FF143 will be used to clock flip flop FF144 which in turn clocks flip flop FF145. FIG. 28 illustrates the generation of signal 6/1 which is turned on by a lengthy carrier detect signal to speed up the automatic gain control. In either event, the Q output of flip flop FF145 provides one input to NOR gate 426 and the -Q output of that flip flop provides one input to NOR gate 427 with the other input for both of those gates being provided by the Q- of flip flop FF144. The outputs from NOR gates 426 and 427 provide inputs to latch circuit 428 which provide output signals phi DOWN and phi UP. The outputs also provide inputs to latch signal 429 whose output signals are phi DOWNA and phi UPA. Signal BIFIL, shown in FIG. 31B, is gated to the negative input of comparator 409 by signal phi DOWNA through transistor T123 or to the positive terminal of comparator 409 by signal phi UPA through transistor T124. Also tied to the negative input of comparator 409 is signal SM0, gated through transistor T122 by the output of NOR gate 432, or signal LOW6 gated through transistor T121 by the output of NOR gate 431. Signal phi UPA- is applied to both NOR gates 431 and 432. Signal 6/1 is applied to NOR gate 432 and signal 6/1- is applied to NOR gate 431. An identical circuit is connected to the positive terminal of comparator 409, except that signal phi DOWNA- is substituted for phi UPA-. Also, the signals applied are SM5 or LG5. These two circuits then determine the output of comparator 409 which is tied directly to latch circuit 411, whose other input is phi DOWN and inverted to latch circuit 412 whose other input is phi UP. Signal UP1 or UP1- is high from latch circuit 412 and signal DOWN1 or DOWN1- is high from signal 411, depending upon the states of those latches. The AND OR gate combination of AND circuits 413 and 414 and NOR circuit 415 provide signal XORAGCOUT which is input to NAND gate 416. The STEP-signal (the inversion of STEP signal of FIG. 27B) clocks flip flop FF150 whose Q output provides the other input to NAND gate 416. The output of NAND gate 416 provides the preset input to four toggle flip flops FF146-FF149 and also provides one input to NOR gate 418. Flip flops FF146-FF149 count and decide whether the gain changes detected are a long enough period to effect the change. The outputs of these flip flops, as shown, are applied to NOR gates 417 and 419. NOR gate 417 presets flip flop FF150 and NOR gate 419 provides another input to NOR gate 418. The third input to NOR gate 418 is signal phi UPA and the output clocks flip flop FF146 which in turn clocks flip flop FF147 and so on. The output from NOR gate 419 also provides one input to NAND gate 420 and one input to NAND gate 421. Signal MUP1 (from FIG. 27A) provides the other input to NAND gate 420, whose output is the signal UP-. Signal MDOWN1 (from FIG. 27A) provides the other input to NAND gate 421 whose output is signal DOWN-. The AGC circuit provides signals to increase or decrease the amplitude of the incoming signal if it is evident that a long enough period is involved and that the magnitude is signficant. The signals developed above are used in the circuitry for causing the actual amplitude changes in the receive filter 18. FIG. 32 is a block diagram of the digital voltage controlled oscillator used in the data recovery loop of the receiver. Divider 450 is shown having a 4 MHz input. Also, a protocol is input. Its end count signal RP6 is sent to variable divider 470 which provides divider 450 with reset signal VCLSR. The output of variable divider 470 provides an input to fixed divider 500 whose outputs are the mixer clocks for the cosine and sine functions. Fixed divider 500 also provides clock signals phi 52, phi 54 and phi 58 to count control 520. The six bit A/D converter 480 has loop error voltage FVIN input and provides six bits out to the count control 520. Count control 520 provides signals D, E and F to the variable divider 470 in response to the 6 bit input from the A to D converter 480. To complete the demodulation of the incoming signal, it must be split into two phases and next described will be that phase shifting circuitry. The control signals utilized in the phase shifter circuitry and not heretofore identified are described in FIGS. 33-35. FIG. 33 illustrates three flip flops FF151-FF153, with flip flop FF151 being clocked by phase phi 5 and the Q output of the flip flops clocking the succeeding flip flops. AND gate 430 receives the Q outputs from flip flop FF151 and signal LBANDR, inverted. AND gate 431 receives the Q output from flip flop FF152 and signal LBANDR. AND gates 430 and 431 provide the two inputs to NOR gate 432 whose output signal is CK125. In FIG. 34, it is seen that signal CK125 is the input signal to the circuit, shown directly clocking flip flop FF157 and, inverted, clocking flip flop 154. The Q output from flip flop 154 provides clocking for flip flop FF155 and connects to the J input of flip flop 157. The Q output from flip flop FF155 provides the J input to flip flop FF156 and the Q- output from flip flop FF155 provides the K input to flip flop FF156. The J and K inputs to flip flop FF157 are provided by the Q and Q- outputs of flip flop FF154, respectively. The Q- output of flip flop 155, inverted, provides the control signal W127. The Q- output of flip flop FF156, inverted, provides control signal W128. The Q output of flip flop FF157, inverted, provides signal CGR. Signal CGR, inverted, provides signal SGR. FIG. 35 illustrates the development of timing signals used in the phase shifting network. AND-NOR circuits 434-437 are made up of two input AND gates which provide the inputs to a NOR gate. Signal W128 provides one AND gate input to circuit 434 and one AND gate input to circuit 435. Signal W128- provides one input to circuit 434 and one input to circuit 435. Signal W127 provides one input to circuit 436 and one input to circuit 437. Signal W127- provides one input to each of circuits 436 and 437. Signal phi 5- provides one input to each of circuits 434-437. Likewise, signal phi 6- provides one input to each of circuits 434-437. The outputs from circuits 434-437 are clock signals phi 5S, phi 6S, phi 5C, and phi 6C, respectively. FIG. 36A illustrates pseudeo random shift register (divider) 450 for providing a fixed count for the digital voltage controlled oscillator of the receiver circuit. Clock generator circuit 449, identical to clock generator circuit 33, provides clocks 1 and 2, for clocking the eight flip flops, making up the shift register, as well as a ninth flip flop that resets the eight flip flops when signal VCLSR is received. Signals VAD and VAD-, and LBANDR and LBANDR- provide control for the PLA of shift register 450. The output signals from the PLA are SP5A, SP5B, RP5A, RP5B, SP6A, SP6B, and through NOR gate 451, the remaining four lines of the PLA, combined and inverted provide signal RP6. FIG. 36B illustrates the development of digital voltage controlled oscillator clock signals phi 5 and phi 6. Input signals SP5A and SP5B are input to OR gate 452, and input signals RP5A and RP5B are inputs to OR gate 457. The outputs from these two OR gates provide inputs to each of AND gates 453 and 456 which are clocked by signal VCLK1, all from the pseudo random shift register 450. The outputs of AND gates 453 and 456 provide inputs to cross coupled NOR gates 454 and 455, respectively, with the output of NOR gate 455 being inverted to provide signal phi 5. Signal phi 6 is generated as shown in a nearly identical manner. Signal RP6 provides one input directly to AND gate 463. Signals SP6A and SP6B provide inputs to OR gate 461 which provides an input to AND gate 462. Both AND gates 462 and 463 receive their second inputs from signal VCLK1. The output of NOR gate 458, which is cross-coupled with NOR gate 459 whose inputs, respectively, come from AND gates 462 and 463, provides clock signal phi 6. Clock signal phi 5 is the input in FIG. 32 for developing various other clock signals. FIG. 36C illustrates signals VCLK1 and RP6 as inputs to AND gate 464. Signal Q168 from flip flop FF168 of FIG. 36A provides one input to NOR gate 466, which is cross coupled with NOR gate 467, whose input is provided by AND gate 464. The output of NOR gate 466 is inverted and provides signal PSETX. FIG. 36D illustrates the digital voltage controlled oscillator three bit divider made up of flip flops FF187, FF188, and FF189. A 4 MHz input into clock generator circuit 465, identical to clock generator circuit 33, provides output clocks 1 and 2 for clocking the flip flops FF187-FF189. Exclusive OR circuit 468 provides the exclusive OR function of the Q and Q- outputs of flip flops FF188 and FF189, respectively, for providing the D input to flip flop FF187. The Q outputs of flip flops FF187, FF188, and FF189 provide one input to each of exclusive OR circuits 472, 473 and 474, respectively. The other input to each of exclusive OR circuits 472-474 are provided by signals D-, E-, and F-, respectively. The development of these signals follows. The output of each of these exclusive OR circuits 472-474 provide inputs to NOR gate 476 with one additional input being signal PSETX from FIG. 36C. The output of NOR gate 476 is signal VCLSR which is applied to the D input of flip flop FF168 of shift register 450, for presetting the eight flip flops forming that shift register. FIG. 37A illustrates the development of signal phi E, phi F, and phi F- for use in the analog to digital converter used in this receiver circuit. Signals APSET and phi 2V provide inputs to flip flop FF190 and NOR gate 467. The output of NOR gate 467 is inverted and clocks flip flop FF190 whose Q output provides one input to NAND gate 477. The other input to NAND gate 477 is signal phi 5. The output from NAND gate 477 provides one input to NOR gate 478 and, inverted, provides one input to NOR gate 479. NOR gates 478 and 479 are cross-coupled with NOR gate 478 providing output signal phi E and NOR gate 479 providing output signal phi F, and inverted, phi F-. FIG. 37B illustrate toggle flip flops FF175, FF176 and FF177 with flip flop FF175 being clocked by signal QF-, flip flop FF176 being clocked by the Q output of flip flop FF175 and flip flop FF177 being toggled by the Q output from flip flop FF176. The Q and Q- outputs from flip flops FF175-FF177 are combined in NOR gates 481-488 so that NOR gates 481-488 each represent a count of the combination of the three flip flops. The outputs of NOR gates 481-486 are signals AD6 and, inverted, AD6- through AD1 and AD1-, respectively. The output from NOR gate 487 is inverted to provide signal LTCH-. The output of NOR gate 488 is inverted to provide signal PSET-. FIG. 37C illustrates flip flops FF178-FF183 for storing a 6 bit binary representation of the analog voltage present. These flip flops are clocked by the output from NOR gates 501-506 whose inputs are provided by signals phi F, and AD6- through AD1-, respectively. The Q output of flip flop FF178 is inverted to provide signal AB6 and the -Q output is inverted to provide signal AB6-. The Q- output of flip flops FF179-FF183 are NORed with signals AD5, AD4, AD3, AD2 and AD1, respectively, to provide signals AC5- through AC1-. These signals are all inverted to provide signals AC5-AC1. Signals AB6 and DAOUT (from FIG. 37D) are exclusively ORed in circuit 498 and provide the K inputs to flip flops FF178-FF183, respectively, and inverted, provide the J inputs to those flip flops. Signals PSET- and phi F are input to NOR gate 496 whose output signal is APSET which, as indicated above, provides an input to the circuit of FIG. 37A and also provides the preset input to flip flops FF178-FF182. Signal LATCH- and signal phi F provide inputs to NOR gate 497 whose output provides an input to each of latch circuits 490, 491, 492, 493, 494 and 495, providing output signals S and S-, M4 and M4-, M3 and M3-, M2 nd M2-, M1 and M1-, M0 and M0-, respectively. If flip flop FF178 is set, then S=1 and S-=0, and so on. These output signals then represent the contents of the flip flops mentioned. FIG. 37D illustrates a capacitive divider network which is used for a successive approximation of a digital value of the input signal. As shown, signal VADH which, in this preferred embodiment is approximately +1 volt, and signal VADL, approximately -1 volt, provide the input the voltage references. Capacitors C160, KC160, K.sup.2 C160, K.sup.3 C160, K.sup.4 C160 are selected by signals AC1, AC2, AC3, AC4, and AC5, respectively, from the K bit counter circuit of FIG. 37B. Another bank of equal size capacitors are activated by the inversions of the signals mentioned above. In one case, the voltage is to be added, and in the other case, voltage is to be subtracted. The constant K is equal to 2 and therefore, if C160=1, then K.sup.4 C160=16. Since the capacitors are connected in parallel, they are additive and therefore the incoming voltage is divided by increments of 1/31, 2/31, 3/31, and so on. A comparison with voltage FVIN, the voltage on which the A to D conversion is done, is made through comparator 515 with the output signal DAOUT used to gate the exclusive OR circuit 498 of FIG. 37C. Therefore, the six flip flops FF178-FF183 count until a comparison is made. Then, on the seventh count from FIG. 37B, the contents of the flip flops are gated out into the latch circuits 490-495. On the eighth count, the flip flops are preset. FIG 38A and FIG. 38B combine to illustrate the development of signals D-, E-, and F- in circuit 520. FIG. 38A illustrates signals M2 and S, M1 and S, M0 and S, and their inversions, combined in exclusive OR gates 516, 517 and 518, respectively. The output from circuit 516 provides one input to NOR gate 526 whose other input is provided by signal phi 58. The output of circuit 516, inverted, provides one input to NOR gate 519 whose other input is provided by signal phi 52-. The output of circuit 517 provides one input to NOR gate 521 whose other inputs are provided by the output from NOR gate 519 and signal phi 54. The output of circuit 517, inverted, provides one input to NOR gate 524 whose other inputs is provided by signal phi 54-. The output of circuit 518 provides one input to NOR gate 522 whose other inputs are provided by the outputs from NOR gates 519 and 524, and signal phi 52. The output from NOR gates 526, 521, and 522 provide the three inputs to NOR gate 523 whose output is signal L/S. In FIG. 38B, signal L/S is seen as an input to PLA 525, both directly and invertd. Signals S and S-, M3 and M3-, M4 and M4- all are input signals to the PLA 525. The output signals from PLA 525 are signals D-, E- and F- which are the gating signals for exclusive OR circuits 472-474 of FIG. 36D. FIG. 39A and FIG. 39B, joined as shown, schematically illustrate the phase shifter, mixer, summing and general filtering circuits for providing output base band signals IC0 and IC9. Signal FRCV, the output of operational amplifier 409 shown in FIG. 29C provides the input to these circuits. Operational amplifier 442 has its positive terminal grounded and its negative terminal receives inputs from various combinations of capacitance based on the selected protocol and also to form an all-pass filter that does not affect the signal magnitude, only the phase, so that a zero phase shift is output from operational amplifier 442. On the other hand, operational amplifier 443 has its input arranged to shift the phase by 90.degree. at its output. The output of operational amplifier 442 is input to buffer filter 446 whose output is signal PS0. Operational amplifier 443 inputs buffer filter 447, and its output is signal PS9. Signals PS0 and PS9 enter identical mixer circuits, except for the application of gate signals for the 0.degree. and 90.degree. conditions. In FIG. 39A, for example, signal PS0 is shown gated by signal phi 6C through transistor T140. Signal SP9 is shown gated through transistor T133 by signal phi 6S. In FIG. 39B, signal PS0 is shown gated by signal phi 6S and signal PS9 by signal phi 5C. An examination of these circuits will illustrate their identity except for the phase of the gating signals. The mixing circuit shown in FIG. 39A store charge on capacitors C152-C154, and C156 and C157. Transistor T132, when gated by signal phi 6 transfers the charge from those capacitors to capacitor C155, thereby summing the signals. Operational amplifier 444, with the associated capacitors, provides a first order filter to produce output signal IC0, the in-phase I channel signal. In like manner, operational amplifier 445 provides output signal IC9, the quadrature phase, Q channel signal. FIG. 40 is a schematic diagram of buffer filter 446 (identical to buffer filter 447). The output of operational amplifier 442 provides the input to buffer filter 446 through capacitor C150, switched through T130 by signal phi 7, which is exactly four times the frequency of phi 3 and phi 4, to avoid any beating of frequencies. Operational amplifier 440 and its associated capacitors and switches provides a switched low pass filter. Operational amplifier 441 and its associated circuitry, including resistor R5, provides a continuous filter whose output is signal PS0. PS0 provides the input to the mixer circuits described above. FIG. 42A illustrates the development of clock signals phi C and phi D. The clock generator circuit 531 which is identical to clock generator circuit 33, receives a 4 MHz input providing clocks 1 and 2 out to clock flip flop FF191 whose Q and Q- output are used to clock flip flop FF192. The Q output of flip flop FF192 clocks toggle flip flop FF193. The Q output of flip flop FF193, inverted, is signal phi C and the Q- output, inverted, is phi D. FIG. 42B illustrates the development of clock signals P68 and P58. As shown, signal phi 58 directly provides one input to a cross coupled NOR gate, and inverted, provides one input to the other cross-coupled NOR gate with the NOR gates providing output signals P68 and P58, respectively. The clock pulses that were used in the transmission of the data to the receiver must be recovered so that accurate timing of the incoming data is assured. As in the transmission process, it is necessary that the data be clocked as close to the center as possible. A clock recovery loop is necessary to extract from the incoming data the clock used in the original transmission of the data. Following is a detailed description of the circuitry in the clock recovery loop. Turning first to FIG. 41, the generation of signals phi 5I, phi 6I, phi 5Q and phi 6Q is shown. Signal IR and signal DBR are applied as inputs to exclusive OR gate 530 which gates signal phi 5- through transistor T135 and, inverted, through transistor T137. Exclusive OR circuit 530 also gates phase phi 6- through transistor T136 and, inverted, through transistor T138. The connection between transistors T135 and T138, inverted, provides signal phi 5I. The connection between transistors T136 and T137, inverted, provides the signal phi 6I. Signals QR and DBR are combined in exactly the same way as described for signals IR and DBR, and as shown, to provide signals phi 5Q and phi 6Q. Signals IR and QR are illustrated in connection with the equalizer circuit to be described, and signal DBR is an output of the clock recovery loop circuit to be described below. FIGS. 42A and 42B are simple circuits illustrating the development of other conrrol signals for use in the clock recovery loop to be described. FIGS. 43A and 43B, joined as indicated, provides a schematic diagram of the clock recovery loop circuit for recovering the clock signal used in the transmission from the data transmitted. As indicated above, accurate recovery of the clock is essential for satisfactory data identification. Base band signals IC0 and IC9 from FIGS. 39A and 39B are received by a low band pass filter made up of operational amplifier 541 and its associated circuitry. Signal IC0 is switched in by signal phi 5I for charging capacitor C158 Signal phi 6I gates a transistor switch to ground from that capacitor. In a similar manner, signal IC9 is gated in through a transistor switch by signal phi 5Q to charge capacitor C154. Signal phi 6Q gates transistor switch connected to capacitor C157 and to ground. The other terminal of each of capacitors C157 and C158 are connected together to one terminal of capacitor C159, at which point the low pass filter becomes common to both IC0 and IC9. Signal phi 5 gates a transistor switch between the one terminal of capacitor C159 and the negative input to operational amplifier 541. Signal phi 6 gates a switch to ground from the one terminal of capacitor C159. The output of the first low pass filter from operational amplifier 541 is signal SUMIN. Signal SUMIN follows two parallel paths, one through an integrator circuit and the other through a by pass of the integrator circuit. The latter path specifically is gated by signal phi C through a transistor switch to one terminal of capacitor C167 and, from that one terminal, by signal phi D through a transistor switch to ground. Signal SUMIN is gated in to the integrator circuit by signal P58 to charge capacitor C161. Both sides of capacitor C161 are selectively grounded by signal P68. Signal P58 switches into the negative terminal of operational amplifier 542. The output of this integrator is the output of operational amplifier 542. When signal EDT is high, indicating no energy detected, it gates a transistor across operational amplifier 542 to disable it. Signal FILOT is switched by signal phi C to charge capacitor C165 which is connected to the previously mentioned capacitor C167 to sum the two signals providing signal SUMOT. The integrator circuit is relatively slow with respect to the entire phase lock loop for recovering the transmitted clock. If the output of operational amplifier 542 became railed, that is, its output equal either one of its inputs, then the phase lock loop would not operate. In the circuit thus far described, by providing the output signal SUMIN from operational amplifier 541 and summing it with the output signal FILOT, the low pass filter overrides the output of the integrator, even when the integrator is railed. Forming an analog divider at the output of operational amplifier 542, to divide the output by 5, in this preferred embodiment, the output of the low pass filter is capable of over-riding that of the integrater. In this case, the phase lock loop will lock. Then the integrator circuit is free to adjust itself relatively slowly to reduce the DC output of the low pass filter to zero, allowing the loop to free run in the absence of an input signal without losing lock. The divide by 5 circuit is a first order filter that receives the output of operational amplifier 542 and includes operational amplifier 543, and includes operational amplifier 543, and associated circuitry. The capacitors are selected to divide so that the first order filter has a gain of 1/5. Voltage reference VCRL2 (minus one volt in this preferred embodiment with respect to the system references) is applied through the transistor switches as shown by way of gating signals phi C and phi D and in combination with capacitor C170, and in combination with the input signal SUMOT sets the center frequency of the voltage controlled oscillator made up of operational amplifiers 544 and 545, comparator 546, flip flop FF195 and associated capacitors and transistor switches as shown. The voltage controlled oscillator (VCO) of this invention is a dual-integrator VCO in which, during any given cycle of oscillation, one half of the circuit is integrating toward the threshold voltage, while the other half is idle with its integrating capacitor fully discharged awaiting the next cycle. This is achieved through flip flop FF195, a toggle flip flop receiving its clock input from the output of comparator 546 signal CL16. The Q output of flip flop FF195 is connected to the gates of transistors T144 and T141. The Q- output of flip flop FF195 is connected to the gate of transistor 145 and transistor T140. Transistor T141 is connected across operational amplifier 545 and its feedback capacitor C167. Transistor T140 is connected across operational amplifier 544 and its feedback capacitor C166. Therefore, if flip flop FF195 is set and Q is high, then transistor T144 is turned on and transistor T141 is turned on. By turning on transistor T141, operational amplifier 545 is removed from the circuit. Therefore, the operational amplifier 544 and its associated components provide the VCO. The output of operational amplifier 544 passes through transistor T144 to the positive input of comparator 546. The negative input of comparator 546 is voltage reference VCRL3 (minus 3 volts in this preferred embodiment). The output of operational amplifier 544 ramps downward and when it reaches 25 volts, the comparator output toggle | ||||||
