View HSP45116A_1399769.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

tm 3-1 HSP45116A numerically controlled oscillator/ modulator the intersil HSP45116A combines a high performance quadrature numerically controlled oscillator (nco) and a high speed 16-bit complex multiplier/accumulator (cmac) on a single ic. this combination of functions allows a complex vector to be multiplied by the internally generated (cos, sin) vector for quadrature modulation and demodulation. as shown in the block diagram, the HSP45116A is divided into three main sections. the phase/ frequency control section (pfcs) and the sine/cosine section together form a complex nco. the cmac multiplies the output of the sine/ cosine section with an external complex vector. the inputs to the phase/frequency control section consist of a microprocessor interface and individual control lines. the phase resolution of the pfcs is 32 bits, which results in frequency resolution better than 0.013hz at 52mhz. the output of the pfcs is the argument of the sine and cosine. the spurious free dynamic range of the complex sinusoid is greater than 90dbc. the output vector from the sine/cosine section is one of the inputs to the complex multiplier/accumulator. the cmac multiplies this (cos, sin) vector by an external complex vector and can accumulate the result. the resulting complex vectors are available through two 20-bit output ports which maintain the 90db spectral purity. this result can be accumulated internally to implement an accumulate and dump filter. a quadrature down converter can be implemented by loading a center frequency into the phase/frequency control section. the signal to be down converted is the vector input of the cmac, which multiplies the data by the rotating vector from the sine/cosine section. the resulting complex output is the down converted signal. the bit position and widths for the outputs of cmac and complex accumulator (acc) are programmable. features nco and cmac on one chip 52mhz version 32-bit frequency control 16-bit phase modulation 16-bit cmac 0.013hz tuning resolution at 52mhz programmable rounding option spurious frequency components < -90dbc fully static cmos applications frequency synthesis modulation - am, fm, psk, fsk, qam demodulation, pll phase shifter polar to cartesian conversions block diagram ordering information part number temp. range ( o c) package pkg. no. HSP45116Avc-52 0 to 70 160 ld mqfp q160.28x28 phase/ frequency control section sine/ cosine section cmac sine/ cosine argument sin cos vector input vector output ri ri microprocessor interface individual control signals data sheet april 1999 file number 4156.3 caution: these devices are sensitive to electrostatic discharge; follow proper ic handling procedures. 1-888-intersil or 321-724-7143 | intersil and design is a trademark of intersil corporation. | copyright ?intersil corporation 2000
3-2 pinout 160 lead mqfp top view 2 1 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 119 120 118 117 116 115 114 113 112 111 110 109 108 107 106 105 104 103 102 101 100 99 98 97 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82 81 41 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 126 127 128 129 130 131 121 122 123 124 125 137 138 139 140 132 133 134 135 136 146 147 148 149 150 151 141 142 143 144 145 157 158 159 160 152 153 154 155 156 ro9 v cc ro8 ro7 ro6 ro5 gnd ro10 ro18 ro17 ro16 ro15 v cc ro13 ro12 ro11 gnd ro14 gnd io6 io5 io4 io3 io2 io1 v cc io0 ro19 gnd gnd ro2 ro1 ro0 det1 det0 gnd v cc ro4 ro3 nc clr ofr encfreg enpha c entireg eni modpi/2pi cs gnd clk v cc ad1 ad0 wr c15 c14 c13 c12 c11 c10 c9 c8 gnd c7 c6 c5 c4 c3 c2 c1 c0 nc outmux1 outmux0 gnd oer rnd oerext oeiext oei p a co mod1 v cc tico gnd rbytild mod0 p a ci lo ad pmsel nc peak rin0 v cc sh1 sh0 acc enphreg enofreg binfmt rin6 rin7 rin8 rin9 rin11 rin5 rin4 rin3 rin2 gnd rin12 rin17 rin18 imin0 rin16 rin15 rin14 gnd rin13 rin10 io7 io8 io9 v cc gnd io10 io11 io12 io13 io14 gnd v cc io15 io16 io17 io18 io19 imin18 imin17 imin16 imin15 imin14 gnd imin13 v cc imin12 imin11 imin10 imin9 nc imin8 imin7 imin6 imin5 imin4 imin3 imin2 gnd imin1 v cc rin1 42 HSP45116A
3-3 pin description name number type description v cc 22, 34, 50, 87, 95, 102, 111, 124, 132, 145, 159 - +5v power supply input. gnd 7, 20, 32, 48, 62, 73, 83, 92, 98, 108, 114, 119, 125, 131, 143, 157 - power supply ground input. c0-15 54-61, 63-70 i control input bus for loading phase and frequency data into the pfcs. c15 is the msb. ad0-1 51, 52 i address pins for selecting destination of c0-15 data. ad1 is the msb. cs 47 i chip select (active low). wr 53 i write enable. data is clocked into the input register selected by ad0-1 on the rising edge of wr when the cs line is low. clk 49 i clock. all registers, except the control registers clocked with wr, are clocked (when enabled) by the rising edge of clk. enphreg 27 i phase register enable (active low). registered on chip by clk. when active low, after being clocked onto chip, enphreg enables the clocking of data into the phase register. enofreg 28 i frequency offset register enable (active low). registered on chip by clk. when active, after being clocked onto chip, enofreg enables clocking of frequency offset data into the frequency offset register. encfreg 42 i center frequency register enable (active low). registered on chip by clk. when active, after being clocked onto chip, encfreg enables clocking of data into the center frequency register. enpha c 43 i phase accumulator register enable (active low). registered on chip by clk. when active, after being clocked onto chip, enpha c enables clocking of the phase accumulator register. entireg 44 i time interval control register enable (active low). registered on chip by clk. when active, after being clocked onto chip, entireg enables clocking of data into the time accumulator register. eni 45 i real and imaginary data input register (rir, iir) enable (active low). registered on chip by clk. when active, after being clocked onto chip, eni enables clocking of data into the real and imaginary input data register. modpi/ 2pi 46 i modulo /2 select. when low, the sine and cosine roms are addressed modulo 2 (360 degrees). when high, the most signi?ant address bit is held low so that the roms are addressed modulo (180 degrees). this input is registered on chip by clock. this control pin was included for fft processing. clr ofr 41 i frequency offset register output zero (active low). registered on chip by clk. when active, after being clocked onto chip, clr ofr zeros the data path from the frequency offset register to the frequency adder. new data can still be clocked into the frequency offset register; clr ofr does not affect the contents of the register. lo ad 38 i phase accumulator load control (active low). registered on chip by clk. zeroes feedback path in the phase accumulator without clearing the phase accumulator register. mod0-1 35, 36 i external modulation control bits. when selected with the pmsel line, these bits add a 0, 90, 180, or 270 degree offset to the current phase in the phase accumulator. the lower 14 bits of the phase control path are set to zero. these bits are loaded into the phase register when enphreg is low. mod1 mod0 phase shift (degrees) 00 0 01 90 1 0 270 1 1 180 HSP45116A
3-4 pmsel 39 i phase modulation select line. this line determines the source of the data clocked into the phase register. when high, the phase control register is selected. when low, the external modulation pins (mod0-1) are selected for the most significant two bits and the least significant two bits and the least significant 14 bits are set to zero. this control is registered by clk. rbytild 30 i rom bypass, timer load (active low). registered by clk. this input bypasses the sine/ cosine rom so that the 16 bit phase adder output and lower 16 bits of the phase accumulator go directly to the cmacs sine and cosine inputs, respectively. it also enables loading of the timer accumulator register by zeroing the feedback in the accumulator. p a ci 37 i phase accumulator carry input (active low). a low on this pin causes the phase accumulator to increment by one in addition to the values in the phase accumulator register and frequency adder. p a co 79 o phase accumulator carry output. active low and registered by clk. a low on this output indicates that the phase accumulator has over?wed, i.e., the end of one sine/cosine cycle has been reached. tico 33 o time interval accumulator carry output. active low, registered by clk. this output goes low when a carry is generated by the time interval accumulator. this function is provided to time out control events such as synchronizing register clocking to data timing. rin0-18 2-6, 8-19, 21, 23 i real input data bus. rin18 is the msb. this is the external real component into the complex multiplier. the bus is clocked into the real input data register by clk when eni is asserted. twos complement. imin0-18 1, 138-142, 144, 146-156, 158 i imaginary input data bus. imin18 is the msb. this is the external imaginary component into the complex multiplier. the bus is clocked into the real input data register by clk when eni is asserted. twos complement. sh0-1 24, 25 i shift control inputs. these lines control the input shifters of the rin and iin inputs of the complex multiplier. the shift controls are common to the shifters on both of the busses. acc 26 i accumulate/dump control. this input controls the complex accumulators and their holding registers. when high, the accumulators accumulate and the holding registers are disabled. when low, the feedback in the accumulators is zeroed to cause the accumulators to load. the holding registers are enabled to clock in the results of the accumulation. this input is registered by clk. binfmt 31 i this input is used to convert the twos complement output to offset binary (unsigned) for applications using d/a converters. when low, bits ro19 and io19 are inverted from the internal twos complement representation. this input is registered by clk. peak 29 i this input enables the peak detect feature of the block ?ating point detector. when high, the maximum bit growth in the output holding registers is encoded and output on the det0-1 pins. when the peak input is asserted, the block ?ating point detector output will track the maximum growth in the holding registers, including the data in the holding registers at the time that peak is activated. pin description (continued) name number type description sh1 sh0 selected bits 0 0 rin0-15, imin0-15 0 1 rin1-16, imin1-16 1 0 rin2-17, imin2-17 1 1 rin3-18, imin3-18 HSP45116A
3-5 outmux0-1 71, 72 i these inputs select the data to be output on ro0-19 and io0-19. ro0-19 84-86, 88-91, 93, 94, 96, 97, 99-101, 103-107, 109 o real output data bus. r19 is the msb. these three-state outputs are controlled by oer and oerext. outmux0-1 select the data output on the bus. io0-19 110, 112, 113, 115-118, 121-123, 126-130, 133-137 o imaginary output data bus. i19 is the msb. these three-state outputs are controlled by oei and oeiext. outmux0-1 select the data output on the bus. det0-1 81, 82 o these output pins indicate the number of bits of growth in the accumulators. while peak is low, these pins indicate the peak growth. the detector examines bits 15-18, real and imaginary accumulator holding registers and bits 30-33 of the real and imaginary cmac holding registers. the bits indicate the largest growth of the four registers. oer 74 i three-state control for bits ro0-15. outputs are enabled when the line is low. oerext 76 i three-state control for bits ro16-19. outputs are enabled when the line is low. oei 78 i three-state control for bits io0-15. outputs are enabled when the line is low. oeiext 77 i three-state control for bits io16-19. outputs are enabled when the line is low. rnd 75 i round enable. this input enables rounding of the output data precision from 9 to 20 bits. this input is active ?ow? this input must be tied either high or low. pin description (continued) name number type description out mux 1 out mux 0 ro16-19 ro0-15 io16-19 io0-15 0 0 real cmac 31-34 real cmac 15-30 imag cmac 31-34 imag cmac 15-30 0 1 real cmac 31-34 0, real cmac 0-14 imag cmac 31-34 0, imag cmac 0-14 1 0 real acc 16-19 real acc 0-15 imag acc 16-19 imag acc 0-15 1 1 reserved reserved reserved reserved det 1 det 0 number of bits of growth above 2 o 00 0 01 1 10 2 11 3 HSP45116A
3-6 functional description the numerically controlled oscillator/modulator (ncom) produces a digital complex sinusoid waveform whose amplitude, phase and frequency are controlled by a set of input command words. when used as a numerically controlled oscillator (nco), it generates 16-bit sine and cosine vectors at a maximum sample rate of 40mhz. the ncom can be preprogrammed to produce a constant (cw) sine and cosine output for direct digital synthesis (dds) applications. alternatively, the phase and frequency inputs can be updated in real time to produce a fm, psk, fsk, or msk modulated waveform. the complex multiplier/ accumulator (cmac) can be used to multiply this waveform by an input signal for am and qam signals. by stepping the phase input, the output of the rom becomes an fft twiddle factor; when data is input to the vector inputs (see block diagram), the ncom calculates an fft butterfly. as shown in the block diagram, the ncom consists of three parts: phase and frequency control section (pfcs), sine/cosine generator, and cmac. the pfcs stores the phase and frequency inputs and uses them to calculate the phase angle of a rotating complex vector. the sine/cosine generator performs a lookup on this phase and outputs the appropriate values for the sine and cosine. the sine and cosine form one set of inputs to the cmac, which multiplies them by the input vector to form the modulated output. the outputs of the cmac and acc can be rounded to different bit widths. phase and frequency control section the phase and frequency of the internally generated sine and cosine are controlled by the pfcs (figure 1). the pfcs generates a 32-bit word that represents the current phase of the sine and cosine waves being generated: the sine/ cosine argument. stepping this phase angle from 0 through full scale (2 32 - 1) corresponds to the phase angle of a sinusoid starting at 0 o and advancing around the unit circle counterclockwise. the pfcs automatically increments the phase by a preprogrammed amount on every rising edge of the external clock. the value of the phase step (which is the sum of the center and offset frequency registers) is: where signal frequency is a 2s complement number. the sign bit will set the output vector notation for upper sideband (usb) or lower sideband (lsb) applications. the pfcs is divided into two sections: the phase accumulator uses the data on c0-15 to compute the phase angle, that is the input to the sine/cosine section (sine/ cosine argument); the time accumulator supplies a pulse to mark the passage of a preprogrammed period of time. the phase accumulator and time accumulator work on the same principle: a 32-bit word is added to the contents of a 32-bit accumulator register every clock cycle; when the sum causes the adder to over?w, the accumulation continues with the 32 bits of the adder going into the accumulator register. the over?w bit is used as an output to indicate the timing of the accumulation over?ws. in the time accumulator, the over?w bit generates tico, the time accumulator carry out (which is the only output of the time accumulator). in the phase accumulator, the over?w is inverted to generate the phase accumulator carry out, paco. the output of the phase accumulator goes to the phase adder, which adds an offset to the top 16 bits of the phase. this 32-bit number forms the argument of the sine and cosine, which is passed to the sine/cosine generator. both accumulators are loaded 16 bits at a time over the c0-15 bus. data on c0-15 is loaded into one of the three input registers when cs and wr are low. the data in the most signi?ant input register and least signi?ant input register forms a 32-bit word that is the input to the center frequency register, offset frequency register and time accumulator. these registers are loaded by enabling the proper register enable signal; for example, to load the center frequency register, the data is loaded into the ls and ms input registers, and encfreg is set to zero; the next rising edge of clk will pass the registered version of encfreg, r.encfreg, to the clock enable of the center frequency register; this register then gets loaded on the following rising edge of clk. the contents of the input registers will be continuously loaded into the center frequency register as long as r.encfreg is low. the phase register is loaded in a similar manner. assuming pmsel is high, the contents of the phase input register is loaded into the phase register on every rising clock edge that r.enphreg is low. if pmsel is low, mod0-1 supply the two most signi?ant bits into the phase register (mod1 is the msb) and the least signi?ant 14 bits are loaded with 0. mod0-1 and are used to generate a quad phase shift keying (qpsk) signal (table 2). phase step = signal frequency clock frequency ---------------------------------------------- 2 32 table 1. ad0-1 decoding ad1 ad0 cs wr function 000 load least significant bits of frequency input. 010 load most significant bits of frequency input. 100 load phase register. 1 1 x x reserved. x x 1 x no operation. HSP45116A
3-7 HSP45116A 16 p a ci 32 phase input register ms input register 16 14 phen 0 16 r.enphreg ls input register lsen offset register center frequency register r.pmsel frequency adder a d d e r 2 16 phase register 16 16 32 32 32 32 r.clr ofr 32 32 32 msen clk 32 r.encf reg r.lo ad r.enofreg 0 0 mod(1:0) encode c(15:0) phase accumulator adder 0 phase adder sin/cos argument a d d e r 32 16 msbs 16 16 lsbs p a co r.modpi/2pi msb 15 r.enpha c a d d e r 32 phase accumulator register phase accumulator decoder 2 r.pmsel r.encfreg r.enofreg r.clr ofr r.lo ad r.enphreg r.enpha c r.modpi/2pi pmsel encfreg clr ofr lo ad enphreg enpha c modpi/2pi enofreg clk time increment 32 0 r.entireg 32 32 32 carry out time accumulator register r.rbytild tico adder 32 ad(1:0) cs wr entireg rbytild r.rbytild r.entireg acc r.eni eni r.binfmt binfmt r.sh(1:0) sh(1:0) peak r.peak sine/cosine generator 20 32 32 16 16 16 16 time accumulator outmux(1:0) oerext oer oeiext oei rnd tico p a co rin(18:0) imin(18:0) clk cos sin phase r.acc r.eni r.binfmt r.sh(1:0) r.peak r.acc frequency r e g > r e g > r e g > r e g > r e g > clk r e g > clk r e g > clk r e g > clk clk r e g > r e g > clk clk r e g > r e g > clk r e g > clk r e g > clk p a ci r e g > 0 1 mux 0 1 mux 1 0 mux 0 1 mux 1 0 mux figure 1. HSP45116A block diagram
3-8 HSP45116A 3 adder r1.acc 0 0 16 16 shifter shifter r.sh(1:0) r.eni 19 19 sin 16 cos 16 complex multiplier 33 33 adder adder 0 adder complex accumulator cmac accumulator r.peak 0 growth detect det(1:0) 16 3 fmt 16 4 outmux(1:0) ro(19-16) ro(15:0) r.binfmt oerext oer 16 fmt 16 4 io(19-16) io(15:0) r.binfmt oeiext oei 35 35 20 20 r1.acc r2.acc rnd round 0 round rin0-18 imin0-18 r.rbytild r.eni r.sh(1:0) rnd sin cos phase r.acc r.peak r.sh(1:0) r.eni clk r e g > clk r e g > reg > clk reg > clk reg > clk reg > clk r e g > clk r e g > clk reg > clk reg > clk reg > clk clk reg < reg > clk clk reg < reg > clk reg > clk reg < clk reg > clk reg < clk 0 1 mux 1 0 mux outmux(1:0) see table 4 mux mux 0 1 mux 1 0 mux clk reg < clk reg < 0 1 mux 0 1 mux see table 4 outmux(1:0) figure 1. HSP45116A block diagram (continued)
3-9 the phase accumulator consists of registers and adders that compute the value of the current phase at every clock. it has three inputs: center frequency, which corresponds to the carrier frequency of a signal; offset frequency, which is the deviation from the center frequency; and phase, which is a 16-bit number that is added to the current phase for psk modulation schemes. these three values are used by the phase accumulator and phase adder to form the phase of the internally generated sine and cosine. the sum of the values in center and offset frequency registers corresponds to the desired phase increment (modulo 2 32 ) from one clock to the next. for example, loading both registers with zero will cause the phase accumulator to add zero to its current output; the output of the pfcs will remain at its current value; i.e., the output of the ncom will be a dc signal. if a hexadecimal 00000001 is loaded into the center frequency control register, the output of the pfcs will increment by one after every clock. this will step through every location in the sine/cosine generator, so that the output will be the lowest frequency above dc that can be generated by the ncom, i.e., the clock frequency divided by 2 32 . if the input to the center frequency control register is hex 80000000, the pfcs will step through the generator with half of the maximum step size, so that frequency of the output waveform will be half of the sample rate. the operation of the offset frequency control register is identical to that of the center frequency control register; having two separate registers allows the user to generate an fm signal by loading the carrier frequency in the center frequency control register and updating the offset frequency control register with the value of the frequency offset - the difference between the carrier frequency and the frequency of the output signal. a logic low on clr ofr disables the output of the offset frequency register without clearing the contents of the register. initializing the phase accumulator register is done by putting a low on the lo ad line. this zeroes the feedback path to the accumulator, so that the register is loaded with the current value of the phase increment summer on the next clock. the ?al phase value going to the generator can be adjusted using modpi/ 2pi to force the range of the phase to be 0 o to 180 o (modulo )or0 o to 360 o (modulo 2 ). modulo 2 is the mode used for modulation, demodulation, direct digital synthesis, etc. modulo is used to calculate ffts. this is explained in greater detail in the applications section. the phase register adds an offset to the output of the phase accumulator. since the phase register is only 16 bits, it is added to the top 16 bits of the phase accumulator. the time accumulator consists of a register which is incremented on every clock. the amount by which it increments is loaded into the input registers and is latched into the time accumulator register on rising edges of clk while entireg is low. the output of the time accumulator is the accumulator carry out, tico. tico can be used as a timer to enable the periodic sampling of the output of the ncom. the number programmed into this register equals 2 32 x clk period/desired time interval. tico is disabled and its phase is initialized by zeroing the feedback path of the accumulator with rbytild. sine/cosine section the sine/cosine section (figure 2) converts the output of the pfcs into the appropriate values for the sine and cosine. it takes the most signi?ant 20 bits of the pfcs output and passes them through a look up table to form the 16-bit sine and cosine inputs to the cmac. the 20-bit word maps into 2 radians so that the angular resolution is 2 /2 20 . an address of zero corresponds to 0 radians and an address of hex fffff corresponds to 2 -(2 /2 20 ) radians. the outputs of the generator section are 2s complement sine and cosine values. the sine and cosine outputs range from hexadecimal 8001, which represents negative full scale to 7fff, which represents positive full scale. note that the normal range for twos complement numbers is 8000 to 7fff; the output range of the sin/cos generator is scaled by one so that it is symmetric about 0. the sine and cosine values are computed to reduce the amount of rom needed. the magnitude of the error in the computed value of the complex vector is less than -90.2db. the error in the sine or cosine alone is approximately 2db better. if rbytild is low, the output of the pfcs goes directly to the inputs of the cmac. if the real and imaginary inputs of the cmac are programmed to hex 7fff and 0 respectively, then the output of the pfcs will appear on output bits 0 table 2. mod0-1 decode mod1 mod0 phase shift (degrees) 00 0 01 90 1 0 270 1 1 180 clk sine/cosine generator 20 sin/cos argument 32 32 reg clk r.rbytild mux cos 16 sin 16 16 16 16 16 figure 2. sine/cosine section HSP45116A
3-10 through 15 of the ncom, with the output multiplexers set to bring out the most signi?ant bits of the cmac output (outmux = 00). the most signi?ant 16 bits out of the pfcs appears on iout0-15 and the least signi?ant bits come out on rout0-15. complex multiplier/accumulator the cmac (figure 1) performs two types of functions: complex multiplication/accumulation for modulation and demodulation of digital signals, and the operations necessary to implement an fft butter?. modulation and demodulation are implemented using the complex multiplier and its associated accumulator; the rest of the circuitry in this section, i.e., the complex accumulator, input shifters and growth detect logic are used along with the complex multiplier/accumulator for ffts. the complex multiplier performs the complex vector multiplication on the output of the sine/cosine section and the vector represented by the real and imaginary inputs rin and iin. the two vectors are combined in the following manner: rout = cos x rin - sin x iin iout = cos x iin + sin x rin rin and iin are latched into the input registers and passed through the shift stages. clocking of the input registers is enabled with a low on eni. the amount of shift on the latched data is programmed with sh0-1 (table 3). the output of the shifters is sent to the cmac and the auxiliary accumulators. the 33-bit real and imaginary outputs of the complex multiplier are latched in the multiplier registers and then go through the accumulator section of the cmac. if the acc line is high, the feedback to the accumulators is enabled; a low on acc zeroes the feedback path, so that the next set of real and imaginary data out of the complex multiplier is stored in the cmac output registers. the data in the cmac output registers goes to the multiplexer, the output of which is determined by the outmux0- 1 lines (table 4). binfmt controls whether the output of the multiplexer is presented in twos complement or unsigned format; binfmt = 0 inverts rout19 and iout19 for unsigned output, while binfmt = 1 selects twos complement. the complex accumulator duplicates the accumulator in the cmac. the input comes from the data shifters, and its 20-bit complex output goes to the multiplexer. acc controls whether the accumulator is enabled or not. outmux0-1 determines whether the accumulator output appears on rout and iout. the growth detect circuitry outputs a two bit value that signi?s the amount of growth on the data in the cmac and complex accumulator. its output, det0-1, is encoded as shown in table 5. if peak is low, the highest value of det0-1 is latched in the growth detect output register. the relative weighting of the bits at the inputs and outputs of the cmac is shown in figure 3. note that the binary point of the sine, cosine, rin and iin is to the right of the most signi?ant bit, while the binary point of ro and io is to the right of the ?th most signi?ant bit. these cmac external input and output busses are aligned with each other to facilitate cascading ncoms for fft applications. table 3. input shift selection sh1 sh0 selected bits 0 0 rin0-15, imin0-15 0 1 rin1-16, imin1-16 1 0 rin2-17, imin2-17 1 1 rin3-18, imin3-18 table 4. output multiplexer selection out mux 1 out mux 0 ro16-19 ro0-15 io16-19 io0-15 0 0 real cmac 31-34 real cmac 15-30 imag cmac 31-34 imag cmac 15-30 0 1 real cmac 31-34 0, real cmac 0-14 imag cmac 31-34 0, imag cmac 0-14 1 0 real acc 16-19 real acc 0-15 imag acc 16-19 imag acc 0-15 1 1 reserved reserved reserved reserved table 5. growth encoding det 1 det 0 number of bits of growth above 2 o 00 0 01 1 10 2 11 3 HSP45116A
3-11 figure 3. bit weighting sin/cos input s 1514131211109876543210 -2 0 . 2 -1 2 -2 2 -3 2 -4 2 -5 2 -6 2 -7 2 -8 2 -9 2 -10 2 -11 2 -12 2 -13 2 -14 2 -15 radix point complex multiplier/accumulator input (rin, iin) sh = 00 s msb lsb 1514131211109876543210 -2 0 . 2 -1 2 -2 2 -3 2 -4 2 -5 2 -6 2 -7 2 -8 2 -9 2 -10 2 -11 2 -12 2 -13 2 -14 2 -15 radix point complex multiplier/accumulator output (ro, io) outmux = 00 msb s lsb 191817161514131211109876543210 -2 4 2 3 2 2 2 1 2 0 .2 -1 2 -2 2 -3 2 -4 2 -5 2 -6 2 -7 2 -8 2 -9 2 -10 2 -11 2 -12 2 -13 2 -14 2 -15 radix point complex multiplier/accumulator output (ro, io) outmux = 01 msb lsb 191817161514131211109876543210 -2 4 2 3 2 2 2 1 2 -16 2 -17 2 -18 2 -19 2 -20 2 -21 2 -22 2 -23 2 -24 2 -25 2 -26 2 -27 2 -28 2 -29 2 -30 0 complex accumulator output (ro, io) outmux = 10 msb s lsb 191817161514131211109876543210 -2 4 2 3 2 2 2 1 2 0 . 2 -1 2 -2 2 -3 2 -4 2 -5 2 -6 2 -7 2 -8 2 -9 2 -10 2 -11 2 -12 2 -13 2 -14 2 -15 radix point HSP45116A
3-12 rounding the operation of the HSP45116A is identical to the hsp45116 with the exception of a programmable rounding option added for the data outputs. the added functionality was achieved by using one of the hsp45116s reserved con?uration registers to specify rounding precision and replacing a v cc pin with a round enable ( rnd) input. when rnd is ?igh? rounding is disabled, and the HSP45116A functions as a pin-for-pin equivalent of the hsp45116. when rnd is active ?ow?rounding is enabled. the rnd input replaces v cc on pin 75 of the 160 lead mqfp package as seen in the pinout diagram. the round control register is loaded by placing the round control value on c15-0, setting ad1-0 = 11, setting cs = 0, and forcing a low to high transition on the wr input. the rounding operation is determined by the least signi?ant 8 bits loaded into the control register as shown in table 6. the least signi?ant four bits (c3-0) loaded into the register govern rounding of the real and imaginary outputs of the complex accumulator (acc). the next more signi?ant four bits (c7-4) govern the rounding of the complex outputs of the complex multiply accumulator (cmac). the real and imaginary outputs from the cmac or acc are rounded to the same precision. the rounding is perform by adding a ?ne to the bit position below the least signi?ant bit desired in the output. for example, for a con?uration that rounds to the most signi?ant 20 bits of the cmac output, a ?ne would be added to bit position 2 -14 (see figure 3 for output bit weightings). table 6. rounding control round control register rounding operation c15-8 c7-4 c3-0 unused cmac rounding acc rounding xxxxxxxx 0000 0000 no rounding xxxxxxxx 0001 0001 cmac outputs rounded to most significant 20 bits, bit positions -2 4 to 2 -15 acc outputs rounded to most significant 20 bits, bit positions -2 4 to 2 -15 xxxxxxxx 0010 0010 cmac outputs rounded to most significant 19 bits, bit positions -2 4 to 2 -14 acc outputs rounded to most significant 19 bits, bit positions -2 4 to 2 -14 xxxxxxxx 0011 0011 cmac outputs rounded to most significant 18 bits, bit positions -2 4 to 2 -13 acc outputs rounded to most significant 18 bits, bit positions -2 4 to 2 -13 xxxxxxxx 0100 0100 cmac outputs rounded to most significant 17 bits, bit positions -2 4 to 2 -12 acc outputs rounded to most significant 17 bits, bit positions -2 4 to 2 -12 xxxxxxxx 0101 0101 cmac outputs rounded to most significant 16 bits, bit positions -2 4 to 2 -11 acc outputs rounded to most significant 16 bits, bit positions -2 4 to 2 -11 xxxxxxxx 0110 0110 cmac outputs rounded to most significant 15 bits, bit positions -2 4 to 2 -10 acc outputs rounded to most significant 15 bits, bit positions -2 4 to 2 -10 xxxxxxxx 0111 0111 cmac outputs rounded to most significant 14 bits, bit positions -2 4 to 2 -9 acc outputs rounded to most significant 14 bits, bit positions -2 4 to 2 -9 xxxxxxxx 1000 1000 cmac outputs rounded to most significant 13 bits, bit positions -2 4 to 2 -8 acc outputs rounded to most significant 13 bits, bit positions -2 4 to 2 -8 xxxxxxxx 1001 1001 cmac outputs rounded to most significant 12 bits, bit positions -2 4 to 2 -7 acc outputs rounded to most significant 12 bits, bit positions -2 4 to 2 -7 xxxxxxxx 1010 1010 cmac outputs rounded to most significant 11 bits, bit positions -2 4 to 2 -6 acc outputs rounded to most significant 11 bits, bit positions -2 4 to 2 -6 xxxxxxxx 1011 1011 cmac outputs rounded to most significant 10 bits, bit positions -2 4 to 2 -5 acc outputs rounded to most significant 10 bits, bit positions -2 4 to 2 -5 xxxxxxxx 1100 1100 cmac outputs rounded to most significant 9 bits, bit positions -2 4 to 2 -4 acc outputs rounded to most significant 9 bits, bit positions -2 4 to 2 -4 xxxxxxxx 1101-1111 1101-1111 undefined HSP45116A
3-13 applications the ncom can be used for amplitude, phase and frequency modulation, as well as in variations and combinations of these techniques, such as qam. it is most effective in applications requiring multiplication of a rotating complex sinusoid by an external vector. these include am and qam modulators and digital receivers. the ncom implements am and qam modulation on a single chip, and is a element in demodulation, where it performs complex down conversion. it can be combined with the intersil hsp43220 decimating digital filter to form the front end of a digital receiver. modulation/demodulation figure 4 shows a block diagram of an am modulator. in this example, the phase increment for the carrier frequency is loaded into the center frequency register, and the modulating input is clocked into the real input of the cmac, with the imaginary input set to 0. the modulated output is obtained at the real output of the cmac. with a sixteen bit, twos complement signal input, the output will be a 16-bit real number, on rout0-15 (with outmux = 00). by replacing the real input with a complex vector, a similar setup can generate qam signals (figure 5). in this case, the carrier frequency is loaded into the center frequency register as before, but the modulating vector now carries both amplitude and phase information. since the input vector and the internally generated sine and cosine waves are both 16 bits, the number of states is only limited by the characteristics of the transmission medium and by the analog electronics in the transmitter and receiver. the phase and amplitude resolution for the sine/cosine section (16-bit output), delivers a spectral purity of greater than 90dbc. this means that the unwanted spectral components due to phase uncertainty (phase noise) will be greater than 90db below the desired output (dbc, decibels below the carrier). with a 32-bit phase accumulator in the phase/frequency control section, the frequency tuning resolution equals the clock frequency divided by 2 32 . for example, a 25mhz clock gives a tuning resolution of 0.006hz. the ncom also works with the hsp43220 decimating digital filter to implement down conversion and low pass ?tering in a digital receiver (figure 6). the ncom performs complex down conversion on the wideband input signal by multiplying the input vector and the internally generated complex sinusoid. the resulting signal has components at twice the center frequency and at dc. two hsp43220s, one each on the real and imaginary outputs of the HSP45116A, perform low pass ?tering and decimation on the down converted data, resulting in a complex baseband signal. figure 4. amplitude modulation clk ncom modulated output pfcs 32 center frequency sin 16 sine/cosine generator cmac rin signal input 16 ro 16 d/a xmtr lo figure 5. quadrature amplitude modulation (qam) clk ncom pfcs 32 center frequency sine/cosine generator cmac ro 16 d/a rin 16 imin 16 xmtr lo 16 16 figure 6. channelized receiver chip set hsp43220 ddf HSP45116A ncom cos (wt) sin (wt) sampled input data 0 10mhz ddf output ncom output input 0 20mhz 0 HSP45116A
3-14 absolute maximum ratings thermal information supply voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . +7.0v input, output or i/o voltage applied . . . . .gnd -0.5v to v cc +0.5v esd classi?ation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . class 1 operating conditions voltage range . . . . . . . . . . . . . . . . . . . . . . . . . . . . +4.75v to +5.25v temperature range . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0 o c to 70 o c thermal resistance (typical, note 1) ja ( o c/w) mqfp package . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38.0 maximum junction temperature . . . . . . . . . . . . . . . . . . . . . . .150 o c maximum storage temperature range . . . . . . . . . . -65 o c to 150 o c maximum lead temperature (soldering 10s) . . . . . . . . . . . . .300 o c (mqfp - lead tips only) die characteristics gate count . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26,000 gates caution: stresses above those listed in ?bsolute maximum ratings may cause permanent damage to the device. this is a stress only rating and operatio n of the device at these or any other conditions above those indicated in the operational sections of this speci?ation is not implied. note: 1. ja is measured with the component mounted on an evaluation pc board in free air. dc electrical speci?ations v cc = 5.0v 5%, t a = 0 o c to 70 o c parameter symbol test conditions min max units power supply current i ccop v cc = max, clk frequency 52.6mhz notes 2, 3 - 184 ma standby power supply current i ccsb v cc = max, outputs not loaded - 500 a input leakage current i i v cc = max, input = 0v or v cc -10 10 a output leakage current i o v cc = max, input = 0v or v cc -10 10 a logical one input voltage v ih v cc = max 2.0 - v logical zero input voltage v il v cc = min - 0.8 v logical one input voltage: clk v ihc v cc = max 3.0 - v logical one output voltage v oh i oh = -5ma, v cc = min 2.6 - v logical zero output voltage v ol i ol = 5ma, v cc = min - 0.4 v input capacitance c in clk frequency 1mhz all measurements referenced to gnd. t a = 25 o c, note 4 -10pf output capacitance c out -10pf notes: 2. power supply current is proportional to frequency. typical rating is 3.5ma/mhz. 3. output load per test circuit and c l = 40pf. 4. not tested, but characterized at initial design and at major process/design changes. ac electrical speci?ations v cc = 5.0v 5%, t a = 0 o c to 70 o c (note 5) parameter symbol notes 52.6mhz (-52) units min max clk period t cp 19 - ns clk high t ch 7-ns clk low t cl 7-ns wr low t wl 7-ns wr high t wh 7-ns setup time ad0-1, cs to wr t aws 10 - ns hold time ad0-1, cs from wr t awh 0-ns setup time c0-15 to wr t cws 10 - ns hold time c0-15 from wr t cwh 0-ns setup time wr to clk t wc 710 - ns HSP45116A
3-15 ac test load circuit note: test head capacitance. setup time mod0-1 to clk t mcs 10 - ns hold time mod0-1 from clk t mch 0-ns setup time p a ci to clk t pcs 10 - ns hold time p a ci from clk t pch 0-ns setup time enphreg, encfreg, enofreg, enpha c, entireg, clr ofr, pmsel, lo ad, eni, acc, binfmt, peak, modpi/ 2pi, sh0-1, rbytild to clk t ecs 10 - ns hold time enphreg, encfreg, enofreg, enpha c, entireg, clr ofr, pmsel, lo ad, eni, acc, binfmt, peak, modpi/ 2pi, sh0-1, rbytild from clk t ech 0-ns setup time rin0-18, imin0-18 to clk t ds 10 - ns hold time rin0-18, imin0-18 from clk t dh 0-ns clk to output delay ro0-19, io0-19 t do -12ns clk to output delay det0-1 t deo -12ns clk to output delay p a co t po -12ns clk to output delay tico t to -12ns output enable time oer, oei, oerext, oeiext t oe -8ns output enable time outmux0-1 t md -14ns output disable time t od 6- 8ns output rise, fall time t rf 6- 4ns notes: 5. ac tests performed with c l = 40pf, i ol = 2.0ma, and i oh = -0.4ma. input reference level for clk = 2.0v, all other inputs 1.5v. test v ih = 3.0v, v ihc = 4.0v, v il = 0v; v oh 1.5v, v ol 1.5v. 6. controlled via design or process parameters and not directly tested. characterized upon initial design and after major process and/or changes. 7. applicable only when outputs are being monitored and encfreg, enphreg or entireg is active. wr is always asynchronous when rnd is active. ac electrical speci?ations v cc = 5.0v 5%, t a = 0 o c to 70 o c (note 5) (continued) parameter symbol notes 52.6mhz (-52) units min max equivalent circuit c l (note) i oh 1.5v i ol dut switch s1 open for i ccsb and i ccop s 1 HSP45116A
3-16 waveforms figure 7. input and output timing figure 8. control bus timing clk mod0-1 p a ci tico det0-1 p a co rout0-19 iout0-19 control inputs t ch t cp t pcs t mcs t ecs t ds t cl t mch t pch t ech t dh t do t deo t po t to rin0-19 iin0-19 t wl t wc t wh t awh t aws t awh t aws t cwh t cws clk cs wr c0-15 ad0-1 HSP45116A
3-17 all intersil semiconductor products are manufactured, assembled and tested under iso9000 quality systems certi?ation. intersil semiconductor products are sold by description only. intersil corporation reserves the right to make changes in circuit design and/or spec ifications at any time with- out notice. accordingly, the reader is cautioned to verify that data sheets are current before placing orders. information furnished by intersil is b elieved to be accurate and reliable. however, no responsibility is assumed by intersil or its subsidiaries for its use; nor for any infringements of patents or other rights of th ird parties which may result from its use. no license is granted by implication or otherwise under any patent or patent rights of intersil or its subsidiari es. for information regarding intersil corporation and its products, see web site http://www.intersil.com figure 9. output enable, disable timing figure 10. multiplexer timing figure 11. output rise and fall times waveforms (continued) high impedance 1.5v t oe 1.7v 1.3v high impedance 1.5v t od ro0-19 io0-19 oer oerext oeiext oei t md outmux0-1 ro0-19 io0-19 2.0v 0.8v t rf t rf HSP45116A

▲Up To Search▲

Price & Availability of HSP45116A

	To Download HSP45116A Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .