T H A T Corporation THAT Analog Engine® IC Dynamics Processor THAT 4301, 4301A FEATURES APPLICATIONS · High-Performance Voltage Controlled Amplifier · Compressors · · Limiters High-Performance RMS-Level Detector · Gates · Three General-Purpose Opamps · Expanders · Wide Dynamic Range: >115 dB · De-Essers · Low THD: <0.03% · Duckers · Low Cost: $4.39 (‘000s) · Noise Reduction Systems · DIP & Surface-Mount Packages · Wide-Range Level Meters Description THAT 4301 Dynamics Processor, dubbed “THAT Analog Engine,” combines in a single IC all the active circuitry needed to construct a wide range of dynamics processors. The 4301 includes a high-performance, exponentially-controlled VCA, a log-responding RMS-level sensor and three general- purpose opamps. crest factors up to 10. One opamp is dedicated as a current-to-voltage converter for the VCA, while the other two may be used for the signal path or control voltage processing. The combination of exponential VCA gain control and logarithmic detector response — “decibel-linear” response — simplifies the mathematics of designing the control paths of dynamics processors. This makes it easy to design audio compressors, limiters, gates, expanders, de-essers, duckers, noise reduction systems and the like. The high level of integration ensures excellent temperature tracking between the VCA and the detector, while minimizing the external parts count. The VCA provides two opposing-polarity, voltage-sensitive control ports. Dynamic range exceeds 115 dB, and THD is typically 0.003% at 0 dB gain. In the 4301A, the VCA is selected for low THD at extremely high levels. The RMS detector provides accurate rms-to-dc conversion over an 80 dB dynamic range for signals with 18 19 17 11 14 VCC OA1 20 - SYM VCA OUT IN + 12 13 OA3 + EC+ EC- 15 THAT4301 1 RMS IN 16 + IT OUT CT 2 5 OA2 - GND VEE 4 9 10 8 6 7 Figure 1. Block Diagram (pin numbers are for DIP only) THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com Page 2 Rev. 04/10/02 SPECIFICATIONS 1 , 2 Absolute Maximum Ratings (T A = 25°C) Positive Supply Voltage (VCC) +18 V Power Dissipation (PD) (TA = 75°C) Negative Supply Voltage (VEE) -18 V Operating Temperature Range (TOP) Supply Current (ICC) 700 mW Storage Temperature Range (TST) 20 mA 0 to +70°C -40 to +125°C Overall Electrical Characteristics Parameter Min Typ Max Units Positive Supply Voltage Symbol VCC Conditions +7 — +15 V Negative Supply Voltage VEE -7 — -15 V Positive Supply Current ICC — 12 18 mA Negative Supply Current IEE — -12 -18 mA Thermal Resistance qJ-C — 140 — °C/W SO-Package VCA Electrical Characteristics 3 4301 Parameter 4301A Symbol Conditions Min Typ Max Min Typ Max Units IB(VCA) No Signal — 30 400 — 30 400 pA Input Offset Voltage VOFF(VCA In) No Signal — ±4 ±15 — ±4 ±15 mV Input Signal Current IIN(VCA) or IOUT(VCA) — 175 750 — 175 750 mArms -0.4 0.0 +0.4 -0.4 0.0 +0.4 dB Input Bias Current Gain at 0V Control G0 EC+ = EC– = 0.000V TA = 25°C (TCHIP @ 55°C) Gain-Control Constant -60 dB < gain < +40dB Gain-Control TempCo EC+/Gain (dB) EC+ & SYM 6.4 6.5 6.6 6.4 6.5 6.6 mV/dB EC-/Gain (dB) EC- -6.4 -6.5 -6.6 -6.4 -6.5 -6.6 mV/dB DEC / DTCHIP Ref TCHIP= 27°C — +0.33 — — +0.33 — %/°C -60 to +40 dB gain — Gain-Control Linearity Off Isolation EC+=SYM=-375mV, EC-=+375mV 110 Output Offset Voltage Change Gain Cell Idling Current Output Noise DVOFF(OUT) 2 — 0.5 2 % — 110 115 — dB Rout = 20kW 0 dB gain — 1 3 — 1 3 mV +15 dB gain — 2 10 — 2 10 mV +30 dB gain — 5 25 — 5 25 mV — 20 — — 20 — mA IIDLE en(OUT) 0.5 115 20 Hz-20 kHz Rout = 20kW Total Harmonic Distortion THD 0 dB gain — -96 -94 — -96 -94 dBV +15 dB gain — -85 -83 — -85 -83 dBV VIN= 0 dBV, 1 kHz 0 dB gain — 0.003 0.007 — 0.003 0.007 % 1. All specifications subject to change without notice. 2. Unless otherwise noted, TA=25°C, VCC = +15V, VEE= -15V; VCASYM adjusted for min THD @ 1 V, 1 kHz, 0 dB gain. 3. Test circuit is the VCA section only from Figure 2. THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com THAT 4301 Dynamics Processor IC Page 3 SPECIFICATIONS 1 , 2 (Cont’d.) VCA Electrical Characteristics 3 (Cont’d.) Parameter Symbol Total Harmonic Distortion (cont’d.) THD Conditions Min 4301 Typ Max Min 4301A Typ Max Units VIN = +10 dBV, 1 kHz 0 dB gain — 0.03 0.07 — 0.03 0.07 % –15 dB gain — 0.035 0.09 — 0.035 0.09 % +15 dB gain — 0.035 0.09 — 0.035 0.09 % VIN = +19.5 dBV, 1 kHz 0 dB gain — — — — 0.05 0.09 % minimum THD -2.5 0 +2.5 -2.5 0 +2.5 mV VOUT= +10 dBV, 1 kHz Symmetry Control Voltage VSYM RMS Detector Electrical Characteristics 4 Parameter Symbol Conditions Min Typ Max Units Input Bias Current IB (RMS) No Signal — 30 400 pA Input Offset Voltage VOFF(RMS In) No Signal — ±4 ±15 mV Input Signal Current IIN(RMS) — 175 750 mA 6 8.5 12 mA 6.4 6.5 6.6 mV/dB .985 1 1.015 1mA < Iin< 100mA — 0.1 — dB 100nA < Iin< 316mA — 0.5 — dB 31.6nA < Iin< 1mA — 1.5 — dB –20 — 20 % 0.2 dB error — 3.5 — 0.5 dB error — 5 — Iin0 IT= 7.5mA EO / 20log(Iin/Iin0) 31.6nA< IIN< 1mA Input Current for 0 V Output Output Scale Factor TA= 25°C (TCHIP »55°C) Scale Factor Match (RMS to VCA) -20 dB < VCA Gain < +20 dB 1mA<Iin (DET)<100mA Output Linearity fIN= 1kHz fIN = 100 Hz, t = .001 s Rectifier Balance 1mA< Iin < 100mA Crest Factor 1ms pulse repetition rate Maximum Frequency for 2 dB Additional Error Timing Current Set Range Filtering Time Constant Output Temp. Coefficient Output Current — 10 — Iin ³ 10mA — 100 — kHz Iin ³ 3mA — 45 — kHz Iin ³ 300nA — 7 — kHz 1.5 7.5 15 mA mV IT Voltage at IT Pin Timing Current Accuracy 1.0 dB error IT = 7.5mA -10 +20 +50 ICT/IT IT = 7.5mA 0.90 1.1 1.30 t TCHIP = 55°C DEo / DTCHIP Re: TCHIP = 27°C — 0.33 — %/°C IOUT –300mV < VOUT< +300mV ±90 ±100 — mA ( 0.026)CT IT s 4. Except as noted, test circuit is the RMS-Detector section only from Figure 2. THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com Page 4 Rev. 04/10/02 Specifications 1 , 2 (Cont’d) Opamp Electrical Characteristics 5 OA1 Parameter Symbol Conditions Min OA2 Typ Max Min OA3 Typ Max Min Typ Max Units VOS — ±0.5 ±6 — ±0.5 ±6 — ±0.5 ±6 mV Input Bias Current IB — 150 500 — 150 500 — 150 500 nA Input Offset Current IOS — 15 50 — 15 50 N/A nA Input Voltage Range IVR — ±13.5 — — ±13.5 — N/A V RS<10k — 100 — — 100 — N/A — 100 — — 100 — — Input Offset Voltage Common Mode Rej. Ratio CMRR Power Supply Rej. Ratio PSRR VS=±7V to ±15V Gain Bandwidth Product GBW (@50kHz) — 5 — — 5 — AVO RL=10k — 115 — — 110 — Open Loop Gain RL=2k Output Voltage Swing N/A VO@RL=5kW — VO@RL=2kW SR Total Harmonic Distortion THD ±13 — — N/A Short Circuit Output Current Slew Rate N/A 100 — — 5 — — 125 — — 120 — MHz ±13 — — ±14 — V N/A — — ±13 — V — 4 — — 4 — — 12 — mA — 2 — — 2 — — 2 — V/ms 1kHz, AV=1, RL=10kW — 0.0007 0.003 — 0.0007 0.003 — 0.0007 0.003 % 1kHz, AV=–1, RL= 2kW N/A N/A — 0.0007 0.003 % Input Noise Voltage Density en fO=1kHz — 6.5 10 — 7.5 12 — 7.5 12 nV Hz Input Noise Current Density in fO=1kHz — 0.3 — — 0.3 — — 0.3 — pA Hz 5. Test circuit for opamps is a unity-gain follower configuration, with load resistor RL as specified. +15V R5 VCA SYM 50K -15V SIGNAL IN C1 C2 R1 47pF R4 300K 20K0 1% 47uF R3 R2 51 20K0 1% +15V C7 SYM OA1 VCC C8 C3 47uF IN + 100n VCA OUT EC+ EC- SIGNAL OUT OA3 + THAT4301 VEE 100n -15V R6 RMS IN + OA2 OUT - Ct It GND 10K0 1% C6 R7 2M00 1% 22uF C4 10uF RMS OUT Ec- -15V Figure 2. VCA and RMS detector test circuit THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com THAT 4301 Dynamics Processor IC Page 5 SO Pin Pin Name RMS In 1 3 EC- 16 24 IT (ITime) 2 4 VCA In 17 25 No Connection 3 5 OA1 Out 18 26 RMS Out 4 6 OA1 -In 19 27 CT (CTime) 5 7 OA1 +In 20 28 OA2 -In 6 9 No Connection 1 OA2 Out 7 10 No Connection 2 OA2 +In 8 11 No Connection 8 GND 9 12 No Connection 14 VEE 10 13 No Connection 15 VCC 11 18 No Connection 16 OA3 Out 12 19 No Connection 17 VCA Out 13 20 No Connection 21 SYM 14 22 No Connection 29 EC+ 15 23 No Connection 30 Pin Name DIP P in DIP P in SO Pin Table 1. Pin Connections Ordering Information E 1 F B G Plastic DIP 4301P 4301PA Plastic Surface Mount 4301S Inquire A 0-10 H C D J K N L B I I 1 M 0-15 P C D O J E F A H G ITEM A B C D E F G H I J K L M N O P MILLIMETERS 24.8 Max. 24.2 +/-0.2 6.4 +/-0.2 7.62 +/-0.25 2.54 +/-0.15 0.46 +0.15 -0.1 1.0 +/-0.15 1.5 Typ. 0.98 Typ. 1.5 1.75 3.25 +/-0.15 4.7 Max. 0.51 Min. 2.8 Min. 0.25 +0.15 -0.05 INCHES 0.98 Max 0.95 +/-0.008 0.25 +/-0.008 0.30 +/-0.01 0.10 +/-0.006 0.02 +0.006 -0.004 0.04 +/-0.006 0.06 Typ. 0.04 Typ. 0.06 0.07 0.13 +/-0.006 0.19 Max. 0.02 Min. 0.11 Min. 0.01 +0.006 -0.002 Figure 3. Plastic dual in-line package outline ITEM A B C D E F G H I J MILLIMETERS 15.4 +/- 0.3 7.5 +/- 0.2 10.3 +/- 0.4 0.4 + 0.1 - 0.05 1.0 Typ. 0.85 MAX. 2.3 +/- 0.15 0.15 +/- 0.1 0.8 0.2 + 0.1 - 0.05 INCHES 0.60 +/- 0.012 0.29 +/- 0.008 0.41 +/- 0.016 0.002 +0.004 -0.002 0.039 Typ. 0.033 Max. 0.09 +/- 0.006 0.006 +/- 0.004 0.031 0.008 +0.004 -0.002 Figure 4. Plastic surface-mount package outline THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com Page 6 Rev. 04/10/02 Representative Data Figure 5. VCA Gain vs. Control Voltage (Ec-) at 25°C Figure 6. VCA 1kHz THD+Noise vs. Input, -15 dB Gain Figure 7. VCA 1kHz THD+Noise vs. Input, +15 dB Gain Figure 8. VCA 1kHz THD+Noise vs. Input, 0 dB Gain Figure 9. VCA THD vs. Frequency, 0 dB Gain, 1Vrms Input Figure 10. RMS Output vs. Input Level, 1 kHz & 10 kHz Figure 11. Departure from Ideal Detector Law vs. Level Figure 12. Detector Output vs. Frequency at Various Levels THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com THAT 4301 Dynamics Processor IC Page 7 Theory of Operation THAT 4301 Dynamics Processor combines THAT Corporation’s proven Voltage-Controlled Amplifier (VCA) and RMS-Level Detector designs with three general-purpose opamps to produce an Analog Engine useful in a variety of dynamics processor applications. For details of the theory of operation of the VCA and RMS-Detector building blocks, the interested reader is referred to THAT Corporation’s data sheets on the 2150 Series VCAs and the 2252 RMS-Level Detector. Theory of the interconnection of exponentially-controlled VCAs and log-responding level detectors is covered in THAT Corporation’s application note AN101, The Mathematics of Log-Based Dynamic Processors. The VCA — in Brief THAT 4301 VCA is based on THAT Corporation’s highly successful complementary log-antilog gain cell topology, as used in THAT 2150-Series IC VCAs, and the modular 202 Series VCAs. THAT 4301 is integrated using a fully complementary, BiFET process. The combination of FETs with high-quality, complementary bipolar transistors (NPNs and PNPs) allows additional flexibility in the design of the VCA over previous efforts. Input signals are currents to the VCA IN pin. This pin is a virtual ground, so in normal operation an input voltage is converted to input current via an appropriately sized resistor (R1 in Figure 2, Page 4). Because dc offsets present at the input pin and any dc offset in preceeding stages will be modulated by gain changes (thereby becoming audible as thumps), the input pin is normally ac-coupled (C1 in Figure 2). 2), which is adjusted for minimum signal distortion at unity (0 dB) gain. The VCA may be controlled via EC-, as shown in Figure 2, or via the combination of EC+ and SYM. This connection is illustrated in Figure 13. Note that this figure shows only that portion of the circuitry needed to drive the positive VCA control port; circuitry associated with OA1, OA2 and the RMS detector has been omitted. R5 50K Positive Control In C1 C2 47pF R1 Signal In 47uF VCA SYM 20K0 1% R3 R4 300K OA1 + IN SYM OUT VCA EC- R2 20K0 1% 51 EC+ OA3 + Signal Out THAT4301 VCC VEE IN It RMS OUT Ct GND + OA2 - Figure 13. Driving the VCA via the Positive Control Port While the 4301’s VCA circuitry is very similar to that of the THAT 2150 Series VCAs, there are several important differences, as follows: The VCA output signal is also a current, inverted with respect to the input current. In normal operation, the output current is converted to a voltage via inverter OA3, where the ratio of the conversion is determined by the feedback resistor (R2, Figure 2) connected between OA3‘s output and its inverting input. The signal path through the VCA and OA3 is noninverting. 1) Supply current for the VCA is fixed internally. Approximately 2mA is available for the sum of input and output signal currents. (This is also the case in a 2150 Series VCA when biased as recommended.) The gain of the VCA is controlled by the voltage applied to EC–, EC+, and SYM. Gain (in decibels) is proportional to EC+ – EC-, provided EC+ and SYM are at essentially the same voltage (see below). The constant of proportionality is –6.5 mV/dB for the voltage at EC–, and 6.5 mV/dB for the voltage at EC+ and SYM. 3) The control-voltage constant is approximately 6.5 mV/dB, due primarily to the higher internal operating temperature of the 4301 compared to that of the 2150 Series. As mentioned, for proper operation, the same voltage must be applied to EC+ and SYM, except for a small (±2.5 mV) dc bias applied between these pins. This bias voltage adjusts for internal mismatches in the VCA gain cell which would otherwise cause small differences between the gain of positive and negative half-cycles of the signal. The voltage is usually applied via an external trim potentiometer (R5 in Figure 2) The signal current output of the VCA is internally connected to the inverting input of an on-chip opamp. In order to provide external feedback around this opamp, this node is brought out to a pin. 4) The input stage of the 4301 VCA uses integrated P-channel FETs rather than a bias-current corrected bipolar differential amplifier. Input bias currents have therefore been reduced. The RMS Detector — in Brief The 4301’s detector computes rms level by rectifying input current signals, converting the rectified current to a logarithmic voltage, and applying that voltage to a log-domain filter. The output signal is a dc voltage proportional to the decibel-level of the rms THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com Page 8 Rev. 04/10/02 value of the input signal current. Some ac component (at twice the input frequency) remains superimposed on the dc output. The ac signal is attenuated by a log-domain filter, which constitutes a single-pole rolloff with cutoff determined by an external capacitor and a programmable dc current. As in the VCA, input signals are currents to the RMS IN pin. This input is a virtual ground, so a resistor (R6 in Figure 2) is normally used to convert input voltages to the desired current. The level detector is capable of accurately resolving signals well below 10 mV (with a 10 kW input resistor). However, if the detector is to accurately track such low-level signals, ac coupling is normally required. The log-domain filter cutoff frequency is usually placed well below the frequency range of interest. For an audio-band detector, a typical value would be 5 Hz, or a 32 ms time constant (t). The filter’s time constant is determined by an external capacitor attached to the CT pin, and an internal current source (ICT) connected to CT. The current source is programmed via the IT pin: current in IT is mirrored to ICT with a gain of approximately 1.1. The resulting time constant t is approximately equal to 0.026 CT/IT. Note that, as a result of the mathematics of RMS detection, the attack and release time constants are fixed in their relationship to each other. The dc output of the detector is scaled with the same constant of proportionality as the VCA gain control: 6.5 mV/dB. The detector’s 0 dB reference (Iin0, the input current which causes 0 V output), is determined by IT as follows: Iin0= 9.6 mA It The detector output stage is capable of sinking or sourcing 100 mA. Differences between the 4301’s RMS-Level Detector circuitry and that of the THAT 2252 RMS Detector are as follows: 1) The rectifier in the 4301 RMS Detector is internally balanced by design, and cannot be balanced via an external control. The 4301 will typically balance positive and negative halves of the input signal within ±1.5%, but in extreme cases the mismatch may reach ±15%. However, a 15% mismatch will not significantly increase ripple-induced distortion in dynamics processors over that caused by signal ripple alone. 2) The time constant of the 4301’s RMS detector is determined by the combination of an external capacitor (connected to the CT pin) and an internal, programmable current source. The current source is equal to 1.1 IT. Normally, a resistor is not connected directly to the CT pin on the 4301. 3) The 0 dB reference point, or level match, is not adjustable via an external current source. However, as in the 2252, the level match is affected by the timing current, which, in this case, is drawn from the IT pin and mirrored internally to CT. 4) The input stage of the 4301 RMS detector uses integrated P-channel FETs rather than a bias-current corrected bipolar differential amplifier. Input bias currents are therefore negligible, improving performance at low signal levels. The Opamps — in Brief The three opamps in the 4301 are intended for general purpose applications. All are 5 MHz opamps with slew rates of approximately 2V/ms. All use bipolar PNP input stages. However, the design of each is optimized for its expected use. Therefore, to get the most out of the 4301, it is useful to know the major differences among these opamps. OA3, being internally connected to the output of the VCA, is intended for current-to-voltage conversion. Its input noise performance, at 7.5nV Hz, complements that of the VCA, adding negligible noise at unity gain. Its output section is capable of driving a 2 kW load to within 2V of the power supply rails, making it possible to use this opamp directly as the output stage in single-ended designs. OA1 is the quietest opamp of the three. Its input noise voltage, at 6.5nV Hz, makes it the opamp of choice for input stages. Note that its output drive capability is limited (in order to reduce the chip’s power dissipation) to approximately ±3 mA. It is comfortable driving loads of 5 kW or more to within 1V of the power supply rails. OA2 is intended primarily as a control-voltage processor. Its input noise parallels that of OA3, and its output drive capability parallels that of OA1. THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com THAT 4301 Dynamics Processor IC Page 9 Applications The circuit of Figure 14, Page 9, shows a typical application for THAT 4301. This simple compressor/limiter design features adjustable hard-knee threshold, compression ratio, and static gain1. The applications discussion in this data sheet will center on this circuit for the purpose of illustrating important design issues. However, it is posslble to configure many other types of dynamics processors with THAT 4301. Hopefully, the following discussion will imply some of these possibilities. verting stage. If, for some reason, more than 0 dB gain is required when the VCA is set to unity, then the resistors may be skewed to provide it. Note that the feedback capacitor (C2) is required for stability. The VCA output has approximately 45 pf of capacitance to ground, which must be neutralized via the 47 pf feedback capacitor across R2. Signal Path As mentioned in the section on theory, the VCA input pin is a virtual ground with negative feedback provided internally. An input resistor (R1, 20kW) is required to convert the ac input voltage to a current within the linear range of the 4301. (Peak VCA input currents should be kept under 1 mA for best distortion performance.) The coupling capacitor (C1, 47 mf) is strongly recommended to block dc current from preceeding stages (and from offset voltage at the input of the VCA). Any dc current C1 into the VCA will be 47uF modulated by varying +15 gain in the VCA, showing THRESHOLD CCW up in the output as R11 R12 “thumps”. Note that C1, 383K 1% 10K R10 in conjunction with R1, CW 2M00 1% will set the low fre-15 R8 quency limit of the cir4k99 1% cuit. IN The VCA output is connected to OA3, configured as an inverting current-to-voltage converter. OA3‘s feedback components (R2, 20 kW, and C2, 47 pf) determine the constant of current-to-voltage conversion. The simplest way to deal with this is to recognize that when the VCA is set for unity (0 dB) gain, the input to output voltage gain is simply R2/R1, just as in the case of a single in- +15 C7 100n C8 100n C3 R13 10K +15 R5 50K 20K0 1% -15 R9 R4 300K 10K0 1% CR2 R3 CR1 51 R2 OA1 + IN EC- SYM OUT EC+ VCA OA3 + OUT THAT4301 -15 R6 IN IT RMS OUT CT 10K0 1% R14 1K43 1% 22uF GND + OA2 R16 R7 2M00 1% 4k99 1% C4 10uF C5 100N -15 R17 R15 CCW 47pF C2 20K0 1% VCC C6 CW VCA SYM R1 VEE 47uF COMPRESSION The VCA gain is controlled via the EC– terminal, whereby gain will be proportional to the negative of the voltage at EC–. The EC+ terminal is grounded, and the SYM terminal is returned nearly to ground via a small resistor (R3, 51 W). The VCA SYM trim (R5, 50 kW) allows a small voltage to be applied to the SYM terminal via R4 (300 kW). This voltage adjusts for small mismatches within the VCA gain cell, thereby reducing even-order distortion products. To adjust the trim, apply to the input a middle-level, middle-frequency signal (1 kHz at 1 V is a good 590K 1% 10K0 1% +15 GAIN CW R18 10K CCW -15 Figure 14. Typical Compressor/Limiter Application Circuit 1. More information on this compressor design, along with suggestions for converting it to soft-knee operation, is given in AN100, Basic Compressor Limiter Design. The designs in AN100 are based on THAT Corporation’s 2150-Series VCAs and 2252 RMS Detector, but are readily adaptable to the 4301 with only minor modifications. In fact, the circuit presented here is functionally identical to the hard-knee circuit published in AN100. THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com Page 10 Rev. 04/10/02 choice with this circuit) and observe THD at the signal output. Set the trim for minimum THD. RMS-Level Detector The RMS detector’s input is similar to that of the VCA. An input resistor (R6, 10 kW) converts the ac input voltage to a current within the linear range of the 4301. (Peak detector input currents should be kept under 1 mA for best linearity.) The coupling capacitor (C3, 47 mf) is recommended to block dc current from preceeding stages (and from offset voltage at the input of the detector). Any dc current into the detector will limit the low-level resolution of the detector, and will upset the rectifier balance at low levels. Note that, as with the VCA input circuitry, C3 in conjunction with R6 will set the lower frequency limit of the detector. The time response of the RMS detector is determined by the capacitor attached to CT (C4, 10 mf) and the size of the current in pin IT (determined by R7, 2 MW and the negative power supply, –15V). Since the voltage at IT is approximately 0 V, the circuit of Figure 14 produces 7.5 mA in IT. The current in IT is mirrored with a gain of 1.1 to the CT pin, where it is available to discharge the timing capacitor (C4). The combination produces a log filter with time constant equal to approximately 0.026 CT/IT (~35 ms in the circuit shown). The waveform at CT will follow the logged (decibel) value of the input signal envelope, plus a dc offset of about 1.3 V (2 VBE). This allows a polarized capacitor to be used for the timing capacitor, usually an electrolytic. The capacitor used should be a low-leakage type in order not to add significantly to the timing current. The output stage of the RMS detector serves to buffer the voltage at CT and remove the 1.3 V dc offset, resulting in an output centered around 0 V for input signals of about 85 mV. The output voltage increases 6.5 mV for every 1 dB increase in input signal level. This relationship holds over more than a 60 dB range in input currents. Control Path A compressor/limiter is intended to reduce its gain as signals rise above a threshold. The output of the RMS detector represents the input signal level over a wide range of levels, but compression only occurs when the level is above the threshold. OA1 is configured as a variable threshold detector to block envelope information for low-level signals, passing only information for signals above threshold. OA1 is an inverting stage with gain of 2 above threshold and 0 below threshold. Neglecting the action of the THRESHOLD control (R12) and its associated resistors (R11 and R10), positive signals from the RMS detector output drive the output of OA1 negative. This forward biases CR2, closing the feedback loop such that the junction of R9 and CR2 (the output of the threshold detector) sits at -(R9/R8) RMSOUT. For the circuit of Figure 14, this is –2 RMSOUT. Negative signals from the RMS detector drive the output of OA1 positive, reverse biasing CR2 and forward biasing CR1. In this case, the junction of R9 and CR2 rests at 0 V, and no signal level informaion is passed to the threshold detector’s output. In order to vary the threshold, R12, the THRESHOLD control, is provided. Via R11 (383 kW), R12 adds up to ±39.2 mA of current to OA1‘s summing junction, requiring the same amount of opposite-polarity current from the RMS detector output to counterbalance it. At 4.99 kW, the voltage across R8 required to produce a counterbalancing current is ± 195 mV, which represents a ±30 dB change in RMS detector input level. Since the RMS detector’s 0 dB reference level is 85 mV, the center of the THRESHOLD pot’s range would be 85 mV, were it not for R10 (2 MW), which provides an offset. R10 adds an extra –7.5 ma to OA1‘s summing junction, which would be counterbalanced by 37.4 mV at the detector output. This corresponds to 5.8 dB, offsetting the THRESHOLD center by this much to 165 mV, or approximately -16 dBV. The output of the threshold detector represents the signal level above the determined threshold, at a constant of about 13 mV/dB (from [R9/R8] 6.5 mV/dB). This signal is passed on to the COMPRESSION control (R13), which variably attenuates the signal passed on to OA2. Note that the gain of OA2, from the wiper of the COMPRESSION control to OA2‘s output, is R16/R15 (0.5), precisely the inverse of the gain of OA1. Therefore, the COMPRESSION control lets the user vary the above-threshold gain between the RMS detector output and the output of OA1 from zero to a maximum of unity. The gain control constant of the VCA, 6.5 mV/dB, is exactly equal to the output scaling constant of the RMS detector. Therefore, at maximum COMPRESSION, above threshold, every dB increase in input signal level causes a 6.5 mV increase in the output of OA2, which in turn causes a 1 dB decrease in the VCA gain. With this setting, the output will not increase despite large increases in input level above threshold. This is infinite compression. For intermediate settings of COMPRESSION, a 1 dB increase in input signal level will cause less than a 1 dB decrease in gain, thereby varying the compression ratio. The resistor R14 is included to alter the taper of the COMPRESSION pot to better suit common use. If a linear taper pot is used for R13, the compression ratio will be 1:2 at the middle of the rotation. However, 1:2 compression in an above-threshold compressor is not very strong processing, so 1:4 is often preferred at the midpoint. R14 warps the taper of R13 so that 1:4 compression occurs at approximately the midpoint of R13‘s rotation. THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com THAT 4301 Dynamics Processor IC The GAIN control (R18) is used to provide static gain or attenuation in the signal path. This control adds up to ±130 mV offset to the output of OA2 (from -V+ R 16 R to -V- R 16 R ), which is approximately 17 17 ±20 dB change in gain of the VCA. C5 is used to attenuate the noise of OA2, OA1 and the resistors R8 through R16 used in the control path. All these active and passive components produce noise which is passed on to the control port of the VCA, causing modulation of the signal. By itself, the 4301 VCA produces very little noise modulation, and its performance can be significantly degraded by the use of noisy components in the control voltage path. Overall Result The resulting compressor circuit provides hard-knee compression above threshold with three essential user-adjustable controls. The threshold of compression may be varied over a ±30 dB range from about –46 dBV to +14 dBV. The compression ratio Page 11 may be varied from 1:1 (no compression) to ¥:1. And, static gain may be added up to ±20 dB. Audio performance is excellent, with THD running below 0.05% at middle frequencies even with 10 dB of compression, and an input dynamic range of over 115 dB. Perhaps most important, this example design only scratches the surface of the large body of applications circuits which may be constructed with THAT 4301. The combination of an accurate, wide-dynamic-range, log-responding level detector with a high-quality, exponentially-responding VCA produces a versatile and powerful analog engine. The opamps provided in the 4301 enable the designer to configure these building blocks with few external components to construct gates, expanders, de-essers, noise reduction systems and the like. For further information, samples and pricing, please contact us at the address below. THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com Page 12 Rev. 04/10/02 Notes THAT Corporation; 45 Sumner Street; Milford, Massachusetts 01757-1656; USA Tel: +1 (508) 478-9200; Fax: +1 (508) 478-0990; Web: www.thatcorp.com