MF10BJ

MF10 MAXIM | Alldatasheet

Document overview

• Manufacturer or author: Provided By ALLDATASHEET.COM(FREE DATASHEET DOWNLOAD SITE)
• PDF pages: 16

Technical content

‘The MF 10 isa dual 2nd order, switched capacitor, state © @ ‘No External Capacitors Required v variable filter. Each of the two filter sections uses two ~” switched capacitor integrators and an op amp to. @ Low Sensitivity to External Component Variation generate a second order function. The location of the 8 Poles (and thus the center frequency and Q) is deter- Excellent Frequency and Q Stability mined by the frequency of an external clock and 2 to 4 external resistors. No external capacitors are used. Easily Cascaded for Multipole Filters Each of the two filter sections of the MF10 can generate all standard 2nd order functions: bandpass. lowpass, @ Filter Frequency Set by External Clock Frequency highpass, notch (band-reject), complex zeroes and alipass functions. Three of these functions are simul- * 9:2% Clock to Center Frequency Ratio Accuracy taneously available. The frequency of the 2nd order poles is accurate to +0.2% and the Q is accurate to ° Highpass, Lowpass, Bandpass, Notch, and Allpass within 2%. . Fourth order filters can be made by cascading thetwo @ ‘Up to 3 Simultaneous Filter Function Outputs 2nd order filter sections of the MF10, and higher order filters can easily be made by cascading more MF 10s. © Up to 30kHz Operation The excellent accuracy and stability of MF10 based filters eliminates the complex, costly tuning normally @ Easy to use—Design Directly from the Data Sheet required in the production of high order (multipole) filters. Design equations for Butterworth, Bessel, # Monolithic, Low Power CMOS Design Chebyshev, and Cauer (Elliptic) filters are provided. This versatile device is used for a wide range of filtering PART TEMP. RANGE PIN-PACKAGE applications such as: MFTOBN Oto 470°C 20 Lead Plastic DIP | Tunable active fiters Adaptive Filtering MFIOCN 0°10 470°C 20 Lead Plastic DIP] ulti-pole filters Phase locked loops = sa Anti-aliasing filters Signal Processing/ METOBWP OC to +70°C_ 2 Lead Wide SO_| nditioning METOCWP O°Ct0+70°C 20 Lead Wide SO_| MF10BJ O°C to +70°C 20 Lead CERDIP* MFI0CJ O°Ct0 470°C 20 Lead CERDIP"_| MFI0CC/D O°C to +70°C Dice" . Contact Factory —_———— Typical Operating Circuit Pin Configuration ror vw ; ag fa = G] 3] 8P5 é nari, Fs] warning n : "6 E] Aaaxian [17] Ws MAKI suf] MFIO Esty vo miro vou se 1080 vor [3] fis] vo {7H OROER. 20 LOWPASE io Fe soxnes uTTERWoRTH FLTER cus [a fi] ous DIPISO. MAXIM Maxi integrated Products 1 For free samples & the latest literature: http:/www.maxim-ic.com, or Phone 1-800-998-8800

° ABSOLUTE MAXIMUM RATINGS. ‘Stresses above those listed under "Absolute Maximum Ratings” may. ‘cause permanent damage to the device, These are stress ratings only and functional operation of the device at these or any other conditions al those indicated in the operational sections of the specifications isnot implied. Exposure to absolute maximum rating conditions for extended periods may affect device reliability. ELECTRICAL CHARACTERISTICS (compiete Fitter) PARAMETER “___conoinions MIN TYP ax | UNITS Frequency Range _| fo < Q's 200 kHz: a) _ 30 - Hz | Clock to Center Frequency Ratio. to.K/to | MF10B_ Pin 12 High, Q = 10 49.94 + 0.2% 20.6% MF10C_ fo x Q = SOkHz, Mode 1 49.94 + 0.2% 215% MFI0B_ Pin 12 at Mid Supplies 99.35 20.2 6 _ MF10C_ Q = 10, fo x Q = SOkHz, Mode 1 99.35 + 0.2% Q Accuracy (Q Deviation iter) Pin 12 High (~50:1) +10 ppm/eC Pin 12 Mid Supplies (~100:1) +100 ppm/°C fo Temperature Coefficient fo X QS 100kHz, Mode 1 External Clock Temperature Independent _— — fo X Q'S 100kHz, Q Setting +500 ppm/°C Q Temperature Coefficient Resistors Temperature Independent OC Low Pass Gain Accuracy Mode 1, RY = R2 = 10k Es 2 * Crosstalk — 50 _ dB Clock Feedthrough _ [0 | mv Maximum Clock Frequency _ [is | MHz _Power Supply Current _ _ 10 ma | ELECTRICAL CHARACTERISTICS (internal op Amps) (Vg = 25V. Ty = +25°C) [panameren | CONDITIONS mn [tye [max | unis Supp Voage ~ - [ss v | Voltage Swing (Pins 1, 2, 19, 20) RL = 5k. | MF108_ 40 +41 i} Vv [_Metoc. - ra 239_ a Voltage Swing (Pins 3 and 18) R= 3.5ka MFIOB_ +40 24) v MF10C_ 138 13.9 v _ es ae <a) Output short Circuit Current | Source 3 | mA Sink . - 15 ma Gain Bangwidth Product _ 25 MHz Slew Rate _ __ 7 WS 2 KI

Feix/Fo vs @ Foux/Fo v8 Q = oo ° zw z a 0s Pa a a0 oy 7 7 0 tn 7 = cy oma 0 oma. 0 Table 1, PIN DESCRIPTION sane [rm Section DESCRIPTION ae Section| DESCRIPTION 8 B Le 20 | These are the lowpass, bandpass, CLK 11 |Clock inputs for each switched land noteh/alipass/highpass capacitor building biock The loutputs of each 2nd order section. duly cycle should be close BP 19 [The LP and BP outputs can typi- to 50% to allow the op amps ically sink 1mA and source 3mA. the maximum time to settle, Ihe N/AP/HP output can typically particularly when the clock N/AP/HP 18 __|sink 1.5mA and source 3mA. frequency is above 200kHz INV 17 _|INV is the inverting input of the 50/100/CL This three-level input pin selects| jsumming op amp of each filter. lone of three MF 10 operating This i 7 lconditions, When the 50/100’CL pin used Tinmodes 1A § and 66, pin is Connected to Vo" the ratio] This pin must be driven with a between clock frequency and low source impedance. center frequency is 50:1. with this pin at mid supplies (i.e., Swe [The Swe input controls a switch Janalog ground with dual supplies! connecting one of the inputs of ithe clock frequency to center the filter's 2nd summer — either {frequency ratio is 100:1. Tying Ito analog ground (Save low} or the pin low activates a simple ito the low pass output (Swe current limiting circuitry which lhigh). The Sav input controls [nalts normal filtering operation the configuration of both sec- land reduces the supply current tions of the MF 10. by 70%. var 7 [Analog and digital positive Vv 13 [Analog and digital negative | ‘supply inputs. Ve 14 supply inputs. These pins are Va 8 These pins are internally internally connected. V4" and 1 iconnected through the MF10's Vo should be derived from the ! substrate and therefore V," and same power supply source. | Vp" shou'? 3u derived vr3m the AGND is Analog Ground. This pin should jsame power voi se be connected to the system gro-. i 7 jLevel sri an Tnisprconsus und for dual supply operation or the digital input threshold level driven to mid supply for single ‘of the clock inputs, CLKa and supply operation. The non- [CLKa, With the level shift pin at inverting inputs of the filter op [OV and with +5V power supplies, ‘amps are internally connected |the clock inputs are TTL com- to the AGND pin, therefore patible. With the level shift pin AGND should be well bypassed Jeonnected to Vp" the clock "put thresholds are approxi- mately 2V above Vo MAXIM

outputs, but is not measured at these outputs. plex zero pair. If f,is different from fo, and Q, is. Figure 3. Highpass Filter Terminology

Dual Universal © Switched Capacitor Filter ——___—General Description oe mm om am] The MF10 is a switched capacitor (sampled data) filter. a oO CC Oo ob qi While the time domain approach most accurately ‘ describes the MF10's transfer functions, time domain we ED o = calculations are cumbersome and most circuit de- P>+-@ > > ra) signers are more familiar with the frequency domain approach used in designing RC active filters. Fortun- s 4 ately, the MF10 closely emulates RC active filters when son alo the sampling frequency is much higher than the frequency band of interest. The operation of the MF10 can then be cessribcd in terms of the fequency cine teva | NOM domain with reasonable accuracy. Specifically, each of swore | Um the two sections of an MF 10 can be treated as asecond ! order state variable filter. The similarity between the | MF10 and the classic state variable filter allows the use xt is oftheextensiveliterature available on thedesignofand | “=O ro order state variable filters. | | ! The RC second order state variable filter (Figure 5) oO i requires 3 op amps, 7 resistors, and 2 capacitors. This I filter lacks the frequency stability and tunability of the » von] | 1 MF10 switched capacitor filter. The MF10 excels in cus DJ these areas because the center frequency of a switched Stock capacitor filter is determined by the frequency of the clock, which, if crystal controlled, can achieve a a! stability of a few parts per million over the entire Operating temperature range, Having the center fe- ? quency controlled by an external digital clock fre- ” > quency also simplifies tuning of the filter since it is owe > &Q-1-> > easier to accurately control a variable modulo divider than it is to precisely vary the time constant of an RC toc ye Oe he Oe integrator. Yo Yr NAMM He ~ The MF10's maximum guaranteed operating clock frequency is 1MHz, corresponding to a 20kHz maxi- Figure 4. Block Diagram of the MF10 mum filter center frequency with a 50:1 clock to center frequency ratio, and a 10kHz maximum center fre- quency with a 100:1 clock to center frequency ratio. 2), For modes 2 and 3 only. determine the value of Des 4 using the available external cl frequency and ——_——- Filter Design the selected value for R2. (The center frequency of Simple 2nd Order Bandpass Filter Design mode 1 and 1A is determined solely by the external All modes except mode 6 offer a 2nd order bandpass clock frequency.) response. The simplest circuit, mode 1A uses only two 3) Determine the value of R3, using the desired Q external resistors, but is limited to low Q operation by and the previously determined values of R2 and R4. puiput swing imHations. Mode or high GO boninoes 4) Determine the value of R1 required to obtain the functions. The center frequency of modes 1 and 1A is desired filter gain. Getermined solely by the external clock frequency Modes 2 and 3 are also suitable for bandpass filters, and are easier toimplementin some applications since ‘Table 2. MODE SELECTION the center frequency is controlled by both theexternal [FiteRTveE_ | —=S=s”~=“‘éaODESCSC~*d clock frequency and a resistor ratio. See Table 2. Towpess Second order bandpass filter functions are charac- Highpass 33a, 6A terized by Q, center frequency, and gain (or amplitude 7 response). Resistor selection should follow these steps, ___ Bandpass 41, 1A, 2,8 34.4, using the design equations for the selected mode: Notch 1) Pick a value for R2, typically 10 to 100k2. Alipass: 46 MAXIM

© _Table 3A. NORMALIZED LOWPASS FILTER PARAMETERS Sets ere | ee ae | | = fm [ao [om To [om Toa fm 7 oa {om Ta | [0 | o7o7 | +27 | os7r | 120 [over | 1231 | oes | coor | is20 | Poa sea ae [Soars (se oe | 1,000 0.691 1,300 1.341 1,069 1.708 0.941 The normalized frequencies for the Butterworth and Bessel filters are for a -348 frequency of 1Hz. The Chebyshev and Elliptic normalized frequencies are for filters whose amplitude response passes from the ripple band to the stopband at 1Hz. Simple Lowpass Filter Design Simple Highpass Filter Design Use mode 6 or 6A if a single pole lowpass filter is Use mode 3 or 3A to implement 2nd order highpass desired (such as the odd pole in an odd-ordered fiers and mode 6 fora single pole highpass titer complex filter). Single pole resistor values are deter- Second order hi 7 ie b { ’ ighpass filter functions are charac- mined using the equations for modes 6 and 6A: terized by Q, cutoff frequency, and gain (or amplitude 1) Select a value for R2, typically 10 to 100 kn. response). Resistor selection should follow these 2) Determine R3, using the selected value of R2, _St8BS. using the design equations for the selected the available external clock frequency, and the ‘ desired cutoff frequency. 1) Pick a value for 2, typically 10 to 100 kn. 3) Determine the value of R1 to obtain the desired 2) For modes 3 and 3A, determine the value of R4 gain. using the available external clock frequency and Modes 1 and 1A, with a fixed clock to cutoff frequency, the selected value for R2 are the simplest 2nd order lowpass configurations. 3) Determine the value of R3, using the desired Q Modes 2 and 3 aliow turing of the cutoff frequency by and the previously determined values of R2and R4. chan justin either changing the clock frequency or adjusting 4) Determine the value of R1 required to obtain the. - desired filter gain. Second order lowpass filter functions are charac- terized by Q, cutoff frequency, and gain (or amplitude Multl-pole Filter Design response). Resistor selection should follow these The two 2nd order filter sections of the MF10 can be Steps, using the design equations for the selected cascaded to obtain a 4th order (4 pole) filter response. mode: - Several MF10s can be cascaded to get very high order 1) Pick a value for R2, typically 10 to 100 kn. filter responses. Unlike filters based on RC time con- 2) For modes 2.and 3 only, determine the value of __ stants, MF10-based filters usually do not require tuning Ré using the available external clock frequency __of each section since the Q and center frequencies are and the selected value for R2. (The cutoff fre- precisely controlled by the external clock frequency quency of mode 1 and 1A is determined solely by and the ratio of external resistors. the external clock frequency.) ‘The information included heres for the most common ; types of multi-pole filters: Butterworth, Bessel, Cheby- 3) Determine the value of R3, using the desired shev, and Elliptic or Cauer. The design information e 'y + given is for a 1Hz lowpass filter. However this filter can 4) Determine the value of R1 required toobtainthe —_be transformed to any desired filter type and frequency desired filter gain. using the following steps: 6 —-——eeeeSeSeSMIAXKIM

Table 3B, CAUER OR ELLIPTICAL FILTER PARAMETERS z STOPBAND EDGE FREQUENCY _ 1s 'STOPBAND EDGE FREQUENCY _ 20 STOPBAND EDGE FREQUENCY _ 30 PASSBAND EDGE FREQUENCY PASSBAND EDGE FREQUENCY PASSBAND EDGE FREQUENCY a PASSBAND RIPPLE = 0.5dB PASSBAND RIPPLE = 0.Sdb PASSBAND RIPPLE = 0.54B J ert ts tl tal) rete! © (4B) 1.266, 0.969 0.803 1247 0.737 1675 227 3 0.707 | Real Pole 28 0.693 | Real Pole HN2 0.682 | Reol Pole. 1.592 2.143 3.3233 3.478 4902 | seg 7.647 1.03 3.922 1,031 1.031 3.032 2.332, 3.251 5.008 0.426 | Real Pole 0.393 | Real Pole 0.375 | Real Pole 0.759 1.754 0.725 1,723 0.705, ATM 1) Identity the type of transfer function (lowpass, ; . Hotse’ onndsese fete): the type of (eaters real pole that should be implemented using mode 6 ora (Butterworth, Bessel, Chebyshev, etc.); the num- Simple RC section. ber of poles, and the cutoff frequency. Lowpass to Highpass Transformation. 2) Determine the normalized lowpass filter fre- The cutoff frequency and Q of each lowpass section is quency and Q of each filter section, using Table3A _"ansformed using these equations: or 3B. fa(highpass) = ————— 3) If a multi-pole transfer function other than fn(lowpass) lowpass is desired, perform the filter type trans- Q(highpass) = Q(lowpass) formation, as described below. Lowpass to Bandpass Transformation. 4) Denormalize each filter section frequency, tn, _If the ratio between the upper and lower -3dB cutoff by multiplying the fn by the actual desired cutoff or frequencies is greater than 1.5, the best way to make center frequency. the desired bandpass fiter isto cascadea lowpass titer 5) Select a mode of operation for each filter 27d @ highpass filter. The lowpass filter's cutoff tre- Section. Mode 3's suitable for mostfiters, includ- _Quency should be set to the desired bandpass upper ing Bessel and Chebyshev. Butterworth filterscan ie pane eal the highpass end A be implemented using either mode 3 or mode 1. Ruaney. snow De set to the desired bandpass filter Elliptical filters can beimplemented using modes2 [wer Cutoff frequency. ; and 3A. Modes 6 and 6A provide a single, real pole For very narrowband filters, several sections with sgededtorodd-ordered transferfunctions.allpass identical center frequencies can be cascaded, When and complex zeroes can be generated in modes 4 identical bandpass filters are cascaded, the Q of the and 5. resultant filter is g)Select a clock frequency. The ratio of clock a= TFET frequency to center Mrequency can Pe aiiwing where Qs the Q of each individual filter section, B is Tee ee Srany conveniently available clock fre. the bandwidth of each individual fiter section, and n is tency that ie approximately 20 to 200 times the tne number of identical sections cascaded. See table ¢ desired cutoff or center frequency. Table 4. CASCADING. ‘ , IDENTICAL BANDPASS FILTER SECTIONS 7) Determine the resistor values for each filter NUMBER OF section, using the design procedures given in the WENTICN, BANDPASS oe sections above for simple 2nd order bandpass, SECTIONS lowpass and highpass filters. 5 70 ime Tables 3A and 3B gives the normalized filter frequency 2 0.6448 1155.0 and Q for each second order section of a multi-pole 3 0.5108 1:96. filter. Filters with an odd number of poles have one 4 0.435 B 2300 entry with “real pole” in the Q column. This denotes a § 0.386 8 2600 MAXIM

° ________ Application Hints section's summer be grounded and the input to the 4 1) The maximum output swing is typically to one eection’s sexta soe tt Son can te contested uw within 1V of either supply rail. Check the peak By, ourptnglownass output connection made wa amplitude response gains, Hogp, Hour, Honp, and the g (or 1g) input = the input signal level to ensure that the outputs will A . not be driven beyond their maximum outputswing _8) If large input voltage signals are applied to the filter, range. This caution particularly applies to mode 1A the BC otsat voltages of the MF10 may cause when used with high Q values. The section labeled output clipping. For a more detailed discussion see "Circuit Dynamics” is included in the description the section on DC Oftsets. of each mode to clarify the relationship between 9) For best results, the positive and negative supplies the various filter parameters and the output ampli- should be bypased to AGND with a 10uF tantalum tude peaking at the filter outputs. The lower Q and 0.1uF ceramic capacitors. sections of cascaded filters should precede the Mode 1A seciions with high Q. Thistedtices the possibility Non-inverting Bandpass, 2) The absolute values of resistors arenoteritical,only This minimum component count configuration uses the ratios between resistors directly affect filter oniy Iwo external fosistors, The peak gain at the operation. The absolute values must be high inverting bandpass output is equal fo the © times the enough so that the output drive currents do not _ input voltage, so this circuit should only be used forlow approach the limits of 3mA source currentand 1A applications. The ratio of bandpass center frequency sink current. At the other extreme, resistor values to clock frequency is fixed at either 50:1 or 100:1, as should not be so high that stray leakage currents Selected by the 50/100/CL input ' and stray capacitances have a significant effect on " circuit operation. Design Equations 3) Selecting 100:1 operation doubles the number of f f samples per output cycle, and halves the number of fo= Bx or Se ‘output steps compared to 50:1 operation. On the other hand, 50:1 allows higher frequency operation _ R3 (20kHz max vs. 10kHz max), and also offers better Q= p> center frequency stability (+10ppm/*C vs. £100ppm/*C). Howp=-1 4) The minimum frequency of operation is limited by R3 the rate of discharge of the internal switched Hoept = ~ Bp capacitors. The droop rate at the output of the integrators will be approximately O.1mV/ms. This a limits the lower value of clock frequency to about Hoare = 1 (non-inverting) 100Hz for reasonable accuracy, corresponding toa center frequency of 1Hz using the 100:1 mode. Cireult Dynamics 5) For the best accuracy in setting the center fre- quency, use the corrections shown in the Typteat Hosp = Q (this is the reason for the low Q. Operating Characteristic graphs. uneee grap hs aid recommendation) in the correction for the slight interaction = or high clock frequency, Q, and center frequency. Howp (peak) = x Hower (for high Qs) 6) As with all sampled data systems, high frequency ————_—— components of the input signal above half the clock vm | rate will be aliased. In particular, input signal com- ponents with frequencies near the clock rate will i oe eas a generate difference frequencies that may fall within oO. a, 9G. the passband of the lowpass and bandpass filters. Ln Since the ratio of clock frequency to center free | O > 2) > > quency is approximately 50:1 or 100:1,asimpleone | en | pole passive RC filter will be sufficient filtering in Nomnverrna | | many cases. In many other cases the input signal vod will itself be band-limited and will not require —— © — | additional filtering. |i 7) The Save input controls the source of feedback into “ | the three input summer of both sections of the MF10. If your design requires that the input of one =a @ CU MMAXKIM

Like Mode 1A, fp is fixed at foix/50 or foix/100. The Thecircultof mode 2iscreated by addingresistorR4to = “WH gain at all three outputs is inversely proportional to the _the circuit of mode 1. This fourth resistor causes the = value of R1; and unlike Mode 1A, high Q bandpass ratio of the bandpass center frequency to clock fre- 8 filters can be bullt without exceeding the output swing quency, to be less than the fixed 50:1 or 100:1 ratio of range of the bandpass output amplifier. The notch and mode 1. Stated another way, R4 allows the center bandpass center frequencies are identical. The notch _frequency of the bandpass filter to be tuned to a higher output gain is the same above and below the notch —_ frequency while maintaining a constant clock fre- center frequency. quency. The notch frequency remains at fou4/50 or foun/100, making mode 2 suitable for elliptic highpass Design Equations filters, where the complex zero pair (fnoich) must be lower than the complex pole (fo). Design Equations il f [,, Re f [Re fo= CLK 1+ oe fouk, / 44 Re froten «fo °* 700 * Ra! “50% 1 * Ra a= B83 t= 1K g, tok ; ga F8, [TRE R2 n= 00 or B58 : en hard __R2 R2 Hore = - 82 _ Re or RT Horp=——Bt : Hopp: - 3 our Re » Mose "Ri Hoge = -83 R4 lose = - ae - 2 R2 H oe Ri Hon(as =O - ay lont (as f — 0) = ia Ra fe R2 Hon(att = fork = - 82 2) Rt Hone (at t= £6) = BS x Houp 2 Ry Circuit Dynamics Cireult Dynamics Hoge = Hop x Q = Hon x Q Hosp = 0 \\fFoip Hon, @,/Fon, x Hon, Ho pipeak) = @x Hop (if the DC gain of the LP outputis lose = QHote x Hon, = Q,/Hon, x Hon, too high, a high Q value could cause clipping at the lowpass output resulting in gain non-linearity and Me distortion at the bandpass output). 4 i Me ae Pen ve Oo ~— O a Qo so) s007 aman on sony Tain Tsay a ery nt Von Yin, LJ w% Sap se Figure 8 Mode 1 Figure 9, Mode 2 MAXIM SSSSSSSSFSFeFeFeFFFeFSSsSsSsSSsSsSSSSSsSSSsSsSsssssssssssese—F

° Mode 3 Mode 3A = Highpass, Bandpass, and Lowpass Highpass, Bandpass, Lowpass, and Notch Lu, _This mode is a sampled time (Z transtorm) equivalent Similar to mode 3. this mode adds an external op amp. of the classical 2nd order state variable filter. In this This op amp creates a notch output by summing the versatile mode, the ratio of resistors R2 and R4 can —_ highpass and lowpass outputs of the MF10. The ratio of move the center frequency both above and below the resistors Ry, and R, adjusts the notch frequency, while fcux/50 and fci«/100 values. Mode3iscommonly used R2and R4 adjust the bandpass center frequency. Since to make multiple pole Chebyshev filters with a single —_ the notch (zero pair) frequency can be adjusted to both clock frequency. A small (10-100pF) capacitor in par- above and below fo, mode 3A is suitable for both allel with R4 may be needed to avoid Qenhancement. —_ lowpass and highpass elliptic or Cauer filters. in multi- pole elliptic filters only one external op amp i needed. . se the inverting input of the internal op amp as the Desig Equations summing node for all but the final section of the filter. fo= fon x [fe or foun x ee Design Equations f Re fi [Ro _ R83, [Re fo= (OK x [FE oy fork, [RZ a> Bx 700 * Vind % “50% VRa = f3, [Re Howe = - B Q- fo* Ra = fou fe fouk Jz Hose = -B footch = “FOX af Rt OF “Bg x RE = -f2 = -B4 = -83 Hour = -B Honp= - Ry Howr= - pq Hose= ~ Ay R R Circuit Dynamics Hon (at t= fo) =|@ (BS Hour - 82 Hov)| RQ . Hop = Hove ( Ra Fa) Hon (as 1 ~ 0) = BS x Hour Hove(peak) = 2 x Hou Hopp = Qa/Honp x Hove Hon at t= fou). Rex Hone Honp(peak) = Q x Hone k] — —— a re Yoort om mo Hea St a, e-= 4 oo fs a] ma HA 8H Be thal mn jn oo aT 109 2S @ pp wal | f a ron . SAPS J scr Figure 10 Mode 3 Figure #1. Moge 9A yo MAKIN

allpass gain at -1. quency response than can be achieved with mode 4. Figure 12. Mode 4 Figure 13. Mode 5

f, (cutott frequency) = “Bek x (B) or a x(B2) for calculating output DC offsets.

52 Vosz= charge injected offset plus op amp offsat

Figure 15. Mode 68 Figure 18. MFIO Offset Mode

2 JALAL

Figure 18. 4th Order, 2kHz Lowpass Butterworth Filter

foux = 100kHz with pin 12 high. 1) Let R2 = 10k0. R2=100K. and RS = 130K. - and using 200kH7 for fox, R4 is calculated as 25.87k!. The output of the first section, LP (pin 1), isthe input The closest 1% resistor value of 26.1k01 is chosen. ratio between the clock frequency and the lowpass Usin = 1. cutoff frequency. values for R2 and R4, R3 is calculated as 79k/). 4th Order Chebyshev Lowpess Fitter The closest 1% value, 787k01 is used. with the following specifications: example Rit is 10k. The filter uses a 200kHz clock for both sections. 19.1k0. Figure 19. 4th Order Chebyshev 5kHz Lowpass Filtor -

94 SSSsSsSsSsSSssSssssss CAA AXKIM

= —— se u : I ' iu t= tee ssscheat ie A ~ | ee) TR ei Oy yc ee Bi al ve | ™ oT te} tsi Mi sseste aon at = eae TS oe PA i rsy

Package Information

: eo aa | e174 * [arfoow | = [ose |] TH a3 [az [ors | ois | ane | aad | rae leer 2} 4 h 0.065 yn sale ots fh Pe pose toses[ rae bea | L [oico | - [ese] = | 5 “ ” [ex os00 [=a o—_lertihe ae “ej eon 0.150 | 290 | 36 =[_INCHES — (MILLIMETERS Plastic DIP [pra] om |rweai wc un HA : PLASTIC = [P10 | @ Jose [0300 ae | 901] | Pp [D | 16 Jo74s [0765 [18.92 [19.43] PACKAGE [spre lomeosetees top| [w[o yes [ais [res [28.06 [213] mpbod. Man reserves bo rn sige fe ciety and soanscabons uber nencans a ehe

16 Maxim Integrated Products, 120 San Gabriel Drive, Sunnyvale, CA 94086 (408) 737-7600

© 1996 Maxim Integrated Products Printed USA MAXIM is 2 registered trademark of Maxim Integrated Procucts.