analog devices high accuracy, low cost, fet -input operational amplifier features low cost: $4.45 (100's, j) low ib: 25pa max (k) low vas: 20m v max (k) low vas drift: 25/lvtc max (k) high differential input voltage capability: :t20v product description the ad540 is the lowest cost, high accuracy fet-input op amp available which provides the user with low bias currents, high overall performance, and accurately specified predictable operation. the device offers maximum bias currents as low as 25pa, offset voltages below 20mv, maximum offset voltage drift below 2511v/c and a minimum gain of 50,000. all devices are free from latchup and are short-circuit protected. no external compensation is required as the internal 6db/octave roll-off provides stability in closed loop applications, the ad540 is suggested for all fet~input amplifier require- ments where low cost is of prime importance. its performance is comparable to modular fet op amps, but its ic construction reduces the price significantly below that of modules. all versions of the ad540 are supplied in the hermetically- sealed, 8-pin, to-99 package. the ad540j and ad540k are specified for 0 to +70c applications, while the ad540s is offered for operation over the full military temperature range of -55 to +125c. product highlights 1. the ad540 op amp meets specified input bias current and offset voltage values after full warm-up. conventional high speed ic testing does not allow for self-heating of the ch~p due to internal power dissipation under operating conditions. i nformation furnished by analog devices is believed to be accurate and reliable. however, no responsibility is assumed by analog devices for its use; nor for any infringements of patents or other rights of third parties which may result from its use. no license is granted by implica- tion or otherwise under any patent or patent rights of analog devices. 2. the bias currents of the ad540 are specified as a maximum for either input. conventional ic fet op amps generally specify bias currents as the average of the two input currents. 3. unlike many fet-input op amps, the ad540 allows a maximum differential input voltage of :t20vdc. standard "bootstrapped" fet-input op amps permit maximum differential input voltages of only about :t3v. 4. offset nulling of the ad540 is accomplished without affecting the operating current of the fet's and results in relatively small changes in temperature drift characteristics. the additional drift induced by nulling is only :t2.0/lv 1c per millivolt of nulled offset, compared to several times this for other ic fet opamps. 5. the gain of the ad540 is measured with the offset voltage nulled. nulling a fet-input op amp can cause the gain to decrease below its specified limit. the gain of the ad540 is fully guaranteed with the offset voltage both nulled and unnulled. 6. to maximize the reliability inherent in ic construction, every ad540 is stored for 40 hours at +150c, temperature cycled from -65 to +125c and receives a high impact shock test. route 1 industrial park; p.o. box 280; norwood, mass. 02062 tel: 617/329-4700 twx: 710/394-6577 west coast tel: 213/595-1783 mid-west tel: 312/297-8710 ~-- obsolete
specifications (typical @ +25c and vs = :!::15vdc unless otherwise specified) model open loop gain (note 1) vout = :!::10v, rl > 2kil ta = min to max output characteristics voltage @ rl = 2kil, ta = min to max voltage@ rl = 10kil, ta = min to max short circuit current frequency response unity gain, small signal full power response slew rate, unity gain input offset voltage (note 2) vs. temperature vs. supply, t a = min to max input bias current either input (note 3) input impedance differential common mode input voltage range differential (note 4) common mode common mode rejection, vin = :!::10v power supply rated performance operating quiescent current temperature range operating, rated performance storage price (1-24) (25-99) (100-999) note: 1. open loop gain is specified with vos both nulled and unnulled. 2. input offset voltage specifications are guaranteed after 5 minutes of operation at t a = +2soc. 3. bias current specifications are guaranteed after 5 minutes of operation at t a = + 25c. for higher temperatures, the current doubles every looc. 4. defined as voltage between inputs, such that neither exceeds :!::lov from ground. * specifications same as ad540]. ** specifications same as ad540k. specifications and prices subject to change without notice. pin configuration tab outune dimensions dimensions shown in inches and (mm). ininep~t~ng ( 2 v+ ref plane 0.335 (8.501 0370] 0.02910.741 0.04511.141 0.028{0.711 0.034 10.861 offset v- offset ~l top view ~ v- to.' to-99 -7- ad540j ad540k ad540s 20,000 min 50,000 min ** 15,000 min 25,000 min ** :!::10v min (:!::13v typ) * * :!::12v min (:!::14v typ) * * 25ma * * 1.0mhz * " 100khz * * 6.0v //lsec * * 50mv max 20m v max ** 75/lv/oc max 25/lv/oc max 50/lvtc max 400/lv /v max 300/lv n max ** 50pa max 25pa max ** 1010illl2pf * * 101 1 illl2pf * * :!::20v * * :!::10v min (:!::12v typ) * * 70db min * * :!::15v * * :!::(5 to 18)v * * 7ma max (3ma typ) * * -- 0 to +70c * -55 to +125c -65 to +150c * * $6.45 $8.95 $14.95 $5.70 $7.20 $11.95 $4.45 $5.95 $ 9.95 obsolete
applying the ad540 the ad540 is especially designed for low cost applications involving the measurement of low level currents or small voltages from high impedance sources in which bias current can be a primary source of error. input bias current contributes to error in two ways: (1) in current measuring configurations the bias current limits the resolution of a current signal; (2) the bias current produces a voltage offset which is proportional to the value of input resistance (in the case of an inverting configuration) or source impedance (when the non-inverting "buffer" connection is used). the ad540 fet-input operational amplifier is, therefore, of use where small currents are to be measured or where relatively low voltage drift is necessary despite large values of source resistance. applications considerations bias currents. most ic fet op amp manufacturers specify maximum bias currents as the value immediately after turn-on. since fet bias currents double every 10c and since most fet op amps have case temperature increases of 15c to 20c above ambient, initial "maximum" readings may be only ~ of the true warmed-up value. furthermore, most ic fet op amp manufacturers specify ib as the average of both input currents, sometimes resulting in twice the maximum bias current appearing at the input being used. the total result is that 8x the expected bias current may appear at either input terminal in a warmed-up operating unit. the ad540 specifies maximum bias current at either input after warm-up, thus giving the user the values he expected. improving bias current beyond guaranteed values. bias currents can be substantially reduced in the ad540 by decreasing the junction temperature of the devices. one technique to accomplish this is to reduce the operating supply voltage. this procedure will decrease the power dissipation of the device, which will in turn result in a lower junction temperature and lower bias currents. the supply voltage effect on bias current is shown in figure 1. ,.0 1 ..s>!; 0.9 1 , !z 0.8 w a: - ~ i;; 0.7 0 - ~ :' 0.6 '" . c> 0.5 w 0 n '" ~ '" 0.4 :;; ~ 0.3 z 0.2l 0 5 10 supply voltage - ! vs-volts 15 figure 1. normalized bias current vs. supply voltage operation of the ad540k at :t5v reduces the warmed-up bias current by 70% to a typical value of 8pa. a second technique is the use of a suitable heat sink. wakefield engineering series 200 heat sinks were selected to demonstrate this effect. the characteristic bias current vs. case temperature above ambient is shown in figure 2. bias current has been normalized with unity representing the 25c -- free air reading. note that the use of the model 209 heat sink reduces warmed-up bias current by 60% to 10pa in the ad540k. 1.0 0 0.9 ..s'1~ !z - 0.8 w" ~ ~ 0.7 0 !;( ~ ~ 0.6 '" 0 c z 0.5 w:>: n i- ~ 0.4 :;; 0 ~,'f 0.3 wakefield model numbers- 0.2 0 5 10 15 20 25 30 case temperature above 25c-oc vs ! 15v figure 2. normalized bias current vs. case temperature both of these techniques may be used together for obtaining lower bias currents. remember that loading the output can also affect the power dissipation. ~ 100 i '" h i- z w a: a: ::> 0 u) " jjj 10 i- ::> a. ~ 25 50 ambient temperature ta-oc vs= ! 15v 70 figure 3. input bias current vs. temperature n~1 -5 0 10 figure 4. bias current vs. common mode voltage. vcm-volts input considerations. unlike some fet-input operational amplifiers, the ad540 accommodates differential input voltages of up to :t20v without any degradation in bias current. in certain time-dependent applications, such as charge amplifiers and integrators, large differential input voltages temporarily occur which may exceed the rated value of a typical fet op amp (approximately :t3v differential). -':/.. / i / / / / v / obsolete
by utilizing un-bootstrapped fet's at the inputs, the ad540 assures the user of expected performance at large differential input voltages without the use of protective diodes or resistors. offset voltage drift. most commercially available ic fet op amps are nulled by adjusting the fet operating currents, causing the offset voltage temperature coefficients to vary 3 to 61lv / c per millivolt of offset nulled. thus a conventional fet op amp with 20mv initial offset, when nulled may display an additional offset drift of 60 to 120ilv/c, in addition to its nulled value. the ad540 achieves nulling without disturbing the operating currents of the fet's, thus allowing a substantial reduction in drift. figure 5 graphically displays the offset drift perform- ance of the ad540, nulled and unnulled. as can be seen, nulling the device can result in either positive or negative offset drifts given by the slope l1 vosl l1 t. the nulled curves represent the maximum changes in drift, indicating perform- ance considerably better than many other ic fet op amps which null vos by varying the operating currents of the fet's. 1.9 1.7 1.5 > e 1.1 i ~ n i- 0: 0 ~ n 0 i- .1 " ~ -.1 ~ -.3 :;; a: 0 -.5 "- .5 temperature loci typical nulled typical unnulled " -.7 > -.9 -1.1 -1.3 -15 -1.7 -19 figure 5. vos vs. temperature noise performance. the noise spectral density vs. frequency for the ad540 is given in figure 6. the curve shows approximately 300nv/vhz at 10hz, declining in a lif fashion (llf for power, livf for voltage) to approximately 12nv/vhz at higher frequencies. current noise in the ad540 is approximately 0.001pa/vhz at low frequencies. above 300hz, the current noise generated by the op amp increases at a 3db/octave rate, determined by wencin, where en = spectral noise density and cin = input capacitance. in most practical applications, the current noise from source or feedback resistors will be larger than the low frequency current noise from the amplifier. at high frequencies, the total circuit curre.nt noise is equal to wenc, where c is the sum of all input and feedback capa{;itors. in well-shielded circuits, c is usually 10 to 100pf, so that the wenc can be a significant factor. thus the user should attempt to minimize c. > 10k (ij ~ " ;! 1k'---- ~~ q.> '" c w 100 '" (5 z c . 10 1 10 100 lk frequency -hz figure 6. noise spectral density vs. frequency dynamic performance. the ad540 is internally compensated to achieve a -3db bandwidth of 1mhz (see figure 7). at unity gain the full power bandwidth is 50khz minimum, and typically 100khz. slew rates are 3v illsec minimum and 6viilsec typical (see figure 8 and figure 9). 11 1.0 look 10k 1m 100 lk 10 frequency -hz figure 7. small signal gain vs. frequency +10 0 "0 > +5 i 0 > i w cj 0 '" .... ~ > -5 ::> q. .... ::> " -10 012345678910111213 time-/.is voltage follower step response figure 8. voltage follower step response 30 25 0 "0 > i 20 ::> q. .... ::> 0 " 15 '" ~ 0 10 " '" ~ 10 100 1k 10k frh1uency -hz 10()k 1m figure 9. p-p output vs. frequency ---- --- 106 los > ;; i 10' z ;;;: '"' q. 103 0 0 ...j z 102 w q. 0 10 // \ \ / \ / / \ \ i rl =2kn i i cl=o , vs=t15v \ / " 1/ " vs=t15v ; v s=t9v \ rl =2kn ,\ ; obsolete
|
