1/4 this application note explains how to implement the ts971 as a microphone pre-amplifier for an electret condenser microphone (ecm). this type of microphone has one of the best price to perfor- mance ratio on the market. microphone pre-amplifiers are very common in to- days appliances, digital appliances have adopted and kept the typologies used in analog ones. this block is helping to interface the microphone to the a/d converter by buffering, filtering and amplifying the microphone signal. 1 - device presentation many low noise amplifiers exist in the market. old- ies (but still goodies!) are mainly dual ones like lm833, mc33078. but few are able to reach low voltages and are not available in todays smallest packages. the ts97x is a family including single, dual and quad low-noise operational amplifier. it features excellent audio characteristics: low distortion (0.003% thd @ f=1khz) and a 4nv/sqrt(hz) equivalent input noise voltage with a 1/f corner @ 100hz. thanks to those characteristics, it helps keeping an optimal signal to noise ratio, a critical point at the entry of the audio amplification chain! these devices also allows a higher fidelity thanks to a 4v/ m s slew rate and 12 mhz gain bandwidth product. this enables the amplifier to cope with quick variations of the input signal well over the audio bandwidth. the family is available in compact packages like sot23-5 for ts971 or even the thin and rather compact package like tssop for ts972/4. this allows them to be used in portable and miniature digital appliances like pda or cellular phones and also in thin notebook computers. table 1: acoustic units reference table 2 - microphone considerations preliminary knowledge of acoustic intensity (in watt/m 2 ), acoustic pressure (in pascal or pa) and sound pressure level or spl (in decibels or db) is important. you can report to table 1 for more in- formation. ecm microphones follow more or less the same characteristics, however gain and surrounding components may vary from one model to another. acoustic intensity acoustic pressure sound pressure level activity i in w/m 2 p in pa spl in db 1.e+02 200 140 ear damage 130 1.e+00 20 120 110 night club, factory floor 1.e-02 2 100 90 1.e-04 0.2 80 70 conversation 1.e-06 0.02 60 50 1.e-08 0.002 40 30 recording studio 1.e-10 0.0002 20 10 i ref = 1.e-12 p ref = 0.00002 0 minimum level of audition, reference level ) db ( ) i i ( log 10 spl ref = ) db ( ) p p ( log 20 spl ref = AN1534 application note ts971 based electret condenser microphone amplifier march 2002 by r.cittadini & f.poulin
AN1534 - application note 2/4 we selected a popular model from panasonic: the wm-60a series. its an omni-directional micro- phone with the following main characteristics: q operating: from 2 to 10v. q sensitivity: -44db +/- 5db (0db=1v/pa). q impedance: less than 2.2k w q s/n ratio: more than 58db q current consumption: 0.5ma max q recommended load resistor: r l : 2.2k w the sensitivity of the microphone defines its gain as per the following formula: so with a -44db sensitivity, we can conclude that the gain of the microphone is 0.0063v/pa or 6.3mv/pa. with this value, we can get a good idea of the output voltage of the microphone. it would be around 12.6 m v for a quiet room (2mpa or 40db) and would reach approximately 6.3mv for the climax of a symphonic orchestra (1pa or 110db). this sound data is with a source at 1 meter from the microphone. this reference is mandatory, the distance between the microphone and the audio signal is illustrated by the acoustic intensity (in watt/m 2 ). lets take the example of a conversation. its equivalent to roughly 20mpa or 60db spl at 1 meter. so an acoustic intensity of 1 m w and 126 m v at the output of the microphone. the intensity de- creases with the squared value of the distance be- tween the source and the microphone. so for a distance of 5cm, you would get a value of 400 m w. as per the table 1 formulas, we can calculate the "equivalent spl value at 1m": 86db, then we get the acoustic pressure: 0.4pa which gives us the output of the microphone: 0.4 x 0.0063 = 2.52mv (distance divided by 20 and output voltage increased by the same ratio). we can summarize these considerations into the following checklist: q what type of signal do you want to amplify? q how powerful?at what distance? q what are the minimum and maximum of each above parameters? with these values, you will be able the calculate the microphones output voltage range and be able to choose the right gain of the amplifier here- after. also, if you want to implement a noise canceling function, you can also choose another type of mi- crophone called bi-directional microphone or noise canceling microphone. 3 - components calculation lets look now on how to implement such an am- plifier with ts971. you can refer to schematic on figure 1 hereafter. weve chosen a non-inverter typology to exploit to the best the low noise char- acteristics of the device. indeed, with an inverter configuration, the input resistor adds significant noise to the application. first, lets look on the behavior in dc mode. the first goal is to polarize the electret condenser mi- crophone. by using r 1 and r 2 , we can polarize it around vcc/2 as per below formula:. the only criteria is that this current must remain below 0.5ma over the supply range (otherwise, you can increase r1 value). r 1 is also acting together with c 1 as a filter for the power supply line of the microphone. then in ac mode, c 1 is fixing the gain of the microphone by allowing only r 2 to act (and not r 2 +r 1 as c 1 is equivalent to a short circuit to the ground). and r 2 must equal r l =2.2k w for the microphone weve chosen. in ac mode, this type of microphone can be simplified and compared to a current source in parallel with r 2 , hence a voltage source. then to avoid extra offset drift due to bias current mismatching, following resistor values need to comply with the following rule: the second step is, still in dc mode, to polarize the reference pin of the ts971. its the inverting pin here that will be set at vcc/2 by the r5 and r6 bridge. c4 adds here additionnal filtering of this reference voltage. this configuration allows the bi- asing or the "centring" of the signal at mid-supply voltage. hence it allows to maximize the swing within the supply voltage range. this bias voltage just needs to be kept within v icm range. this means v icm or common mode input voltage must be at least 1.15v inside the supply voltage rails, i.e. from vdd+1.15 to vcc-1.15v. ) pa / v ( 10 g ) 20 y sensitivit ( mike = ) a ( ) r r ( 2 vcc i 2 1 mike pol + ? - ) ohms ( r r r r r r r 6 5 6 5 3 4 8 + + + ?
AN1534 - application note 3/4 figure 1: ecm microphone pre-amp. with ts971 this bridge can also be supplied by an asics "vref out" pin or by another operator of the op-amp connected as a buffer (using the ts972, a dual op-amp, this can be implemented easily). an important note on the impedance of the ampli- fier is that, in ac mode, r 4 is equivalent to the in- put impedance of the amplifier stage. it must not be too small to avoid the collapsing of the micro- phone signal! the coupling capacitor (c 6 ) makes this application universal, however, you could omit it when attack- ing an a/d converter. in this case, you only have to adapt the bridge set by r 5 & r 6 to match the in- put voltage range of the converter. thanks to its rail-to-rail output, the ts971 simplifies the pro- cess. were coming now to the filter definition, when looking at figure 1 , we can see three filters: two high pass and a low pass. each has a 6db/octave attenuation factor. the high pass filter is built by c 5 , r 4 and also r 2 . the theoretical formula to calculate f cl or the lower cutoff frequency (here approximately 79 hertz) is the following another high pass filter is made by c 7 and r 7 . the cutoff frequency is better set at a lower value than f cl to have a stronger reduction (i.e. -12db/oc- tave) of low frequencies (here 59hz). then for the low pass filter (optional), to calculate f ch or the higher cutoff frequency (here approx- imately 10.7khz), we have the following formula:: the next step is to configure g or the gain of the amplifier (here 90 or +39db): ts971 r8 82k r7 820r r6 100k r5 100k r1 1k r2 2.2k + c7 3.3u + c1 10u + c6 1u c5 100n + c2 10u vcc = 2.7v min. vcc c8 180p electret condenser microphone output + c4 2.2u + c3 4.7u r3 15k r4 18k r 100r c 47p optionnal emi filter emi emi () ) hz ( c r r 2 1 f 5 4 2 1 cl + p ? ) hz ( c r 2 1 f 7 7 2 cl p ? ) hz ( c r 2 1 f 8 8 ch p = microphon e the of gain ) r r r ( ) r r 1 ( g 2 4 4 7 8 + + ?
AN1534 - application note 4/4 this represents the gain of the amplifier in the "non-filtered" bandwidth area. r 2 and r 4 have to be taken into consideration as they establish a voltage divider. changing r 7 value helps to modi- fy the gain without being forced to change the val- ues of the other components (apart from the high pass filter with c 7 ). what we refer to as the "gain of the microphone" is the gain we talk about in the "microphone consid- erations" chapter. you need to consider the signal fed to the microphone to avoid saturation or insuf- ficient gain. to be completely exhaustive, if you use c 6 , you have to consider that it creates together with the impedance of the load z l a high pass filter with cutoff frequency defined like above (replacing r 4 with z l and c 5 with c 6 . 4- special note on emi as you can notice on figure 1, we have suggested an optional electro magnetic interference filter (cutting frequencies above 33.8mhz). with cellu- lar phones being a constant company to our daily life, high power rf interferences need to be prop- erly managed into sensitive amplification devices. this filter has to be implemented close enough to the microphone and we also have to pay attention on not having too long connecting wires (anten- nas!). ideally this application should be wired on a double sided pcb with one side acting as ground plane. 5 - components calculation so, in this example, we have a gain value of 90 or +39db and we have globally a band pass filter with f cl ~60-80hz and f ch =10.7khz. concerning the overall gain, we also have also to consider the stage after the pre-amplifier. if its a codec or a/ d converter, it will most probably have its own am- plification (usually +20db gain). all together, this gives the below chain for the chapter 1 example: this configuration is well adapted to battery pow- ered equipment as the overall maximum con- sumption of this application is 3.3ma (0.5ma for the microphone and 2.8ma max for ts971). stage \ v out at 1 meter at 5 cm microphone output 0.13mv (126 m v) 2.5mv ts971 output (+39db) 0.11mv (11.34mv) 0.23v codec output (+20db) 0.11v 2.27v information furnished is believed to be accurate and reliable. however, stmicroelectronics assumes no responsibility for the consequences of use of such information nor for any infringement of patents or other rights of third parties which may result f rom its use. no license is granted by implication or otherwise under any patent or patent rights of stmicroelectronics. specificati ons mentioned in this publication are subject to change without notice. this publication supersedes and replaces all information previously supplied. stmicroelectronics products are not authorized for use as critical components in life support devices or systems without express written approval of stmicroelectronics. ? the st logo is a registered trademark of stmicroelectronics ? 2002 stmicroelectronics - printed in italy - all rights reserved stmicroelectronics group of companies australia - brazil - canada - china - finland - france - germany - hong kong - india - israel - italy - japan - malaysia malta - morocco - singapore - spain - sweden - switzerland - united kingdom - united states ? http://www.st.com
|
