4-1 tm an9903.1 1-888-intersil or 321-724-7143 | intersil and design is a trademark of intersil corporation. | copyright ?intersil corporation 2000 unislic14 and the texas instruments tcm38c17 quad combo reference design using the unislic14 and the texas instruments tcm38c17 quad combo the purpose of this application note is to provide a reference design for the unislic14 and texas instruments tcm38c17 quad combo. the network requirements of many countries require the analog subscriber line circuit (slic) to terminate the subscriber line with an impedance for voiceband frequencies which is complex, rather than resistive (e.g. 600 ?) . the unislic14 accomplishes this impedance matching with a single network connected to the z t pin. the tcm38c17 quad combo has a programmable receive output ampli?r to adjust the output gain into the slic. the output ampli?r gain is programmed with two simple resistors. transhybrid balance is achieved via the tcm38c17 gsx ampli?r. discussed in this application note are the following: 2-wire 600 ? impedance matching. receive gain (4-wire to 2-wire) and transmit gain (2-wire to 4-wire) calculations. transhybrid balance calculations. reference design for 600 ? 2-wire load. reference design for china complex 2-wire load. impedance matching impedance matching of the unislic14 to the subscriber load is important for optimization of 2 wire return loss, which in turn cuts down on echoes in the end to end voice communication path. it is also important for maintaining voice signal levels on long loops. impedance matching of the unislic14 is accomplished by making the slic s impedance (z slic , figure 1) equal to the desired terminating impedance z 0 , minus the value of the protection resistors (z tr = z o ). the formula to calculate the proper z t for matching the 2-wire impedance is shown in equation 1. equation 1 can be used to match the impedance of the slic and the protection resistors (z tr ) to any known line impedance (z o ). figure 1 shows the calculations of z t to match a resistive and 2 complex loads. example 1: calculate z t to make z tr = 600 ? in series with 2.16 f. r p = 30 ? . z t = 108k ? in series with 0.0108 f. note: some impedance models, with a series capacitor, will cause the op-amp feedback to behave as an open circuit dc. a resistor with a value of about 10 times the reactance of the z t capacitor (2.16 f/200 = 10.8nf) at the low frequency of interest (200hz for example) can be placed in parallel with the capacitor in order to solve the problem (736k ? for a 10.8nf capacitor). example 2: calculate z t to make z tr = 200 + 680//0.1 f r p = 30 ? . z t = 28k ? in series with the parallel combination of 136k ? and 500pf. z t 200 z tr 2r p () ? = (eq. 1) z t 200 600 1 j 2.16x10 6 ----------------------------------- 2 () 30 () + ?? ?? = (eq. 2) z t 200 200 680 1j 680 0.1 () x10 6 + -------------------------------------------------------- - 2 () 30 () + ?? ?? = (eq. 3) figure 1. impedance matching tip ring v rx v tx z o v 2w vs z slic z tr v tr + - + - z t z t ptg floating z t r p 30 ? r p 30 ? resistive z o = z tr = 600 ? z t = 200(600 - 2*30) z t 108k ? z t complex z o = z tr = 200 ? + 680//0.1 f z t = 200(200 - 2*30)+200(680) // 0.1 f/200 z t 28k ? 136k ? 500pf intersil unislic14 z t complex z o = z tr = 600 ? + 2.16 f z t = 200(600 - 2*30) + 2.16 f/200 z t 108k ? 10.8 nf 736k ? application note september 2000 authors: don lafontaine, chris ludeman
4-2 slic in the active mode figure 2 shows a simpli?d ac transmission model of the unislic14 and the connection of the tcm38c17 to the slic. circuit analysis of the unislic14 yields the following design equations: substitute equation 5 into equation 6 substitute equation 7 into equation 8 substitute equation 9 into equation 10 substituting -v 2w /z l into equation 11 for i m and rearranging to solve for v 2w results in equation 12. where: v rx = the input voltage at the v rx pin. v a = an internal node voltage that is a function of the loop current detector and the impedance matching networks. i x = internal current in the slic that is the difference between the input receive current and the feedback current. i m = the ac metallic current. r p = a protection resistor (typical 30 ? ). z t = an external resistor/network for matching the line impedance. v tr = the tip to ring voltage at the output pins of the slic. v 2w = the tip to ring voltage including the voltage across the protection resistors. z o = the line impedance. z tr = the input impedance of the slic including protection resistors. v a = i m 2r s 1 80k --------- - 200 z tr 2r p () 5 (eq. 4) v a i m 2 ------- z tr 2r p () = (eq. 5) v rx 500k ------------ - - v a 500k ------------ - =i x node equation (eq. 6) at unislic14 v rx input i x v rx 500k ------------ - - i m z tr 2r p () 1000k ----------------------------------------- = (eq. 7) i x 500k - v tr +i x 500k = 0 loop equation (eq. 8) at unislic14 feed amplifier and load v tr 2v rx i m z tr 2r p () = (eq. 9) v 2w -i m 2r p +v tr = 0 loop equation (eq. 10) at tip/ring interface v 2w i m z tr 2v rx = (eq. 11) 2w 1 z tr z l ---------- - + ?? ?? ?? 2 v rx = (eq. 12) v rx tip ring z tr e g v tr i m v tx r p r p + - + - - + v rx + - i m z o figure 2. unislic14 simplified ac transmission circuit and tcm38c17 + - 500k r s + - 500k r s z t 500k 500k 500k 500k ptg + - i x v a = i m (z tr -2r p ) i x i x + - i x + - + - + - i m + - 20 ? 20 ? 1/80k = 200 (z tr - 2r p ) a = 1 i x 2 r int 20 ? r int 20 ? v 2w + - pwro+ r f r 1 anlgin+ anlgin- r 2 texas gsx v in pcmin pcmout aref + - - + + - + - instruments tcm38c17 gsr pwro- 100nf 5 intersil r a1 r a2 floating unislic14 v tx application note 9903
4-3 receive gain (v in to v 2w ) 4-wire to 2-wire gain is equal to the v 2w divided by the input voltage v in , reference figure 3. the gain through the tcm38c17 is programmed to be 1.0 using equation 13, where v in = v pcmin = v pwro+ = v rx . the input and output gain adjustments are discussed in detail in pcm codec / filter combo family: device design-in and application data [1]. the maximum output (gain =1) can be obtained by maximizing r 1 and minimizing r 2 (figure 2). this can be done by letting r 1 = infinity and r 2 = 0, as shown in figure 3. the receive gain is calculated using equation 12 and the relationship z t = 200(z tr -2r p ). equation 14 expresses the receive gain (v in to v 2w ) in terms of network impedances. notice that the phase of the 4-wire to 2-wire signal is 180 o out of phase with the input signal. transmit gain across unislic14 (e g to v tx ) the 2-wire to 4-wire gain is equal to v tx /e g with v rx = 0, reference figure 2. from equation 9 with v rx = 0 substituting equation 16 into equation 15 and simplifying. by design, v tx = -v tr , therefore, a more useful form of the equation is rewritten in terms of v tx /v 2w . a voltage divider equation is written to convert from e g to v 2w as shown in equation 19. rearranging equation 19 in terms of e g , and substituting into equation 18 results in an equation for 2-wire to 4-wire gain that s a function of the synthesized input impedance of the slic and the protection resistors (z tr ). notice that the phase of the 2-wire to 4-wire signal is in phase with the input signal and that the gain will always be less than one because of the protection resistors. transmit gain across the system (v 2w to v pcmout ) 2-wire to 4-wire gain is equal to the v pcmout voltage divided by the 2-wire voltage v 2w , reference figure 3. v pcmout is only a function of v tx and the feedback resistors r a1 and r f equation 22. this is because v in is considered ground for this analysis, thereby effectively grounding the v pwro+ output. g pcmin pwro ? () r 1 r 2 + 4r 2 r 1 4 ------ - + ?? ?? ------------------------------ = (eq. 13) g 4-2 = v 2w v in ----------- - =-2 z o z o +z tr -------------------------- - 2 z o z o z t 200 --------- - 2r p + ?? ?? + ----------------------------------------------- = (eq. 14) e g z l i m 2r p i m v tr + + 0 = loop equation (eq. 15) v tr i m z tr 2r p () = (eq. 16) e g i m z l z tr + () = (eq. 17) g 2-4 = v tx e g ---------- = i m z tr 2r p () i m z l z tr + () --------------------------------------- - z tr 2r p () z l z tr + () --------------------------------- = (eq. 18) v 2w = z tr z tr z l + ------------------------ ?? ?? ?? e g (eq. 19) g 2-4 = v tx v 2w ----------- - = z tr -2r p z tr ----------------------------- z 0 -2r p z 0 ------------------------ - = (eq. 20) g 24 v pcmout v 2w ---------------------------- - = (eq. 21) figure 3. receive gain g(4-2), transmit gain (2-4) and transhybrid balance tip ring intersil v rx v tx r p r p z o pwro+ 30 ? 30 ? v 2w + - anlgin+ anlgin- r a2 texas v in pcmin pcmout aref + - + - + - gsr pwro- 100nf + - v tx + - ptg floating r a1 r f z t z t unislic14 instruments tcm38c17 application note 9903
4-4 an equation for the system transmit gain is achieved by substituting equation 20 into equation 22. to achieve a transmit gain of one (v pcmout / v 2w ), make r f = z o and r a = (z 0 -2r p ). actual values of r a1 and r f were multiplied by 100 to reduce loading effects on the gsx opamp. transhybrid balance g(4-4) transhybrid balance is a measure of how well the input signal is canceled (that being received by the slic) from the transmit signal (that being transmitted from the slic to the codec). without this function, voice communication would be dif?ult because of the echo. the signals at v pwro+ and v tx (figure 3) are opposite in phase. transhybrid balance is achieved by summing two signals that are equal in magnitude and opposite in phase into the gsx ampli?r. transhybrid balance is achieved by summing the pwro+ signal with the output signal from the unislic14 when proper gain adjustments are made to match v pwro+ and v tx magnitudes. for discussion purpose, the gsx ampli?r is redrawn with the external resistors in figure 4. the gain through the gsx ampli?r from v pwro+ is set by resistors r a2 and r fr .the gain through the gsx ampli?r from v tx is set by resistors r a1 and r f . transhybrid balance is achieved by adjusting the magnitude from both v pwro+ and v tx so their equal to each other. reference design of the unislic14 and the tcm38c17 with a 600 ? load the design criteria is as follows: 4-wire to 2-wire gain (vp cmin to v 2w ) equal 0db 2-wire to 4-wire gain (v 2w to vp cmout ) equal 0db two wire return loss greater than -30db (200hz to 4khz) rp = 30 ? figure 5 gives the reference design using the intersil unislic14 and the texas instruments tcm38c17 quad combo. also shown in figure 5 are the voltage levels at speci? points in the circuit. impedance matching the 2-wire impedance is matched to the line impedance z 0 using equation 1, repeated here in equation 24. for a line impedance of 600 ? , z t equals: the closest standard value for z t is 107k ? . transhybrid balance (z l = 600 ? ) the internal gsx ampli?r of the tcm38c17 is used to perform the transhybrid balance function. transhybrid balance is achieved by summing two signals that are equal in magnitude and opposite in phase into the gsx ampli?r. from equation 23, repeated here in equation 26, a transmit gain (v pcmout / v 2w ) of one is achieved if we make r f = z o and r a1 = (z 0 -2r p ). actual values of r a1 and r f were multiplied by 100 to reduce loading effects on the gsx op-amp. closest standard value for r f is 60.4k ? closest standard value for r a1 is 53.6k ? the tcm38c17 receive gain is programmed to 1.0 by maximizing r 1 and minimizing r 2 resistor values (figure 2). the gain from pwro+/v rx through the slic at v tx is 1.1 (eq. 31 in the intersil unislic14 data sheet). to achieve transhybrid balance from the pwro+ pin to pcmout set r a2 = r a1 x 0.9. closest standard value for r a2 = 48.6k ? . speci? implementation for china the design criteria for a china speci? solution are as follows: desired line circuit impedance is 200 + 680//0.1 f receive gain (v 2w /v pcmin ) is -3.5db transmit gain (v pcmout /v 2w ) is 0db 0dbm0 is defined as 1mw into the complex impedance at 1020hz ? p = 30 ? v pcmout v tx r f r a1 ---------- = (eq. 22) g 24 v pcmout v 2w ---------------------------- - = z 0 -2r p () z 0 ----------------------------- - r f r a1 ---------- = (eq. 23) gsx - + r a1 r f anlgin+ v gsx figure 4. transhybrid balance circuit anlgin- r a2 v tx v pwro+ z t 200 z tr 2r p () ? = (eq. 24) z t 200 600 60 () ? 108k ? == (eq. 25) v pcmout v 2w z 0 -2r p () z 0 ------------------------------------------ - r f r a1 ---------- = (eq. 26) r f z 0 = 600 100 () 60k ? == (eq. 27) r a1 z 0 -2r p () = 600 - 60 () 100 () 54k ? == (eq. 28) r a2 53.6k () 0.9 () 48.24k ? = = (eq. 29) application note 9903
4-5 figure 6 gives the reference design using the intersil unislic14 and the texas instruments tcm38c17 quad combo. also shown in figure 6 are the voltage levels at speci? points in the circuit. these voltages will be used to adjust the gains of the network. adjustment to get -3.5dbm0 at the load referenced to 600 ? the voltage equivalent to 0dbm0 into 811 ? (0dbm0 (811 ?) ) is calculated using equation 30 (811 ? is the impedance of complex china load at 1020hz). the gain referenced back to 0dbm0 (600 ?) is equal to: the adjustment to get -3.5dbm0 at the load referenced to 600 ? is: the voltage at the load (referenced to 600 ? ) is given in equation 33: impedance matching the 2-wire impedance is matched to the line impedance z 0 using equation 1, repeated here in equation 34. for a line impedance of 200 + 680//0.1 f, z t equals: z t = 28k ? in series with the parallel combination of 136k ? ( closest standard value for is 137k ? and 500pf (closest standard value for is 470pf). to achieve a 4-wire to 2-wire gain (v pcmin to v 2w ) that is equivalent to 0dbm(600 ? ) at the complex load, the gain through the tcm38c17 (v pcmin to v pwro+ ) must equal - 2.19dbm (0.60196v rms ). the gain through the tcm38c17 will then equal -2.19dbm (0.60196v rms ) divided by the input voltage 0dbm (0.7745v rms ). this gain is equal to 0.777. the gain through the tcm38c17 (v pcmin to v pwro+ ) is given in equation 13 and repeated here in equation 37. setting the gain equal to 0.777 we can now determine the value of the gain setting resistors r 1 and r 2 . selecting the value of r 1 to be 49.9k ? ,r 2 is calculated to 5.27k ? . (note: the value of r 1 +r 2 should be greater than 10k ? but less than 100k ? .) figure 5. reference design of the unislic14 and the tcm38c17 with a 600 ? load impedance pcm bus 0dbm0 (600 ? ) 0.7745v rms 0dbm0 (600 ? ) 0.7745v rms 0dbm0 (600 ? ) 0.7745v rms 0dbm0 (600 ? ) 0.7745v rms 0dbm0 (600 ? ) 0.7745v rms v gsx g4-2 g2-4 tip ring intersil v rx v tx r p r p z o pwro+ 30 ? 30 ? v 2w + - anlgin+ anlgin- r a2 texas v in pcmin pcmout aref + - + - + - instruments tcm38c17 gsr pwro- 100nf + - v tr + - ptg floating r a1 r f = 60.4k ? z t z t -0.915dbm0 (600 ? ) 0.6970v rms 107k ? 48.6k ? 53.6k ? unislic14 0dbm 811 ? () 10 v 2 811 0.001 () ----------------------------- - log 0.90055v rms == (eq. 30) gain 20 0.90055v rms 0.7745v rms -------------------------------------- log 1.309db == (eq. 31) adjustment 3.5dbm0 1.309dbm0 + 2.19 db == (eq. 32) 2.19 dbm 600 ? () 10 v 2 600 0.001 () ----------------------------- - log 0.60196v rms == (eq. 33) z t 200 z tr 2r p () ? = (eq. 34) z t 200 200 680 1j 680 0.1 () x10 6 + -------------------------------------------------------- - 2 () 30 () + ?? ?? ? = (eq. 35) z t 200 140 ? () 200 680 ? 0.1 f 200 --------------- - || ? + ? = (eq. 36) g pcmin pwro () r 1 r 2 + 4r 2 r 1 4 ------ - + ?? ?? ------------------------------ = (eq. 37) 0.777 r 1 r 2 + 4r 2 r 1 4 ------ - + ?? ?? ------------------------------ = (eq. 38) application note 9903
4-6 the closest standard value for r 2 is 5.23k ? . transhybrid balance (z l = 200 + 680//0.1 f) the internal gsx ampli?r of the tcm38c17 is used to perform the transhybrid balance function. the voltage at v tx is equal to v pwro+ times the 2-wire to 4-wire gain of the slic, as the gain from v rx to v 2w = 1.0. the 2-wire to 4-wire gain is calculated using equation 20, repeated here in equation 40. z o for the china complex load at 1020hz is equal to 811 ? . the voltage at v tx is equal to v 2w times the 2-wire to 4-wire gain. the design speci?ations require the gain from v 2w to v pcmout equal 0dbm. this results in an output voltage at v pcmout of 0.51769v rms or -3.5dbm0 (600 ? ) . (note: -2.19dbm0 (811 ? ) = -3.5dbm0 (600 ? ) ). the gain from v tx to v pcmout is calculated in equation 43. setting this gain equal to r f / r a1 enables the value of r f and r a1 to be determined (equation 44). if r f = 49.9k ? then r a1 = 53.72k ? closest standard value for r a1 is 53.6k ? the gain from v pwro+ to v pcmout , to achieve transhybrid balance is calculated in equation 45. setting this gain equal to r f / r a2 enables the value of r f and r a2 to be determined (equation 46). if r f = 49.9k ? then r a2 = 58k ? closest standard value for ra1 is 57.6k ? reference [1] website www.ti.com/sc/docs/psheets/abstract/apps/slwa006.htm figure 6. reference design of the unislic14 and the tcm38c17 with china complex load impedance -2.19dbm0 (600 ? ) -2.19dbm0 (600 ? ) 0.60196v rms -2.19dbm0 (600 ? ) 0.60196v rms g4-2 g2-4 v pwro+ v pcmout 0.60196v rms 0dbm0 (600 ? ) 0.7745v rms 0.51769v rms -3.5dbm0 (600 ? ) tip ring intersil v rx v tx r p r p z o pwro+ 30 ? 30 ? v 2w + - anlgin+ anlgin- unislic14 r a2 texas v in pcmin pcmout aref + - + - + - instruments tcm38c17 gsr pwro- 100nf + - v tr + - ptg floating r a1 r f z t z t 28k ? 57.6k ? = 49.9k ? 53.6k ? 137k ? r 1 r 2 470pf 5.23k ? 49.9k ? -2.857dbm0 (600 ? ) 0.55742v rms r 2 r 1 0.222 2.108 -------------- - ?? ?? 49.9k ? 0.105 () 5.27k ? == = (eq. 39) g 2-4 = v tx v 2w ----------- - = z 0 -2r p z 0 ------------------------ - (eq. 40) g 2-4 = 811 ? -60 ? 811 ? --------------------------------- - 0.92601 = (eq. 41) v tx = 0.60196 () 0.92601 () 0.55742v rms = (eq. 42) g vtx vpcmout = 0.51769 0.55742 --------------------- 0.92872 = (eq. 43) r f r a1 ---------- = 0.92872 (eq. 44) g vpwro vpcmout + = 0.51769 0.60196 --------------------- 0.86000 = (eq. 45) r f r a1 ---------- = 0.86000 (eq. 46) application note 9903
4-7 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 www.intersil.com application note 9903
|
