Chapter The Desig of Passive Itermodulatio Test System Applied i LTE 600 Gogli, Wag Cheghua, You Wejue 3, Wa Yuqiag 4 Abstract. For the purpose of measurig the passive itermodulatio (PIM) products caused by the oliearities of passive devices, the paper proposes a desig of PIM test system which ca be applied i log term evolutio (LTE) 600. By reflectig the combied sigals geerated by two power amplifiers, the test system is able to capture ad aalyze PIM products accurately. Furthermore, the test system is composed of trasmitter, receiver ad cotrol system. By the adoptio of advaced RISC machies (ARM)--S3C640, the test system achieves a excellet performace. Fially, experimetal results show that the test system has good reliability ad accuracy. Keywords: itermodulatio (PIM); log term evolutio (LTE) 600; advaced RISC machies (ARM). Itroductio Log term evolutio (LTE) is the ext step forward i cellular 3G service, which covers frequecy bad from 698MHz to 690MHz []. However, passive itermodulatio (PIM) is a commo pheomeo i commuicatio system. Whe two or more frequecy sigals exist i the same passive device, they would geerate passive itermodulatio products [, 3]. Ideed, if the PIM products fall ito the receiver s bad, they will make system SNR declie ad affect the capacity ad quality of the commuicatio system. Sice the LTE 600 has bee used i orther America, it will be widely adopted all over the world. Therefore, the paper puts forward a desig of a portable passive itermodulatio system to test the PIM products i LTE 600 system. Gogli () No.9, Yudao Street, Baixia District, Najig, Jiagsu Provice, Chia e-mail: gogli988@gmail.com
Gog Li. PIM Theory for LTE 600 Due to the oliearities of passive devices, it is quite commo to geerate PIM products whe two or more sigals are mixed [4]. Assume that the compoets of the oliear system have the followig trasfer fuctio: V [ V ( t)] kv ( t) out i = i = I the equatio(), Vout[ Vi( t)] is the voltage of output sigals, Vi( t) is the (.) voltage of iput sigals ad k is coefficiets related with the properties of compoets. We assume the two iput sigals are siusoidal sigals with the amplitudes beig X X f ad, frequecies beig f ad φ, phases beig φ ad respectively. Vi ( t) = X cos( πft + φ) + X cos( πft + φ) (.) Accordig to the biomial theorem, we ca deduce the followig equatio: m m m m m i() = cos ( π + φ)cos ( π + φ) m= 0 V t C X X ft ft (.3) m m m m I order to simplify the equatio, we defie: r = C X X. By substitutig equatio (3) ito equatio (), we ca obtai the relatioal expressio betwee the iputs ad outputs: m m m out i = i = π + φ π + φ = = m= 0 (.4) V [ V ()] t kv () t kr cos ( ft )cos ( ft ) Based o the characteristics of trigoometric fuctios, cos π ft could be represeted i two differet coditios: cos πft= [ C ( ) k + C cos( k) πft] is eve k= 0 (.5) [ k = C cos( k ) πft ] is odd k= 0 By covertig the multiplicatios to additios ad subtractios, we kow that itermodulatio products are geerated at frequecies described by the followig equatio: f( PIM) = mf ± f, where m ad are itegers, either zero, positive or egative, ad the sum ( m + ) defies the order of the itermodulatio product.
Chapter The Desig of Passive Itermodulatio Test System Applied i LTE 600 Accordig to the distributio of LTE 600 system s spectrum, the receivig bad rages from 545 MHz to 580 MHz, while the trasmittig bad rages from 60 MHz to 695 MHz. As we all kow, f ad fare about 600 MHz i the LTE 600 system, they would geerate itermodulatio products of differet frequecies. Whe the order of the frequecy is eve, such asf ± f, f ± f, f ± 3f, the PIM products are too low or too high to affect LTE 600 system. However, whe the order is odd, such as3f, 3f, f ± f, f ± f, the PIM products might fall ito the receiver system ad affect the whole commuicatio system. The trasmittig bad rages from 60MHz to 695MHz ad the receivig bad rages from 545MHz to 580MHz. Thus, we ca calculate the 3rd, 5th ad 7th PIM products respectively: PIM3mi = fmi fmax = 60 695 = 545MHz PIM3max = fmax fmi = 695 60 = 770MHz PIM5mi = 3fmi fmax = 3 60 695 = 470MHz PIM5max = 3fmax fmi = 3 695 60 = 845MHz PIM7mi = 4fmi 3fmax = 4 60 3 695 = 395MHz PIM7max = 4fmax 3fmi = 4 695 3 60 = 90MHz The bad distributio is showed as follows: Figure.. The bad distrubutio of PIM products, RX ad TX Accordig to the Figure, the 3rd, 5th ad 7th PIM products fall ito the trasmittig bad ad receivig bad. Take the 3rd PIM products as a example, the rage of the 3rd PIM products is [545MHz, 770MHz]. Therefore, the itersectio of the 3rd PIM products rage ad receivig bad is [545MHz, 580MHz]. Sice the receivig bad gets a small power ad the trasmittig bad lauches a large power, the PIM products iterfere with the high-sesitivity receivig bad. I fact, the PIM products will decrease the performace of the receivig bad sigificatly. Subsidized projects of ordiary uiversities postgraduates scietific iovatio pla i Jiagsu Provice. (CXLX_00)
4 Gog Li.3 The Desig of PIM Test System Due to the PIM products low power, it is quite difficult to detect. Although some methods have bee used to test PIM products, there is still o uiform iteratioal measuremet stadard at preset [5]. The IEC (Iteratioal Electrotechical Commissio) issued IEC 6037-999, which states the itermodulatio level measuremet of RF coectors, coector cable assemblies ad cables. Figure.. The diagram of PIM test system The priciple of basic tests is as follows, which is show i Figure : merge sigal f ad f of the same power ito the combier, the put the combied sigal ito the device uder test. If the device is dual-port, a termial is eeded to coect the other port. The level of tuig or itermodulatio by itself should be at least 0dB lower tha the level produced by devices uder test. Fially we ca measure the PIM products of f ± mf with a calibrated receiver. Accordig to the priciples of IEC, the paper desigs a portable passive itermodulatio test system suitable for LTE 600 system. The test system is composed of three parts: trasmitter, receiver ad cotrol system, the frame of the system is simplified i Figure 3. Figure..3 The overall framework of PIM test system I the trasmitter, widebad sythesizer composed of ADF4350 geerates two radio frequecy (RF) sigals. The, the two sigals are itroduced ito power amplifier. Cosiderig the sigals after amplifyig are too large, digital cotrol atte-
Chapter The Desig of Passive Itermodulatio Test System Applied i LTE 600 uators are applied to stabilize the sigals at 43dBm. Fially, the two sigals are combied ito diplexer s Tx port ad the reflected sigal comes out from the Rx port. I the receivig part, the reflected sigal first passes through the bad-pass filter to lead the frequecy compoet to fall ito the receiver s badwidth pass. Due to the PIM products are too low to detect, they are mixed with F 3 produced by local oscillator ad the output sigals cotai FIM, F3, F3 ± F. After passig IM through bad-pass filter, the itermediate frequecy sigal F3 F is remaied IM ad it is elarged by amplifier. At last, the PIM level is detected by the system. The cotrol system chooses Widows CE 6.0 as the software platform ad S3C640 as the mai processor to implemet power cotrol, carrier cotrol, data process ad aalysis, test data storage, soud, etwork, USB master/slave, etc. [6,7]. As for the hardware part, it is composed of two layers, the core board ad backplae: the core board itegrates S3C640 processer, FLASH, SDRAM, while the backplae provides may periphery iterfaces. The S3C640 processer of Samsug is a RISC processor based o ARM architecture, which has a excellet performace compared with previous processers. Figure..4 The diagram of system software As for the software part, it adopts Widows CE 6.0 to implemet a complete software platform as well as various hardware iterfaces. The block diagram is show i Fig.4.We use Widows CE 6.0 kerel to realize the coversio betwee widows ad BSP system ad we also carry out customized kerel for the etire system. What s more, the commo peripheral drives, such as LCD drive, microwave sythesizer driver ad shared driver betwee AD ad touch scree are eeded to support the system.
6 Gog Li.4 Experimetal Results I order to affirm the accuracy ad reliability of the istrumet, we use the test system to measure ateas PIM products [8]..4. Istrumetal Noise Level Testig The oise level cosists of phase oise of local oscillators ad amplifier oise of receivers. At first, we eed to test the oise level of the PIM test system proposed i the paper to evaluate its performace. I fact, we coect Port ad Port with 50 ohm load ad close two RF source sigals ad the itermodulatio level measured i the process is the oise level of the test system. Figure..5 The test result of istrumet oise level The result measured by the system software is show i Figure 5 ad the height of the bar chart represets the PIM level value with the uit of dbm. Obviously, we ca get the average of istrumet oise level, which is approximately - 40dBm ad ca meet the requiremet of the istrumet oise level testig..4. Atea Testig Whe two-port or multi-port devices work i two high-power sigals simultaeously, two sigals are mixed i passive compoets of the commuicatio system, which will geerate PIM products. The vast majority of passive compoets such as ateas, cables, duplexers, filters ad directioal couplers ca use the test system to obtai the PIM level. Durig the test, we coect the test ateas with Port ad Port of PIM test system separately. We test the 3rd PIM products ad
Chapter The Desig of Passive Itermodulatio Test System Applied i LTE 600 give the results i two coditios: tighteig ad looseig the mechaical iterface. At the begiig, we tight the atea ad take 3rd PIM products as a example to make a further aalysis. We assume the rages of the two RF sigals are f [60, f(max)] MHzadf [ f (mi),695] MHz respectively. I the first sweepig frequecy test, we fix f at the maximum poit of the trasmittig bad ad we ca get the followig equatio: * f(max) 695 = 580MHz. I the secod sweepig frequecy test, we fix f at the miimal poit of the trasmittig bad ad we ca get the followig equatio: * 60 f(mi) = 580MHz. Therefore, i the 3rd test system of LTE 600, the rages of the rages of the two RF sigals are f [60, 638] MHz ad f [660, 695] MHz respectively. I the test, we fix fat the frequecy of 695 MHz ad f sweeps from 60 MHz to 638MHz. Later, we fix f at the frequecy of 60MHz ad f sweeps from 695MHz to 660MHz ad the result is show as follows. Before the test, we loose the test lie iterface which is coected with the itermodulatio test system ad result is show below. Figure..6 The test result of atea PIM level From the test curve of Figure..6(left), we ca get several values related with the atea. Obviously, the yellow lie represets the typical value, which is - 0dBm. The gree lie shows the real-time poit i the coordiate of the first sweepig test ad the red lie shows the real-time poit i the coordiate of the secod sweepig test. What s more, the maximal value of PIM products is show i the figure. Whe f is 60MHz ad f is 660MHz, we get the maximal value -4dBm at the frequecy of 573MHz i the receivig bad, which is less tha the stadard value -0dBm. Accordig to the results, we ca coclude that the stability ad precisio of the test system meet the requiremets. From Figure..6(right), we ca see clearly that the PIM level of atea is much larger tha the compliace idex. Obviously, we ca draw a coclusio that the PIM products are sesitive to exteral coditios. Ideed, chages of exteral coditios will geerate great PIM level.
8 Gog Li.5 Coclusio Sice the itermodulatio product is vital i commuicatio system, the measuremet of the itermodulatio product is of great importace. Due to the complexity of the PIM mechaism, it ca t be calculated i theory directly ad previous devices are too large to carry. Therefore, the test system proposed i the paper is suitable to measure PIM products sice it is portable ad easy to operate. Fially, the PIM test system is proved to be accurate ad reliable..6 REFERENCES. B. Deats ad R. Hartma, Measurig the passive-im performace of RF cable assemblies, Microw. RF. vol. 36, 997, pp. 08..Bolli P., Selleri S., Pelosi, G. Passive itermodulatio o large reflector ateas, Ateas ad Propagatio Magazie. vol. 44, Oct. 00, pp.3-0. 3.Bolli P., Selleri S., Pelosi G., Passive itermodulatio o large reflector ateas, J. Ateas ad Progatio Magazie, vol 44, Issue 5, Oct 00, pp. 3-0. 4.Chie Chou ad Chue-Tsai Su, A computer-etwork-supported cooperative distace learig system for techical commuicatio educatio, IEEE Trasl.o Professioal Commuicatio, vol. 39, Dec. 996, pp.05-4. 5.J.Saford, Passive itermodulatio cosideratios i atea desig, Atea ad Propagatio Society Iteratioal Symposium, IEEE Digest, vol.3, 8th Jue, 993, Michiga, USA, pp. 65-654. 6.J. Herie, A. Christiaso ad W.J. Chappell, Predictio of passive itermodulatio from coaxial coectors i microwave etworks, J. IEEE Tras. Microw. Theory Tech. vol. 56, Ja, 008, pp.09. 7.Lui.P.L, Passive itermodulatio iterferece i commuicatio systems, Electroics & Commuicatio Egieerig Joural, vol., issue 3, Jue 990, pp. 09-8.S.Abeta, Toward LTE commercial lauch ad future pla for LTE ehacemets(lte- Advaced), Proc. IEEE Symp.Commuicatio System(ICCS 00), IEEE Press, 00, pp. 46-50, doi:0.09/iccs.00.5686367.8.