Passive intermodulation interference in communication systems

Transcription

1 ~ ~~ Passive intermodulation interference in communication systems n multifrequency communications environments, such as land mobile radio sites, satellite earth stations, ships and surveillance aircraft, passive intermodulation products (PMP) generated by nonlinear materials and metallic contacts can cause serious radio interference. This problem is well known and a wide range of coaxial cables, connectors and materials have been investigated. The paper gives an ovenkv of passive intermodulation interference in communication systems. t describes briefly the theory of intermodulation, types of passive nonlinearities, mechanisms responsible for the generation of PMP, guidelines for minimising PMP generation and techniques for locating PMP sources. Previous investigations of PMP are summarised and the design of PMP measurement systems is discussed. An cxaniple of PMP measurement is also included. 1 ntroduction ntermodulation products are spurious signals generated when t\vo or more signals mix in devices \vith nonlinear characteristics. ntermodulation becomes an interference problem \vhen the intermodulation products by P. L. Lui generated fall into the passband of receivers operating nearby. n a radio environment, there arc three main arcas \vhere intermodulation can occur: (a) the transmitter output stages, due to the nonlinearity of the poiver amplifier circuits (b) the receiver input stages, due to the nonlinearity of the mixer and radio frequency (RF) circuits (c) in nonlinear materials and nonlinear metallic contacts, such as corroded and/or loose contacts in coaxial cables, Lvaveguides, connectors, inulticouplers, \viie fcnces, toiver and mast assemblies. While it is ucll recognised that the intermodulation interference (M caused by transmitters and can be minimised b> adequate system isolalion, the M caused by nonlinearities in nictallic contacts and materials cannot be suppressed by the same technique. Over the last forty years, passitze M has causcd attention owing to the increase in radio communication services and the atrendant increasc in the passive M problem. This can be seen from the number of articles on passive intermodulation published over the last t\velve inon t lis. 2 PMP generation To understand the generation of nonlinear U 0 r r G (? r c i Z z c P), frequency 1 Output spectrum of a nonlinear passive component ELECTROCS & COMMUCATO EGlEERG JOURAL JUE 1990 O

2 cspeiienced order of intermodulation This is illustrated in Fig. 2, which sho\vs the relationship bet\veen the number of oddorder inband intermodulation products and the number of transmission channels.' Because the amplitudes of the inrerrnodulation products generally fall off \vith increasing order, thc lowerorder products are more likely to cause by strong signals. can have amplitudes large enough to cause serious interference. This usually happens in satellite and iiiarinc communications, \vhere highan sni it t e rs and hvn vise arc collocated. M is not from the evenorder products as they usually fall \cell outside the reccilc bands. Mali!. of the oddodelproducts can also be discountcd lor the same reason. Only those products that fall into the rcceivc bands can cause M. 3 Types and mechanisms of passive nonlinearities Vel\. often the term 'rusty bolt' effect is used to describe thc passkc M problem, although cuperience has sho\vn that most M problems are caused by loose rnctallic joints rather than iustv ioints. To understand the pas number of channels 2 Relationship between number of channels and number of inband intermodulation products passive intermodulation products (PMP) and the basic theo~y ol intermodulation, consider a simplified case \vhere a nonlineat passive component is excited by tlvo unmodulated signals of rcquencies ', and fs V,=V,cos(2nfr)+Vzcos(2n/~i) (1) \\.here V, is the combined input signal, and Vi and V, are the amplitudes of the t\r% esciting signal\. The transfer [unction ot the nonlinear component may be represented by an nthorder potver series:' V<,=KiV,+K2V: +K,V: +... (2) lvherc V,, is the output signal, and K,, Kz and K, are coefficients \\ hich depend upon the properties of the nonlinear component. Subsriruting eqn. 1 into eqn. 2 and solving gives the spectrum of V,,, ntermodulation producls arc generated at frequencies described by the follo\ving equation: t(intermodulation) =1711, tnfz (3: \vhcre in and are integers, either icro, positive or negative, and the sum (n ) defines the order of the intermodulation product. The spectrum of V,,, sho\vn in Fig. 1, consists of the t\vo exciting signals, 1, and /?, as \vel1 as man) neuzly generated harmonic and intermodulation signals. A linear component produces no such additional signals. So far only a simple, t\volrequencv case has been considered; in multifrequency environments, the number of products increases ven rapidly \vith the increase in number of transmission channels. kilo\\ the mechanisms and types of passive noiilinearitics, Basicall) there are t\vo types of passive non li n eari t ies in coni niu n icat ion systems: contact nonlineariries and mat ciial non i n eaiit ies. The lo rni er refer to an! contacts thar have a nonlinear cu rren t/volt age behavioul; loose, osidised and contaminated metallic joints arc bpical esamples. The latter relei. to bulk niareiials, such as fcrroniagnctic materials and carbon libres, that have nonlinear Aectrical characteristics. The characteristics ol these nonlinearities have been measured and tj,pical response cui~es ol nonlinear contacts and nonlinear hysteresis effects are shown in Fig. 3. The actual nicchanisnis responsible for the nonlineaiclfccts are coniples and not \vel1 understood, especially in the case of nonlinear contact mechanisms. Over the years, the following mechanisms have been identified as being responsible for the generation of electron tunnelling and semiconductor action through thin oxide layers separating conductors at metallic contacts 1 10 ELECTROCS & COMMUCATO EGEERG JOURAL JUE 1990

3 0 microdischarge bet\veen microcracks and across voids in metals nonlinearities associated \vith diit and metal particles on metal suifaces high current densities at contacts 0 nonlinear rcsisli\,iiy ol carbon fibres 0 nonlinear hysteresis effects in ferromagnetic materials. n addition to these mechanisms, poor \vorkn~anship can cause loose conncctions. cracks in metals and ovidisaiion ai joints, uhich in turn generate PMP. n practice, a combination of these mcc han ism s ma! bc re spoil s i blc lor PMP generation. 4 Guidelines for minimising PMP generation Once the types and mechanisms ul passite nonlincaiities are knotin, there are a number ut steps \\ hich can be takcn io minimise the generation of PMP in communication systems:sy dummy loads, circulators, isolators and some semiconductor devices 0 achieve good isolation between the highpower transmit signals and the lowlevel reccive signals by filtering and physical separation 0 frequency planning should take account of the higherorder products as they can be potential interference signals in some communication systems. n addition to these general guidelines, careful planning, good \vorknianship, stringent quality control and a high standard ol maintenance arc cqually borne in mind that no communication sysiem is coniplctcly free from PMP, although proper attention to deiail during the design and construction stagcs can make a substantial rcduction in level. 5 Location of PMP sources When the M caused by passivc nonlinearities becomes intolerable remedial action is required. This usually involves some combinatior of the approximate location of PMP sources by excitation and measurement techniques and the recognition of likely PMP sources. A variety of location techniques") have been developed and the type of tcchniquc used depends upon the location ol the PMP sources. Generally there are t\vo main areas \vhere PMP sourccs can occur: (a ) \vi1 h i n sysrcm co mpon en s, including niulticouplcrs. coarial cablc5, connectors. \raveguides and antennas (b) in the eutcrnal environment, such as suppoll stluctures, icnver and mast components, \virc lences and any neaib! memllic objects. n the lirst case. a technique nonlinear materials should nul be used in 01' near ihc current paths. f foi sonic reason ihc! ha1e to be there, the! should be coaicd \\it11 lineal materials kccp the cuireiii densilies lo\\ in the conduction paths by using largei conductors or having bigger contact areas bet\\een metals minimise metallic contacts, espcciall~ loose contacts and rotating joints. thcse cannot be a\oided. rhen pio\ide insulaiors or a1tcrnati1.e current paths at the coniacis orjoints. Also ininimisc the cvposuie of loose contacts, rough surfaces and sharp edges to radiated signals D keep thermal variations to a minimum as the expansion and contraction of metals and materials can create nonlinear contacts D use bonded joints if possible, but make sure that thcsc joints are good and have no nonlinear materials. cracks, conlainination or corrosion avoid having iuning SC'C\~S 01' nioxing parts in the current paths. Keep all joints and contacts clean and tight. and if possible keep them free from vi brat ion D cable length, in general, should be minimised and the use of quality and o\v PMP cables is D essential minimise the use ol nonlinear components, such as lumped ~ m l voltage, V magnetising force, H 3 Response of passive nonlinearities: (a) contact nonlinearity (nonlinear /V curve); (b) ferromagnetic nonlinearity (nonlinear hysteresis curve) <LECTROCS & COMMUCATO EGEERG JOURAL JUE

4 proposed by a Swiss PTT research Alternatively, the location ol team may be used." t is based on PMP sources can be determined the pulse reflectometer principle by adding attenuation at various and is designed for locating PMP locations along the feeder. n sources in [ceders \vith lengths of accordance with the power law up to 150 m. The system sends two relationship, the change in PMP RFpulse signals to the feeder and level will indicate the location of measures the returned the sources. Recognition of the intermodulation pulse that is likely PMP sources is perhaps th generated by the PMP source. The most useful and underestimated detector, which comprises a video approach, since some familiarity detector and an oscilloscope, \vith the more typical culprit provides the information on the components can reduce the location of the PMP source and location time quite substantially. has a length resolution of less than t is well recognised that 1 m. 1 metallic joints ila reflector duplexer PMP a b Fix filter receiver 4 Simplified block diagrams of passive intermodulation product measurement systems: (a) radiating: (b) and (c) nonradiating DUT=device under test antenna can be potential PMP sources.l7li To locate such sources, a novel method that employs a microwave holography imaging technique2 X.19 mav be used. This technique has shown promising results and may also be used to evaluate reflector antenna pcrformancc. When the PMP sources are in the external enlironment, the most videly used technique is O search the suspected areas with a small loop antenna connected to a portable radio receiver or spectiuin analyser. The detcctor is usually tuncd to the inteference signal, although related harmonic and intermodulation signals may be used. When thc directionfinding operation is carried out in a high lield strength area, care must be taken to ensure that there is no generation ol intermodulation products in the detector. This can be achicved by adding additional screening and filtering to the detecroi. 6 Review of previous work Over the last fony years, many investigations have been made into passive intermodulation interlcrence. Although they \veie carried out by different research groups lor different communication systems, many oc the results and conclusions are vev similar. n this Section, a brief reyiew of research \vork conducted between 1947 and 1990 is presented. 11 covers mainly the measuicment and characterisation ~f PMP sourccs and does not include the analysis and modelling A passive M; the detection and location of PMP sources has already been covered in a revieiv paper.l(l The review of xperimental \vork is summarised in two tables and the related references are given for further reading. Table 1 shows the types of :ommunication systems studied ind Table 2 shows the types of naterials and components studied. 7 PMP measurement systems Most experimental studies of 3assive nonlinearities require PMP measurement systems. Jenerally these systems can be Aassified into either nonradiating jr radiating types. The former are uitable for investigating items such as nonlinear materials, :onnectors, coaxial cables, filters md waveguide components: they are normally housed in a screened room or laboratorv, arc terminated by a matched load and ideally radiate no energy. The latter types are suitable for investigating 12 ELECTROCS & COMMUCATO EGEERG JOURAL JUE 1990

5 diating ctnicturez zuch as itennas, feeds and structural imponents, and are usually tuated in an anechoic chamber open field test site. The jnradiating system is more idely used because the user has xter control over the test irameters and cnvironmcnt. n,ntrast, the radiating system can 2 affected by the local signal ivironment. However, for certain sts a radiating system is jsential. Measurement systems can inher be classified according to ie test frequency, such as audio, idio or microivave. Although the jnfigurations. components and st equipment of these systems iffer, the basic design is usually +sed on the tirofrequency iethod. i.e. t\ro signals are used excite the sample and the itermodulation signals generated e filtered and measured. Fig. 4 >o\rs the basic configurations of )me t\rofrequency radiating and onradiating measurement stems. A nonradiating measurement stem, based on the design of a HF measurement system, is 1oum in Fig. 5. This proposed :stem incorporates a feir new :sign ideas and may be used for ieasuring nonlinear materials, ietallie joints and components at Table 1 Previous studies of passive intermodulation interference: tmes of communication svstems Environment and system groundair communication systems cable television ASA launch pad communication systems surveillance aircraft communication systems ship communication systems landbased radio communication systems satellite, microwave and RF communications systems radio frequencies. n this system, tivo RF signals, produced by frequency synthesis generators, are amplified by linear amplifiers. HighQ bandpass cavity filters are used in each transmit arm to remove unwanted signals and provide isolation between the two sources. A broadband 3 db coupler is used as a signal Reference Chamberlin4 Kel1ar3 Frazie?? W0ody,3~.~~ Shands Mieth? Betts,34.35 Wats~n,~~,~~ Blake,41 Mason:> El~ner,4),~4 Salisbury s Lui,i.5355 Gardiner:o Ho:O Sturton,48 Betts,49 Bevan,sO B~ley,~ Fudge52 Hall,2 Chapman,6 Rootsey? Hoeber? Kumar? Siegenthaler, COX,^^ uding, Kong,i4 Higa, Guenzer,l6 Bond, Aspden,g,iY Gardiner>o Kudsia,zl Elsner,22 Bayrak,23 Sanli:4.26 Arazm,25 Arnin,z7.*8 Martin:Y Young,3o Lee3637 combiner, giving a good match and an additional 2030 db isolation. A dummy load for the coupler is provided by a long length of coaxial cable, such as RG214U or RG58C/U. The combinedsignal lcvel and voltage standing \rave ratio (VSWR) are monitored by a throughline power meter. A lowloss notch cavity power amplifiers 25 lyj+ :er :, f2 +@signal generators 1 fl 8 power meter H& 25 DUT HcouplerM 20dB load H cavity filte; test chambar (2f,f1) fl Proposed passive intermodulation product measurement system UT=device under test; connection=rg214u coaxial cable: load=rg58c/u or RG214U coaxial cable LECTROCS & COMMUCATO EGEERG JOURAL JUE :

6 Table 2 Previous studies of passive intermodulation interference: components and materials Component and material waveguide components waveguide components waveguide components waveguide components feeds reflectors reflectors reflectors multicouplers duplexers coaxial cables metallic contacts and coaxial cables metals, connectors and coaxial cables connectors metals, connectors and coaxial cables connectors metals and metallic contacts carbon fibres, metals and coaxial cables carbon fibres, metals and metallic contacts metallic contacts metallic contacts metals and metallic contacts metals and metallic contacts ter tuned to the desired ntermodulation signal is placed >et\veen the power meter and the est chamber to suppress the "esidual intermodulation signal. The output signals from the test :hamher are sampled by a 2roadband 20 db coupler and 3rered by another set of high4,andpass cavity filters. The signal.s detected by a test receiver and he control of the test equipment and the processing of the test esults are implemented by a microcomputer system. Zonnections are made in double braid, silverplated RG2 14U maxial cables and silverplated type connectors. To reduce radio nterference, sensitive equipment is housed in a screened room. 1 Design considerations The design of a consistent, iccurate and low PMP ncasurement system is not an :asy task because of the dynamic ange requirement and the ibiquitous nature of ionlinearities. During the last hree vears, two Frequency band 8 GHz 11 GHz 6 GHz 2 GHz UHF 2 GHz 7 GHz UHF HF 1.55 GHz HFUHF MHz 15 MHz UHF 100 MHz8 GHz MHz Reference Chapman: Rootsey7 Siegenthalerl cox12 uding13 Kongl4 Higals Guenzer,l6 Bond,l7 Aspden2,18.19 Gardiner20 Kudsiazl Elsnerz2 Bayra k, 23 Sanli,24.*b Ara~m,~~ Ami11,27,~~ Martin29 Young30 Woody,31 Shands32 Kellar33 Bett~3~.35 Lee36.37 Wats~n~~,'~ H0,2.40 Bevan50 Shands46 Frazier47 Lui.2 55 measurement systems for laboratory and field investigation of passive M problems have been developed by the author.".i5 The laboratoty system was developed for measuring small metallic plates and irregularly shaped structural components. The field system, \vhich consists ol a newly built 8 m tall radio tower, t\vo folded dipoles and most of the laboratotj system equipment. \vas built to study PMP behaviour under conditions similar to those of a radio site environment, and to evaluate various PMP detection techniques. The details of these systems will not be discussed here as they are available in the published literat~re.s~.~s However, three types of components that are important in developing PMP measurement facilitics will be discussed: RF signal combiners and samplers, test chambers and dummy loads. RF signal combiners and surnplers There are various ways of combining and sampling RF signals. deally thc signal combiner and sampler should have low insertion loss, high isolation between ports, low VSWR at ports, good linearity, sufficient bandwidth, high power rating and low cost. n practice, there is always a tradeoff between these lactors and it is very difficult to achieve all of them in a single design. The study of PMP measurement systems has revealed that there are three common ways of combining and sampling RF signals: (a) the use of Tpieces and quarter \cavelength coaxial cables.a",5"," This is a simple and lowcost method, hut the length of the cables is critical and a careful design is needed to achieve maximum signal isolation (b) the use of a combination ot hybrids, multicouplers and duplcxers.l7 3" 3637 This method is not as often used because ol the higher cost involved (c) the use of directional couplers \cith various degrees of coupling." 32 '5 These provide good signal isolation and are available for many bands, but there is always a 3 db loss at the combiner. The use of duplexer and directional couplers in PMP measurement systems is sho\vn in Fig. 4 and 5. n practice, the choice of method depends upon the test requirements. Test cliartibrr.~ f the test sample is not a component that can be connected directlv into the measurement system, then some kind of test chamber is required. Matching is a very important factor in test chamber designs because reflected power can damage sensitive equipment. Further, experience has shown that a VSWR of 1.5 at the input port of the tcst chamber can change the measured PMP level by as much as 10 db?h The t\vo test chambers described in the following paragraphs are designed to match the characteristic impedance of the system and to handlc a variety of test samples. The first design is based on a rectangular coaxial line structure in which the plate sample forms the centre conductor. Thc design curve and formula uscd are available in published literature.2 With dimensions of 11 5~4.8 em for the box and 6.5x0.2 cm for the sample, as shown in Fig. 6a, a 50 box with a response as shown in Fig. 6c can be achieved. 14 ELECTROCS & COMMUCATO EGEERG JOURAL JUE 1990

7 t has a typical VSWR of 1.1 S and an insertion loss of 0.5 db from 50 to 500 MHz. This design has a flat frequency response over a very [vide band and is only suitable for plate samples. n practice. it is more convenient to choose a suitable box (Fig. 7) and then design the sample size to give the desired characteristic impedance. For larger and irregularly shaped samples a different design, based on a halfwavelength transformer, is uscd. h. The sample forms the centre conductor ol the transmission line and the length of the test chamber is equal to half the average \vavelength of the input signals. The test chamber can be in the form of a cylinder, rectangular or square box, but it must have an opening for accessing the sample. A 1 m long cylindrical test chamber. as shoivn in Figs. 6b and 8, has been developed. ts matching performance, shown in Fig. 6c, is a function of the input frequencies and sample size. A smooth transition of eneru at the sample and connector interlace is essential to minimise mismatch in 1he test chamber design ~12?. louds n nonradiating PMP measurement systems, dummy loads arc iiacd 10 absorb the transmitted poxver. t is important that the dummy loads should not generate significant PMP. Commercially available dummy loads, \vhich are adequate for many applications, are not normally suitable for PMP measurement. To select a dummy load, there are a number of factors, such a5 linearity, power rating, screening performance, attcnuation, cobt, \\eight and \&me, [rhich need to be considered. Linearity is the prime requirement as it is the main contributor to the residual intermodulation level. Attenuation is directly proportional to frequency: a highattenuation cable either extends the lower frequency limit or reduces the required length. However, the power handling capability of a lossy cable is usually hver than that for cables with lower atienuation consrant. A variety of lumped and distributed loads have been tested. The test results and test setup are shown in Fig. 9. t can be seen that the distributed loads arc more linear than the lumped loads. The RG58C/U coaxial cable is better than the URM67 coaxial cable and a long length of coaxial cable m. : E f+hf faf 0 S = 10 v) C. DUT + b frequency. MHz 6 Characteristics of test chambers: (a) 50 R test chamber: (b) halfwavelength test chamber (=V2= 1 m off= 150 MHz); (c) frequency response 50 R test chamber: halfwavelength test chamber 7 PMP measurement facilities a signal combiner: b power meter: c plate sample: d test chamber C ELECTROCS & COMMUCATO EGEERG JOURAL JUE

8 is more linear than a combination of lumped and distributed loads. The RG58C/U is a good choicc among many types of coaxial cables because of its linearity, light weight and low cost. However, if cost is not an important factor, then the doublescreened, silverplated RG214U may be considered. Silverplated steel cables should not be used because the nonlinear hysteresis effects in steel can generate a significant level of PMP. 9 PMP measurement A wide range of components and materials, as shown in Table 2 have been measured. n this Section, an example of PMP measurement on a variety of large structural components is presented. The PMP measurement system used is shown in Fig. 8 and the test resultsy5 are shown in Fig. 10. Measurements were made of the thirdorder PMP generated by twc equal signals at highband. Two types of components in common use on antenna towers and masts were chosen for the measurements: galvanised mild sleel angle irons shilar to those used for tower construction, and mastsupporting components such as mildsteel ropes, metallic chain5 and mildsteel ropes with shackles and thimbles. These were chosen to provide a comparison between joints in typical structural components and had different contact areas and different degree5 of corrosion. All the galvanised samples had been exposed to the weather for many years and so thc corrosion \vas natural. The plots given in Fig. 10 are fl f2 E 70 5 Load L1 Load L2 L al._ al?? r 0 : 90 U e cn E 52 $ 110 al h combined input level, dbm 9 Test results of lumped and distributed loads for thirdorder intermcdulation (2f2 f,) with f, = 150 MHz and f2=155 MHz 116 ELECTROCS & COMMUCATO EGEERG JOURAL JUE 1990

9 mean results obtained from many tests. The results show that the PMP and input level relationships vary between 1.6 and 3 db/db. Type 'a' and 'b' samples generated the lowest PMP level. This signal level can be regarded as the system's noise floor and is due to the intermodulation products generated by the measurement system, metallic contacts and hysteresis effects in mild steels. Mildsteel rope is noisier than a corroded tight joint and substantially noisier when fitted with a shackle and thimble. Angle iron with a loose joint generates significant levels of PMP and the metallic chain is clearly to be avoided. These results suggest that components that have loose and/ or small contact areas can give rise to significant levels of PMP. This agrees with many field studies where the most significant passive M sources located were found to be loose joints rather than corroded joints. Such a conclusion would probably justify a more appropriate selection of passive components. t might also be necessary to reassess the effectiveness of using chemical agents, as suggested by some resear~hers,40.~~ to reduce the passive M problem caused by corroded joints. 10 Conclusions This paper has outlined the results of various research and development projects concerning passive intermodulation in communication systems. The basic theory of intermodulation, the types and mechanisms of passive nonlinearities, ways to minimise the passive M problem and techniques for locating PMP sources have been described. Some of the design considerations in developing PMP measurement systems have been discussed and an example of PMP measurement has been given. Engineers often overlook the fact that many passive components and structures are potential PMP sources, and it is hoped that this review paper will increase their awareness of the passive M problem. Passive intermodulation should always be considered as an important factor in the design and operation of communication systems. References LU. P. L.: 'A study of inkxmodulation interference due to nonlincarities in metallic structures'. PhD Thesis. University of Kent, UK. February 'Passive intcrmudulation products in m E $ c 3? a g 90. * $ E v) combined input power level. dbm 10 Relationship between combined input power level and passive intermodulation product level for thirdorder intermodulation product (2fzfi) with f ~= 150 MHz and fz= 155 MHz a galvanised mildsteel angle iron without bolted joint; b galvanised mildsteel angle iron with tight joint: c galvanised mildsteel angle iron with rusty tight joint; d galvanised mildsteel angle iron with loose joint: e galvanised mildsteel rope: f galvanised mildsteel rope with shackle and thimble: g mildsteel chain antennas and related structures'. EE Colloquium Digest o. 1989/94, June METH, W. B.: 'A costeffectivc solution to measurement of hullgenerated intermodulation interference on US aw shim'. ational Svmposium. on Electromagnetic compatib& USA, May 1989, pp CHAMBERL. K.: 'Ouanlilative analysis of intermodulation product interference'. Tans. 1989, EMC31, (3). pp GARDER. J. G., and FUDGE, R. E.: 'Aerials and base station design'. Chap. 4 in HOLBECHE, R. J. (Ed.): 'Land mobile radio systems' (Peter Peregrinus, London, 1985). pp CHAPMA, R. C., et al.: 'Hidden threat multicamer passive component M generation'. AAA/CAS 6th Conf. on Communication satellite systems, Canada, April 1976, pp ROOTSEY. J. V.. cl al.: 'USASCA HT MT terminals programintermodulation study final test report'. WDLTR5243, PhilcoFord Corp.. USA, August HOEBER. C. F.. et al.: 'Passive ELECTROCS & COMMUCATO EGEERG JOURAL JUE

10 intermodulation product generation in high power communications satellites'. AAA 1 th Conf. on Communication satellite systems. USA, March 1986, pp KUMAR. A.: 'Passive M products threaten highpower satcom systems', Microwavc &'z RF, December pp LU, P. L., ef al.: 'A sumey of nonlinear junction (rustybolt effects) detection techniques'. ERE nt. Conf. on Land mobile radio, UK, December SEGETHALER. J., and STAGER, G.: 'The measurement of microwave intermodulation effects on passive components and systems pans'. Conf. on Microwave and optoelectronics, West Germany, March 1988, pp. A5/19 12 COX, R. D.: 'Measurements of waveguide components and joint mixing products in 6 GHr frequency diversity systems'. EEE Trans., 1970, COM18, (l), pp UDG, E.: 'onlincarities of flange connections in transmission lines cartying high RF power', 4th European Microwave Conf., pp KOG, A. K. C.: 'A broadband circular polarised antenna feed with low passive intermodulation products'. Antennas & Propagation; Pan 1, ovember EE Conf. Pub1 o. 69 pp HGA, W. H.: 'Spurious signals gcncrated by electron tunneling on large rcflector antennas'. Proc. EEE, 1975, 63, (2), pp GUEZER, C S 'Comments on "Spurious signals generated by electron tunneling on large reflector antennas"', Proc. EEE, February , pp BOD, C. D., er al.: 'ntermodulation generation by electron tunncling through aluminiumoxide films'. Proc. EEE, December 1979, 67, pp ASPDE, P. L., el al.: 'Microwavc holographic imaging of intermodulation product generation applied to reflector antennas'. Proc. JA, France, pp ASPDE, P. L., c al.: 'Microwave imaging of intermodulation product sources applied to reflector antennas'. Proc. 6th CAP, 1989, pp GARDER, J. G.. el al.: 'Origins and minimisation of intermodulation outputs from mobile radio basc station muhicouplers'. ERE nt. Conf. on Electromagnetic compatibility, 1984, pp KUDSA, C. M.: 'A high power low passive intermodulation (PM) UHF duplexer for space application'. EEE nt. Microwrave Symp. Digest, 1979, pp ELSER, R. F.: 'Comment5 on "Coaxial cables as sources of intermodulation interference at microwave frequencies"', EEE Trans., EMC.21, (), pp BAYRAK. M., and BESO, F. A: 'ntermodulation products from nonlinearities in transmission lines and connectors at microwave frequencies', Proc. EE, 1975, 122, (4). pp SAL. H.: 'onlinear effects in contacts and coaxial cables at microwave frequencies'. MEng Thesis, University of Sheffield, UK, ARAZM, F., and BESO, F. A.: 'onlinearities in metal contacts at microwave frequencies', EEE Trans., 1980, 22, (3), pp SAL, H.: 'onlinear effects at contacts at microwave frequencies'. PhD Thesis, University of Sheffield. UK, AM, M. B., and BESO, F. A.: 27 onlinear effects in coaxial cables at microwave freauencies'. Electron. Leu., 1977, 13, (25), pp AM, M. B., and BESO. F. A.: 'Coaxial cables as sources of intcrmodulation interference at microwave frequencies', EEE Pans., 1978, 20, (3), pp MART, R. H.: 'ntermodulation product generation studies on materials, connectors and structures'. ERE nt. Conf. on Electromagnetic compatibility, April 1978, pp YOUG, C. E.: 'An update on intermodulation generation by RF connector hardware containing ferromagnetic materials'. 9th Annual Connector Symp., pp WOODY, J. A., and SHADS. T. G.: 'nvestigation of intcrmodulation products generated in coaxial cables and connectors'. RADCTR82240, Georgia nstitute of Technology, USA, September SHADS, T. G., et al.: 'ntermodulation interfercncc gcncrated in coaxial cables and connectors'. ational Symp. on Electromagnetic compatibility. USA, 1984, pp KELLAR, B. S.: 'Measurement of intcrmodulation products generated by corroded or loose connections in CATV Systems'. Technical Paper Annual CTA Convention, 1984, pp BE'TS. J. A.. and EBEEZER. D. R: 'Generation of intermodulation interference due to nonlinear effects in the nearfield regions of multipletransmission communication systems'. Proc. AGARD Conf. on Aerospace tclecommunicalion systems, May 1972, pp BE'TS, J. A., and EBEEZER, D. R.: 'ntermodulation interference in mobile multipletransmission communication systems operating at high frequcncics (330 MHr)', Proc. EE. 1973, 120, (l), pp LEE, J. C.: 'ntcrmodulation measurement in the UHF band and an analysis of some basic conducting matcrials' T , Lincoln Lab., MT, USA, ovember LEE, J. C.: 'ntermodulation measurement and analysis of some conducting materials commonly used in acrospace'. EEE nt. Conf. on Communications, Vol. 2, Junc pp WATSO, A. W. D.: 'The measurement, detection. location and suppression of external nonlinearities which affect radio systems'. ERE Conf. on Electromagnetic compatibility, 1980, pp WATSO, A. W. D.: 'mprovements in the suppression of external nonlineanties ('rusty bolt' effects) which affect naval radio systems'. EEE nt. Symp. on Electromagnctic compatibility, 1983, pp HO. P. S. W.. er al.: 'ntermodulation interfercnce in radio systems'. AGARD (ATO) Conf. on Effects of electromagnetic noise and interference on performance of military communication systems, Portugal, 1987, AGARDCP420, pp. 30,'30/8 41 BLAKE, K. W.: 'External crossmodulation in the 100 Mc/s band', J. EE, 1947, 94, Part 111A, pp MASO, H. P.: 'An investigation into the magnitude of spurious responses produced from a multiple transmitting system operating from a common aerial aboard ship'. EE Convention on HF Communication, ED4. UK, pp ELSER. R. F.. er al.: 'Engineering study for electrical hull interaction'. TR 56013, T Research nstitute, USA, January 1965, pp ELSER. R. F., er al.: 'Environmental intcrfcrcncc study aboard a naval vessel'. EEE Electromagnetic compatibility Symp., 1968, pp SALSBURY, G. C.: 'Topside interference aboard USS Mount Whitney, USS Blue Ridge and USS wo Jima'. aval Electronic Lab., USA, ELC7D 206, Dccembcr SHADS. T. G., and WOODY, J. A.: 'Metalinsulatormetal junctions as surface sources of intermodulation'. Georgia nstitute of Technology, USA, RADCTR8331. February FRAZER. M. J., and MORSSEmE. S.: 'Study concerning nonlinear mixing of RF signals in steel structures'. TR T Research Laboratory. USA, October STURTO, C. H., er al.: 'nvestigation of intermodulation intcrfcrcncc in the Mc/s mohilc radiotelephone senice'. Radio Report o. 50, ew Zealand Post Officc. March BE'TS, J. A., and DEBEY, C. W.: 'ntermodulation measurements on land mobile radio sites'. EE Conf. on Radio transmitters, UK, 1980, pp SO BEVA, H. L.. er al.: 'Suppression of intcrmodulation product generation in materials and structures used in radio communications'. ERE nt. Conf. on Electromagnetic compatibility, UK, 1986, pp BULEY, P., et al.: 'nvestigation of intermodulation products generated by antenna towers at radio sites utilised for land mobile radio scrviccs'. ERE nt. Conf. on Electromagnetic compatibility, UK, 1986,pp. 111l19 52 FUDGE, R. E.: 'Thc reengineering of Mobile Radio Services in the United Kingdom'. PhD Thesis. City University, UK, LU, P. L., and RAWLS, A. D.: 'The field measurement of passive intermodulation products'. EE nt. Conf. on Mobile radio and personal communications. UK, December 1989, pp LU, P L., and RAWLS, A. D.: 'The dcsign and improvement of PMP measurement facilities and the measurcmcnt of PMP in antenna structures'. EE Colloquium on Passive intermodulation products in antennas and related structures, Digest o. 1989/ 94, June pp. 7/17/8 55 LU, P. L., ct al.: 'Measurement of intermodulation products gcneratcd by structural components', Elecfron. Left. 1988,24, (16), pp EE: 1990 Received 16th March 1990 PakLeng Lui is at the Electronic Engineering Laboratories. The University, Canterbury. Kent CT2 7T. UK. He is an EE Member. 118 ELECTROCS & COMMUCATO EGEERG JOURAL JUE 1990

11 易迪拓培训 专注于微波 射频 天线设计人才的培养网址 : 射频和天线设计培训课程推荐 易迪拓培训 ( 由数名来自于研发第一线的资深工程师发起成立, 致力并专注于微波 射频 天线设计研发人才的培养 ; 我们于 2006 年整合合并微波 EDA 网 ( 现已发展成为国内最大的微波射频和天线设计人才培养基地, 成功推出多套微波射频以及天线设计经典培训课程和 ADS HFSS 等专业软件使用培训课程, 广受客户好评 ; 并先后与人民邮电出版社 电子工业出版社合作出版了多本专业图书, 帮助数万名工程师提升了专业技术能力 客户遍布中兴通讯 研通高频 埃威航电 国人通信等多家国内知名公司, 以及台湾工业技术研究院 永业科技 全一电子等多家台湾地区企业 易迪拓培训课程列表 : 射频工程师养成培训课程套装该套装精选了射频专业基础培训课程 射频仿真设计培训课程和射频电路测量培训课程三个类别共 30 门视频培训课程和 3 本图书教材 ; 旨在引领学员全面学习一个射频工程师需要熟悉 理解和掌握的专业知识和研发设计能力 通过套装的学习, 能够让学员完全达到和胜任一个合格的射频工程师的要求 课程网址 : ADS 学习培训课程套装该套装是迄今国内最全面 最权威的 ADS 培训教程, 共包含 10 门 ADS 学习培训课程 课程是由具有多年 ADS 使用经验的微波射频与通信系统设计领域资深专家讲解, 并多结合设计实例, 由浅入深 详细而又全面地讲解了 ADS 在微波射频电路设计 通信系统设计和电磁仿真设计方面的内容 能让您在最短的时间内学会使用 ADS, 迅速提升个人技术能力, 把 ADS 真正应用到实际研发工作中去, 成为 ADS 设计专家... 课程网址 : HFSS 学习培训课程套装该套课程套装包含了本站全部 HFSS 培训课程, 是迄今国内最全面 最专业的 HFSS 培训教程套装, 可以帮助您从零开始, 全面深入学习 HFSS 的各项功能和在多个方面的工程应用 购买套装, 更可超值赠送 3 个月免费学习答疑, 随时解答您学习过程中遇到的棘手问题, 让您的 HFSS 学习更加轻松顺畅 课程网址 : `

12 易迪拓培训 专注于微波 射频 天线设计人才的培养网址 : CST 学习培训课程套装该培训套装由易迪拓培训联合微波 EDA 网共同推出, 是最全面 系统 专业的 CST 微波工作室培训课程套装, 所有课程都由经验丰富的专家授课, 视频教学, 可以帮助您从零开始, 全面系统地学习 CST 微波工作的各项功能及其在微波射频 天线设计等领域的设计应用 且购买该套装, 还可超值赠送 3 个月免费学习答疑 课程网址 : HFSS 天线设计培训课程套装套装包含 6 门视频课程和 1 本图书, 课程从基础讲起, 内容由浅入深, 理论介绍和实际操作讲解相结合, 全面系统的讲解了 HFSS 天线设计的全过程 是国内最全面 最专业的 HFSS 天线设计课程, 可以帮助您快速学习掌握如何使用 HFSS 设计天线, 让天线设计不再难 课程网址 : MHz FC/RFD 线圈天线设计培训课程套装套装包含 4 门视频培训课程, 培训将 13.56MHz 线圈天线设计原理和仿真设计实践相结合, 全面系统地讲解了 13.56MHz 线圈天线的工作原理 设计方法 设计考量以及使用 HFSS 和 CST 仿真分析线圈天线的具体操作, 同时还介绍了 13.56MHz 线圈天线匹配电路的设计和调试 通过该套课程的学习, 可以帮助您快速学习掌握 13.56MHz 线圈天线及其匹配电路的原理 设计和调试 详情浏览 : 我们的课程优势 : 成立于 2004 年,10 多年丰富的行业经验, 一直致力并专注于微波射频和天线设计工程师的培养, 更了解该行业对人才的要求 经验丰富的一线资深工程师讲授, 结合实际工程案例, 直观 实用 易学 联系我们 : 易迪拓培训官网 : 微波 EDA 网 : 官方淘宝店 : 专注于微波 射频 天线设计人才的培养易迪拓培训官方网址 : 淘宝网店 :