10.1 mTKwmIoR

L nseauring wyetm with tel*mwtry in gsnwral oonaiwts of transducers, signal ccnditicning circuit*, an airbcrne mltiplexnsr and a radio frequsncy tmnamitter in ths aircraft, and on ths ground an rf rwosivwr, a dwmltiplwx*r and a d&a wtorags sywtem (usually * tape racordsr); in moat a*** *data (pre)pmo*seing *y?Jtem with displaya ia WddSd *or on-line data an¶lywiW. The te1ematry part Of the systm will be disoussed I* this chapter.

It oonwiata of the mltiplwxwr with the aasooiafed data modulatora, the rf link, the damltiplsxer, the data dwmodulators and the grand meording equipat. The methods of on-lina data pmcsssing till also ba briafly ccneidsred. In ccmscticn with on-line d&s pmcassing, tslamatry has bsccme a powerful msans of incressing the capability and sfficisncy of flight tssting.

Telwmetry of flight teat data haa a numbar of advmtagas c"*r ths n** of on-board moording: a tslsmwby sywtsm has less weight and volume, it i* I*** ssnsitivs to *tireme *mirom*ntal condition* than the m-board ~ccrdsr and it has bettsr quick-lock and on-lin* data pmosasing capabilitias. In *cm* type* of flight teat it wculd be almost impoa*ibl* to ccllsot a tificisnt quantity of data without telemetry. The u** of a second telemetry link from the ground-station to the aircraft (tele- oonmmnd link) can provide further ~mprcvement of the fli&t teat efficiency.

On the other hand there am a few drawback*: ths rang* ia limited by tbs physical characteristic* of wave pmpagation, and there are problsma with tha mcmti,,g of on-beard a,,t*n,,s* and with dmponts of data mcapticn dus to fading inths r&c fmqusncy channel by multiple pmpagaticn. Detailad ocmparisms srw given in Ref. 1 and 81 wet) alao Ohaphr 8. The tslsmstry chin is shown in Figure 10.1.

Figure 10.1 Blook diagrsm of a t*lem*try *yat*m

The airborne aystsm consists of the mltiplex*r, tha rf tnmmittar and tha on-boar.3 antanna. Tbs radiated *l*ctromgn*tio WV* inducss an rf voltage in ths mesiting antanna. Tbia vcltage i* ampli- fied snd filtered by the mcaiver in crdsr to alimimts mwmt*d signal* and noise. The detitiplsrsr d*ccmpa**s the recsiv*r output signal and mccv*re the original data. Wmlly ths data pmcsseing equipment conv*rk* the racsived data into the proper form for th* UBOI. x) Some modulation mathods m&e urn* of the well-known trads-off between signal-to-noise ratio and bandwidth. They obtain better signal-to-noise ratios at the output of the dsmadulator by manipulating bandwidth ad shape of the spectrum. As s rule it will be advantageous to use a modulation n&hod which oocupies 811 the available ohanne bandwidth. 10.3

The usual mcdulaticn methods *=a shown in SPECTRAL CARRIER Figures 10.3 end 10.4. We have to distln- DENIITYt IPECTRUMOFTHE goirish betuesn the continuous modulation

SPECTRN the frequency domain by its spectrum (awe DENSITY t Fig. 10.3). For the pulss modulation methods, the parea*t*,ra of a pulse oarrier (amplitude, duration) are ccntmlled by the signal fm rl,-l,-AF '< Gdf,+f,+aF voltage (awe Fig. 10.4). In this cam the

FREWZG modulated carrier can bs displayed more P FREQUENCYYOcuL*TION IFM) clea*1y in *ha *ml* domain by Its time *lJnctmn.

The time function of the modulated carrier for amplitude modulation (AMlis:

} . CC8( WC . t) (10.1) with the carrier wa”B UC. CO8( w,.t). The ncrmalizsd signal time fun&ion *(t) is bounded by + 1. The range of the modulation factor m usually is m&l. The spectrum cf the modulated wave consists of one line at the csn-isr frequency oca"d an upper and lower sideband. The upper sideband is ob- AMLI- tained by shifting the signal spsctnun by luCE w. along ths w-axis. The low** sideband is the image of the upper sideband, symmatri- aalto "c. Therefcre, the bandwidth of the AM spedmm is twice the bandwidth of the origin.41 wiglIe. epctrm (see Ref. 1, 2, 4 ad 5).

The main drawback of AM i8 that most of the power cf the modulated way* i* n- &red for the carrier. In ths dsmcdulatcr the oarrisr is ussd only for stitching the Aid wave in order to r*cov*r ths demcdulatsd signal. It contains no signal infczmaticn. Since ii is passable to obtain the witch- ing signal for the demodulation pmcess in Figure 10.4 IUse modulation msthcds other waya, - e.g. by mmipulating the sidebands -, it i* mere efficient to suppress the carrier before transmission (double-sideband suppraee- ad-carrier modulation (DSB)). Howevsr, ths hardvars of DSB ia more ccmplex than that of the simple AM. In the cas* cf AM and DSB all the signal infcxmation ia contained in each of the two sidebands. Tnerefcre, one sideband may bs suppressed by filtere and the sir&*-sid*be.nd suDDz.ass*d-carrier modulation (SSB) i* obtained. The bandwidth of SSB is equal to that of the signal,and half ths bandwidth of AM and DS3. 'I%* signal-to-noise ratio is equal to that of DSB but, b*cau** of ths smaller bandwidth, an* can handle twice as many data channels in a given frequency range as lnth AM and DSB. The main drawbacks cf SSB ar* ths high degree of hardware complexity and ths lack of M: responss. Th* spectra of DSB end $23 may easily be derived from ths Ay ape&rum dieplayed in Figure 10.). 10.4

For a deeper understanding ci modulation it may be wcrth mentioning that a close connection erista with the sapling theorem, wiuch is discussed in Chapter 6. lie aemme a signal speotmm with an upper frequency limit f.. When the" the oartier frequency fC IS less *hII 2 I.' thsrs will be flaqusncisa at which ths unmcdnlatad and the modulated signals overlap. This will cause aliaaing error* (or ermm of aommisicn as they are called In Chapter 6) when the signal is demodulated. The AM pmce88 may be regarded as the sampling of a signal s(t) by the ains wave.

Frepusncy modulation (EWE (Fwxe 10.3b) IS a wideband modulation method which me&es use of an ex- tended bandwdth zn order to ~mprcve the signal-to-noise r&u,. Because of the relatively simple hardware FM IS of grsat importance for flight testmg. In FM the frequency of the carrier wave is modulated in the following way fm * fc + AF.S(t) (10.2)

whsre fc is the frequency of the modulated carrier, AF is the frequency deviation and B(t) is the ncnnal- lzed signal (see above). It can be shown (see Ref. 1, 2, 3, 4 and 5) that the increase of the sqm,l-tc- nciss ratio is pmperticnal to the ratio M-F (10.3) I M ie the 80 called mcdulatlon lnder and f. tile highest frequency ccmponsnt of the Signal spectrum (8.0 Fig. 10.3b). Unfortunately,the bandtidth cf the FM-spectrum also increases aa e linear function of AF, thus limiting the obtainable gain because of the general restnotions cn available bandwidth. For FM subcarrxers mcdulat~cn lndxces of 5 are used in practice for fllgbt testing.

A special case of FM is the phase modulation (PM). T%is ia accomplished by letting the signal a(t) control the carrier phase instead of the carrier frequency. Its special feature is a preemphasia which incrsasss the amplitude of the signal spectrum linearly with frequency.

In pulse amDlituda modulation (PAM) (Fig. 10.4a) the signal a(t) ~8 sampled at discrete points 1" time. This pmcass is described in detail in Chapter 6. All considerations of that chapter apply directly tc PAM.

Fulse duration modulation (FDML ('Pig. 10.4h)

E4K?L is obtained by ccnvetiing the amplitude of the PAM samples into a pulse duration. Thus a train t 2 n of pulses with variable width is generated, and the d.yne.mic range of the signal is transformed fmm the amplitude domain to the time domain. The minimum value of the signal s(t) corraspcnds to the shortest pulse dureticn) the maximum value oarresponds to the longest pulse duraticn. The amplitude of the PDM pulse train rsmains constant. PDM, 88 FM, is a wideband mcdulaticn method, It is only used for simple ayeterns (sectmn 10.5).

For the sake of completeness, pulse Doei- ticn modulation, or pulse phase modulation (PFM , is also displayed in Fig. 10.40, though it is *YPLINDE/ not a standard modulation method in telemetry. In this 0~80 the relative psitmn of * pulse is controlled by the signal. A time reference El” I/ is required for demodulation. u ” Due to the great technclc,$cal progress in integrated cirouits, the use of pulse code mcdu- letion (PCM~haa become important during the last few yeam.In this method the PAM samples are 10.5 ooded in l %r.wd~ of R pubes, wing only the levslm 0 and 1. Being a digital method, any mwimd aocuracy mn be obtained by pmpsr ohoioe of the word length II. Besides, PCMWea axcellent "ae of the law of exobangeability between signal-to-noise ratio and bandwidth. FCMis widely used in time domain multiplexing my.tams. Serial fomat XX oar, be derived from a PAM signal by a" analog-to-digital oon- verter with a aerial output. The olook fmquenoy for the A,D cometier rmpt be A-times the PAM sempling fnpuw,oy. This ob"iowly #ho"8 the inamase in b-width. A, the mrmbsr of bits for eaoh sample, is d&.lmined bytba xquirad mplitude nsaoltiion. On the other hand it im alear, that PCA ie less ssnsi- tive to emn dua to noiaa, beoawe only two levsl~ of the signal are possible. Thala are various formst. for encoding the two l-la 0 and 1. In Mg. 10.46 the "non return to mm-ohange (BRZ-C)" for- ut hae been used. For further details 8oe Seotions 5.3.5 and 9.5 and Ref. 6. ITyre 10.5 illustrates the tuo multiplexing methods uasd in praoticel frequenoy-ditision multi- plexing ad tine-divieion nultipldng.

The method of fmqwncy-diviaion multiplexing (Fig. 10.5a)is generally used with oontinuou~ modu- lation nethodB, .wh aa AM, WE‘, SSB or FM. Eaoh d&s signal modulates a subcarrier with a different frsquenoy. By proper seleotion of the mbcsrrier frequenoiss, overlapping of the modulation spaotra oan be avoided. At a noeiver end of the teasmission link the individual suboarrlsrs am separated by band- paas filters. The data signbla are reoovued by demodulation of the mboarrisra. On the other hand, time-division multiplsxingie generally used with pulse modulation methoda, such aa PAM, PDY. PRI and pcy. Figure 10.5b ia based on PAM. A commutator, e.g. a rotary sntoh, samples n different data signals with the same sampling frepuenoy fo, but at consecutive points in time. Thus, a train of non-xerlapping pulesm ia generated whioh w.y be decommutatsd by a aynohmnously running switoh at the reoeiver end of the transmission than- nel. In ordw to obtain the required ayn- chmnism, a synobmniaation signal must be / inserted in the alse train, whioh oan be deteoted by the decommutator and can be wad for aynohmnizing the position of the stitch. ThsnZore, the eynchmnisation signal umally im given a value outside the aynamio range of data signals. Beaides ma!cLngpossible the efficient uee of the available transmission ohannal bandwidth, time-division multiplexing allows s simple exchange bstwsen data channel bendwidth and the number of obsnnsls by auperoomamtation and auboomutation teohni- qnem. Both methods are shown in Fig. 10.6. Superoommutation increases the sampling frequency of a ohanne by seapling the data signal more than once per frame. This can be dons by paralleling some oh-elm of the commutator. Obviously, the number of than- nele dsoraa#en. The example in Pig. 10.6a showa the substitution of 24 ohannele with a ~ampli-ng frequency of fo by 10 ohannsls with different sampling frequencies ranging fmm f. to 4.fo. Suboommutation means the dsoresae of the ohannel sampling frspusnoy by aubsti- tuting one ohannel of the main frame sampl- ing fnpuenoy f,, by n channela of a aub- frame using a sampling frequency fJn. The Fig. 10.6 Eramplss of euperoommutation and exeapls in Fig. 10.6b show the inomsss auboommutation principlea 10.6

The FM FM telemetry *,ptem me* freouenop Aivision multiplexinp. The data si~~~als modulate mboar- riers with FM (standardised fremency deviation + 1,506 or 15% with optional wideband channels). The mximum number of ohanneln 19 21 (for the P-band rf channels limited to 19). The subcarrier fre- quencies am looated between 400 Hz and 165 kHz. This is true for the proportional bandwidth *yetem, in which the bandwidths of the subcarrier channels inoraaae proportionally to the subcarrier frecuenoy. The constant bandwidth sptem also provides 21 data ohannels (15 channels for the P-band frewencies). The mbcsrrier frequencies are located here between 16 kHz and 17h Hz. The data bandwidth of all ohan- nels is equal. Consequently, the signma delay in 811 channels is equal and the time correlation between the channel* is preserved (see aleo Chapter 3). The wailable hardware is proven and reliable. This ia especially true for airborne subcarrier oscillators and for groun* subcarrier discriminators. WI fur- ther t~yetem using frequencg multiplexing have been developed: the DSB/FM qyatem snd the SSB/FM *yetem. Both systems m&e better use of the available transmitter power than the standardised F%!/FPl *yetem. Because of the lack of DC response of the data channel, however, SSS/lW 1s not mAted for penera sppli- cations. DSB/FM has excellent prqerties. especially f-or hi@ frRgU*nCy L¶t*. But, beoause Of historical reasons and the perfect technolow of the FM/FM technique, DSB/FM is not in widespread use. Besidae using the rf channel efficiently, the PAM/FM e.ystem has some f'u~ther advmtaps vhioh may recommend it* u*e. Being a time multiplex system it Dan accommodate a large number of data channels with highly different bandwidths by using: mpercomutation and subcommutation techniques. The teohnolaqy of PAM multiplexem has made preat prom-e** in the l*st few y**r*. At present a 16 channel multiplexer on one chip is already available and further progress in medium scale integration of PAM circuitr,y c*n be expected. In modem flight teatinr, PCM/FM system* *re more and mom used. PC% provides the *me flexibility aa PAM, with the poesibility of attaining: higher accuraoies. TIem accurate anslog-ta-dietal converters are available in integrated circuitry. Thus, the drawbacks of complexity and price will be reduced. Hitherto apeoial decommutators were used in most ground stations. Recause mmputers *m mm-e and more used, decmmtation is often done by a oomputer along with certain on-line data processing. The bit synchronizer, however, which has to detect the bit sequence in the noisy background, should preferably not be integrated into the computer. All three telemetry systems mentioned above make use of 174 in the rf channel. This ia the onlq method standardised by IRIG. The other continuous modulation methods 88 AM, DSB, SSB, which m*y *lea be used for rf modulation, *re of little importance.

Mixed systems *m often used in practice, because only few fiiqht test siqnnals have a wide bendwidth, while the majority has B narrow bandwidth. Because of past technological difficulties in praducinr: cmm- tatars and decomutators with nufficiently high sampling rates, these high frequency rlata signals were transmitted by frequency multiplexed channels. On the other hand economical reasons and the moderate mm- her of available frequencg-muitlplexsd ohannele rewire the use of time multiplexing for the low frewuen- cy data br, for instance, the PAM/PM/FM system. Due to the recent progm** in high 8ped integr*ted Fir- cuita, uideband PAM and PCM channels are currently available. Therefore, strai~htfomard time multiplex- ing is preferable because of the hi& flexibility (mperoammtation and suboommt*tion) to&her with the effioient UBB of the *vailable bandwidth.

finally, the inaccuracxes introduced by the processes of modulation and mltiplexinp: met be con- sidered. Errors, oritinatin.. n from h*r&cwe imperfections (e.g. *em and gain drift of amplifiers, non- 10.7

linear distortion) should bs distinguished from errore originating fmm pouliaritiam of the mathods used in the mymtmn (e.g. aenaitivity to noise in the trawniaaion -01). The first alama of wren im mbjeof to the teoi,nolo@oal progress, vbsrsas the latter olass muat be n@ed am inbsnnt to the SYstem. Aa mentioned above, wideband modulation methods incorporate a gain in signal-to-noimm ratio in no far aa transmission oh-al noise ia anowned. Baoause of the twofold modulation in tslamatq, the system performance is oharaoterieed by the overall bandwidth repinsme& and the cQn.sl-to-noise ratio at the output of the demultiplexar. Therefore, the question for the optimal sysbam ial wbiob oombination of two modulation gmossaes assures the maximumdata bandwidth (marlmum infolpation rate) snd meets B given rf bandwIdth limitation as well aa a resuirsd ao-oy. It haa been fwnd (see Ref. 9)tbat m ia the optilnum ayetem in the oaee of low to medium ao~uraoy rapuirementa (1 to 3 %), while pnr/aw is be& with high aco"x%oy raquirements (aoouraoy bsttartban + 0,5 6).

On the other band the bardmrre of the W/IN ayetam is very refined bsoause of the long period of usage and the hi& level of knowhow in production. Tbsrefors, the first-mentioned -up of emra is relatively low in i+X,M. As PAMtaobnology wqrassaa, Pw/nc till be usad inorsasingly in the naar future. 10.3 THE aaD FRDDXCY LINK

(a) The sleotrioal length of an ante- must be a aignifioant fraotion of the wavelagth for reasonable radiation sffioienoy. The small size rewind in airborne applioations indicates wavsleagtha smaller than 3 m (frequenoiaa greater than 100 We). (b) Tne high bandwidth &red for data tranBrdssion is only avrilable at high freqwmoies. Aooording to IRIcl StandanTs (se.. Ref. 6) three frquenoy bands are available for telemetry: In the range of 816 - 260 MHz (P band) are 62 cbannele with 500 leas bandwidth eaob, in the range of I435 - 1540 I‘& (I band) are 1M) channela with 1 )(Az bandwidth eaob and in the range of 22C0 - 23WMSs (Sbsnd)pr. 89 ohannels with 1 WHz bandwidth eaob =I.

Figure 10.7 Simplified propagation dssi- ob.srt

=I The rime of the P band "as only alloved until Janusry let, 1970. "nfo~t~ately, bo,,avsr, the tr,,xv"itt.re and reoeivem for the L band and S band are at pn~ant &ill muoh moxw viw than tbo~e for the P band. Besidaa, the aiss and the wei@ of the airborne L band and S band transmitterm M higher oom- pared to the P band eqtigw,,,t. Wirily beoaulre of theme reasons the suitoh-r to the L band end 5 band channel8 has not yet been completed. 10.8

the hiper the frequsnoy of (~1 elaotmmagnetic wave, the nom its pmpagstion resembles that of light. The newable range between the airborne and the gmnnd terminals is limited by the line of sight if, in a first approximation, the diffraotion is negleoted. For L and S band frspuenoiea this oan be done with good appmximation. Within the line of sight the formula far wave pmpagstion in free space is a good appmximation, provided the heights above ground of the transmitting and the receiving antennas am at least several wavelengths.

The line of sight Do oan be celoulated by uei the formula Do - 2.28 ( hl +v;;) P 00.4) !&are Do - the dintawe in !a hl - the height of the transmitting antenna (in the arcraft) in feet h2 D the height of the receiving antenna above the ground in feet. Forthme different hslghts h2, thxs formula haa been displayed in Fig. 10.7 by the dashes lines. (iood data reception c&n only be expected for distanaea D between aircraft and ground station satisfying the condition D&Do (10.5)

Within this range the required radiated transmitter power oan be estimated by using the solid OUVBB in Fig. 10.7. Two auxiliary parameters a1 and b2 am used. In a logarithmic scale we have

oT*p**% tj - 10.1og Lp *E

b2 - 71 + 2O.log D (10.7)

oT is the gain of the tmnemittxng antenna, L the product of losaes (aable, mismatahing, minima of radiation pattern), PA and PE are the transmitted pawer and the retired minimum mceiver input power mspeatively for B given telemetry system, $ th e effective area of the receiving antenna in square meters and D is the distance between the antennas in km. h2, measured in decibels, indicates the atten- uatxon of the transrmtted wake ae a fun&Ion of the tistancs D. In order to obtain good moeption, this attenuation must be overoome by bl, which contains those parameters which am indspndent of the distance D.

Thus a supplementary condition to equation 10.5 is

a,& b2 (10.8)

The plot of eq. (10.6) in Fig. 10.7 is based on the parameters GT = 1; L - 101 PE - 4.10 -% ( oorrsct for FM/FM systems)+ s - 1 a2 and treasmitter powers PA of 0.1 W, 1 W and 10 U. This is mugbly squiv-alsnt to the practical conditions. Often instead of the effective ama of the receiving antenna its gain CR is specified. Then, AR can be oalculated by using the formula

x is the wavelength. The UBB ofFig. 10.7 will be made clear by the following trample: For a height of the aircraft of 1COO feet and a height of the receiving antenna of 3.3 m (10 feet), the line of sight is 80 !a. With the above-mentioned assumptions we have at this tistancs a safety margin of 5 db even with a transmitter power of only 0.1 W. This safety margin may be sufficient when using a gmund station with diversity reception capabilities (see below). Otherwise, a transmitter power of 1 W would be repired in order to o~*roome fading effeots, wluch are not taken into acoount in Fig. 10.7.

One of these fading effeots is the 80 called fading due to multioath ~~ropwation. Sometimes the antenna not only receives the wave coming directly from the tmnsmxttar but also a reflected wave (reflsoted from the ground or fmm buildings), the phase of which 18 shifted with mspad to the dxraot wave. Depending on the heights and the distances of the antennas the phase shift may reach 180'. Thus the received signal power may from time to time deomasa considerably. 10.9

1t should be mentioned also that the curve b,(D) at the limit of the line of sight is too Optimistic. Beyond this limit the wave-attenuation increases very rapidly. Therefore, little or nc increase in range can be obtained even if the transmitter pcver is substantially increased. On the other hand, the example shows that within the line cf si&t reliable ccmmunicaticn with relatively low transmitter power is possible.

An effective way to cverccme the difficulties caused by fading is to use diversity reception tech- niques. This technique uses two rf receivers (see Fig.lO.1). Each reoeiver is comnscted to a separate antenns, which picks up waves of different pclarizaticn (polarieaticn diversity), of different frequen- cies (frequency diversity) or at different looaticns (space diversity). The drawback of the frequenoy diversity method is that it repuxres the additional bandwidth for a complete second rf channel. A pre- detection or pcstdetectvan ccmbiner at the output cf the receivers combines the wei&ted sum of both input simls. The weighting ccefficxnts are ccntrclled by the siml-to-noise ratio of the two inputs, in such a way that the coefficient approaches 1 for a good signal and 0 for a bad eignal. Because Of the different ccnditicns of wave reflection it is not very probable that a loss of signal cccups at both in- puts simultaneously. The probability of pod recepticn, therefore, is substantially increased.

Tne on-board transmitting antennas are mounted in a fixed position with respect tc the vehicle. The axes cf these antennas may, therefore, have different directions with respect tc the ground during a flight. Therefore, an cwidwecticnal radiation pattern is desirable but wnnct be achieved exactly. The free- space radiation patterns of the antenna systems that oan be used an aircraft alwws show minima. In addi- tion, the aircraft itself distcrts the radiation pattern. Parts other than the antenna may reradiate *i&s (parasitic radiation) which will interfere with the original eignal. Furthermore, at sore atti- tudes of the aircraft, the wing or the fuselage may intercept the radiation to the ground station (shadcw effects). Therefore, the layout of the airborne antennas must be dcne very carefully, taking into account the expected attitudes of the aircraft during fli@t.

The airborne antennas are oftenh/4-stub or blade antennas, which have a linear polarisation. Most of the grcund receiving antennas are circularly pclarieed in order tc obtain a reception independent of the orientation of the airborne antenna. Turnstile or helical antennas azw used for the P band frequen- ties; circularly pclarieed feeds with a parabolic reflector are preferred for the L band end S band flwpencies. For the latter frequenciaa higher pins OR nnrst bs used than for the P band frequencies in order to get the sexe effective antenna wea +. This is shown by eq. (10.9). The wavelength h of the L band ie about l/6 and of the S band abcut l/Y of that of the P band, respectively. As a result of the hi&e? gain the directivity is higher, 80 that prcblems with manual tracking of the receiving antennas occur, especially with aircraft which fly at hi& speed near the ground antenna (hi@ angular velocity). Therefore, autcmatic tracking (e.g. monopulse tracking) cften is preferred, despite the higher complex- ity of the ground station.

At the receiving end of the telemetry link double suparhetercdyne reoaivers with plug-in techniques are mostly used. Therefore, the receivers can easily be matched to different telemetry systems by chccsing the proper tuning heads, intermediate frequency filters, and rf demodulators. In order to obtain good receiver input sensitivity (lcw value of PE in equation (10.6))the noise figure of the receiver must be kept lcw. A low-noise preamplifier situated at the antenna is recclmnsridable, if it ia not possible to locate the antenna near the r+zeiver. The lcw nci8e fi@n% of the preamplifier then detarmines the ncise ii- of the receiving system. Reception of L band and S band frequencies CM be done with special tuning heads, which are available for stantidired telemetry receivers. Alternatively, a frequency dcwn- converter may be mcunted at the antenna which converts the L band or S band irequencies to the P band. Then the existing P band equipment can still ba used.

Most telemetry receivers can be equipped with en accessory unit for postdetection or prsdetecticn diversity combining. Both methoda exe treated in the next section in scmo more detail.

10.4 EW”NE RECORBINO*ND DISPLAY

The recovand data signals must be diaplmd for quick lock in the evnd station and stored fcor subsequent data processing. Quick-look display usually is dons with pointer inst-nts as well 8s with strip-chart recorders and X-Y recorders. Dsta signals containing frequency components higher than 100 BB must be recorded by oscillographic recorders. Theme paper recorders have the advanty of being both a 10.10

display and a storage equipment. On the other hand, the storage capacity does not meet the hi& require- mentm of modern fli@t tests. Moreover, the stored information can only be converted back into an eleotri- cal signal with great difficulty. Thus, subsequent data processing is restricted.

In this connection ma@xt~c tape recorders have excellent properties (see Chapter 9.4, Ref. a), and they are therefore standard equipment in telemetry ground ststlona. A few years ago postdetection recording was mainly used. m which the frequenoy-multiplexed a~gnal (FI@W system) or the t~me-~lti- plsrsd sq,al (PDy/FM, PA&V, and PCWFM system) at the output of the demodulator ie reoorded on one traok of the tape recorder in the direct mode or in the FM mode. Tie derultiplering is done during playback. This method allows the recording of a great many data channels. It is also poeaible to do the demnltiplezing on-line, 80 that the individual data signals are immediately recorded on different tracks of the tape recorder. This method ie limited by the marimwn number of tracks that can be recorded simultaneously in the ground reoeiving station.

Recently, padetection recording has become more naportant. The svailab~lity of recorders with continuous recording oapabllltaes up to frequenoias of 2 MHz allows the direct recording of the re- ceiver intermediate frequency (third 12.1 manmum YOO Liz+ see Ref. 6) prior to demodulation. The main advantage is that the ope~et~on and maintenance of pound statvans is simplified, especially for those using several telemetry methods (iJ@X; PDvFMi PAvFMi PCM/FM). This ia the case, bcause the recording method is the same regardless of the type of multiplexing used. In postdetectIon recording, on the other hand, the recording method (direct recording or FM recording) must be matched to the type of multiplexing. In addxtlon, rt 1~ advantageous to record the signal as early in the transmission chain 8s posszble. The optimaatlon of the playback ohsnnel with respeot to the signal-to-noise ratlo (e.g. varying the I.F. bandwidth and consequently the demodulator threshold) need not be done prior to reception but can be attempted later by repetitive trial and error. In other words, in postdetection recording the I.F. bandwidth must be ohosen before the test on the baszti of an expeoted rf signal bandwidth. Normally a higfler standardieed I.F. bandwidth 18 used. Under bad reoeiYing oonditions, however, the sipal-to-nolse ratlo S/N may became too low and consequently a loss of data can oocur. In predeteotion reoordxng the I.F. bandwidth oan be decreased before playback in such a way, that the S/N ratio is raised to a oerta~n extent.

10.5 coKPAFmcm OF IRIG-STANDARDI~D *mmTRrmy sYsTEMs

A comparison of the IRIC standardised telemetry systems xs made in Table 10.1 on the basx of 5 different principles. IRIC standardlzed telemetry systems a~ well-establlehed, proven and available on the market at reasonable prices. They are mutable for almost all measuring problems.

Accvaoyr The accuracy of modern analog systems, when carefully adjusted, approaches + 1 $ (see Ref. 8). The aoouraoy of the twofold multiplexing 1n the case of the PAM/FdFM system can be + 2%. PCM/FM 18 capable of almost my required a~~uraoy. The only limitation ,B the aoouraoy of the transducers and the A/D converters "Bed in the eystem.

Maximum number of data ohannele: AB discussed in Section 10.2 the systems using time dx"lelon multiplex- ing have the highest oapaolty in number of channele.

Maximum information rats (13.1; In Table 10.1 the 13. is given xn b~t/sec. In the ease of the analog systems the I.R. is calculated from the slpiflcant amplitude resolution and the sampling theorem (see Ref. 8). The PNFM system and the PCM/FM system can handle the hzgflest I.R., followed by the F,$'%M system. Comparatively, PAM/FM/FM and PDVFM can handle only very poor information rates.

Flexibility; Flexibility is very good for eystems u8mg tun-dL"lalon mult~pler~ng (see Section 10.2). Subcommutation and supercommutation can be used to adapt the system to requirements for the numbw of ohannels and for frequency ~eponse. In the ease of PCM/FM xt is also possible to we different word len&hs for the Individual parameters; the indl"idual channels can thereby be adapted to the accuracy of each parameter.

Utilisation of radio frequency power and bandwidth: These aspects are covered in detail XI Ref. 9. It has been found, that 1x1 the case of high aocuracy reqnrements PC@! 18 the best method for transmit- ting data with B certain informatIon rate through a band-lImited rf channel with m~nimvm rf power. In the case of moderate accuracy requirements PAWFM requires minmurn bandwIdth. The widespread W/R4 method is not as good and the PAVJFVFM method has only poor features in this respect. In the latter Table 10.1 Comparison of IRIG - Stendardiaed Telemetry systems

Rinclples of I FM-FM II FM-FM III PAM-FM-FM I” PAM-FM " FIN-FM VI KM-FM oom*arison proportional constant bandwidth bandwidth

+1% 22% 21% +l% Limited by accuracy &'*,= TM= 5) of transducers or A/D converter* only

19 (VHF band) 15 (m band) 128 without sub- see III go WIthout sub- Depends on word len@h. 21 (UHF band) 21 (UHF bana) commutation. Can comtation. Frame length mar. 2048 be eqmded most Can be expanded bit. In practice, all effectively by mO*t effective? requirements c2n be met subcommtation by subcannrluta- by super- and subcom- tion tation

Maximum tata1 VHF band: Nearly tile sane 16103 g VET band: t-ST band: information rate 65.103 =(M-5)* a* I (channel F)= 3.6.105 ait 2.105 bit set set set (sum Of all 270.103 q&l) OEF Land: channels) sac 1.2.106 bit UHF band: set 110.103 *M-5) set 460.103 =(M=l) set

Flexibility Noderate (fmed see I Good Good Good Excellent (sub- and (number of chan- subcarrier fre- (low data fm- (low data fm- suprcomtation, IlelSi channel quency allocation; quencies pre- quencies p-e- variable word len@h) cat-o** freq.) only a few maria- supposed) supposed) tmns in channel bandwidth)

Utilization of "oderate See I Poor Good Good Good radio frequency (can be tolerated (optm.1 for (optimal for hi& pmer and radla as the power r.?- moderate accu- accuracy requrements) frequency band- quiremnt and the racy require- width bandwidth is low ments) because of the lox information rate)

YHF band: P tend UHF band: Land S band -- M _ Modulation Index of FM carrier m Channel with the highest bandwidth of the FM proportional bandwidth *ubcamier pattern 10.12

cam, however, the poor features ten be tolerated, as the system carlYes only a low iniarmtion rats. Normally it is used in addition to B F@‘M system. Then, the features of the combined system BP* dster- mined by that of the FM/FM method.

10.6 ON-LINK D*T* PllOcESSINC

*s tht data nmaaured I” the aircraft are mmediately available in the ground station. te1emtry makes it poeslble to observe the measured parameters while the fligw ie in progress. ““Cl recently this on-lins processing and display was mainly used m critical phase* of the flights only. In ths latest flight tests with mllltary end civil prototypes telemetry and associated on-lina proosssing i* being used in all flight test phases and for the majority of the pareasters. Although on-board record- ing is *till used 88 a standby in case telemetry data ar* lost, the=* is B tsndenoy to do most of the analyem from the on-line displays fed by telemetry. Thus, modifications to ths flight test programms oan be made during flight. Exper~enoe has shown that thi* can reduoe the number of flights required In a test programme by 3% or more.

The genera1 conslderatlons about data processing a** discussed in Chapter 12. Her-3 a few remsrke will be made about on-line computing.

Andog oomputina methods are very useful means of on-line data processing, *specially if the number of channels I* not v*ry large. They combine high speed of computation with good adaptability of the hard- ware to individual problems. The acouracy is usually not very hi&. But in many was** a. sufficxnt accuracy can be obtained, especially for quick-look display. An example for a relatively simple applica- tion is the computation of lndlcated airspeed, true airspeed and Mach number from the measured data, total pressure, static p~***u** and temperature. About *even operational amplifiers, two rraltipli*r* and two square-root function generators are reqmred.

Diaital computing has a number of advantaps over analog computing, far instanca:

- the computation accuracy can be as hi& as I* justified by the accuracy of ths input data - integration oan be done wIthout draft (analag integration shows * drift which increases with time) - storage specifications are better (quick act***, arbitrarily long dwatlon) - the dletal computer can be more readily used for m&ing logical deoiaiona, such as dsttcting that a signal or a oombinatlon of szgnals has exceeded a c*rtaxn limit value - in a more sophisticated application several measured parameter* can ba used e.8 an input to a model& the output of the model 18 comparad to other measured parameters and if the difference is too largs a special pro~mme 18 nut,ated - the computer oan also be used for several tasks in the ground station (*wh 88 dscormzutation of PC% signals) and for automatic control task* in the ground *tatIon (automatio oontral of the receiving antenna, autometic **arch pattern* with high-directivity antennas, *witch-over to autotraoking when aoquisition is obtained etc.).

A problem with dxgital on-line computng is that the time required for all computations must be lees than the time between two successive data samples. Even very fast computers reaoh this limit vsry qu~clily when handling complex problems.

Hybrid comwtzng may be, to some extent a eolution to this problem. By combining an analog computer and a digital computer, the computing programme can be divided into two parts, making optimal II** of the advantages of both methai’s. The interface between the two computers consists essentially of malog-to- distal and dlptal-ta-malog converters. The programming of a hybrid computer ia, however, very dlffloult.

10.7 REFERENCES

1. M.H. Nlchale Radio Telemetry, John Wiley and Sons, I”c.,Nev York, N.Y., L.L. Rauch 2nd Fdxtzo”, 1956.

2. H.L. Stiltz Aerospace Telemetry, Volume 1, Prentxce-Hall, Inc., Englewood Cliff*, (Editor) N.J., 1961.

3. H.L. st11tz Aerospace Talemetry, Volume 2, Prsntica-Hall, Inc., hglsuood Cliffs, (EdLtor) NJ., 1966. 10.13

4. L.E. Foster Telemetry SyBtem., John Wiley and Sons, Inc., NW York, N.Y.. 1965.

5. E.L. Oruenbrg Handbook Of Telemetry and Remote control, McGraw-H~ll, Inc., New York, 1967. (Editor-in-Chief)

6. Telemetry Standards, IRTC Document 106-11.

7. A. Bearer The Role of Telemetry in the Modem In-Flight Measuring Techmque, presented at the 26th AWARDFlight Mechanics Panel Meeting, Paris, 1965.

8. A. Becker Comparison of Systems, presented at the 30th AOARDFlight Mechanics Panel B. cartung Meeting, Montreal 1967, AGAFtDConference Proceedings No. 32. H. Meyer

9. Aeronutronic Inc., Division of Ford Co., Newport Beach, Calif. Telemetry Syetem Study, Final Report, Vol. I - III, AS1 Publ~catmn No "-743, 1959. (S mary in Ref. 5,Chapter %Section 4).