HP-885A-Application-Note-19 电路图.pdf

《HP-885A-Application-Note-19 电路图.pdf》由会员分享，可在线阅读，更多相关《HP-885A-Application-Note-19 电路图.pdf（24页珍藏版）》请在收音机爱好者资料库上搜索。

1、. . . -. ! . _ . _ - , , . ;. . . . - : . .- I ,. , . . _.,. . SERVICE NOTES HEWLETT-PACKARD COMPANY 1501 PAGE MILL ROAD PALO ALTO, CALIFORNIA, U.S. A. CABLE HEWPACK TELEPHONE DAVENPORT 6-7000 885A-l A TECHNIQUE FOR CALIBRATING PHASE SHIFTERS TO HIGH ACCURACY by E. F. Barnett INTRODUCTION The genera

2、l principle of the following method of calibrati_ng phase shifters has been described in the literature 1-3, but the method 1s not well known and some of the details of the present version are believed to be novel. Because of the complexity of the apparatus and procedure, it is useful mainly for cal

3、ibrating a primary standard of phase by means of which other phase shifters can be calibrated using a relatively simple comparison technique. Section I gives the general theory of the method. Section II gives the de- tails of the procedure used in calibrating Hewlett-Packard 885A series waveguide ph

4、ase shifters for use as phase standards. Section III gives an analysis of the e:-rors of the method and the precautions which should be taken to minimize them. I. GENERAL THEORY The transmission-type bridge circuit used in the calibration procedure is shown in Figure 1. A directional coupler splits

5、the signal from the source into two parts, one of which passes through two phase shifters and the other through a single phase shifter and a variable attenuator. A second direc- tional coupler brings the signals from the two branches together at the de- tector. Enough well-matched padding is used to

6、 make mismatch errors negligible. Simple tee junctions can be used as power dividers in place of the directional couplers. However, such junctions do not decouple the two branches of the bridge, so that more padding attenuation is necessary with a consequent loss in sensitivity of the bridge balance

7、. A null can be obtained at the detector by adj us ting the variable attenuator in the lower branch and any one of the phase shifters. Because there are three phase shifters, there are many different combinations of settings of the phase shifters for which the bridge will be balanced. As will be sho

8、wn, it is possible to exploit this multiplicity of balance conditions to calibrate any or all of the phase shifters in the bridge accurately knowing only the approximate calibrations of two of them to begin with. distribution -b, c, f, g RadioFans.CN 收音机爱 好者资料库 SIGNAL SOURCE x I :m DETECTOR Figure 1

9、. The method depends on the possibility of measuring small phase changes with considerable accuracy by means of the corresponding changes in the detector output when the bridge is nearly, but not quite balanced. Two signals arriving at the detector through the different branches of the bridge are re

10、presented by vectors VA and VB in Figure 2. Figure 2. The resultant signal at the detector is represented by V c VA + VB Suppose that the bridge is initially balanced so that Ve 0 and VA= -VB. If a phase shift is then introduced into one arm of the bridge, then the phase of -VB will be shifted relat

11、ive to VA but its magnitude will remain the same. Hence the phase shift ljl can be determined from the signal level at the detector. If /ljl I I (0), cp11 0) = actual phase shifts of I and II corresponding to nom- inal phase shift 0 e 1 (0), en (0) =errors in the nominal calibrations of I and II. By

12、 definition we have the following relations: cJi = 0 + e 1 n (0) = 0 + e 11 (0) The following basic procedure, repeated for a sufficient number of values of the parameters, gives the data necessary for the calibration. (1) Let phase shifters I and II be initially at settings 01 and 02 respec- tively

13、. Adjust phase shifter Ill and the variable attenuator to obtain a null at the detector. Referring to Figure 2, this corresponds to VA = -VB, V C : O. (2) Change the phase shift of Ill by about 1800, adjusting it so that the power at the detector is a maximum, This corresponds to VA = VB, Ve = 2V A

14、Use this power level as a reference level. (3) Change Ill again until the signal at the detector is some small frac- tion (accurately determined) of the reference level established in Step (2). The bridge is then nearly, but not quite, balanced, and we have J Ve I = P jVA ( where p 1. The choice of

15、pis determined by the order of magnitude of the errors expected in the phase cali- bration. Typically, p : O. 1. (4) Denoting the angle between VA and -VB by iV0, we have iVo p radians -+ -+ In Figure 3 the vectors OA0 and OB0 represent the signals VA and -VB after the completion of Step (3). Shift

16、I and II through nominally the same angle as, so that they read a1 +as and a2 +as respectively. The signals VA and -VB are now represented by OA and oB in Figure 3. The new value of the phase angle between them is denoted by ljl. If there were no errors in the phase shifter calibrations, the two sig

17、nals would be (5J. and OB with the same phase angle iVo between them as before. We have, -3- RadioFans.CN 收音机爱 好者资料库 B Figure 3. Therefore, -o = EI (e1 + es) -EI (e1) -en (e1 + es) -en (e1) J -0 (e1, e2, es) (2) The quantity o can be found from the change in the RF signal at the detector produced by

18、 Step (4) relative to the reference level estab- lished in Step (3). Let D denote this change measured in decibels. We have, D o (radians) : p (10 20 -1) ( 3) D :! O. 115 p D if 20 1 To calibrate the phase shifters we must measure the quantity o (81, e2, 8s) by the foregoing procedure for a number o

19、f values of Si. 82, es which are whole multiples of some exact submultiple of 2tr. There are many possible variations in the method. One possibility is to set 81 : 0 and measure o (0,8 z., 8s) where 2Trk 82 = N, k : 1, 2, N e _ 2TTn s -, N n : 1, 2, N The correction of phase shifter I for 8 ing rela

20、tion: 211n = N can be found from the follow- N-l 2trk x 0 (0, r:r, k:O ,11 zk J -4- 0 (27rn d 1. h Solving for e I -W-) an cance ing terms in t e last summation gives: N-1 27rn 1 2rrk 27rn E I-w-) = E 1( 0) + N L: 0 ( 0, N N) k=O n-1 +.!. L N k=O (4) In the case of a microwave phase shifter, which i

21、s typically several wave- lengths long, the choice of the zero setting is arbitrary. Therefore, we can set q O) = EII (O) = 0 without loss of generality. The last term in Equation 4) will vanish if the phase shifter is of a type such that the settings e and e + 2rr are physically identical. Even if

22、this is not the case, however, the value of each of the terms in the summation can be found by a simple modification of the procedure for measuring o e1, e2, es) After performing Steps (1) -(3) of this procedure, shift II from nominal phase 0z to e 2 + 27r. In the notation used before we now have, 4

23、i -lji 0 = e II ( e 2 ) -e II ( e 2 + 2 rr) : on ( e 2) ( 5) We can now rewrite Equation (4): 27rn 1 N-1 27rk 2llIl n-1 o II (2k ) o(O, 1 (6) eI(-w-) = N L -L k=O N N Nk:O The calibration corrections of II can now be found from the relation (7) Alternatively, the roles of I and II in the calibration

24、 process can be re- versed, and the corrections of II found from the relation N-1 1 : - -L N k:O c5(2k, o, n-1 27rn) + .!:._ L N N k:O (8) (9) It is often advantageous in practice to take enough measurements so that e I and e II can be evaluated independently by means of Equations ( 6) and (8), afte

25、r which Equation (7) can be used as a check. If the phase shifter is to be calibrated at a large number of points, time can be saved by using a variation of this method consisting of two or more successively finer subdivisions. Let ea, eb be points at which phase shifters I and II have been calibrat

26、ed by the first subdivision. Let it be required to calibrate the interval between ea and eb at M equally spaced points. The calibration corrections are given by the following equation, which can easily be verified: -5- m q(0a + m60) El (Sa)+ M El (Sb -Sa) .1e = M-1 +.!. !: oeea, ktle, m0) M k=O m-1

27、1 !: c5(0, ke, ab -ea) M k=O (10) 211 Since ea and eb are multiples of N the interval of the first subdivision, E 1( Sa) and E 1( 0,p -Sa) are known quantities. For the special case where ea = 0 eb = 2 we have w 21Tm m 211 1 M-l 2 Tile 2 ITm EI(MN) = M E1(N)+M k;O o(O, MN MN) m-1 - 1 1: c5(0, . ) JM

28、 k=O .1.v.1.Lll .LI (11) For m = 1, 2, M, Equation (11) gives the calibration corrections for 0 e + E1(2;n) r (O 21Tm 21Tn) -u , X!N -6- (13) (14) - II. CALIBRATION OF HEWLETT-PACKARD 885A SERIES WAVE- GUIDE PHASE SHIFTERS FOR USE AS PHASE STANDARDS A block diagram of a practical setup for calibrati

29、ng waveguide phase shifters is given in Figure 4. The waveguide circuit is drawn approxi- mately to scale, and the Hewlett-Packard model numbers of each of the components (for X-band) are given in parentheses. The bridge circuit is basically that shown in Figure 1. A signal generator supplies 10 mw

30、or more of R. F. power with 1000 C. P. S. amplitude modu- lation. The power level is monitored by means of a directional coupler, detector and voltmeter. A 3 db directional coupler is used as a power divider at each end of the bridge. In the lower branch of the bridge simple 10 db pads are used to i

31、solate phase shifter I. In the upper branch two 10 db directional couplers are used as isolating attenuators around phase shifter II. The unused port of each coupler is terminated with a well matched load. The reasons for this difference between the lower and up- per branches are given in the discus

32、sion of mismatch errors in Part III. An uncalibrated variable attenuator is placed to the left of the auxiliary phase shifter (III) in the upper branch. The attenuator in the lower branch actually need not be accurately calibrated. However, it is desirable that it have the high resolution and small

33、variation of phase shift for small attenuation values characteristic of the Hewlett-Packard 382A series at- tenuators. Since 3600 phase shift is produced by 1800 rotation of the center section of a rotary waveguide phase shifter, there are two orientations of the center section corresponding to each

34、 dial reading. Since it is necessary to distinguish between these two orientations, the covers of phase shifters I and II should be removed so that the center sections can be directly ob- served. A reference mark should be made to show the orientation of the center section corresponding to one o0 se

35、tting of each phase shifter. The 885A series phase shifters have a friction clutch which permits the phase shift to be varied while the dial is clamped at a fixed reading. If it is desired to use one of the phase shifters as a reference standard after calibration, the clutch should be eliminated to

36、prevent any possible error due to the clutch being shifted accidentally. This can be done by putting a dowel pin through the clutch sleeve into the shaft. It has been found adequate to calibrate the 885A series phase shifters at intervals of 300. The general scheme of the calibration procedure is fi

37、rst to determine the corrections to the dial calibrations at multiples of 900 and then to interpolate within each quadrant to find the corrections at multiples of 300. In the notation of Part I this corresponds to N 4, M = 3. The detailed calibration procedure is given below. Typical data from an ac

38、tual test run following this procedure are given in the Appendix. -7- I I R. F. SIGNAL GENERATOR 10 00 C. P. S. AMP. MOO. (686A) SIA) (X375A) e 1000 C. P. S. VO LT METER (4158) PHASE - IOdb DIR. COUPLER SHIFTER IOdb DIR. COUPLER . r - II - - 8ARRETTER I (X752C) (X752C) (X910A) (X910A) PHASE (X885A)

39、(X4858 ) .JI SHIFTER MOUNT) . : m (X375A) I (X375A) . . (X885A) I PHASE CALIBRATED -. - . SHIFTER - - - IOdb DIR. COUPLER - 3d b DIR. COUPLER - 70 db) so that a negligible amount of power reaches the detector through the lower branch of the bridge. Set the attenuator in the upper branch at about 5 d

40、b. Adjust the power from the source and the input attenuator so that the detector is operating in the square law region, but well above the noise level of the voltmeter. Set phase shifter II to o0, and adjust the voltmeter reading to some convenient reference level by means of the gain control. Turn

41、 II through two complete cycles of phase shift (cor- responding to one complete rotation of the center section), and record the variation of the R. F. level at the detector at multiples of 300. Repeat this procedure for phase shifter III. The insertion loss of II at phase setting e will be denoted b

42、y An (0), and that of III, by AIII ( 0). 2. Balancing the Bridge Adjust the input attenuator so that the R. F. level at the detector is at least 40 db above the minimum detectable signal. Set the attenuator in the lower branch at 0 db. Set phase shifters I and II at o0, noting the orientation of the

43、 center sections. Adjust phase shifter III and the attenu- ator in the lower branch alternately until a null is obtained at the detector. The component of the signal at the detector reaching it through the upper branch of the bridge is now equal in magnitude and opposite in phase to the component re

44、aching it through the lower branch. In the balance con- dition the attenuator in the lower branch should be in the range from 2 to 8 db. Record the dial reading of III in the balance condition. 3. Setting a Reference Level with Signal Components in Phase Turn phase shifter III through approximately

45、180 from the balance condi- tion, adjusting it to maximize the detector output. The two components of the signal at the detector are now in phase and still equal in magnitude. Adjust the input attenuator so that the power level at the detector is at a convenient reference level, low enough so that t

46、he detector is operating in the square law region, but well above noise. 4. Setting the Bridge to 5. 73 of Phase Unbalance Adjust the input attenuator to increase the power level entering the bridge by exactly 26 db. This can be measured by means of the detector used to monitor power, either by rely

47、ing on the square-law characteristics of a barretter or by using a calibrated attenuator in front of the detector. -9- Adjust phase shifter III so that the detector output is returned to the ref - erence level established in Step 3. There are two settings of III satisfy- ing this condition. Choose t

48、he setting for which a slight increase of the . phase reading on the outer scale of numerals decreases the detector out- put. This choice will result in the corrections having the right signs if the measurements are made as described in the following steps. Record this phase setting of III, which wi

49、ll be denoted by e0 The bridge now has a small and accurately known degree of phase unbal- ance. The component signals at the detector are related as shown in Fig- ure 2. The resultant voltage Ve has a magnitude one-tenth that of the components VA and VB, so that the phase difference between these com- ponents