A COMPARISON OF METHODS FOR EVALUATING THE TEST ZONE PERFORMANCE OF ANECHOIC CHAMBERS DESIGNED FOR TESTING WIRELESS DEVICES

Transcription

1 A COMPARISON OF METHODS FOR EVALUATING THE TEST ZONE PERFORMANCE OF ANECHOIC CHAMBERS DESIGNED FOR TESTING WIRELESS DEVICES Jame D. Huff John C. Mantovani Carl W. Sirle The Howland Company, Inc Atwater Court, Suite 07 Buford, Georgia 3058 USA Abtract The two mot common tet method ued to evaluate wirele tet chamber are the Ripple Tet Method tandardized by CTIA - The Wirele Aociation and the Field Senor Method tandardized by the 3rd Generation Partnerhip Project (3GPP). Both method ample the magnitude of the illuminating field at fixed patial point in the Tet Zone to determine the magnitude of the ripple in the tet zone. Thi ripple data i then tatitically proceed to determine the expected meaurement uncertainty attributable to chamber reflection at a given frequency. The trength and weanee of each of thee evaluation method are dicued in detail. Tet reult uing both method to evaluate a ingle chamber are preented. A third wirele tet chamber evaluation method i alo decribed. In thi method a erie of Total Radiated Power (TRP) meaurement are made on an antenna with the antenna poitioned at variou patial location in the tet zone. If meaured with a perfect plane wave, each TRP meaurement hould produce the ame reult regardle of the patial location of the antenna. Variation in meaured TRP relate directly to meaurement uncertainty caued by deviation of the incident wave from a perfect plane wave. Keyword: CTIA, 3GPP, Ripple Tet, Field Senor, TRP, Wirele Tet Chamber, Meaurement Uncertainty. Introduction An uncertainty budget for wirele meaurement will contain many contributor. Typically, one of the larger contributor will be the meaurement uncertainty due to extraneou ignal that add in and out of phae with the direct ignal. In order to have a meaningful undertanding of the meaurement uncertainty, it i neceary to accurately meaure the level of extraneou ignal in the tet zone and convert that level to an uncertainty contribution. Two method for tet zone characterization have been tandardized for wirele tet ytem. The oldet and mot accepted i the method defined by the CTIA Tet Plan for Mobile Station Over the Air Meaurement. More recently the 3rd Generation Partnerhip Project (3GPP) tandardized a econd method uing a broad band field enor. Each method ha it advantage and diadvantage, but both method require converting the meaured ignal level into an uncertainty contribution. It i alo poible to meaure the uncertainty contribution directly by maing a erie of TRP meaurement with a probe antenna poitioned at variou point in the tet zone. In concept the three method hould provide the ame uncertainty contribution. Thi paper preent experimental data collected in a typical wirele tet chamber uing all three method and compare the calculated meaurement uncertainty. 2. The CTIA Ripple Tet The method defined by the CTIA ue dipole and loop antenna that have highly ymmetrical pattern. The antenna are required to have an azimuth plane pattern ymmetry of <+/-0. db. The probe antenna are poitioned at a erie of defined location within the tet zone and the received ignal i recorded a the phi or theta axi i rotated. The reulting pattern will how a ripple a the extraneou ignal phae in and out with the direct ignal. Becaue the antenna are highly ymmetrical, any ripple in the pattern i aumed to have been caued by an extraneou ignal.

2 Phi pattern are made with the theta axi at 90 degree. The probe antenna i poitioned at ix different poition within the tet zone. Thee poition are hown in figure. The probe i rotated about the phi axi and the field in the tet zone i ampled at 2 degree increment. A typical phi axi meaurement etup i hown in figure 2 below. Theta Axi Structure if Applicable -Y +X Meaurement Antenna at Phi = 0 +Z -Z (Phi Axi) Meaurement -X (0,0,+) Antenna at Phi = 80 (-,0,0) 50 mm ± 3 50 mm ± 3 (0,0,0) +Y (Theta Axi) (Theta Axi) 50 mm ± 3 (0,+,0) (0,-,0) (+,0,0) 50 mm ± 3 Theta Axi Tet Two 50 mm offet along each orthogonal (0,0,-) axi for a total of 7 po. Rotate Theta axi mm ± 3 50 mm ± 3 at Phi = 0 and 80, or -65 to 65 at Phi = 0 for each poition. Support Label Point (X,Y,Z). Structure at Theta = 80 Meaurement Antenna at Theta = 90 Phi Axi Tet Three poition along Phi axi (center and one each ± 50 mm axial offet) with one 50 mm radial offet at each poition for a total of 6 poition. Rotate Phi axi 360 for each poition with Theta = 90. Label Point (R,Z) 50 mm ± 3 (0,+) (+,+) 50 mm ± 3 (0,0) (+,0) 50 mm ± 3 (0,-) (+,-) (Phi Axi) (Phi Axi) (Phi Axi) Figure 3 Theta Axi Tet Poition Support Structure at Theta = 80 Figure Phi Axi Tet Poition Figure 4 Typical Theta Axi Scan Setup The CTIA method i eentially a variation of the claical pattern comparion method. Becaue the probe antenna pattern are highly ymmetrical, the variation in the pattern are eay to interpret. Figure 2 Typical Phi Axi Scan Configuration Theta pattern are made with the phi axi at 0 degree. The probe antenna i poitioned at even different poition within the tet zone. Thee poition are hown in figure 3. The probe i rotated about the theta axi and the field in the tet zone i ampled at 2 degree increment. A typical theta axi tet etup i hown in figure 4 below. 3. The 3GPP Ripple Tet The tet method defined by 3GPP ue a broadband field enor. The field enor i poitioned at even different poition in the tet zone a hown in figure 5 and the received ignal level i recorded at the frequencie of interet. The field enor meaure the received ignal in three orthogonal component which are then ummed internally to provide a ingle compoite output. The radiation pattern of the field enor i aumed to be omni-directional o that variation in the ignal received at the different poition can be interpreted a variation in the field of the tet zone. A typical field probe meaurement etup i hown in figure 6.

3 Figure 5 3GPP Field Poition contribute an error to each data point, but the magnitude of the error will depend upon the relative phae of the direct ignal to the extraneou ignal. The TRP i calculated by converting the meaured value into value of EiRP, weighting them by the ine of the theta angle, and averaging them. There i a tendency for the individual error to average out, reulting in a TRP meaurement that ha a lower uncertainty than the individual data point. However, determining a valid meaurement uncertainty for the overall TRP meaurement i omewhat complicated. The CTIA tart by defining a tatitical uncertainty value referred to a the urface tandard deviation (SSD). The SSD ue the meaured ripple tet reult to determine a Type A uncertainty value for a theoretical iotropic radiator placed anywhere within the quiet zone. The SSD i defined a Figure 6 Field Setup for 3GPP Meaurement 4. Deriving the Uncertainty Contribution from Ripple Tet Reult Having meaured the amplitude variation (ripple) in the tet zone, the next tep i to calculate the meaurement uncertainty that thi ripple generate in a TRP or TIS meaurement. For a ingle point meaurement, it i fairly traightforward to calculate the meaurement bound when an extraneou ignal interfere with the deired ignal. L MAX = 20log(e D +e X ) db L MIN = 20log(e D -e X ) db Where L MAX,MIN are the max and min ignal level e D i the level of the direct ignal e X i the level of the extraneou ignal A typical TRP meaurement would conit of data point taen every 5 degree in phi and theta for a total of 264 data point. The extraneou ignal will N p ( p ) = = in( θ ) ( N ) 0 p p where i an individual amplitude meaurement point, converted to linear unit, and p i the average of the amplitude meaurement aociated with a ingle meaurement can. The tandard uncertainty contribution, u(x), due to the amplitude ripple i then given by the maximum of all the ( p ) value of the 26 phi and theta axi can. u x) = 0 log( + max( j ( p ))) db ( The 3GPP procedure i lightly different from the procedure ued by the CTIA. The 3GPP calculate the uncertainty directly from the even reading taen from the broadband field probe. Where = N N ( P i P) i= S i the tandard deviation N i number of meaurement poition P i db average of all P i i Pn or P mea n Pn 2 2

4 The 3GPP tet plan alo provide a mean of obtaining a more accurate uncertainty calculation by ampling the field amplitude while canning the probe in phi and theta. The average ample tandard deviation i then calculated a follow: S freq = π I J π in( Θi ) + 2IJ 2 i, j, hor 2IJ i= j= i= j= I J i, j, ver in( Θ ) i can were made with the probe antenna in the precribed poition. The meaured data wa proceed per the procedure in the CTIA tet plan and the wort cae data elected a the meaurement uncertainty contribution. The reult are hown in Table & 2 below. Where I i the number of angular interval in elevation, J i the number of angular interval in azimuth and Θ i the theta angle of meaurement. i i, j, pol There are obviouly ignificant difference between the approach of the CTIA and 3GPP. 5. Uing TRP Meaurement to Determine Meaurement Uncertainty A more traightforward approach to determine the meaurement uncertainty due to extraneou ignal i to imply mae TRP meaurement at elected location in the tet zone with near omni-directional antenna. For intance if meaurement are made at ix poition in the tet zone uing both loop and dipole, one would end up with 24 different value of TRP. The effect of the extraneou ignal in the chamber will how up a variance in the meaured value of TRP. A typical wirele OTA TRP meaurement conit of meaurement made at 5 degree tep in phi and theta. Meaurement are made with both horizontal and vertical polarization. To calculate TRP, the raw data i converted to value of EiRP, weighted by the ine of theta, and averaged. Included in the meaured TRP i the integrated effect of the extraneou ignal. There are everal approache one could tae to define the meaurement uncertainty from the reult of the TRP meaurement. The mot traightforward would be to define the meaurement uncertainty a the tandard deviation of the 24 TRP meaurement. 6. Comparion of Tet Reult Single frequency data wa taen in the ame chamber uing the three different tet procedure decribed above. The frequency elected wa 240MHz. The tandard CTIA meaurement procedure wa ued to generate ripple tet data. Both phi axi and theta axi Poition Phi Axi Scan 240MHz P-P Ripple (db) u(x) (db) (x,y,z) in HP VP HP VP 0,0, ,0, ,0, ,+6, ,+6, ,+6, Max Poition Table CTIA Phi Axi Ripple Tet Reult Theta Axi Scan 240MHz P-P Ripple (db) u(x) (db) (x,y,z) in HP VP HP VP 0,0, ,0, ,0, ,+6, ,-6, ,0, ,0, Max Table 2 CTIA Theta Axi Ripple Tet Reult

5 The field probe ued in the 3GPP meaurement procedure wa the Model HI-6005 from ETS Lindgren. It calibration wa current and it i pecified to have an iotropic deviation of le than +/-0.5 db. The meaurement were made at even different probe poition per the 3GPP tet plan. The reult are hown in Table 3 below. Data Poition (Normalized db) (x,y,z) in HP VP 0,0, ,0, ,0, ,+6, ,-6, ,0, ,0, Standard Deviation Table 3 3GPP Field Data A total of 24 TRP meaurement were made uing both a Howland VA00 loop and a Howland VA00 dipole poitioned at different location in the tet zone. Data wa taen at 5 degree increment in phi and theta. All 24 meaurement were made at 240MHz. The computed TRP value are hown in Table 4 below. Antenna Pol Poition (x,y,z) in TRP (db) Delta Dipole VP 0,0, Dipole VP 0,0, Dipole VP 0,0, Dipole VP 0,+6, Dipole VP 0,+6, Dipole VP 0,+6, Dipole HP 0,0, Dipole HP 0,0, Dipole HP 0,0, Dipole HP 0,+6, Dipole HP 0,+6, Dipole HP 0,+6, Loop VP 0,0, Loop VP 0,0, Loop VP 0,0, Loop VP 0,+6, Loop VP 0,+6, Loop VP 0,+6, Loop HP 0,0, Loop HP 0,0, Loop HP 0,0, Loop HP 0,+6, Loop HP 0,+6, Loop HP 0,+6, Standard Deviation 0.43 Table 4 TRP Meaurement Reult

6 7. Summary & Concluion In a perfect world one would expect all three tet procedure and the appropriate uncertainty calculation to give the ame reult. However, apparently thi i not a perfect world. The reult are ummarized in Table 5. The 3GPP and the TRP meaurement reulted in eentially the ame meaurement uncertainty etimate. However, the CTIA ripple tet predict a ignificantly maller meaurement uncertainty. Maximum Amplitude Variation CTIA Tet Plan 3GPP Tet Plan TRP Meaurement 2.27dB.3dB N/A u(x) 0.2dB 0.43dB 0.43dB Table 5 Predicted Uncertainty Contribution All three method have their advantage and limitation. The CTIA procedure require tuned loop and dipole maing the meaurement proce fairly time conuming if multiple band mut be meaured. On the other hand the 3GPP procedure ue a broadband field probe which mae taing data at many frequencie over a wide range of wirele band fairly eay. The dipole and loop ued in the CTIA procedure have a ymmetry requirement of +/-0. db. Thu even a perfect chamber can how a 0.2 db ripple. The field probe ued in the 3GPP procedure ha a pecified Iotropic Deviation of +/-0.5 db. Hence the 3GPP procedure could how a variation of db even in a perfect chamber. Neither the CTIA nor the 3GPP tet plan give a very detailed or convincing dicuion of how the amplitude variation meaured in the tet zone are converted to meaurement uncertainty. The TRP meaurement technique i attractive from it implicity. It doe not require any pecific characteritic of the probe antenna other than nearomni performance. Meaurement can be made at multiple frequencie over broad bandwidth. Thi procedure capture all of the effect of meauring a DUT with a le than perfect plane wave. Thi would include the effect of mutual coupling, phae taper, amplitude taper, polarization variation and extraneou ignal. 8. Reference [] Tet Plan for Mobile Station Over The Air Performance, Reviion 2.2, CTIA Certification Program, November 2006 [2] ETSI TR V.4. (200-2), Electromagnetic compatibility and Radio pectrum Matter (ERM); Uncertaintie in the meaurement Of mobile radio equipment characteritic; Part [3] 3GPP TR V7.0.0 ( ), 3rd Generation Partnerhip Project; Technical Specification Group Radio Acce Networ; Meaurement of radio performance for UMTS terminal in peech mode (Releae 7) [4] Microwave Antenna Meaurement, Second Edition; Lyon, Holli, Clayton; Scientific-Atlanta, Inc.; 970 The CTIA tet procedure require a minimum of 3,300 data point to characterize the ripple in the chamber. In comparion the 3GPP tet procedure only require 4 data point (7 poition at 2 polarization) at any given frequency.