US Council of EMC Laboratories [USCEL] Technical Issues having Significant Cost Implications for EMC Laboratory Owners/Operators

Transcription

1 US Council of EMC Laboratories [USCEL] Technical Issues having Significant Cost Implications for EMC Laboratory Owners/Operators Presented to the ACIL CAS EMC Committee On 17 August 2009 [Note: includes corrections dated 27 August 2009 By: Harry H. Hodes Principal EMC Engineer President & CEO Acme Testing Co. For and on behalf of the US Council of EMC Laboratories [USCEL] 1

2 Three Topics ANSI C63.5(2006) is invoked in the soon to be issued ANSI C63.4(2009). Note: ANSI C63.5(2006) is soon to be accepted by CISPR as containing acceptable Cal Procedures Site validation above 1 GHz: - FCC (ANSI C63.4(2009) - EU (CISPR :2007 +A1: A2:2008, invoked in CISPR 22: ) - Japan (VCCI 2009 Rules which take effect in May 2010). Effects of Tables on Radiated Emissions Uncertainty 2

3 Topic # 1: ANSI C63.4(2009) invokes ANSI C63.5(2006) ANSI C63.4(2009) [Due out later this yr] invokes ANSI C63.5(2006) as the only allowable standard for antenna calibration. Massive Technical Effects on EMC Lab Operations, including: - NSA and VSA OATS/Chamber Site Qual Procedures, - Procedures used to obtain Antenna Cal Services for Product Testing, - Procedures used to obtain Antenna Cal Services for NSA/VSA, - Choices of Antennas to use (BiConLog Antennas are now very problematic, especially w.r.t. Cal Uncertainties). New EMC Engineer/EMC Technician training will be needed to address how ANSI C63.5(2006) will impact Lab Operations. Significant Cost Impacts: - assessment of the effects of the change of the Antenna Cal procedures on Pass/Fail decisions on Products. - assessment of the effects of the change of the Antenna Cal procedures on Pass/Fail decisions on OATS/Anechoic Chambers. 3

4 Topic # 1: ANSI C63.4(2009) invokes ANSI C63.5(2006) ANSI C63.5(2006) - Is a complicated Standard. We (ACIL Member Labs) are most concerned with the Standard Site Method (SSM) (i.e., the 3- Antenna Method) portions of the Standard. - Introduces a whole new set of terms and concepts, The most important of which are: * Free-Space Antenna Factor (FSAF) * Near-Free-Space Antenna Factor (NFSAF) * Geometry-Specific Antenna Factor (GSAF) - Figure G.2 is a Key Figure in understanding ANSI C63.5(2006). 4

5 Topic # 1: ANSI C63.4(2009) invokes ANSI C63.5(2006) ANSI C63.5(2006) The Antenna Factors used when testing Products are different from those used in performing NSA / VSA Site Qualifications! All Antenna Factors to be used for Product Testing are obtained from Antenna Cals that are performed at D = 10m, using a 2 m antenna height (Near-Free-Space conditions for Bicons & LogPers), in Horizontal Polarization only. For Bicons (but NOT for BiConLogs or LogPers), the correction factors that are to be applied to NFSAFs to obtain AFs for NSA/VSA have been theoretically determined (and have been verified experimentally). These Bicon Correction factors (for 30 MHz to 200 MHz) are given in the Tables in Annex G of C63.5(2006). Important Note: there are different correction factor tables for 50 Ohm Baluns and for 200 Ohm Baluns. Be sure to use the correct table [see Figure G.3 in C63.5(2006) to get an understanding of this]. For all other antennas (i.e., BiConLogs or LogPers), the Geometry-Specific Correction factors are not accurately determinable by theoretical means, and thus must be measured this means that the antenna pairs are required to be Calibrated at both H & V Pols at all of the heights and distances used in making the NSA/VSA Measurements. Also note that one antenna in each pair has to be designated for Transmit and the other for Receive. 5

6 Topic # 1: ANSI C63.4(2009) invokes ANSI C63.5(2006) Top-level Implications of using ANSI C63.5(2006) New NSA/VSA Spreadsheets will need to be implemented at each EMC Lab, because the Normalized Site Attenuation Equations have changed! New Purchase Order Templates for Antenna Cals will be needed. Labs may need to Cal the Antennas the old and new ways in order to establish continuity for NSA/VAS Site Quals, as well as for Pass/Fail product determinations. (The differences can be >+/-2 db between old and new AFs). 6

7 Topic # 2: Site validation above 1 GHz: Radiated Emissions Test Site (OATS/RF Anechoic Chamber) Validation above 1 GHz by means of the Site VSWR [SVSWR] Method is being mandated in many parts of the world: - FCC (ANSI C63.4(2009) - EU (CISPR :2007 +A1: A2:2008, invoked in CISPR 22: ) - Japan (VCCI V-3/ Rules which take effect in May 2010). The SVSWR Method requires a set of exotic Antennas (as specified in CISPR : A1:2007 +A2:2008), and, extremely rigid and dimensionally stable mounting mechanisms. The currently available choices are made by the following firms: Seibersdorf /Austrian Research Centre [Models: POD16 for 1 6 GHz, POD 618 for 6-18 GHz] Schwarzbeck Mess - Elektronik [Models: SBA 9119 for 1 6 GHz, SBA 9112 for 3-18 GHz] 7

8 Topic # 2: Site validation above 1 GHz: Required Transmit Antenna Pattern (E-Plane) 8

9 Topic # 2: Site validation above 1 GHz: Required Transmit Antenna Pattern (H-Plane, 1-6 GHz) 9

10 Topic # 2: Site validation above 1 GHz: Required Transmit Antenna Pattern (H-Plane, 6-18 GHz) 10

11 Topic # 2: Site validation above 1 GHz: Site validation above 1 GHz (continued) - The SVSWR Method can be implemented using a Sig Gen and a Spectrum Analyzer using 50 MHz frequency steps OR by using a Vector Network Analyzer [VNA] with (nearly continuous) sweeps. Site validation by the SVSWR method evaluates a given test volume for the specific combination of site, receive antenna, test distance (described in Subclause of CISPR ), and absorbing material placed on the ground plane, if needed to meet the criterion of Clause of CISPR : A1: A2:2008. Influences of the receive antenna mast located as used for the site validation tests, and permanently-fixed objects in the test volume (such as a permanently-installed turntable), are evaluated by and included in this site validation procedure. Removable objects, such as a removable test table, are not required to be in place during the site validation tests if their influence is to be evaluated separately using the additional procedures of Clause 5.8 of CISPR : A1: A2:2008. The purpose of the SVSWR procedure is to check for the influence of reflections that may be incident upon an EUT of arbitrary size and shape placed within the test volume. 11

12 Topic # 2: Site validation above 1 GHz: Site validation above 1 GHz (continued) The SVSWR Method procedures are exhaustively detailed in Clause 8 of CISPR :2007 +A1:2007 +A2:2008. [Note: the key figures are Figures 17 and 18]. Measurements are required to be made at the Center, Front, Left, Right, and possibly the Rear Positions on the Turntable, at two heights (i.e., at h 1 = 1.00 m above the bottom of the test volume, or at the middle of the test volume, whichever is lower, and at h 2 = the top of the test volume, if h 2 is >= 0.5 m above h 1 ). At each position (i.e., Center, Front, Left, Right, Rear), data is collected at H Pol and V Pol at a total of six measurement locations, as follows: Reference 2 cm rearward of reference 10 cm rearward of reference 18 cm rearward of reference 30 cm rearward of reference 40 cm rearward of reference Measurements are made at 50 MHz intervals (or at finer intervals), from 1 GHz to 6 GHz for the EU and VCCI, and to 18 GHz for the USA. 12

13 Topic # 2: Site validation above 1 GHz: S VSWR Measurement Positions (Horizontal Plane) 13

14 Topic # 2: Site validation above 1 GHz: S VSWR Measurement Positions (Vertical Plane) 14

15 Topic # 2: Site validation above 1 GHz: S VSWR Criterion The SVSWR is directly related to influences of undesired reflections. Acceptance criterion for site validation from 1 GHz to 18 GHz is: S VSWR 2:1, or S VSWR, db 6.0 db 15

16 Topic # 2: Site validation above 1 GHz: SVSWR Test Method Problems The Tx Test Antennas and the associated mounting jigs are EXTREMELY expensive (~$12 - ~$16k) and take a long time (i.e., days ARO) to get. The Tx Test Antennas setup tolerances for position (X, Y, and Z) and Pitch, Roll, and Yaw are EXTREMELY critical (fractions of a cm make a BIG difference at 18 GHz). To accomplish tests at 3m and 10 m distances, it typically takes 2-3 days of really tedious effort if using a Sig Gen and Spectrum Analyzer, but only 1 to 1.5 days if using a Vector Network Analyzer. The quantity of data collected during an S VSWR Test is massive! Automation (of the test data collection and test data reduction processes) is a practical necessity. If you use older instrumentation, you may need to use a 3 rd party software package. (More $$$). 16

17 Topic # 2: Site validation above 1 GHz: S VSWR Test Method Problems (continued) The test method is a technical compromise it under-samples a sine wave-like chamber response at 6 points. Here is a Quote from Clause of CISPR :2007 +A1:2007 +A2:2008: The measurement points chosen for and contained in the preceding procedures are intended to provide an overall measure of the SVSWR of the test site across the frequency range of 1 GHz to 18 GHz. Note however that the peak SVSWR may not always be captured using the procedures of or at any specific frequency f. Therefore, statements about SVSWR compliance based on measurements at any single frequency should be avoided. However, the peak found by the above procedures within adjacent octaves (0,5f to 2f) is typically representative of the worst case SVSWR for all frequencies inclusive in the band. In cases where more accuracy of the SVSWR result is desired at a single frequency, the above method can be improved by measuring more than six locations along the lines shown Figure 17 and Figure 18. The additional data collection points should be spaced unequally, and chosen based on a distance translation of the source antenna (or field probe in the reciprocal SVSWR method) using quarterwavelength steps at the frequency of interest. 17

18 Topic # 3: Effects of Tables on Radiated Emissions Uncertainty Basic Test Method (per CISPR : A1:2007 +A2:2008 Clause 5.9) 18

19 Topic # 3: Effects of Tables on Radiated Emissions Uncertainty Basic Test Method (per CISPR : A1:2007 +A2:2008 Clause 5.9) (continued) 19

20 Topic # 3: Effects of Tables on Radiated Emissions Uncertainty Basic Test Method (per CISPR : A1:2007 +A2:2008 Clause 5.9) (continued) 20

21 Topic # 3: Effects of Tables on Radiated Emissions Uncertainty Basic Test Method (per CISPR : A1:2007 +A2:2008 Clause 5.9) (Continued) 21

22 Topic # 3: Effects of Tables on Radiated Emissions Uncertainty Basic Test Method (per CISPR : A1:2007 +A2:2008 Clause 5.9) (Continued) 22

23 Topic # 3: Effects of Tables on Radiated Emissions Uncertainty Problems Problem: The required method is detailed in Clause 5.9 of CISPR : A1:2007 +A2:2008, but the data is not included in the Uncertainty Budgets for Ucispr in the current version of CISPR However, an Amendment to CISPR is in work to include the effects of the Table on the Radiated Emissions Uncertainty. Additionally, this Amendment may extend this method to 6 GHz (or 18 GHz), which will mean that one or more additional Antennas will be required! (Yet more $$$$). Problem: The required Antenna for this test is a Biconical Antenna (covering the frequency range of at least 200 MHz to 1 GHz) that has an overall length of less than 0,40 m (measured from Tip to Tip across the Birdcage Elements). This is a loose specification! Question: Will comparable results be obtained using Bicons of different Birdcage sizes? The ETS-Lindgren 3180 (which is claimed to be suitable for this requirement and costs ~$2500) is 33.6 cm from tip to tip). The Schwarzbeck Model SBA 9113B (cost unknown) is claimed to suitable for this requirement, is 14 cm from tip to tip. The Schwarzbeck Model VUBA 9117 (cost ~$2110) is claimed to suitable for this requirement, but is 44 cm from tip to tip [i.e., is 4 cm too long to be Compliant]. 23

24 Topic # 3: Effects of Tables on Radiated Emissions Uncertainty Problems Problem: Fixtures! The question is: How do you install the Test Antenna above the Table in such a way that the mechanical support equipment used will not cause reflections (or cause enough attenuation) that could dominate the resulting measurement. Problem: This requirement imposes an overhead cost on every EMC Lab. Unfortunately, the required Antenna is useless for ordinary testing. (You will only use the Antenna one time for each kind of Table that you need to evaluate). 24

25 The End (Thank you for listening!) 25