Comments on OET Lab Policy Document

There have been 19 comment(s) made on this document:

Brian Scarpelli commented on 2012-06-29 16:38:40.343:
The Telecommunications Industry Association (TIA) hereby submits input to the Federal Communications Commissions (FCC) Office of Engineering and Technology (OET) on draft Laboratory Division Knowledge Database (KDB) publication 865664 (Is there any additional guidance for Specific Absorption Rate [SAR] measurements that addresses devices operating in the 100 MHz to 6 GHz range?) (KDB 865664). Specifically, TIA submits the following input for OETs consideration: Repeatability Issue: o TIA hereby expresses substantial concern about the operation and impact of this KDB. We recognize FCCs interest in insuring quality testing by Labs but believe that concern about accurate quality will not be addressed by this KDB sufficiently enough to warrant the extreme lab disruption it will create. TRPCs primary concerns are twofold: (1) the increased test time that will be created for many products, and (2) the new issues that will need to be addressed regarding the treatment of the repeated test values. o That test time will be substantially increased is evident from the requirement of repeatability. There are three tiers for testing for the highest measured SAR value: tier one has no requirement to repeat; tier two requires two extra tests; tier three requires at least three extra tests. For tier two products, there is a reasonable likelihood that the testing could trigger a SAR value in the tier three range, thereby increasing that products testing from two tests to at least three more tests on the new highest value. The impact of the testing, then, will be to cause the SAR values to spiral higher within the uncertainty range and substantially increase testing. o The above example demonstrates that new guidance will be needed for how to treat repeated tests. For example, all tests subsequent to the initial test could be treated only as verifiers to substantiate the SAR value, and, therefore, would not be a value used on the grant even if one or more were higher than the verified value but within the uncertainty range. Another approach could be to do a simple of average the values, which could result in a value on the grant that does not tie to any measured value. The complexity of the issue will require significantly more direction than now is contained in the KDB. o TRPC does not believe that the added complexity will render more reliable outcomes. To the extent that the FCC does not find a Labs data to be reliable, more unreliable data will not address the issue. Rather, additional indicia of reliability are required. Generally, new requirements for System Validation and resolution tolerance of post processing algorithms for scan resolution used in SAR measurements are included. Resolution tolerance may be requested to support marginal test results. More time is needed to understand what the FCC interprets as marginal or unacceptable resolutions and what this means. On page 4, the new information and formatting required for SAR Reports will require a major overhaul of test reports including the requirement to place SAR compliance results on a summary page at the front of report where space is limited. It is not clear why a new report format is needed or what the impact will actually be, but it will require costly changes in established report software and seems likely to increase the amount of time it takes meet the additional reporting requirements. On page 10, the new proposed SAR measurements required for repeatability will significantly increase test times and change past practices regarding measurement uncertainty. It appears that repeatability is replacing the well-established approach based on measurement uncertainty information for certain SAR levels, and is thus inconsistent with the FCC guidance to use IEEE 1528-2003 standard. On page 10, bullet 3(c) SAR scan procedures are provided. TIA believes that a KDB inquiry should not be required when using the state of the art SAR methods described in these KDBs and automatically implemented in the the current state-of-the-art SAR measurement systems. Please consider removing this requirement. On page 11, there are new requirements for System Validation and Resolution Tolerance of post-processing algorithms for scan resolution used in SAR measurements. Resolution Tolerance may be requested to support marginal test results. We recommend that FCC delay implementing this KDB until the industry has more time to understand what the FCC interprets as marginal or unacceptable resolutions and what this means. We also recommend that FCC follows existing international standards. Any new requirements should be discussed in the standard committee to get acceptance of other international experts. On page 12, Three criteria are specified for the measurement of dipoles during system validations and verifications: o 1g SAR must be within 10% of SAR target o Extrapolated peak SAR at phantom surface must be within 15% of calibration value, - OR - o within 15% of value stated in IEEE Standard 1528-2003. The need for verifying all three criteria for every dipole measurement seems excessive. Rather, checking that the 1g SAR value is within 10% of the calibrated target (as per the system performance check specified in the IEEE Standard 1528-2003) should be sufficient to verify SAR measurement system continues to operate within expectations. On pages 12-14, where system validations and verifications are provided, the entire section is somewhat confusing, and it appears that it creates a significant additional burden on test labs (e.g. the extensive CW and modulated validation measurements), but seems completely beyond the requirements of IEEE Standards 1528-2003 and -2011. Further clarification of the rationale and intent would be useful. On page 14, we note that the requirement to perform system verification per each probe calibration point is not a requirement of IEEE Standards 1528-2003 or -2011. Rather, these standards allow system performance checks at a frequency that is within ±10% of the compliance test mid-band frequency. As the purpose of system performance checks is to verify that the measurement systems characteristics have not changed in an unacceptable manner, it would seem that the IEEE methods adequately accomplish this end. We respectfully request that the FCC explain the need for this new procedure. Given the potential impact of the proposed KDBs on test time, lab capacity, and even product design, we request that OET determine and announce a reasonable transition period for implementation of the KDBs once finalized. TIA members recommend that a transition period of at least ninety days in order to mitigate the impact that such extensive changes to testing protocols will have. TIA has previously requested an extension of the due date for comments on draft KDBs as critical to industrys ability to provide thoughtful comments. In order to facilitate review of industrys concerns, TIA may submit comments to selected KDBs, subject to supplementation, after June 30, 2012. We therefore respectfully submit this comment to draft KDB 865664, and urge the Commission to act consistent with the above. View attachment associated with this comment

Ying Wang commented on 2012-06-29 14:09:31.14:
We don't see any real value for the new requirements being proposed. Suggest FCC to adopt Test Reduction and Power Scaling Procedure in Annex K and L of IEC 62209-2: 2010. This would greatly alleviate the burden on manufacturers.

Mats Hansson commented on 2012-06-29 08:49:04.42:
SAR measurement repeatability and uncertainty The repeatability requirement is adding additional test time but not any added value in test results. Our normal testing procedure is to test a large number of samples from the prototype production units. Units used for testing must be the same hardware status as the mass production units and be mechanically and electrically stabile. The 10 highest SAR samples are selected and tested to ensure correct mechanical stability and RF power tuning. The 5 highest SAR samples are then selected for SAR compliance testing. Testing is normally performed on multiple samples in order to have an effective testing. Embedded battery products require between 30-60 minutes of charging time to ensure a fully charged battery. Its our experience that the repeatability in testing is quite good inside the same lab. Using the same test system and operator provides a repeatability of <5% from the original value. Between different systems and operators we have a repeatability of <10%. This is also an internal requirement for us to ensure an accurate testing regardless of what lab or operator is being used. We also use the same criteria for our inter lab comparisons. We do not see any added value in the repeatability requirement and proposed that this is removed. We would rather see a lab qualification where the lab can prove the accuracy in a separate testing if FCC has a concern on the labs qualifications.

Steve Liu commented on 2012-06-26 10:51:46.916:
Thank you for your work in updating these KDBs. PCTEST submits comments regarding this draft publication in the attached file. Thank you. View attachment associated with this comment

John Lewczak commented on 2012-06-25 10:46:27.063:

[KDB 865664 D01, SAR Reporting Requirements; Page 14]: System validations & verifications - The requirement to perform system verification per each probe calibration point is not a requirement of IEEE Std 1528-2003 or -2011.

The requirements of the IEEE standards allow system performance checks at a frequency that is within ±10% of the compliance test mid-band frequency.

The purpose of system performance checks is to verify that the measurement system has not drifted or changed in an unacceptable manner. This is accomplished by following the IEEE requirements. Please clarify the purpose and necessity to perform system verification for each calibration point.

John Lewczak commented on 2012-06-25 10:45:19.736:

[KDB 865664 D01, SAR Reporting Requirements; Pages 12-14]: System validations & verifications - This entire section is entirely confusing, creates a huge burden on testing labs, & is completely beyond the requirements of IEEE Std 1528-2003 and -2011. Please clarify the intention and goal of the extremely large number of CW and modulated validation measurements.

Furthermore, the entire section (especially for modulated validation) is unclear & confusing. A seminar or similar training is most definitely needed and appropriate.

John Lewczak commented on 2012-06-25 10:43:55.066:

[KDB 865664 D01, SAR Reporting Requirements; Page 12]: System validations & verifications - Sub-section 1), instructs that The measured 1g and 10g SAR levels should correspond to approximately 0.1, 0.8, 1.6, and 2.0 W/kg.

It is not possible for both the 1g and 10g values to simultaneously measure the same value. We believe the intention is for the 1g SAR level to correspond to the levels stated, NOT the 10g SAR levels. Please clarify and update.

John Lewczak commented on 2012-06-25 10:41:56.313:

[KDB 865664 D01, SAR Reporting Requirements; Page 12, Reference Dipoles]: System validations & verifications - Three criteria are specified for the measurement of dipoles during system validations and verifications:

1g SAR must be within 10% of SAR target

Extrapolated peak SAR at phantom surface must be within 15% of calibration value - OR - within 15% of value stated in IEEE Std 1528-2003.

Please explain the need & purpose for verifying all three criteria for every dipole measurement. Checking that the 1g SAR value is within 10% of the calibrated target (as required in IEEE Std1528-2003, system performance check) should be sufficient to verify on an ongoing basis that the system is operating within expectations.

John Lewczak commented on 2012-06-25 10:38:26.35:

[KDB 865664 D01, SAR Reporting Requirements; Page 10, Bullet 3c]: SAR Scan Procedures - Page 10, Bullet 3c] A KDB inquiry should not be required when using the state of the art SAR methods described in these KDBs and automatically implemented in the SAR measurement equipment. Please clarify or remove this requirement.

John Lewczak commented on 2012-06-25 10:35:28.136:
[KDB 865664 D01, SAR Reporting Requirements; Page 7]: SAR Scan Procedures - Instructions ... area scan resolution must be less than or equal to the corresponding dimensions of the device surface being measured. will result in too numerous points when the smaller dimension is used for both the x and y scan resolution. Thin but long surfaces should be allowed to have the surfaces long dimension scan resolution set to a standard separation and not be required to use the thin surface dimensions. Please clarify this section.

John Lewczak commented on 2012-06-25 10:34:02.376:

[KDB 865664 D01, SAR Reporting Requirements; Page 7]: SAR Scan Procedures - Instructions when measured peak is closer than ½ of the zoom scan volume dimension from the edge of the area scan region, the measurement must be repeated by shifting and expanding the area scan region, to ensure all peaks are away from the are scan boundary do not allow for the current state of the art in SAR measurements. Current SAR measurement technologies automatically expand the size of the zoom scan to ensure that the peak SAR location is sufficiently captured to calculate the appropriate 1g/10g SAR value. Also, this requirement is not achievable when testing the edges of most current cellular phones for the mobile hotspot SAR measurements. Many devices are thinner than ½ of the zoom scan volume. Please modify this statement to allow the newer scan methods.

John Lewczak commented on 2012-06-25 10:31:55.71:

[KDB 865664 D01, SAR Reporting Requirements; Page 10]: Repeatability and Uncertainty - Separation distance from probe calibration point to phantom surface

Pages 6-7, Table:

≤ 3 GHz

> 3GHz

Maximum distance from closest measurement point (geometric center of probe sensors) to phantom surface

5 ± 1mm

1/2*d*ln(2) ± 0.5mm

However, page 3 specifies: The closest measurement point from the phantom surface must be ≤ 5-mm for measurements ≤ 3 GHz; and must be ≤1/2*d*ln(2) for measurements > 3GHz.

Please harmonize the requirements in table and the text.

John Lewczak commented on 2012-06-25 10:28:49.88:
[KDB 865664 D01, SAR Reporting Requirements; Page 6, Bullet 2a] The KDB does not specify what to do if the tissue parameter(s) are exactly the target value. Please clarify.

John Lewczak commented on 2012-06-25 10:24:21.143:

[KDB 865664 D01, SAR Reporting Requirements; Page 14]: SAR Compensation - Current verbiage suggests that SAR values will be compensated both up and down according to the deviation of measured tissue parameters relative to the target tissue parameters. Should SAR really be scaled DOWN to account for tissue parameters away from target values?

John Lewczak commented on 2012-06-25 10:18:34.996:

[KDB 865664 D01, SAR Reporting Requirements; Page 10]: Repeatability and Uncertainty - We have significant concerns regarding the proposed repeatability requirement set out in the subject KDB 865664. Our assessment of the benefits and burdens of the proposal leads to the conclusion that the significantly detrimental impact on the time for testing is not balanced by a meaningful benefit. In fact, the repeat tests will not provide sufficiently increased reliability to warrant adoption of the methodology. Moreover, the increase in testing complexity will present a substantial challenge to day to day operation and could inject ambiguity into reported SAR values.

1.    Increased Test Time

The proposed methodology will increase test time for several reasons. First, of course, the physical act of taking the product off the robot, replacing the battery with a newly charged battery and then repositioning it for a repeat test will take time. For products with embedded batteries, the phones battery will actually have to be recharged prior to each repeated test. Given that modern products are moving to embedded batteries, the time will be increased for a substantial part of every manufacturers portfolio.

Another impact on test time will result from the timing of the tests. Since a lab will not know what tests will need to be re-done until the entire frequency band (probe calibration point) has been tested, it could be days after the initial test was performed before a repeated test can be made. At that point, the recreation of the original test conditions i.e., probe tissue, operator may not be possible. Efforts at such recreation will be time consuming.

1.    Increased Testing Complexity

Motorola has serious concerns about the increase in testing complexity that the requirement will engender.  The complexity will lead to both ambiguous SAR findings due to the normal variations in testing and to potentially extended testing procedures.

For example, a product may be tested once and measure 1.18 W/kg. Under the KDB, a second test would be required. However, if that second test measures 1.21 W/kg, then at least three more test are needed. Under this situation, the tester now has five SAR values in the report. This gives rise to the question whether the value on the grant should reflect the mean, the highest value, or the last value measured. Since more than three measurements may be required for the higher values (i.e., the language reads that the test should be repeated at least three times), the problem could be made even more complex and ambiguous.

Motorola recommends that the repeated tests be treated as verification that the value to be submitted i.e., the initial test is reliable. The repeated values, therefore, would not be treated as values appropriate for the grant.

1.    Reliability Will Not Be Substantially Increased

Repeatability is not justified without meaningful increases in reliability. In the present case however, there already are substantial indicia of reliability for the measurements. Test labs are certified by third party certification bodies. The FCC identifies standard test equipment, software, tissues, and practices common to all certified labs. Deviations from any of the above are required to be addressed via the PBA process, or by requiring the filing to be reviewed by the FCC. Labs currently provide detailed Uncertainty Analyses in their reports, which capture the relevant variables affecting the testing.

In addition, we note that repeated tests are not deemed to be needed for reliability for other FCC requirements, such EMC or power levels.

Given the existing levels of reliability above, there is no basis for requiring repeated measurements. We note further that if the FCC does not have confidence in a specific lab, repeated measurements will be an ineffective means of establishing confidence.

Kai Niskala commented on 2012-06-25 06:25:34.66:
KDB 865664 D01 SAR measurements 100MHz to 6GHz: SAR Measurement Requirements: Narrowed liquid parameter windows set by these requirements can be very hard to achieve in some cases. Considering overall uncertainty budget of SAR measurement, the gain in uncertainty is minimal compared to practical efforts. SAR Scan Procedures: Some stringent spatial resolution requirements will make SAR scans very time consuming. At the same time mobile phones have more and more bands and communication systems available. To balance the effort needed, it would be desired that FCC would seriously consider accepting some of the widely used and well proven fast-SAR procedures. We believe that correctly applied fast SAR methods for searching maximum SAR configurations would not endanger the stringency of the overall test procedure in finding the maximum SAR of the device. Simultaneous transmission SAR measurement: It should be clarified that systems having different duty cycle cant be measured correctly when transmitters are transmitting simultaneously. These cases should be measured separately and be combined in post-processing. SAR Measurement Repeatability: Repeating highest measured SAR configurations test does not offer any benefit considering overall uncertainty budget, but only increases measurement time. Practice has shown that measurement repeatability is very good, one would be getting more or less the same result. Especially repeating measurement already on the very low levels starting from 0.4W/kg is not necessary considering very low possibility for exceeding regulatory limit. We propose that trigger level for repeat measurements should be raised at least to 0.8 W/kg or even higher. From practical point of view it can very time consuming requirement as one might need to conduct a new system check to make one extra repeat measurement. Dipole Calibration: Requirement of less than 20% deviation in return-loss dB value is not feasible technically. It should be sufficient to fulfill -20dB return-loss specification. When dipole meets -20dB criteria, even very high numerical return-loss dB value deviation has very little impact on inserted power, 1% at maximum by definition. In practice inserted power variation is negligible when meeting -20dB requirement. SAR System validation: It could be more clear if FCC would give list of communication systems falling under requirements for high PAR (>5dB) signals. Reporting and Documentation Considerations Information for SAR reports: 5) Is this same KDB865664 more correct reference for repeatability results, not KDB447498?

Kai Niskala commented on 2012-06-25 06:21:51.92:
KDB 865664 D01 SAR measurements 100MHz to 6GHz: Information for all test reports: 8) d) i): states that KDB tracking numbers should not be included. For some devices, extensive discussions are conducted prior to the start of testing; surely, in these case, referencing the KDB tracking number would help traceability of the procedures carried out. Information for SAR reports: 7) b): it is very welcome that this section calls for only the maximum plot in each band, mode and exposure configuration, rather than every plot as is currently required. However, there is no mention of the requirement of z-axis scans; is this an intentional omission?

Mark Briggs commented on 2012-06-14 14:47:15.713:
Please refer to attached document for comments. View attachment associated with this comment

David Case commented on 2012-06-12 11:05:34.516:
In Cisco's review of the draft KDB, we raise a concern on the requirements of repeating tests when the SAR exceeds various specific thresholds. Whether there is a requirement to repeat it once or several times, repeating of the tests to demonstrate compliance will increase both cost to the manufacturer as well as increase time. Given some devices no longer have replaceable batteries , we will be impacted additional time which will result in more cost when at a 3rd party do to time needed to re charge the battery. At this time the FCC does not make us repeat tests when our spurious emissions are within 1 or 2dB of the limit, we only need to demonstrate compliance to the limit based on the data provided. Given that the testing could very well be done in an accredited lab, the test data taken by the lab should be sufficient to provide proof of compliance. Given the overall burden of compliance and penalties of non compliance rests with the manufacturers and given test labs risk their accreditation and reputation if the data they take is poor, the decision to confirm the data is correct is one best left up to the test engineer and the manufacturer upon review of the data instead of mandated by the FCC. The use of accredited labs is encouraged by the FCC and as such the acceptance of test data from a test lab especially an accredited one should rest with the labs credentials on file and not with the number of repeatable tests on the product done by a lab versed in testing to the applicable standard or guidance.

Note: It is important to understand that the staff guidance provided in the KDB is intended to assist the public in following Commission requirements and does not constitute rules. Accordingly, the guidance is not binding on the Commission and will not prevent the Commission from making a different decision in any matter that comes to its attention for resolution.

Office of Engineering and Technology