As more and more devices are designed and deployed with wireless functionality, the need for proper testing and evaluation is more critical than ever. Unlike testing a computer or a peripheral transmitter, radio testing can be more complicated. For that reason, it is imperative that the lab doing the testing be familiar with transmitter testing.
Testing systems is much easier than it was a few years ago. In the early 1990s, we were developing the first series of products and the test methodologies at the same time. Today, the FCC and other agencies have published guidelines and test procedures for testing a majority of Title 47 Part 15 devices.
Initial Review
In my opinion, one of the most critical phases of wireless testing is determining what you are testing, what the requirements are, what standards or limits apply, and how the DUT relates to those requirements. In this phase, the manufacturer and test lab play their respective roles.
The manufacturer should supply the test lab with a complete overview of the product including a manual, the modes of operation, and the standards to which the device should be certified.
The test lab should review this information before testing. After the final report is written and submitted to the FCC or a telecommunications certification body (TCB), it is too late to find out that the product operates at a second higher output power with a second modulation. Also, the test lab should not be afraid to point out possible holes in the test plans or suggest alternate test procedures to use as part of the test suite.
The manufacturer and test lab should prepare a test plan based on how the device will be tested and to what standards. Both the test lab and the manufacturer must review the product to all applicable requirements.
For example, for a Part 15 wireless device, the test lab should verify that the device meets all applicable requirements including Part 15.15. This part stipulates that the end user must not be able to configure the device to operate in modes in violation of the FCC regulations. This means that the standard software, not the test software, and the manual should not provide instructions so the user can configure a radio for channels outside the U.S. bands of operation nor should there be a user menu that allows this. This review is critical.
The FCC recently issued a clarification on crystals in Part 95 radio control (R/C) transmitters which stated that allowing the user access to the crystals puts the device in violation of Part 15.15 and ineligible for certification. This affected several R/C systems that had been granted approval under the TCB program. I understand that the FCC is rescinding the grants on these devices.
The FCC also is addressing an issue related to some 2.4-GHz access points and unauthorized software. Some access points have been certified with software that allows the end user to configure the system to operate on frequencies it is not certified for, or worse, outside the approved FCC frequency band of operation. The situation is summarized in Table 1 (see below).
Table 1. 2.4-GHz Access Point Allowed/Disallowed Frequency Ranges
There also should be a review of the various operational modes of the wireless device, and the test software should support operation on every channel, power level, modulation, and data rate for test purposes only. Unless the software supports this, formal compliance testing cannot be done. Although the test software should allow configuration for all channel sets and power levels, the tester and reviewer must verify that, in the production models, the user does not have the ability to configure the operation in modes outside of the FCC regulations.
Test Procedures
Today, there are many test houses that have automated software for portions of the radio testing. The correction factors automatically are calculated and the limits presented on the screen. All you do is push a button.
Test engineers have it much easier than we did 20 years ago when we performed meter corrections and calculated the correction factors ourselves. Instead of looking for the signal, we listened to it via a set of headphones. But what if automation should get in the way of the actual radio test?
I like automated testing, but it concerns me when engineers rely too much on it. I worry when I do not see a current copy of the Code of Federal Regulations Title 47 or an equivalent radio standard within reach of the test engineer nor the relevant public notices, reports and orders, and a binder with the latest FCC interpretations on testing.
From time to time, the FCC regulations including test procedures do change, whether by FCC rulemaking, a public notice, or an interpretation. These changes—sometimes minor, sometimes not—affect how the product is tested. So, it is possible for the test engineer to rely too much on the test software and not spend enough time reviewing the results to see if the data actually complies with the regulations and interpretations.
A good example of this happened a few years ago when a test house evaluating one of our 900-MHz direct-sequence spread spectrum (DSSS) radios measured the power via an analyzer. Since it was below 1,000 MHz, the automatically controlled system recorded the measurements in quasipeak. Yet the requirement for transmitters is measurement in peak. The FCC caught the mistake during the approval review.
Another good example was a number of the early Bluetooth devices that failed certification because they were not properly tested. The FCC had issued an interpretation that the one scan mode that Bluetooth could be configured in removed it from being a frequency-hopping device and classified it as a direct sequence or hybrid device, which needed processing gain. This ruling required the processing gain data to be provided before the devices could be certified.
This problem was not discovered until the manufacturer filed for certification and the FCC raised the question. However, with the removal of the requirement for processing gain testing for digital transmission systems (DTS), this no longer is an issue.
Also, industry standards may not provide the testing guidelines that complete the requirements of the FCC. As an example, for audio-level testing, the FCC requires up to 5 kHz for audio response; the Telecommunications Industry Association/Electronics Industry Association (TIA/EIA) 601 stops at 3 kHz. So although the test methodology from the TIA standard is to be used, if you do not test to the requirements as stipulated by Title 47 certification request form (47CFR), you may not be able to certify your product.
The Correct Test Methodology
What amazes me when reviewing some test reports approved by a TCB or occasionally the FCC is finding approvals issued although there are glaring mistakes. Again, I am concerned when I do not see reference that the device was tested in all various modes and modulations in the report.
A good example is an 802.11(g) wireless fidelity (Wi-Fi) device. At the very least, there are three sets of modulations to verify: the binary phase shift keying (BPSK) modulation for 1 and 2 Mb/s, the complementary code keying (CCK) modulation used to reach 5.5 and 11 Mb/s, and the orthogonal frequency division multiplexing (OFDM) modulation to reach 54 Mb/s. The various modulations can affect the maximum transmitter output, specifically the transmitter power measurements.
For instance, you cannot use the test procedure for OFDM in the public notice addressing 5-GHz unlicensed national information infrastructure (U-NII) devices to test an OFDM device operating at 2.4 GHz. The Part 15.247 specification calls out peak transmitter requirements, and the Part 15.407 power can be measured using the time averaging methodology in the public notice.
Using the 5-GHz OFDM test procedure in the public notice not only will produce a lower transmitter power reading for the test report for a 2.4-GHz 802.11g device, but also not comply with the requirements spelled out in that rule section. Also, systems that pass the band-edge emissions with CCK modulation at 11Mb/s may fail at the same power level with OFDM at 54 Mb/s because of the modulation envelope differences between OFDM and CCK.
Although in most cases the highest data rate may produce a worst-case emission, some research on testing these products for RF exposure has demonstrated that the lower data rates actually have produced higher specific absorption rate (SAR) numbers in some cases as well as some higher out-of-band emissions. Again, this possibly can be related to differences in the modulation and transmitter duty cycles.
Then there are the power measurements that were performed with the modulation turned off. Power measurements need to be done for each modulation at the maximum power for the correct antenna combination requirement. In theory, the data rates should not affect power output, but quick verification never hurts.
Another issue is band-edge emissions where the bandwidth of the test instrument is critical. I have been at several labs that outright failed a passing product because the band-edge measurement requirements were not used as described in FCC Report and Order 96-8 on spread spectrum.
I also have seen test reports where 4-W effective isotropic radiated power (EIRP) systems that met band-edge emissions limits for restricted bands by 20 dB only passed because the bandwidths were set wrong, and the reviewer did not catch it.
Another testing and reviewing error occurs when verifying that the peak-to-average ratio for a Part 15.247 device does not exceed 20 dB. Even if both peak and average readings are in compliance, the system will not be in compliance if this ratio is exceeded (Table 2, see below).
Table 2. The Effect of the Part 15.247 Peak-to-Average Limit
This will be more critical when ultra-wideband devices start being evaluated. Labs not only will need to thoroughly study the Part 15 subpart F rule section, but also the various reports and orders, rulemakings, notice of proposed rule makings (NPRMs), and any technical opinions on the subject of ultra-wideband devices. I suspect until ultra-wideband products become more mainstream like 802.11 devices, the full understanding of the test methodology will be in the hands of a few experts.
Conclusion
The best advice I can give on testing any product, especially wireless, is to review all of the requirements including public notices, reports and orders, and interpretations. This along with a thorough review of the product will allow the lab to develop a plan and properly test the product to all the applicable requirements to obtain radio type approval.
About the Author
David A. Case, NCE, NCT, is a senior regulatory engineer for the Cisco Systems Corporate Compliance Group. He is a member of numerous technical committees including the Executive Committee of the TCB Council and the NARTE Board of Directors and chair of Technical Committee 8 (Telecom) of the Information Technology Industry Council. The former EMC test engineer served as part of the U.S. delegation to the 2003 World Radio Conference and the U.S. delegation to CITEL PCC II Conferences that addressed global and regional spectrum issues. Cisco Systems, Corporate Compliance EMC Standards and Operations, Stonegate Corporate Park, 4125 Highlander Parkway, Richfield, OH 44286, 330-523-2139, e-mail: [email protected]
FOR MORE INFORMATION
on Title 47 CFR
www.rsleads.com/311ee-179
on FCC Reports and Orders ET 96-8 and ET 99-231
www.rsleads.com/311ee-180
on ANSI C63.4 2000
www.rsleads.com/311ee-181
Return to EE Home Page
Published by EE-Evaluation Engineering
All contents © 2003 Nelson Publishing Inc.
No reprint, distribution, or reuse in any medium is permitted
without the express written consent of the publisher.
November 2003