UHF RFID Reader Designs: Achieving FCC Part 15 Certification

April 3, 2011

I was watching this video the other day on “How to Build an RFID Reader“; while it’s interesting to see how compact and prevalent UHF RFID reader modules have become, they missed one of the biggest challenges which is achieving certification from the regional authority to operate the reader in question.

This post will focus on deciding how to implement a UHF EPC GEN 2 RFID reader design (902 to 928 MHz) and, if necessary, the steps needed to get it certified for use in the United States by the Federal Communication Commission (FCC) under Code of Federal Regulations (CFR) 47 part 15.

Module or System-on-chip design?
If you plan on developing an RFID reader the first consideration should be whether to design your own from scratch by using an RFID reader-on-a-chip like the Impinj Indy series reader chips or if you should integrate a ready-made, pre-certified module like those from Thingmagic.

Cost considerations
A pre-built and certified RFID reader module costs on the order of $250 to $300 and requires an additional $50 or so in additional parts (i.e. power supply, connectivity module, heat sink, casing, etc.) to have a ready-for-market product. An RFID reader-on-chip typically costs somewhere around $40 and usually includes a Bill of Materials (BOM) which can be used to build a complete module. A good electronic contract manufacturer should be able to build a complete module using this BOM for around $100 each. In either case, there will be an additional cost for antennas.

The decision needs to be made based on a calculated profit point consisting of two key factors: volume and time-to-market. If your volume is going to be in the hundreds or even low thousands, it probably does not make sense to go through the long development, test and certification cycle of a reader-on-chip. Development time using a reader chip is on the order of at least 6 months and can be a year or more depending on how much modification you choose to do to the supplied firmware and BOM; a reader module will be significantly faster but the tradeoff is in terms of flexibility in form-factor design and implementation.

Some specific costs associated with taking a reader through design, test and certification are:
1. Test equipment – while an FCC-approved lab will be doing all the final test, you will need some to verify design and prototype board build as we as for pre-compliance testing. this will typically include at least a spectrum analyzer and Vector Network Analyzer (VNA).

2. Certification costs (for various regions): certification for FCC costs around $6000. If you plan to sell in other regions you will need to pay for certification for each of those regions; Canadian certification costs about the same as FCC, whereas to get certification for your device to operate in the European market (ETSI) you will need to allow for around $10,000. Reader modules are typically pre-certified and do not require additional testing but you would need to confirm this with the manufacturer.

Control and Flexibility
Going with a from-scratch design allows you more freedom in terms of form factor and implementation. Often reader chip manufacturers will supply the MAC source code to do a complete custom build of control firmware giving you the ultimate in design and implementation flexibility. That being said, it is wothwhile to confirm that a reference design with a Bill of Materials (BOM) is included in the purchase of the reader chip development kit.

Speaking of implementation, be sure to factor in software development; confirm with reader chip or module manufacturer that a Software Development Kit (SDK) is available in the programming language(s) and platform that you prefer developing in.

Steps for FCC Approval
Should you decide to move forward with a design implementation, here’s some details on getting FCC part 15 approval once you have functional prototypes made:
1. Lab selection – my advice is to select a full service test lab which will file all the paperwork for you in addition to the technical testing. Basically a one-stop-shop whereby you submit your reader and some paperwork and in return you have an FCC certification number and ready to go to market. Here is the link to search the dtabase of FCC accredited test labs:

2. Registration (Federal Registration Number)
After selecting a lab, you will need to register your device and get a Federal Registration Number (FRN), this can be completed on-line using this link: https://fjallfoss.fcc.gov/coresWeb/publicHome.do

3. Once you have the FRN, you will need to obtain a Grantee code using form 159

Submitting Device for Testing
When submitting your device to the testing lab for certification, you will need to include:
a. Letter appointing the test lab – lab should be able to provide an example of this
b. FCC ID code of the unit (first three digits are grantee code, last 14 are up to you)
c. Sketch of FCC label location and dimensions
d. Block diagram showing all clock oscillators and their frequencies – reason being that one key test (and the most difficult to pass) is that for spurious emissions. This test is performed starting at the lowest frequency radiated by the device, notice I did not say transmitted but, rather, radiated – so this includes clock frequencies.
e. Full schematics
f. User’s manual
g. A brief, non-technical description of the device and method of operation
h. A sample of the device for testing and photos

Design Confidentiality
All the documentation submitted for FCC approval is a matter of public record. You can file with the FCC to request confidentiality under 47 CFR 0.459

For additional information, you may wish to contact the FCC directly.
Federal Communications Commission
Office of Engineering and Technology
7435 Oakland Mills Road
Columbia, MD 21046
Phone: (301) 362-3000
E-mail: labhelp@fcc.gov


WiMax in Portland OR – High Speed but at what Cost?

January 17, 2009

WiMAX has arrived in Portland, Oregon via the Sprint/Clearwire  service named “Clear” so I thought I would share some interesting highlights of the WiMAX roll-out thus far. 

WiMAX  from Sprint is proving to provide amazing download speeds as shown in this video report from New York City

However, there are concerns especially in those areas where the 3.5 GHz band is being utilized. One example is the CBS-owned broadcaster KYW in Philadelphia. Something was interfering so badly with the network feed coming off its C-band satellite downlink that the picture and sound were freezing up several times a minute — not a good thing in primetime. It turns out it was the interference emanated from a WiMax service provider testing equipment in the 3650-3700 MHz band that the FCC authorized for WiMax and other wireless broadcast access services just last year.

The problem for broadcasters using the C band to import programming (which is basically every one of them in the United States) is that the upper end of the 50 MHz band abuts the lower end of the satellite C-band downlink frequencies, which run from 3700 MHz to 4200 MHz.

While Sprint is using the 2.5GHz band far away from the C band broadcast frequencies,  both Verizon and AT&T  have won an FCC-auctioned licence to use the 3.5 GHz band. 

Of course the 2.5 GHz band is not entirely in the ‘clear’; some Portlanders are wondering if it is to blame for the recent rash of XM radio outages. XM radio operates two satellites, cleverly nicknamed “Rock” and “Roll” which operate between 2333 and 2344 MHz; when you factor in intermodulation distortion, this could indeed be a factor. The other problem could be that XM finally reduced the power output of their terrestrial repeaters which have been frequently cited by the FCC as broadcasting above authorized power levels. WiMAX vendors have been howling for this to happen for some time and, if it now has, XM could be finding themselves unable to provide the same level of coverage. 

What are your thoughts on WiMAX and on the RF spectrum congestion in general? 

In future posts I had thought to touch on some things that came up in my research such as “4G – WiMAX vs. LTE”, “Is RADAR interference really an issue?” and “How smart radios (i.e. Cognitive/Software Defined Radios) deal with RF spectrum congestion”. Of course, I am always open to suggestions!