Return to SSS Online home page
SSS Online is Your SS, RF & Wireless E-zine!


Topical Menus:

   SSS Online Ezine
   Spread Spectrum
   RF Topics
   Wireless Topics
   Design Topics
   Tech Notes & Tips
   Design Tools
   Software Downloads
   Reference Material
   Ham Topics
   Fun & Games
   Links
   Assorted Topics
   Community Topics
   Site Info
   Home


This site is managed by: Visit Pegasus Technologies
Pegasus Menu
Contact Us


Search:

Google

Search site
Search Web
Leave a Comment

Sign our Guestbook



Visit our Sponsors:
















































































































































































































































































































Spread Spectrum Scene Online

Issue 12, Summer 2004




 

Inside This Issue:



Review our Previous Issues

SSS is proud to present our Twelfth Online issue.

We are now soliciting ideas and articles for our thirteenth issue, which is tentatively scheduled for late fall. Please send your comments and suggestions to:

 
What's New At Pegasus Technologies and SSS Online

Pegasus Technologies and SSS Online on the Move

We're going to be moving! We started out with one room in our home, and have since grown to the point where we have people working in 4 different locations within a twenty mile radius. This is not ideal, and has made communications a continual challenge! So, about a year ago we started looking for a permanent space where we could all work together. We have just signed a lease agreement for a new building to be constructed on a 4.6 acre parcel in the Roane Regional Industrial and Business Park, less than a mile away from I40 close to both Knoxville and Oak Ridge, Tennessee. It's a very beautiful site, as you can see from the picture. We are working with architects and builders to come up with a final building design, one that's expandable so if we continue to grow as we have been, we won't have to move again. We hope to move in around the end of the year, although that may be a bit optimistic.

Our new building site
The site for our new office


We Helped Record History!

Did you see the video of Spaceship One in its historic flight into space on June 21, 2004? The pictures of the takeoff and landing taken at the runway were transmitted back to the America by Air production center over a Pegasus Technologies wireless video link design! It worked very well, as you can see in this video clip.

Changes to Pegasus Tech's Module Lineup

This has been a really busy time since our last issue. We've got several new RF modules up on our RF Module Sales page. Take a look at our new LOW POWER versions of the PTSS2003, which offer ultra-cool operation and extreme battery life. These are available in the 902-928 MHz band, at at 868 MHz. These two new modules use the Xemics XE1202 chip, and have an onboard microcontroller. Most of our modules have FCC and Canadian certification, and the 868 MHz version has been tested to the RTTE standard for use in the European market. Development kits for these modules are also available. Improved production efficiencies and increased volume have allowed us to lower our prices again! New modules are in development, including a 433 MHz version of the PTSS2003 and a brand new 2.4 GHz module with onboard microcontroller.

You can purchase these modules and accessories through our online store, with secure checkout available that accepts Visa and Mastercard. Of course, you can also go the phone-in route and get your questions answered at the same time. We can customize these modules as well, and offer full technical support for getting them incorporated into your application.

Spread Spectrum Scene Completes its 9th Year Online

We celebrated by finishing putting up all the old paper copies from the early days, back in 1992 and 1993. Take a look at our Past Issues Index and check out all those great old paper classics!

We're always looking for great technical content for this site. If you have something you'd like to have published, please let us know. We can't pay cash, but do trade advertising space for articles and papers from time to time, and also provide continuing advertisement through prominent biographies and links that stay with the article forever. drop us a line at . We're always happy to hear your comments on any aspect of our website!




Return to Contents

-->  

Introduction to the Firewire Protocol


By Danny Simpson

firewire connectors


Danny Simpson is a Contributing Editor of SSS Online, and a Senior Associate of Pegasus Technologies, Inc., a provider of electronic RF and advanced electronic design services and RF modules. A natural engineer, Danny is always fascinated by finding out how things work, and provides SSS Online with the benefits of his research.

Overview:

Communications and interconnectivity, so important to the human animal, is more achievable today than at any time in history. With the full flowering of the electronic age, fast and direct digital communications between computers, peripherals, and a variety of embedded devices are vital.

During the past 20 years, digital communications have been growing at a staggering rate. The first popular digital communications protocol, EIA RS-232, has served well for many years and is still in widespread use. However, as higher volumes of data needed to be transferred faster and more reliably, many other digital protocol methods have been developed. The Universal Serial Bus (USB) has become one of the most widely used methods of digital serial data transfer during the last 5 years. USB connectors are now installed on most new PCs and many new peripheral devices intended to be attached to PCs.

Firewire is another digital protocol whose development began well before USB, but currently its use has not been as widespread. Firewire is a low cost, high performance, digital serial interconnect that can be used for many digital items, such as audio-visual devices, consumer PCs and home networks, and today its popularity is growing. The purpose of this article is to serve as an introduction to Firewire; it will not go into the many protocol details. For a similar introduction to USB, see our earlier article, An Overview of the Universal Serial Bus.

Many of the newer PCs now ship with both Firewire and USB serial interfaces installed. As a general rule, USB can be designed and manufactured into products for less money, but Firewire does have several advantages. Unlike USB, Firewire has no need for a server or central point to perform transactions. Firewire is a peer-to-peer relationship as opposed to USB, which uses a master/slave relationship. There are many ongoing and sometimes heated discussions over which protocol is better. Both support hot 'plug-and-play,' where a device can be plugged into its respective bus and then enumerated before being used. However, because of the differences between the peer-to-peer and master/slave protocols, this is handled quite differently. Each Firewire node is exposed to all data traffic on the bus. Unlike USB, the Firewire bus is daisy-chained; there are no hubs to contend with. The current data bandwidth (under the latest Firewire and USB specifications) is fairly equal with Firewire being slightly ahead, but there are differing overhead data requirements. Ongoing development should continue to raise both standards to higher levels.

Both Firewire and USB devices may obtain operating power from their respective buses within their bus power limitations. The Firewire specification currently has higher power availability limits. There are many differences between the two high-speed protocols and each tends to serve differently for differing needs. Both serve well for peripheral devices connected to computers, although USB is currently the more popular of the two for such devices. Firewire seems to be more popular in the higher bandwidth devices such as consumer and industrial higher-resolution video products. Unlike RS-232 where you could simply transmit raw data, both Firewire and USB protocols are very stringent on how they are to be used. The transmitted data is arranged into formatted packets that contain many details along with the data that they may be carrying. Despite the complexity of the protocols, however, they are literally invisible for the person making use of an end product that utilizes either protocol.

Of interest, there is a mechanism called a 'device bay' that can use either Firewire or USB for attaching high speed serial bus devices to PC desktop systems by means of a 'hot docking' arrangement. Microsoft, Intel, and Compaq developed this specification. Using this system, devices such as hard drives or CD ROM drives may be added to a computer without powering down. The proper device drivers are then dynamically loaded to support the newly installed drive device.

Firewire History:

Firewire development first began in 1986, and the initial specification was completed in 1987. The standard was adopted in 1995 as IEEE-1394. Cable length is specified as a maximum 4.5 meters with a characteristic 110 ohms impedance using differential mode.

Initially, there were problems with devices not working properly together on the Firewire bus. Supplement IEEE-1394a has since been issued, adding new features and better performance as well as fixing these initial problems. Another supplement, IEEE-1394b, is currently being developed. It is primarily aimed at supporting even higher speed data transfers (up to 3.2Gbs) and longer distances (up to 100 meters with the use of CAT5 cable, but limited to 100Mbps). It will be backwards-compatible with the two earlier Firewire specifications. Because of speed constraints, IEEE-1394b will more than likely be available only in newer ICs that have the PYS (physical) and Link layers combined. IEEE-1394 is based upon the IOS / IEC 13213 (ANSI / IEEE1212) specification commonly referred to as 'Control and Status Registers (CSR) Architecture.' It defines a set of basic features that can be used by a variety of buses.

Firewire is a registered trademark of Apple Computer Corporation. It is generally considered to have been developed by Apple, although there have been reports that it was the result of a developmental process between several companies. Before using Firewire in their computers, Apple Computer made use of Small Computer System Interface (SCSI) for their Mac computer hard drive interfacing. Eventually, both as a cost-cutting exercise and as a way to share fewer common interfaces between devices, Apple did away with SCSI serial and parallel interfaces on their newer Macs and replaced them with Integrated Drive Electronics, USB, and Firewire serial interfaces.

In addition to Apple, Sony began using a four-wire version of Firewire in their iLink interface, which does not supply power for devices. Yamaha is now using Firewire for their mLAN protocol for high-speed music transfers. Many other manufacturers are now designing products to work with Firewire. Firewire's first main entry was in video applications. A large percentage of the new consumer-based video products that are being designed today, such as video cameras and digital video recorders, are being designed to work with Firewire.

Firewire Below the Surface

Unless otherwise mentioned in this article, the transfer rate of 400Mbps (megabits per second)is assumed for discussion. The Firewire bus currently allows a scalable performance depending on bandwidth requirements such as 400 Mbps, 200 Mbps, and 100 Mbps, and has the following features:
  • Hot plug and play enumeration
  • Currently 400Mbps (six wires)/ 800Mbps (9 wires) Higher speeds currently being developed are 1600Mbps and 3200Mbps
  • Both Asynchronous (time not guaranteed) and Isochronous (real time) transfers
  • Peer-to-peer communications
  • Low cost
  • Power supplied to the interface (6 and 9-wire interfaces)
  • Networking using standard networking protocols

Firewire devices can send data directly to each other without the intervention of a host. Unlike USB, a host is not used at all, since Firewire makes use of peer-to-peer communications. Firewire supports 'hot insertion' plug-and-play. When a new device is inserted or removed from the Firewire bus, devices on the bus are re-enumerated. Since it is a peer-to-peer relationship, arbitration must be performed to determine who will get bus usage. Arbitration is the process of gaining ownership of the bus for a period of time. There is a fairness interval algorithm that helps determine which node will get the bus, so that the bandwidth guarantees made for both asynchronous and isochronous modes can be adhered to. It ensures that all nodes have their turn to use the bus. There is also a detection mechanism to identify and report nodes that may be 'stuck' in a loop. A monitoring mechanism on each Physical port monitors incoming power.

Some pre-defined classifications of Firewire devices use 'unit architecture' details such as identical data protocols, device internal register implementation, and addressing. Some of the benefits of this 'design specification' are better inter-operability between other similar devices and simplification of the entailing software design.

Nodes and Addressing

There can be 1024 buses in a single Firewire implementation, and each serial bus can support up to 64 nodes. Therefore, Firewire can support up to 64K nodes. Since each of the 64K nodes has addressable memory space of 256 terabytes, Firewire's total address space is a huge16 exabytes (64K nodes x 256 terabytes) and is based on the 'big endian' memory-addressing convention. The Firewire specification additionally includes both cable and backplane implementations of Firewire.

Firewire addressing works as follows. Firewire uses a 64 bit-addressing scheme for a total of 16 exabytes. The highest order of 10 bits defines one of 1024 target buses. The next order of 6 bits addresses the 64 nodes on that bus. The remaining 48 bits define the 256 terabytes memory addressing set aside for each node. This node address space is divided up into blocks by the standard for special purposes, including the initial Control and Status Registers (CSR) register space, each node's private memory space, and initial units space.

As mentioned at the beginning of this article, Firewire nodes utilize the CSR Architecture register implementation. This allows a common arrangement to be set up which improves interoperability between different platforms, facilitates moving between different bus types, and allows for more consistent software implementation. Also, CSR defines various ways of handling different types of system failure and error handling as well as automatic configuration. CSR additionally defines a standard method for configuring a local configuration ROM. This ROM is used to help locate the proper drivers for modules, nodes, or units as well as to assist in identifying local properties as they are being enumerated upon connection to the bus. The CSR Architecture also defines a Message Broadcast protocol in which a node can send messages to multiple nodes simultaneously over node address 63, which is reserved for this purpose. An Interrupt Broadcast message mode is also defined.

Firewire Data Transfer Methods

Firewire supports both asynchronous and isochronous methods of data transfer. In both cases, the bus bandwidth is based on 125us time intervals (8000 times/second) called cycles. Both asynchronous and isochronous transfers can be intermixed within these cycles.

Figure 1: Firewire Packet Diagram

Firewire Packet Diagram


Asynchronous transfers communicate to a single node based on a unique 64-bit address. Asynchronous mode does not require a continuous usage of the bus, but does allow fair usage of the bus over time. For asynchronous error checking, a Cyclic Redundancy Check (CRC) is performed and if an error is determined to have occurred, a transfer retry is attempted. Most asynchronous transfers require acknowledgment back to the data initiator, including whether the data was received correctly. The Firewire specification defines error codes that are to be returned from the responder (the receiver of the data and command) back to the requestor (the originator of the data and command). Asynchronous packets consist of seven different packet formats:
  1. Read request
  2. Write request
  3. Read response
  4. Write response
  5. Lock request
  6. Lock response
  7. No data
Lock is used as a way to make a request for bus resources from a manager node. The requesting node guesses if a resource is available, sends a lock request packet for the resource (for example, channel 10) and if the requester is correct that the resource is available, that resource is locked so that no other node can obtain it.

Isochronous transfers use a pre-defined 6-bit channel number to broadcast to one or more nodes that have been assigned to a particular channel address. This mode is for transfers that are required at constant intervals, such as ongoing video transmissions. This mode carries a higher bus priority over the asynchronous transfer mode. Isochronous transfers do not require confirmation of data delivery. In this mode, the data sender is known as the 'talker' and the receiver of the send data is referred to as the 'listener'. Isochronous transfers require initial negotiations with the 'isochronous resource manager' to obtain bandwidth that will be needed for the continuing data transfer. Isochronous packets consist of only one packet format:
Data block
Communications using both asynchronous and isochronous transfers are performed using packets of data. Asynchronous data packets can contain destination address, the source ID, type of transaction (read, write, or lock), response code (from a previous request), the actual data, CRC, acknowledge message, and acknowledge parity. Isochronous data packets are simpler containing channel number, type of transaction (read, write, or lock), the actual data, and CRC.

Firewire uses a software layering method for communications over its bus. There are three layers used in isochronous mode and four layers used in asynchronous mode. These three or four layers are used to communicate with what can be considered a fifth, or Application layer, which is the actual program or device that the user is working with.

Since the asynchronous mode requires more 'handshaking' than the isochronous mode, an extra layer, called the Translation layer, is required to support this protocol. The same three layers used in isochronous mode are utilized in asynchronous mode along with this additional layer. These layers are:
  1. Physical layer - The actual interface for transferring data between the Firewire device and the Firewire bus. This layer provides the necessary signal timing and mechanical interface. It is also used for bus configuration when a device is either added or removed from the bus. During bus configuration, each node is uniquely identified. This physical layer is generally implemented with an IC such as Texas Instrument's TSB41LV03 or Philips PDI1394P23 along with the necessary software. Newer ICs are available that contain both the Physical and Link layers.

  2. Link layer - Provides the 'assembly' and 'disassembly' of the packets that are transferred over the bus. For outgoing packets, the encoding occurs here. For incoming packets, decoding of addresses or channels occurs at this level. CRC error checking occurs here for incoming data. For isochronous transfers, the isochronous program/driver communicates directly with this layer, bypassing the Transaction layer. This Link layer is generally implemented with an IC such as Texas Instrument's TSB12LV22 or Philips PDI1394L40 along with the necessary software. Newer ICs are available that contain both the Physical and Link layers.

  3. Transaction layer - Used exclusively for asynchronous data transfers. This layer supports the CSR Architecture specification for the request-response 'handshaking' protocol.

  4. Bus Management layer - Provides initial node configuration as well as overall management of a node. This layer works in conjunction with other nodes on the Firewire bus to form a 'global' management system when required. This global system handles items such as defining isochronous channel numbers and bandwidth, and handling available bus power and optimizing bus bandwidth during packet transfer.

Each 'requestor' synchronous data transaction is followed by an acknowledgement by the receiving 'responder' for each packet transferred. And in turn when the 'responder' handles the actual request back to the 'requestor', the original 'requestor' sends back an acknowledgement back to the 'responder'. The previously mentioned Transaction layer handles these 'handshaking' functions.

Figure 2: Diagram of Firewire Layers

Diagram of Firewire Layers


Negotiating for Bus Usage

When any node needs to perform a transaction over the bus, it must use arbitration for bus usage. Firewire's design works to make data transactions as efficient as possible. It can multiplex requests from multiple nodes and respond simultaneously to each request. It also may use 'concatenated transactions' and 'unified transactions' to help eliminate usage inefficiencies.

Concatenated transactions can eliminate the problem caused when a slow response from a node causes it to lose its place in the bus line. For example, if one node is waiting on a response from a slow-responding node, transactions between other nodes can occur during this waiting period. Then, when the slow node is ready to return the acknowledgement message, it might have to start the bus arbitration process all over again, which can waste bandwidth. If a node is relatively fast, it can actually return its acknowledgement response with its data response concatenated afterwards to help eliminate the possibility of losing its place on the bus and having to re-arbitrate for bus bandwidth.

A unified transaction occurs when a write transaction has the acknowledge and response completion codes coming back from the 'responder' combined into one simple packet. However, a read transaction cannot work as a unified transaction, as the 'responder' must include the requested data, which cannot be part of an acknowledgement packet, and the completion status as well.

A single device that connects to the Firewire bus is called a 'module'. A module can have one or more 'nodes' in the module, which are logical connections. A node may be further broken down into more logical 'units'. In other words, a 'module' can contain multiple 'nodes', which in turn may contain additional 'units'. This flexible hierarchy simplifies system design. An actual physical connection is made through a module's (device's) 'port'. When a multi-ported node receives a packet, it is synchronized to the device's local clock and retransmitted to its other 'local' units.

Hardware

There are several connector types that can be used when connecting to the Firewire bus. The original 1394 specification mandated a 6-pin connector that can also carry power. As with USB, when using the power capabilities of Firewire, the power pins are a bit longer than the signal pins, giving a 'make first/break last' power connection as the Firewire connector is being inserted and removed. The 1394a supplement added the option of a smaller 4-pin connector, which does not have the capability of sourcing or receiving power to Firewire devices. Sony makes use of this connector in their video cameras. The cables in both cases have outer shields as well as shields on each of the two twisted wire pairs.

The Firewire 1394/1394a 400Mbps data transfer is performed bi-directionally over two twisted-wire pairs in a half-duplex manner. One wire pair is used for data transfer and the other is used for synchronization. The Firewire 9-wire configuration is to be used with the 800Mbs specification. If a 6 or 9-wire system is used, another pair of wires carries power to or from Firewire devices. A Firewire device can either be a power source for other devices connected to the bus, or it can receive its power from the Firewire bus.

There are differing mechanisms in the different Firewire specifications that determine when a device has been connected or removed from the Firewire bus. The Firewire specification also gives much attention to items such as speed signaling to determine data delivery rates as well as various types of control/status signaling. Both differential and common mode signaling are used to generate various special internal functions, much like interrupts occur in microprocessors. Firewire utilizes NRZ encoding to help reduce the frequency component in the serial transmission. The recovered clock is generated on the receive end and is obtained by data being exclusive OR'd with the data strobe.

Firewire can supply voltage to power devices connected to the bus. This voltage can range from 8 to 33VDC at a maximum of 1.5amps. There is a detailed power state management section in the Firewire specification that specifies how control can be managed for the power states within nodes, or even within units inside nodes. There are a total of four power-on states that can control the power within each node as well as four additional power states that control the power of individual units within that node. Nodes that are in a low power state can be notified to wake up. A Node can alert the power state manager of its current power state. One node can control another node's power functionality.

Engineering notes:

There are a number of developmental resources on the Internet to assist in the development of a product that is being designed to operate on the Firewire bus. Most of the manufacturers who produce ICs for use with Firewire have many reference designs available. Some of our favorite links are listed in the box below.

Firewire Links

1394 Trade Association
Texas Instruments Firewire
Philips Firewire (now NXP)
Fujitsu Firewire
NEC Firewire
Agere Firewire (Now LSI 1394 series)
Apple Developer Connection
Microsoft Firewire Support





Return to Contents

 


How-to: Debugging a Low-Powered RF Transmitter



by Jim Pearce
Pegasus Technologies, Inc.


We were recently commissioned to redesign a remote control transceiver operating in the 902-928 MHz license-free US band. The client had an operable unit, but its output spectrum was so ugly that the possibility of passing FCC certification testing was nil.

The original board had been layed out by a competent PCB engineer but he was not knowledgeable about RF layout issues. There was little in the way of ground planes, no via grids to connect a top and bottom ground pours to the inner ground plane, no controlled impedance traces, etc.

We addressed these and other problems by re-laying the board, taking good RF performance into account. Soon we had the revised board in our hands, populated and ready to check out.

The firmware in one of the boards was set to make the board continuously transmit at a frequency in the middle of the band. This allowed us to easily verify the spectral purity and power of the RF output signal. By replacing the normally soldered monopole antenna with a short section of micro co-ax (I like to use Digikey part number 229-1007 along with an SMA adapter) we were able to measure the output power into 50 ohms by simply connecting the co-ax to the input of a spectrum analyzer.

This is where the fun started! The power was low --- way too low. The transmitter had been designed for an output power of about +20 dBm or 100 mW but we were only seeing about +6 dBm or 4 mW. So where does one start to see why the power output is so low in a situation like this?

The first thing we noted was the power supply current. It was what we would expect for full power operation, so we were convinced that the power was being generated - it was just not making its way to the antenna connection.

Jim debugging a PCB
Debugging a board in the lab
  Next we isolated the RF output stage-by-stage. We connected more of the micro co-ax to various spots in the output chain, starting with the output of the RF integrated circuit. We were careful to break the connection to the next stage by removing a coupling component, such as a coupling capacitor. We were also careful to be sure that there wasn't a DC path to ground on our output tap. We would then fire up the transmitter with our tap connected to the input of the spectrum analyzer and measure the output power.

This particular transceiver consisted of an RF integrated circuit, SAW filter, RF power amplifier, transmit / receive switch, and a discrete component L-C low-pass filter. We moved the micro co-ax tap to the output of each of these stages to check the output power levels. Everything looked fine; even the output of the T/R switch was well over +20 dBm. This focused our attention on the L-C filter since the power was fine ahead of it but lousy after it.

We placed micro co-ax on the input and output sides of the L-C filter so that we could measure its frequency response and insertion loss. We used a sweeped frequency signal generator for this test, since our spectrum analyzer with the internal tracking generator was in use on a different project.

This rapidly showed that the filter, which was supposed to have a 950 MHz 3dB cut-off, had a 600 MHz cut-off instead! We changed the values of a couple of inductors and one capacitor and we got the sort of low-pass response that we were expecting.

We removed all of the micro-coax (except for the output connection), replaced the coupling capacitors, and turned the transmitter on. The output was now just what it should have been and the spectrum looked nice and clean.

The time from first power to good output power was only a couple of hours.




Return to Contents

  Article divider


FCC Modifies Rules for Unlicensed Devices
by Karen Edwards, Pegasus Technologies

US Capitol


On July 12, 2004, the FCC issued a Report and Order (R&O) modifying Parts 2 and 15 of its rules for unlicensed devices and equipment approval. The Notice of Proposed Rulemaking was released September 17, 2003. By this modification, the FCC adds flexibility that it hopes will encourage continued innovation in broadband technology and services.

Since unlicensed devices form the bread and butter of Pegasus Technologies' work, we have been following this rulemaking very closely. We reported on the proposal in our October 2003 issue and provided comments to the FCC. We didn't get everything we wanted, but we were one of the commenters named several times in the R&O, so we know for sure that our comments were considered — and that's gratifying!

The revised rules will:
  • Permit the use of advanced antenna technologies ("smart antennas") with 2.4 GHz spread spectrum devices (updated Section 15.247);
  • Modify the replacement antenna restriction for Part 15 unlicensed devices;
  • Allow RF transmitters to be marketed separately outside of transmission systems, as long as each complete system that the transmitter can be used with has been tested;
  • Permit digital modulation systems authorized pursuant to Section 15.247 of the rules to use the average measurement of output power like U-NII devices authorized under Sections 15.401 - 15.407; and
  • Modify the channel spacing requirements for frequency hopping spread spectrum devices in the 2.4 GHz band in order to remove barriers to the introduction of new technology that uses wider bandwidths.
A brief summary of each of the proposed changes, the comments, and the final rules are summarized below.

Advanced Antenna Technologies

Previously, the rules provided only for use of omnidirectional and point-to-point antennas, and set limits on the amount of directional gain that may be used by each type.

The proposal would have permitted sectorized or phased array antenna systems that must be capable of forming at least two discrete beams with a total simultaneous bandwidth not greater than 120 degrees, regardless of the number of beams formed. In addition, the proposed rules would allow these types of antennas to operate within the higher limits currently allowed for point-to-point antennas, with an upper limit that the aggregate power transmitted simultaneously on all beams could not exceed 8 dB above the limit for an individual beams. Finally, the FCC proposed that the transmitter output power must be reduced by 1 dB for each 3 dB that the directional antenna gain of the complete system exceeds 6 dBi.

Commenters were generally favorable on this proposal (although we were not, fearing that low-power applications in the 2.4 GHz band might be affected by the proliferation of relatively high-power Wireless Internet Service Provider (WISP) signals that are likely to result from this change). There was concern that the rules be flexible enough to allow different technologies other than the ones specifically mentioned. Opinions were mixed on the various limitations in the proposed rules.

Based on the comments, FCC's final rules will permit advanced antenna systems, including sectorized and adaptive array systems, to operate with an aggregate transmit output power up to 8 dB above the limit for an individual beam. However, the total EIRP on any beam may not exceed the EIRP limits for conventional point-to-point operation. The 120 degree restriction has been dropped in favor of the new EIRP limits. Existing advanced antenna systems that have already received an equipment authorization are grandfathered.

Replacement Antenna Restrictions

Before this new rulemaking, Part 15 required that intentional radiators be designed in such a way that no antenna other than the one supplied can be used with the device. To this end, the use of standard antenna jacks or electrical connectors is prohibited. This was an attempt to ensure that the Part 15 emission limits aren't intentionally (or unintentionally) circumvented by replacing the supplied antenna with another one with higher gain.

FCC's proposal was to modify this requirement to require testing with the highest gain antenna of each type that would be used at maximum transmitter output power. Any antenna of a similar type that doesn't exceed the antenna gain of the tested antenna(s) could be used on that transmitter without retesting and approval. The prohibition on standard antenna jacks would remain.

Comments were generally favorable on this proposal (including ours). However, several commenters suggested that both in-band and out-of-band emissions must be considered in determining if an alternate antenna is acceptable. Others argued in favor of eliminating the restriction on standard antenna connectors, saying it was not working as intended and added additional design complexity and unwarranted costs.

FCC has decided to permit the use of a variety of replacement antennas. Revised Section 15.204 will allow multiple antennas of "similar in and out-of-band gain and radiation pattern." Compliance testing for the intentional radiator must be performed using the highest gain antenna that will be used with the device. The manufacturer must supply a list of other acceptable antennas in the product literature. FCC is retaining the requirement for non-standard connectors, fearing that uninformed consumers might accidentally use antennas that would exceed limits, if this rule were eliminated. However, it is removing the Section 15.407(d) requirement that devices operating in the U-NII band must have an integrated antenna. This is in recognition that equipment is increasingly able to operate across multiple unlicensed bands, making it impractical to maintain separate antenna requirements for each band in which a device may operate.

Flexible Transmission System Configuration

As the previous rules required equipment authorization for every combination of a transmission system, it was difficult to mix and match components without extensive retesting. This had an impact on WISPs and limited their ability to design systems that meet the needs of the particular application without the additional time and expense of retesting.

The Commission proposed several changes to the rules to address this problem: (1) changes to permit limited marketing of RF power amplifiers outside of a specific transmission system; (2) provisions that would permit "professional radio systems installers" to configure systems using technically equivalent components without requiring retesting of the specific system selected. This would, of course, put a much greater burden on the technical competency of the systems installers to ensure that their systems meet FCC regulations, but would enable both quicker deployment and lower costs for WISP services.

Commenters, including us, were very concerned that the proposed rules went too far in permitting the use of external RF power amplifiers without testing in WISP systems. There was general agreement that this proposal, if adopted, would open Pandora's box of escalating interference and violation of FCC rules. The Commission considered the comments and modified its proposal. The final rules will allow external RF amplifiers to be marketed separately, but only if they are designed in such a way that they can only be used with a specific system (or systems) that is covered by an equipment authorization, so that the complete system will still be ensured of compliance. Other limitations on output power, and certain notice requirements, are also included in the final rules. Because of these changes, there is no longer a need to define and provide special authorization for "professional radio systems installers," and this part of the proposal was dropped in the final rules.

Measurement Procedures for Digital Modulation Systems

Devices that operate under the Section 15.247 and those that operate under the U-NII rules (Subpart E) are both limited to a maximum power output of 1 watt. However, output power was measured differently for the two types of devices under the previous rules. FCC proposed changes that would make measurements the same for both types of devices - and would use the average measurement of output power (U-NII) rather than the peak emission method previously used for spread spectrum devices.

Most commenters, including us, were in favor of this proposal. There were some concerns that in certain cases the peak emission method would be better. So, in the final rules, the Commission is permitting the part 15.247 devices to choose whether to use the average measurement of output power or the peak emission method. If the average power procedure is used, then out-of-band emission must be reduced to 30 dB below the level of the device's fundamental frequency. Additional clarifying changes are being made to the definitions of transmit power.

In its proposal, the FCC had solicited comments on whether it should amend the spectrum occupancy rules for Section 15.247 and U-NII devices to apply the same limits to both types of devices. There weren't many responses, and they were conflicting, so the Commission has decided to leave the spectrum occupancy limits unchanged for the time being.

Frequency Hopping Channel Spacing Requirements

The previous rules for the 2.4 GHz band required a separation of 25 KHz or the 20 dB bandwidth of the hopping channel, whichever is greater. The Bluetooth SIG suggested a modification of the channel separation requirements for FHSS systems, to 25 KHz or 2/3 of the 20 dB bandwidth, whichever is greater. This would permit Bluetooth devices to be capable of data rates of up to 3 Mbps. The FCC proposed to adopt this suggestion for 2.4 GHz systems with an output power limit of no more than 125 mW and fewer than 75 hopping channels.

Commenters were strongly in favor of this proposal, however, most recommended dropping the proposed limit of 75 hopping channels. We also proposed extending this modification to the 915 MHz band. The Commission's final rules will modify the spacing requirement for systems that are limited to an output power of 125 mW using any number of hopping channels. They are not extending this provision to the 915 MHz band, because of the differences in width of the band and insufficient information on the affects that modifying the spacing requirements might have on existing users of the band.

Modular Transmitter Approvals

FCC had proposed to clarify the equipment authorization requirements for modular transmitters. We thought that was a great idea, but it has been scrapped for the time being. The Commission recognizes that there are many complex and evolving issues associated with this subject at the present time, and believes it would be better to wait and collect more information before proposing changed guidelines.

Spectrum Sharing

The FCC asked for comments on how spectrum sharing in the unlicensed bands can be improved, and specifically whether the spectrum etiquette provisions that apply to unlicensed PCS devices should be expanded to cover other types of unlicensed devices. Most commenters, including us, were not in favor of spectrum etiquette provisions. However, Microsoft recommended some spectrum sharing protocols that would require devices to cease transmission if no information was being sent, allow unlicensed devices to transmit only if they can do so without causing interference to other devices already using the channel, and to require unlicensed devices to incorporate dynamic range control which would force it to use the minimum transmit power necessary to complete a communication link.

The Commission has decided not to impose any type of spectrum etiquette for the already heavily-used Part 15 bands. However, they are considering the Microsoft recommendations for any additional spectrum that may be made available for unlicensed operation in the future. This would be the subject of further rulemaking, so be alert!

Special Temporary Authority

The FCC has decided to eliminate the STA provisions in Section 15.7. No one has requested a STA in the last 10 years. Part 5 provisions for experimental licenses are being used instead and seem to be adequate.

Conclusion

There's a lot more in the R&O that is not summarized here, including some information on Part 2 changes, which aren't that numerous, and the actual language of the changed rules themselves. We urge you to review the R&O carefully. Note that the changes will not be effective until 30 days after publication of the new rules in the Federal Register, and that may take a while - so we're not quite there yet!

Related Links




Return to Contents

 


UWB Standards Battle Update



by Karen Edwards, Pegasus Technologies, Inc.


Over a year has passed, and there's no end in sight to the battle that still rages over which of the two leading Ultra Wideband approaches will be selected as the standard. The two candidates are the Multi-Band OFDM approach, which is supported by the Multi-Band OFDM Alliance (MBOA) consisting of Intel and a host of other companies, and the Direct Sequence approach sponsored by Freescale Semiconductor (a Motorola spinoff) and others. In the most recent meeting of the IEEE committee tasked with developing a UWB standard, the Direct Sequence approach narrowly won an initial vote but did not have the 75% necessary for adoption. In several previous meetings, the multi-band approach won, but again, not by enough margin. Both groups are continuing to develop products based on their own approaches, so this may have to be decided in the marketplace.

Keep on top of the latest in this battle by referring often to our UWB News page. We have an automated news feed on this page, and also post information and links whenever anything new and exciting happens in this area. You can see more details about the recent votes there, plus find some great background information on the various approaches, the key players, and more.



Return to Contents

 

Job Search Spotlight
The "Interviewable" Resume

by Linda Matias

Recognized as a career expert, Linda Matias brings a wealth of experience to the career services field. She has been sought out for her knowledge of the employment market, outplacement, job search strategies, interview preparation, and resume writing, quoted a number of times in The Wall Street Journal, New York Newsday, Newsweek, and HR-esource.com. She is President of CareerStrides and the National Resume Writers' Association. Visit her website at www.careerstrides.com or email her at linda@careerstrides.com..


It is rumored that the only word William Shakespeare wrote on his resume was "Available." We'll probably never know if that is true. But it raises an interesting question. How much information is too much and how much is too little when dealing with resume copy?

The resume is a vital piece to any job search. As companies scramble to find the ideal candidate, they use the resume to screen candidates. Done right, a resume builds an instant connection with the reader and helps steer the course of the interview in your favor. If you submit a resume that piques the curiosity of the reader, he or she most likely will ask questions based on the information you provided on the resume as opposed to relying on a pre-packaged questionnaire. That's how you know you have an "interviewable" resume, when it assists in shaping the course of the interview.

The challenge is, How does one create an "interviewable" resume, one that isn't boring or sterile? How does one write a resume that motivates the reader to give you a call?

Write with the employer in mind

Cast aside the belief that the resume is about you - because it isn't. Though the resume is your "story", the heart of it should focus on the needs of the employer. When developing your resume give thought to the person who will be reading it. What are his or her immediate concerns? How will you be able to solve that person's problems?

Though it may be difficult to pin down a company's immediate concerns before an interview, the reality is that organizations recruit candidates for one of the following reasons: they need to replace an unproductive employee, a peak performer was promoted or left, or a new position has been created. A recruiter usually searches for a candidate who will produce certain results, one that is a skilled communicator and has a strong work ethic. If you are able to target your resume toward these key areas, you will, without a doubt, tap into the organization's concerns.

Choose your phrases carefully

Sentence starters and appropriate use of action words all determine whether the resume is "interviewable." Instead of using predictable phrases, think of ways to add punch to your resume. For example, instead of using increased sales by 250% ... write delivered a 250% increase in sales...; instead of using ability to effectively ... write demonstrated ability to effectively ... ; and instead of using reduced costs ... write slashed costs.

When your resume doesn't "sound" like all the others on the recruiter's desk, he or she will take notice. You will be remembered when your resume breaks the monotony of the recruiter's day. Guaranteed.

Have a consistent message

Don't try to become all things to all people. If you are a CEO, don't add a statement that indicates that you are willing to be a Business Manager. If you are a Sales Manager, don't indicate that you are willing to take on a position as a Customer Service Representative. Get the picture? Determine what you are selling (and looking for) before you put one word to paper.

Determine your major selling points

Though you may share the same job title with many other people, your accomplishments and how you carry out your responsibilities are what distinguishes you from all the other qualified candidates. Focus your resume on not only what you did but also how well you did it. By design, what makes you "interviewable" is how you market your strengths on paper.



Return to Contents

Spread Spectrum Scene Online is managed by Pegasus Technologies. We can help you get your quality RF products to market!
Visit Pegasus Technologies
Meet Pegasus Technologies!




Contents SS Glossary Pegasus Tech Navigation Home

  Tel: 865-717-9339   ||   FAX: 865-717-9904    ||   E-Mail:
This site © 1995-2008 by SSS Online, Inc. All rights reserved.
Revised November 25, 2008