![]() Preview / Read the Current Issue of SSSVolume 4 - Number 3 -- Spring 1996Updated 8/8/97[Rumors ] [Decipherings] [Editorial] [New Products] [SS Chipsets Simplify SS Design] [Data Randomizing ...] Part 1
RUMORS & RAMBLINGSLoral Corporation and Bernie Schartz are at it again! They announced in early January that Loral has entered into a series of strategic transactions with Lockheed Martin Corporation. Under the new agreement, Loral's space and telecommunications activities will become an independent company under the new name, Loral Space and Communications Corporation (LSC). Loral's defense electronics, information services, and systems integration operations were transferred to Lockheed Martin for 89.1 billion. LSC will begin business with an estimated value of 82 billion including $700 million in cash and no debt, and with 31% interest in Clobalstar, 33% interest in Space Systems/Loral, and 22% interest in K&F Industries. As the man- aging partner of Globalstar, LSC will continue to focus on the development of the Globalstar system. The system remains on-track and on-schedule for launch in mid-1997 and commencement of operation in 1998. For more information contact Globalstar at TEL: +1 408-473-5872 or FAX: +1 408-473-5750 -- Good Luck Bernie, you may need it! Decipherings"Common sense is genius dressedin its working clothes." -- RALPH WALDO EMERSON EditorialHere it is -- our SECOND online issue! Of course, we're a little late (again) with it! Better late than never they say. The World Wide Web is a most interesting place! Since we went online and eliminated paper, we have been adding stuff to this site in an effort to attract visitors. Most of our new features have been well received. Our hit rate has gone from the low end at 20 or so hits per day in early January to over 200 per day recently. It's not a top 5% site yet -- but we're growing! The Web is so interesting, it never ceases to amaze me what you can find here. To help our readers / visitors find their way around the Web, in their search for valuable information, we have added a number of new SSS Online Subject areas, including:
In addition, every week we update and add new material to the following three Subject areas:
All of these additions were made to help you out! If you would like to see something else added here -- just email us: randyrf@sss-mag.com. We will compare your request with others and probably add something that will solve your problem or appeal to your interests. SSS Online is here to do two things:
New ProductsRockwell 900 MHz Digital Cordless Telephone Chipset Product ReviewRockwell's state-of-the-art 900 MHz Direct Sequence Spread Spectrum DCT solution provides all the components you need for long-range, secure cordless telephone communications--with minimum development time and costs. As an OEM, you'll appreciate Rockwell's complete solution when it translates into lower total system cost and faster time to market. By providing both baseband and RF devices, system integration is minimized. Your customers will appreciate the features--longer range, superior voice quality, higher security and extended talk time. Spread Spectrum is a coding technique for digital transmission which uploads the information of a data signal over a wide frequency range, making the signal look like noise. Data integrity, resistance to interference, and security are ensured using this technique. Also, if the spread spectrum is used, the FCC allows more power to be transmitted--which translates into longer range. The power management of Rockwell's solution (baseband and RF) allows you to design phones with 4 hours of talk time and 7 days of standby time. 16 million random security codes pose quite a challenge for eavesdroppers who want to hear the conversation. Baby monitors and garage door openers will no longer disturb your customers' phone conversations. Plus: No more channel changing. Rockwell's technology does it for you, automatically finding the best channel at any given time. All of this, plus Rockwell's proven support and development tools in place, ensures the lowest total system cost and the quickest time to market. And it's a safe investment because you know Rockwell is a company you can count on to grow with you well into the future.
Rockwell's SS DCT "Block Diagram" SS Chipsets Simplify SS DesignBy Randy Roberts, RF/SS Consulting In the last year or so a number of new chipsets for SS signal processing have been announced. Some of these chips are very good -- some are not so GREAT! How do you choose the GOOD from the BAD? A few years ago CYLINK and O'Niel Comunciations (now out of business) marketed their post detection de-spreading ASICs. I think quite a few people jumped on those chips -- but, were they in for a surprise! Not only did the FCC change Part 15 SS rules, but shipping these products almost gave SS wireless products a bad name! Does the latest generation of SS chipsets really work? Are these chips easy to use? Will these new designs find customer acceptance in the marketplace? This article will attempt to answer some of these questions. SSS Online will try to save you time, trouble and perhaps even some real embarassment with this preview! New SS chipsets are available from IMEC. Stanford Telecom AMI and ATMEL / Wireless Logic also have come out with SS Chipsets for specialized, niche markets. Seattle Silicon and others have had baseband chipsets out for some time. There are a lot of chips on the market, aren't there? We at RF/SS have designed a couple of DS SS radios with the the STEL-2000A and find it a very useful, if not difficult chip to use. It is capable of baseband data rates up to 2.048 Mbits per second (E1 speed) and can do PSK/DPSK/QPSK/DQPSK. While not for newcomers to the SS world, this chip PERFORMS! RF/SS has been talking with Harris and is trying to become a Technical Design Rep for them. They have loaned us an HSP3824 evaluation board, a handful of sample chips and have promised evaluation boards for their other chips. In the next few weeks we will have completed a full evaluation this promising new chipset and we will report our results right here! IMEC has also talked to us about being a sort of Technical Design Rep for them. They have formed a new company, SIRIUS, to market their new Baseband SS chips in the USA.The specs on their chips are superb! They also have promised us a set of evaluation boards for their chips. I'm hoping to complete a thorough evaluation of their stuff in the next month. A recent client of ours has asked us to help him finish his wireless data radio design which uses the ATMEL / Wireless Logic digital cordless chipset. We are still busy helping him out, but our first impression is stay away from it! Wireless logic over a year ago promised us evaluation boards, etc. and NEVER Delivered! We do have the complete data package on this chipset -- but their design includes a great deal of "Proprietary" firmware for an 80C51 generic microprocessor needed to supervise / run their chip. I have heard from a number of people that this chipset just does not live up to all the hype that's been published about it. Besides, ATMEL doesn't answer their phone most of the time and DOES NOT support this product. The ONLY support available for the chip is from Wireless Logic in San Jose, when they are not in Taiwan! I've heard great things about the Rockwell DCT chipset reviewed above. But when I tried to get price and delivery information I called four diferent places -- I still did NOT get the information I needed! Of course, I wanted pricing only for quantities up to 5000 per year or so -- I probably didn't even trip their sales least significant bit! Still the Rockwell chipset appears to be a real technical performer for DCT type applications -- you should check it out! QUALCOMM and OMNIPOINT are also developing (or have developed) CDMA SS chipsets, to meet their respective IS standards -- I understand that QUALCOMM is peddling theirs in Japan -- maybe before long they'll both find out there is a market right here! AMI is selling various versions of its basic SX043 SSIC -- a part of its WavePlex family of wireless products. Other than being slightly high priced, these chips seem to be very good! AMI sells a set of evaluation boards for these chips -- so early design familiarity can be gained almost painlessly. Expect to pay $500 or so for the eval boards, but they do include software and are easily interfaced to PCs! Total functionality and flexibility is built in to these chips. You can buy a full duplex transceiver chip or seperate half duplex transmit and receive chips from AMI. This chipset's design is very well thought out and very useful across a wide range of data rates, processing gain and market niches. I predict that if AMI finds the right price point, these chips will be real winners! The job of the SS system designer has certianly been made easier with the introduction of these new SS chipsets. But, as always -- CAVEAT EMPTOR (BUYER BEWARE)!!! Some of the available chipsets are more than a challenge to make work! If I had to recommend a direction for choosing one of these new chipsets -- I would go with an established vendor who has good support and / or a second source! Good luck -- gentlemen, fire up your SS Design Engines . . .
Contributions: Data Randomizing with Pseudo-Noise Coding TechniquesProviding important information for today's SS system designer, this article is based on Mr. C. R. Netherton's original application note which was done in the early 1980s.By C. R. Netherton, Decom Systems The PN generator is an n-stage shift register, employing modulo-2 (half adder) feedback, as shown in Figure 1. NRZ data and the PN generator are multiplied, as shown in Figure 1, to inform the randomizer and undergo the inverse divisidn process, in an identical decoder, to form the derandomizer. Figure 1: PN Data Randomizer & De-randomizer.The data transmission code remains NRZ so that the optimum 0.7 bandwidth to Bit Rate ratio is maintained. The PN encoding process is a 100% efficient code, in that there are no added bits and all transitions occur in phase, with a times one clock. Typically, an 11 bit, 2047 bit (2^11 - 1) PN sequence is utilized in the randomizing process. Other lengths can be used, but, generally, they do not provide any significant performance increases. The arithmetics of the PN data process are straightforward. The commutative property of an "EXOR" process enables the derandomizer to synchronize to the incoming data within (12) data bits, regardless of the initial power up-state of the individual randomizer and derandomizer shift registers. Obviously, nothing is generally gained without some penalty. PN randomizing of NRZ data does suffer from an error multiplication process, which essentially translates into about a 0.5 dB Bit Error Rate loss, versus conventional NRZ. However, when you consider the increase in data packing denisty on tape, the virtual elimination of "baseline gallop" and the increased ability to operate at much lower synchronization thresholds and data transition densities, one must consider the 0.5 dB penalty a small price to pay, especially since other highly publicized codes require substantially higher data bandwidths and bit error degradations. In today's DS system designs these PN randomizing techniques provide a convenient means of meeting the FCC's rules at the small cost of having to re-synchronize the randomizer, if several uncorrected bit errors in a row are generated in the receiver. References: C. R. Netherton, "An Optimum Use of Bandwidth Via Pattern Conditioning", Decem Systems, Inc., 1977, San Marcos, California C. A. Castle & J. H. Stein, "Data Randomizing Shrinks HDR Recording Hardware', Sangamo Weston, Inc., 1977, Springfield, Illinois E. L. Law, Experimental Comparison of Pulse Code Modulation Codes for Magnetic Recording", Instrument- tation Department, Pacific Missile Test Center, 1977, Point Mugu, California D. A. King, "Comparison of PCM Codes for Direct Recording", International Telemetering Conference, 1976, Los Angeles, California
|