Site Home Archive Home FAQ Home How to search the Archive How to Navigate the Archive
Compare FPGA features and resources
Threads starting:
Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Howdy, I am using a Altera 81188 in one of my designs. I was able to acheive downloading of the data via a serial interface to a microprocessor. There were several things that I found out. 1) The pullup resistors are critical for proper operation. They are listed in the Altera FLEX 8000 guide. I kept forgetting these pullups. 2) Approx. 11 extra clock cycles were needed after the data was loaded into the FPGA before the device became functional. One would think that the Bit Blaster does this for you. 3). you can look at the validity of your data to the FPGA by examining the nSTATUS signal. Originally I had the data in the wrong format so it was being downloaded wrong into the FPGA. This signal gave me a clue as to what was wrong. Hope this helps Ralph Watson rwwatson@ti.comArticle: 1926
In Article <michaelk.1161931222A@news.mcgill.ca> lines deleted >Could someone refer me to some references or sites where I can get info >on what LFSR counters are? I'm building a system using the Xilinx 4000 >series that requires fast FSM's and this would help a lot. Also, would >there be any sites that contain VHDL code for these counters? Thanks >in advance. > LFSR counters are Linear Feedback Shift Register COunters, also known as psuedo random number generators because of the apparently random sequence they generate. The sequence is really predictable and repeatable. The counters are constructed by either XOR or XNOR combining selected taps off the shift register and feeding the result back to the input. The sequence goes through (2^n)-1 states before repeating where n is the number of registers in the shift register. The result is a physically small counter with very little combinatorial logic that can run close to the ff toggle rate if laid out right. Xilinx has an app note for these counters on page 9-24 of the 1994 third edition data book. I think there are additional app notes on their web page. The data book app note includes a table showing the correct taps for counters up to 40 bits. You must use the correct taps to get a maximal length sequence. Counts other than the maximal length are easiest to obtain by decoding the desired state(s). Since the count sequence is a little fruity, you will need to determine the state for a particular count. That's easy to do by hand for the smaller counters. A simple program can be used to generate the longer sequences or even the state at a particular count. The four and five bit counters are particularly useful in combination with state machines, as they can be decoded in one level in a xilinx 3K. System speed can be maintained by pipelining the decode. See, you learn something new everyday! Best of luck. -Ray Andraka Chairman, the Andraka Consulting Group 401/884-7930 FAX 401/884-7950 email randraka@ids.net The Andraka Consulting Group is a digital hardware design firm specializing in obtaining the maximum performance from FPGAs. Services include complete design, development, simulation, and integration of these devices and the surrounding circuits. We also evaluate, troubleshoot, and improve existing designs. Please call or write for a free brochure.Article: 1927
I implemented a simple VRAM control signal state machine in a 4003-6 the other day, and just for grins started swapping crystals to see how fast it would run. It ran up to 107 Mhz, at which point I ran out of crystals. Verifying operation consisted of looking at signals with a logic analyzer. The "state machine" is a shift register with decoded outputs, but as a relative FPGA newbie I was very surprised. -- Rainer M. Malzbender Senior Research Physicist rainer@displaytech.com Displaytech, Inc. 303.449.8933 2200 Central Ave. Boulder, CO 80301Article: 1928
Take a look at "http://memec.com/DataSheets/Atmel/Prodg/prodg.toc.html". Scott Evans (chan_isd@ix.netcom.com) wrote: : In article <dimitris-1109951303030001@thor.engn.uwindsor.ca>, : dimitris@engn.uwindsor.ca (Dimitris Phoukas) wrote: : >Does anybody if ATMEL has a WWW server? : >Thanks in advance : > : There is not one yet (it's in the works). : Meanwhile they can be contacted at : e-mail: fpga@atmel.com : phone: 408 436 4119 : Scott Evans -- --- Doug Smith University of Cincinnati dsmith@ece.uc.edu Electrical & Computer Engineering "Fear God, and keep his commandments: for this is the whole duty of man." -King Solomon ---Article: 1929
The web site for CICC for more info is: http://uivlsi.csl.uiuc.edu/~brian/cicc.html ======================================================================= CALL FOR PAPERS IEEE Custom Integrated Circuits Conference .... the leading edge of ICs San Diego, CA May 5 - 8 ,1996 ========================================================================== CICC is the premier conference devoted to IC development. It provides a forum for circuit designers, CAD developers, manufacturers and ASIC users to present and discuss exciting new developments, future trends and innovative ideas. It provides an unique platform and an opportunity for the CAD developers and the designers to interact and exchange ideas for developing leading edge CAD solutions to the many design problems. CICC is sponsored by the IEEE Electron Devices Society in cooperation with the IEEE Solid State Circuits Council. CONFERENCE HIGHLIGHTS Technical Papers: The focal point of CICC: approximately 140 papers addressing a broad range of leading-edge circuits, applications, design techniques and tools will be presented. Instructions for potential contributors are given below. Panels: Discussions and debates by industry leaders on key issues, problems, solutions and potentials of the IC industry. Educational Sessions: A full day of extended tutorials offer a unique and valuable opportunity for practicing professionals to get acquainted with aspects of ASIC development outside of their specialties. Exhibits/Exhibit or Sessions: CICC includes extensive exhibits by ASIC, EDA and test/measurement vendors. Exhibitors are also invited to give presentations on new products and capabilities in special exhibitor sessions. Contact the Exhibits Chairman for details. SUBMISSION OF PAPERS Deadline December 6, 1995 Prospective authors must submit CAMERA-READY papers, up to four pages in length inclusive of all illustrations, charts and tables. Those interested in submitting papers should contact the Conference Manager as early as possible to obtain an author's kit and detailed instructions. The address is: Melissa Widerkehr, CICC, 101 Lakeforest Boulevard, Suite 270, Gaithersburg, MD, 20877, Telephone: (301)-527-0902, Fax: (301)-527-0994. Email: Widerkehr@aol.com ACCEPTED PAPERS WILL BE PRINTED IN THE PROCEEDINGS WITHOUT OPPORTUNITY FOR FURTHER CHANGE. The paper should report original and previously unpublished work, including specific results. Circuit oriented work must include measured experimental results. Deadline for RECEIPT of technical papers is December 6, 1995. Appropriate company and government clearances MUST be obtained prior to submission. After receipt of the author kit, the completed original and 25 copies should be sent to the Conference Manager at the address above. Authors of accepted papers will be notified by mail by January 31, 1996. PAPERS IN THE FOLLOWING AREAS ARE REQUESTED: Analog and Mixed-Signal Circuits and Systems: Data convertors, modulators, filters, high speed analog, low voltage techniques. Mixed analog-digital IC applications, disk read/write channels, RAMDACS. Digital Signal Processing: Image compression, recognition, and enhancement particularly for digital video. Voice coding and recognition. Specialized processing function architectures. Digital filtering, encryption, HDTV, video conferencing, multimedia, graphics controllers, video drivers. Communications: Data, voice and image transmission. Digital, analog modulation, equalization, error correction, coding, switching. SONET/SDH, ISDN, LAN/WAN/ATM and broadband applications. Wireless and RF submissions are encouraged. Custom Application Specific Circuits: Innovative designs for cell-based circuits, full custom ICs and ASIC memories. Novel design concepts, low power low voltage circuits, high-performance circuits, architectures or system applications (automotive, bio-medical, etc.) are of particular interest. Gate Arrays and Programmable Devices: GA, EPLD, FPGA, PAL, PLA devices, circuits and architectures, as well as the CAD tools targeting these devices. Device applications are encouraged. Design Automation - Design Capture:Logic and high-level synthesis and optimization, innovative design capture techniques, frameworks, user interfaces. Design Automation - Physical Design: Silicon compilation, module generation, symbolic layout, automatic place and routing techniques, circuit parameter extraction and circuit verification. Simulation and Modeling:Device or process modeling, circuit, functional, timing or logic simulation with emphasis on analog modeling, multi-chip modules and mixed-signal simulation. Submissions are encouraged in the power estimation and analysis areas for low-power design, RF modeling and simulation; and interconnect and packaging issues for simulationa and modeling. Fabrication and Assembly: Developments in process integration, quick-turn manufacturing, MOS, bipolar, BiCMOS, smart power and GaAs, multichip modules, package modeling, ESD protection, fiber optic tranceivers. Testing and Reliability: Advances in design-for-testability (DFT), fault modeling and grading, IDDQ measurements, parametric characterization, high speed or high frequency measurement techniques and failure analysis. Library Development and Design Methodology: Cell library generation and characterization, reusable and customizable megacells, Algorithim to Silicon design flows. FURTHER INFORMATION For complete author kit instructions, registration information and general inquiries contact the Conference Manager: Melissa Widerkehr, Custom Integrated Circuits Conference, 101 Lakeforest Boulevard, Suite 270, Gaithersburg, MD, 20877, Telephone: (301)-527-0902, Fax: (301)-527-0994. Email: Widerkehr@aol.com Further information can also be obtained from: Bob Cordell, Technical Program Chair, Bellcore, 331 Newman Springs Road, Room 3Z-307, Redbank, NJ 07701, telephone: (908)-758-2963 email: rrc@nyquist.bellcore.com OR Jake Buurma, Conference Chair, 7115 Raich Drive, San Jose, CA, 95120, telephone: (408)-268-2576. email: jake.buurma@taec.com Potential Exhibitors should contact the Exhibits Chair: Brian Fitzgerald, IBM Corporation, 404 Wyman Street, Waltham, MA, 02254, telephone: (617)-895-1338. email: fitzgerald@vnet.ibm.com See the CICC'96 WWW site for more information at : http://uivlsi.csl.uiuc.edu/~brian/cicc.html ==========================================================================Article: 1930
Does anyone know where I can find a SBus DMA (or DVMA) controller for a XC4003A? In fact, I'm not sure whether I need a DMA controller on my SBus slave card or whether the Sparc motherboard's DMA controller can perform all of the DMA transfers. Does anyone know about SBus DMA controllers or where I can find more information on them? Thanks. --- Doug Smith University of Cincinnati dsmith@ece.uc.edu Electrical & Computer Engineering "Fear God, and keep his commandments: for this is the whole duty of man." -King Solomon ---Article: 1931
michaelk@macs.ee.mcgill.ca (Michael C. Kim) wrote: >In Article <43d28f$ij8@paperboy.ids.net>, randraka@ids.net wrote: > >lines deleted > >>The trick is to pay attention to the device architecture when you do the >>design. Keep the combinatorial logic to one level whenever at all possible >>(yes this is possible more than is obvious...pipeline your decodes and state >>machines). Use OHE statemachines, and LFSR counters for timers. Without doing > ^^^^^^^^^^^^^^^^^^^^^^^^ > >Could someone refer me to some references or sites where I can get info >on what LFSR counters are? I'm building a system using the Xilinx 4000 >series that requires fast FSM's and this would help a lot. Also, would >there be any sites that contain VHDL code for these counters? Thanks >in advance. > >Michael > > LFSR stands for linear-feedback shift register. Essentially, an LFSR counter is a shift register with an XOR-style feedback path. Generally, LFSRs count to 2^n - 1 possible states instead of the 2^n for binary counters. By decoding a specific counter state and feeding it back into the XOR feedback, you can make an LFSR cycle through a predetermined series. LFSRs are a bit more trouble to design, but they are fast and very silicon efficient. If you would like additional information, check out the application note on LFSRs located on the Xilinx webLINX site at http://www.xilinx.com/products/appsweb.htm#FPGA Or, if you would prefer an encoded, compressed PostScript file, just send an E-mail to 'xdocs@xilinx.com' with 'send 80015' in the Subject header. -- Steven K. Knapp Corporate Applications Engineer Xilinx, Inc.Article: 1932
Evagelia Diamantakou <evd@sn2.ee.umist.ac.uk> wrote: >Hello , > Does anyone know if I can simulate a sheet which >consists of both symbols taken from Xilinx XC3000 library >and symbols based on VHDL description.If you think I can , >do you know which sequence of steps I sould follow to >simulate the sheet in Mentor Graphics QuickSimII ? >Please e-mail me any replies . >Thanks , > Evagelia Diamantakou . > By using the synthesis tool, generate the xnf file for the block described in VHDL, and then assign the FILE attribute to that block in your schematics. Once you have generated an xnf file for that block, I think you can simulate your design without any additional step. Good luck, Yuce Beser "speaking for myself"Article: 1933
Hi, Could you tell where I can get this VHDL-sensitive editor ? JG -- ********************************************************** Jorgen Gade ABB Industrial Systems AB jgade@hw.seisy.abb.se **********************************************************Article: 1934
Hello , I sent a question with the title "Simulation using XC3000 libraries" . What I forgot to mention is that I would like to do only FUNCTIONAL simulation to a sheet . I know that to do timing simulation , first I need to do place and route which means that VHDL has to be translated to XNF . Is it the same for functional simulation ? Please e-mail any replies . Thanks , Evagelia Diamantakou .Article: 1935
In article <43k049$549@moe.tas.drs.com>, djg@tas.com (David Gesswein) wrote: >In article <43js68$o01@merlion.singnet.com.sg>, <jothi@singnet.com.sg> wrote: >We tried the 128 but it is still pre-production and has several bugs in it >which prevent using it with the XC52xx and possibly all Xilinx chips. > The 17c128 version works fine with XC40xx parts - it does seem to be a bit slower than the OTP version from Xilinx. >We also have used the AT&T reprogramable which did work but does not >reliably load. I think it is power ramp up sensitive, on one chases it >almost always loaded and in another it never would load. The Xilinx EPROM >part always loaded. I don't think they have a 256 out yet either. > Atmel said the 256 version should be out by December.Article: 1936
Just in case you are interested, Altera is having some sort of seminar on their new FLEX 10K family. Altera claims this family has 100,000 usable gates with BIG ram blocks. You can check it out on their WWW site. TArticle: 1937
SECOND INTERNATIONAL SYMPOSIUM ON ADVANCED RESEARCH IN ASYNCHRONOUS CIRCUITS AND SYSTEMS Sponsored by: The IEEE Computer Society Technical Committee on VLSI The IEICE Technical Group on VLSI Design in cooperation with: IFIP WG 10.5 The University of Aizu, Japan AIZU-WAKAMATSU, FUKUSHIMA, JAPAN MARCH 18 - 21, 1996 *********************** * CALL FOR PAPERS * *********************** The Second International Symposium on Advanced Research in Asynchronous Circuits and Systems is being organized as a forum for disseminating advanced research results on asynchronous circuit and system design. Asynchronous systems design is currently enjoying a strong revival, and many clear directions for further research have begun to emerge. A major goal of this symposium is to articulate emerging focal areas through invited papers, and thereby foster research growth. Special emphasis will be given to industrial applications and mature CAD tools. To this end, papers are being sought on all aspects of asynchronous circuit and system design, as well as in the following focal areas (in no specific order of priority): --- Industrial Applications of Asynchronous Design --- Hybrid Synchro-Asynchronous Systems --- CAD Tools for Asynchronous Design --- Testing/Design-for-Testability of Asynchronous Circuits --- Noise Immunity in Asynchronous Digital Circuits --- Low-Power Techniques --- High Performance Asynchronous Circuits --- Hazard-Free Logic Synthesis and Optimization --- Design of Asynchronous State Machines --- Metastability and Related Issues --- Formal Methods for Asynchronous Design --- Petri-Nets, Trace Theory and Related Formalisms --- High-Level Synthesis of Asynchronous Circuits and Systems --- Asynchronous Design in Commercial HDL Environment --- Verification of Asynchronous Circuits and Systems --- Asynchronous Datapath and Processor Design --- Practical Design Examples --- Future Trends (Optical Computing, Neural Networks, Nanotechnology, etc.) Submissions: - ----------- Submitted papers should be no more than 15 pages in 11-point font (or equivalent) with a 60-word abstract, and should include a cover page with authors' physical and e-mail addresses, phone and FAX numbers. Prospective authors should submit papers in one of the following forms: (1) Send six single-sided copies of the manuscript with a cover-page to: Alexander Taubin (Async96) THE UNIVERSITY OF AIZU Tsuruga, Ikki-machi, Aizu-Wakamatsu City Fukushima, 965-80 Japan (2) Email compressed Postscript(tm) files to "async96@u-aizu.ac.jp". Accepted papers will appear in a published proceedings. Important Dates: - --------------- Papers due: October 10, 1995 Notification of acceptance by : December 1, 1995 Final Version by: January 10, 1995 Symposium Officers - ------------------ General Chair Tosiyasu Kunii (The University of Aizu, Japan, kunii@u-aizu.ac.jp) Conference Co-Chairs: Takashi Nanya (Tokyo Institute of Technology, Japan, nanya@cs.titech.ac.jp, tel. +81-3-5734-3041, fax: +81-3-5734-2817) Alex Kondratyev (The University of Aizu, Japan, kondraty@u-aizu.ac.jp, tel. +81-242-37-2557, fax: +81-242-37-2744) Program Co-Chairs: Luciano Lavagno (Politecnico di Torino, Italy, Cadence Berkeley, USA, lavagno@polv2k.polito.it, tel. +39-11-5644150, fax: +39-11-5644099) Alexander Taubin (The University of Aizu, Japan, taubin@u-aizu.ac.jp, tel. +81-242-37-2572, fax: +81-242-37-2744)) Publication Chair: Takeshi Yoshimura (NEC, Japan, yoshi@swl.cl.nec.co.jp) Publicity Chair: Fumiyasu Hirose ( Fujitsu Laboratories Ltd, Japan, hirose@flab.fujitsu.co.jp) Finance/Registration Chair : Kazuaki Yamauchi (The University of Aizu, Japan, yamauchi@u-aizu.ac.jp) Local Arrangement Chair: Yuko Kesen (The University of Aizu, Japan, kesen@u-aizu.ac.jp). Tutorial/CAD booth Chair: Michael Kishinevsky (The University of Aizu,Japan, kishinev@u-aizu.ac.jp) Industry relations: Tsuneo Ikedo (The University of Aizu, Japan, ikedo@u-aizu.ac.jp) Masatoshi Sekine (Toshiba, Japan, sekine@srd.ull.rdc.toshiba.co.jp) US industry representative/chair: Alan Davis (University of Utah, USA, ald@endo.cs.utah.edu) US academy representative/chair: Erik Brunvand (University of Utah, USA, elb@telemark.cs.utah.edu) European representative/chair : Steve Furber (University of Manchester,UK, sfurber@cs.man.ac.uk) Program Committee: Kunihiro Asada (Japan), Graham Birtwistle (UK), Steven Burns (USA) Tam-Anh Chu (USA), Jordi Cortadella (Spain), David Dill (USA) Jo Ebergen (Canada), Ran Ginossar (Israel), Ganesh Gopalakrishnan (USA) Mark Greenstreet (Canada), Mark Josephs (UK), Sadatoshi Kumagai(Japan) Bill Lin (Belgium), Alain Martin (USA), Viacheslav Marakhovski (Japan) Charles Molnar (USA), Steven Nowick (USA), Takuji Okamoto (Japan) Martin Rem (Netherlands), Jens Sparso (Denmark), Robert Sproull (USA) Pasupathy Subrahmanyam (USA), Jan Tijmen Udding (Netherlands), Stephen Unger (USA) Peter Vanbekbergen (USA), Kees van Berkel (Netherland), Victor Varshavsky (Japan) Alex Yakovlev (UK), Tomohiro Yoneda (Japan) -- Kind regards Alexander Taubin THE UNIVERSITY OF AIZU phone +81-242-37-2572 (office) Tsuruga, Ikki-machi, Aizu-Wakamatsu City fax +81-242-37-2744 Fukushima, 965-80 Japan e-mail taubin@u-aizu.ac.jp ---- <A HREF="http://www.u-aizu.ac.jp/~taubin/">------------------------Article: 1938
Evagelia Diamantakou <evd@sn2.ee.umist.ac.uk> wrote: >Hello , > I sent a question with the title "Simulation using XC3000 libraries" . >What I forgot to mention is that I would like to do only FUNCTIONAL >simulation to a sheet . I know that to do timing simulation , first >I need to do place and route which means that VHDL has to be >translated to XNF . Is it the same for functional simulation ? >Please e-mail any replies . >Thanks , > Evagelia Diamantakou . > You do not need to do placement&routing in order to have an xnf file for the block that is described in VHDL. After synthesis, and before placement&routing, you will have an xnf file (generated by the synthesis tool) which you can use for functional simulation. Good luck, Yuce Beser "speaking for myself"Article: 1939
In article <43lu7d$2oc@btmpjg.god.bel.alcatel.be> Yuce Beser <yuce@sh.bel.alcatel.be> writes: >There is a program called "xbloxgen" from Xilinx which is used to instantiate >xblox modules in the vhdl/verilog code. You input which xblox module you want >to instantiate in your code, and it generates the xnf file and the vhdl/verilog >code (component definition, and instantiation lines) for that xblox module. You >can use this tool to instantiate the xblox modules: clk_div, shift, inc_dec, >accum, add_sub, compare, counter, decode, data_reg. Where would I get that program from? Does anybody have pointers? m. -- Michael Gschwind, Institut f. Technische Informatik, TU Wien snail: Treitlstrasse 3-182-2 || A-1040 Wien || Austria email: mike@vlsivie.tuwien.ac.at PGP key available via www (or email) www : URL:http://www.vlsivie.tuwien.ac.at/mike/mike.html phone: +(43)(1)58801 8156 fax: +(43)(1)586 9697 Boycott Whaling!!! Boycott Norway!!! Boycott Norwegian Products!!!Article: 1940
In article <43mj0t$bb8@bcarh8ab.bnr.ca> qzhang@bnr.ca (Qian Zhang) writes: >I had slack time in the range of -30 ns, but the design worked after >fitting. > >Synopsys can still be used as an optimizer to do some common sense >optimization. > >I read that for Xilinx parts, Synopsys maps logic to CLBs, and the >timing reports are more accurate. Yup, FPGA compiler (used for the 4k series) maps to LUTs... I dunno for the 3k series though, which uses the Design Compiler (you can also use DC for the 4k series, not idea on that either...) I have no idea how Synopsys decide which chips are supported by FPGA compiler, and which ones get the DC treatment... As for the accuracy of timing reports: Synopsys seriously underestimates routing delay, so you always seem to end up with slower chips after routing :-( I'd rather have a speedup from routing, as experienced with Altera ;-) m. PS: Hi Qian! -- Michael Gschwind, Institut f. Technische Informatik, TU Wien snail: Treitlstrasse 3-182-2 || A-1040 Wien || Austria email: mike@vlsivie.tuwien.ac.at PGP key available via www (or email) www : URL:http://www.vlsivie.tuwien.ac.at/mike/mike.html phone: +(43)(1)58801 8156 fax: +(43)(1)586 9697 Boycott Whaling!!! Boycott Norway!!! Boycott Norwegian Products!!!Article: 1941
Dear Friends, I have read many posts about LFSR (linear feedback shift register) on this net. I myself by chance had read the Actel's FPGA Design Guide II, December 1992, in which a nice decription of LFSR's application in non-linear counter and random number generator is discussed. I'd like to outline the details: For a linear 4-bit counter, it counts in the following sequence 0000 0001 0010 . . . 1110 1111 0000 But with LFSR, one can get the following sequence: (you can start from any value excepting 0000) 1111 0111 0011 0001 1000 0100 0010 1001 1100 0110 1011 0101 1010 1101 1110 1111 The skill is to feedback the XOR(MSB + bit[n1] + bit[n2]...) to LSB at every cycle. Depending on the size of the register, different n1, n2 ... are needed. Here is a summary: (M=regster size, K=possible value) m=2 k=3 n=1 m=3 k=7 n=2 m=4 k=16 n=3 m=5 k=31 n=3 m=6 k=63 n=5 m=7 k=127 n=6 m=8 k=255 n=3,4,5,7 m=9 k=511 n=5 m=10 k=1023 n=7 m=11 k=2047 n=9 m=12 k=4095 n=1,9,10,11 m=13 k=8191 n=6,10,11,12 m=14 k=16383 n=1,11,12,13 m=15 k=32767 n=14 m=16 k=65535 n=10,12,13,15 Regards, Felix K.C. CHENArticle: 1942
In article <43jihk$6k0@server.cpr.it>, Maurizio Lippi <lippi> writes: > We have been informed by the manuf. AMD that the production of this device will > be stopped in the coming weeks. does anyone know a second source of this pld? > or a similar device? > the major features of this pal are: > - High output current drive capability (64ma Iol) > - Programmable Totem-Pole or Open Drain Outputs > - 200 mV Histeresis > - Programmable Direct or latched Inputs. > > regards, Maurizio Lippi > > +------------------------------------------------------------------------+ > + Maurizio Lippi R&D Division + > + + > + CAEN SpA + > + Via Vetraia, 11 50049 VIAREGGIO (ITALY) + > + Tel. +39 584 388 398 Fax +39 584 388 959 + > + E-Mail: lippi@caen.it URL: http://www.caen.it + > +------------------------------------------------------------------------+ > When we looked for a part with Hystersis, this was the only one we found. YOu post explains why we never got a straight reply from AMD on it. This part has a strange pin-out, so the odds of a drop in are almost zilch. Let me know if you find anything, as we would like a part with Hysteresis. Few PLD suppliers seem to understand Hystersis, or even low power in their rush for ever faster parts.... We are working on this problem - If anyone else could really use a PLD with Hystersis, please email details of PLD Size / Speed / Power / Other features. jim granvilleArticle: 1943
In article <43fgn7$45f@news.Belgium.EU.net>, Jan Decaluwe <jand> writes: >jcooley@world.std.com (John Cooley) wrote: > > >>I personally tend to see hardware when I design hardware, > >This argument is widely used against design methodologies that rely >on higher levels of abstration. I believe it is misleading. It is the >same argument that schematic entry addicts use(d) against RTL synthesis. >However, I feel safe to assert, John, that you're a convicted synthesis user!! > Am I correct if I understand that you indicate that the designer should not tend to see HW while designing, but rather focus on modeling the functionality at as high abstraction level as possible? If so, I could not disagree more. I would say that it is necessary to see hardware architecture when you design. Otherwise, you are incapable of making acceptable design solutions. Performance and product properties are determined by choise of algorithm, architecture (at many levels) and technology. Power dissipation/consumption is a physical activity, and in order to fulfil power requirements I would say it is necessary to see HW while designing. Prediction on power dissipation must be made for all architectural decisions and in order to make such prediction you must see hardware. If not you may end up with drastic violation on important properties. (A cellular telephone with a talk time of 5 min would probably be hopless to sell even for Microsoft :-). Robert TjarnstromArticle: 1944
I just wanted to thank everyone who helped answer my question regarding what LFSR's were. This will prove helpful for me since my design will have to be fast (approx 20 MHz) and I need to eke out all the speed I can get from my smaller modules. Also, the postings on one-hot encoding for FSM's was also helpful. MikeArticle: 1945
twtmail@twt.co.il wrote: > > > I'm using an EPM5016 in a small project. > I'm using 4 i/o pins for 2 NOT gates osc. > > The component is getting very very hot (after about 5 min.) > > Does anybody knows why? > > Thanks in advance. > > R.H > I've used the 5032'2 and they run hot also. This is due to the quiescent current requirements. Also depends on how fast you system clock is running. Check the graph in the data book "Current vs. Frequency", the amount of current times Vcc is approx. the dissapated heat. I've never burned myself on one yet although using the ceramic package, I have got them fast/hot enough to be uncomfortable. RogArticle: 1946
> >We tried the 128 but it is still pre-production and has several bugs in it > >which prevent using it with the XC52xx and possibly all Xilinx chips. > > > > The 17c128 version works fine with XC40xx parts - it does seem to be a bit > slower than the OTP version from Xilinx. > > >We also have used the AT&T reprogramable which did work but does not > >reliably load. I think it is power ramp up sensitive, on one chases it > >almost always loaded and in another it never would load. The Xilinx EPROM > >part always loaded. I don't think they have a 256 out yet either. > > > > Atmel said the 256 version should be out by December. You mean ATMEL & AT&T sell parts which does not work reliably?? I feel you should have a look at the data sheets and design it rather than just replace an IC with pin for pin replacements.Article: 1947
>I run into a problem when designing an Altera flex8452-3 with VHDL. After [...] >I was wondering if anyone else run into the same problem. Your comments will >be much appreciated. > >Regards, > >Louis >c Same problem here. We now ignore the Synopsys timing violations and depend on the timing report from maxplus2. And yes, I worry that someday Synopsys will change the design in a very unproductive way (because it imagines a time violation) and maxplus2 can not straighten it out. ======================================================================== Richard A. Springer UUCP/USENET: ricks@aruba.WV.TEK.COM Tektronix, Inc. US Mail: P.O. Box 1000, MS 63-356 Color Printing & Imaging Div. : Wilsonville, OR 97070, USA (503) 685-3535 FAX: (503) 685-4901 ========================================================================Article: 1948
Folks, I would like to clarify Atmel's Serial EEPROM offering. Atmel offers an EEPROM version of the 17Cxxx series FPGA support serial programming proms. The AT17C65 and the AT17C128 are currently shipping in production volume and the AT17C256 will be sampling late Q4 early Q1. The Atmel EEPROM serials are pin compatable with the OTP EPROM versions currently offered in the market-place and have the added feature that they can be in-system (re)programmed (5V) via an industry standard 2-wire interface (24CXXX like). The parts differ very slightly from the OTP EPROM parts as follows : Cascade mode is not supported on the AT17C65 parts. Cascade mode will be support from End Nov. '95 on the AT17C128 parts. Clock must be low during Reset active. Pin 7/17 is used as the 2-wire interface enable pin (active low) as it is not required for a 12V programming voltage (Vpp). When this pin is high the parts work as a 17CXXX part. The serial parts do support programmable Reset/OE polarity. The parts are supported by (most) industry standard programmers. The AT17CXXX serial parts are design to have the same fundumental performance capability as the OTP EPROM parts. If you would like more information on these parts, please send an e-mail with your SNAIL MAIL ADDRESS to martin@atmel.com and a databook will be yours by return. Martin. ------------------------------------------------------------------------- | Martin Mason | Snr. FPGA Applications Engineer | | Atmel Corp. | (Work) martin@atmel.com | | 2125 O'Nel Drive | (Work2) fpga@atmel.com | | San Jose | (Fax) + (408) 436 4300 | | CA 95131 | (Tele) + (408) 436 4178 | -------------------------------------------------------------------------Article: 1949
We've used 8820As with both download (bitblaster) and serial eproms. It is important that the download cable header be close to the part and traces run directly. More than an inch separation and we could not get reliable bitblaster down loads. Serial eprom downloading seems to be more robust. Good luck!
Site Home Archive Home FAQ Home How to search the Archive How to Navigate the Archive
Compare FPGA features and resources
Threads starting:
Authors:A B C D E F G H I J K L M N O P Q R S T U V W X Y Z