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
Hello I am looking for good IC Schematics to Chip Layout Engineer and am unable to find one. Does any one know of good layout engineer in Silicon Valley. San Francisco Area Thanks bharat 408 567 3435 Bharat.Kurani@add.ssw.abbott.comArticle: 6751
All attempts to install Foundation software to a Jaz drive unsuccessful. If anyone has ever succeeded in doing this I would appreciate learning how it is done. Thanks. Stephen Vale vsrd@airmail.netArticle: 6752
On Mon, 23 Jun 1997 17:56:09 +0200, Jonas Thor <i93-jtr@jota.sm.luth.se> wrote: >I guess this is becuase there are no ORCA macros for two's complement >multipliers...- If so, could anyone please give me a hint how to best >implement a two's complement multiplier in ORCA FPGA? For large multipliers, modified booth triplet is probably the best, with some induced parallalism. If a different multiplier & multiplicand is required every cycle, then each partial double term generated needs a 5 input, two output function, which is neatly done in a single ORCA PFU. (5 input functions are pretty fast in ORCA). This will give you the correct bit shift, and an add/sub decision result. Booth triplet is nice, as it halves the maximum number of additions required when compared with conventional add/pass/subtract implementations, which is often what synthesis gives, although I recall that Synplify does introduce some parallelism. If a fixed value multiplier is used every time, on a varying multiplicand, then pre-applying the booth triplet and creating your own fixed multiplier is very efficient, just as it can be with unsigned multipliers. Induced parallelism helps to reduce the critical path, as of course will pipelining. StuartArticle: 6753
to whom it may concern. I have a problem attaching jedec files compiled from abel software on the synario package to the mentor system. synopsys application hotline assisted me to some extent, but would like to hear from any-one else who had a simlar problem. Thanks DaveyArticle: 6754
Hello, I am arranging a visit to the Boston area for an international group. They are among other things interested in visiting companies involved in ASIC/FPGA development. As this is not my field, I would like to ask this group for ideas. What companies in the Boston area would be interesting to visit? Any information is welcome. Please respond by email. Thanks in advance. Sincerely, Hakan Thyr hakan@worldknowledge.comArticle: 6755
-- -------------------------------------------------------------------------------- Deependra Talla Res: Off: Apt. # R - 635, Graduate Assistant Sugartown Mews, Room. 426A, Tolentine Hall Devon, Dept. of Elec & Comp Engg. PA 19333. Villanova University Villanova PA 19085 Ph: (610) 225 - 0243 Ph: (610) 519 - 4295 email: deepu@ece.vill.edu --------------------------------------------------------------------------------Article: 6756
Steve, APS makes a low cost board with just what you are asking. It can take any of the XILINX 52xx and 40xx series chips and comes with an FPGA installed. It can work both in a PC (ISA BUS) or stand alone. The board has an on board 8255 IO chip, an oscillator socket, prom socket, on board timer chip, status LEDs, and a through hole prototype area, as well as 4 20 pin IDC connectors which bring all the useable pins of the FPGA out to pins. The board comes with schematics, download software, VHDL examples, and C code control program boilerplates. The board can also be purchased with XILINX foundation software in four configurations at VERY LOW PRICES. They can be seen at http://www.erols.com/aaps Steve Martindell wrote: > > I'm looking for a board that would have a Xilinx or Altera FPGA(either > soldered or socketed) with all the FPGA I/O pins brought out to a > connector(s). A board like this would allow me to quickly protype > designs without having to send out to a board-shop. Does anyone > know of a company that makes a product like this? > > thanks, > Steve Martindell > s-martindell@ti.comArticle: 6757
Hi everyone, I have some trouble with that asynchronous peripheral mode. I can download the header and the first two frames, then within the third byte of the next frame /INIT goes low -> frameerror, I think. That occurs always within the 74th byte transfer, independant of the design file (I testet two completely different designs) and independant of the length count (I modified the length count by hand, just for testing). BTW: It's a 4013PQ208-6 The download of both files with the XChecker cable works well. Usually the Xilinx should stop CCLK after detecting a frameerror and forcing /INIT to go low, but CCLK continous after any following byte transfer. I don't use CRC-checking (normal 0110 at end of each frame). Has anybody any idea ??? Or even had the same problem ? Thanx, ErikArticle: 6758
Hi there, I am learning VHDL but I have some troubles. In particular, I need to know where I can find a online manual on the net, and if there are any specific newsgroups on VHDL. Bye. GianpaoloArticle: 6759
Gianpaolo Scassellati wrote: > I am learning VHDL but I have some troubles. In particular, I > need to know where I can find a online manual on the net, and if there > are any specific newsgroups on VHDL. On line manual? Try the VHDL Cookbook http://www.cs.adelaide.edu.au:80/~petera/designers-guide/DG.html Newsgroup? news:comp.lang.vhdl -- David Bishop INTERNET: dbishop@kodak.com | The opinions voiced are mine US MAIL: 1194 West Ave, Hilton NY 14468 | and not my company's. PHYSICAL: 43:17:17N 77:47:37E 281' |Article: 6760
Has anybody used Xilinx 3.3V XC4000XL devices in a mixed 5V/3V system? I am trying to find out if the "5V tolerant" I/O's on the XC4000XL devices can be driven from a low-impedance (i.e. CMOS) 5V source, or can only be driven from a higher-impedance TTL-style 5V source. The published information from Xilinx has been changing as the parts have gone from lab to production. Here is a brief timeline: - The 9/18/96 XC4000 family data sheet says the 4000XL inputs can be driven from a CMOS 5V source "...if the dedicated 5-Volt supply pad(VTT) is tied to 5V." - A "Technical Support Solutions Database" record at the Xilinx web site (record #1942, which Xilinx tech support told me was entered on 2/21/97) states that "VTT connections were anticipated 5 volt connections to allow the 3.3volt XL devices to properly interface the IO to other 5 volt devices. Later testing realized that these were not necessary so, the Pins marked VTT are currently not bonded N.C." (All grammatical errors have been faithfully reproduced.) - A 3/1/97 Xilinx app note entitled "3.3 V and Mixed Voltage Compatible Products" states that "The I/O structures of the XC4000XL FPGAs have been designed to tolerate being driven to a 5 V rail by a low-impedance source. These 3.3V FPGAs can be directly driven by 5 V devices with either TTL or CMOS outputs." - The "Features" section of the 3/25/97 XC4000XL data sheet says the inputs are "5V tolerant", and makes no mention of a VTT pin. - The 5/30/97 version of the XC4000XL data sheet lacks the "Features" cover sheet of its predecessor, but in its absolute maximum ratings table gives VCC max as 4.0V, and VIN max as 5.5V. However, the VIN spec references a footnote that states "Maximum DC overshoot or undershoot above Vcc or below GND must be limited to either 0.5V or 10 mA..." While I suspect this last item was just a copy/paste mistake, it was troubling enough that I looked for some more information on the Xilinx web site. Much of what I found stated that the XC4000XL inputs were TTL compatible, but there was nothing recent which said anything about true 5V CMOS compatibility. I've been in touch with the Xilinx tech support hotline via email and telephone, and so far the information I have received has been ambiguous at best. One of the tech support reps simply emailed back exactly what was written in the data book, despite my pointing out that this was contradicted by subsequent information. Another told me that he asked the XC4000 product manager, who told him that they were truly 5V tolerant (not 3.7V tolerant), but he was unable to provide me with any written documentation or the name of the person to whom he spoke, only that it was the "product manager". So, my question is: Has anybody actually used a 3.3V Xilinx XC4000XL part with true CMOS 5V drivers at the device inputs? Thanks in advance, Mike Coren Staff Engineer SALIX Technologies, Inc. _______________________________________________________________________ ========= =========== SALIX Technologies, Inc. mailto:salix@clark.net ============ 9400 Key West Avenue http://www.salixtech.com ============ Suite 275 TEL: (301) 738-8284 == ==== == Rockville, MD 20850 FAX: (301) 738-8289 == === "...Innovative technology for the 90s and beyond." The statements contained herein are strictly those of their author, and unless otherwise specifically indicated should not be attributed to SALIX or its management. -- ____________________________________________________________________________ ========= SALIX Technologies, Inc. =========== 9400 Key West Avenue, Suite 275 ============ Rockville, MD 20850Article: 6761
Erik Lins wrote: > > Hi everyone, > > I have some trouble with that asynchronous peripheral mode. I can > download the header and the first two frames, then within the third > byte of the next frame /INIT goes low -> frameerror, I think. From my experience with XC4ke parts, there are two things to worry about: 1. Byte download rate. Watch the RDY/BUSY line, which tells you when byte "n" has been absorbed and the device is ready for byte "n+1." I have found the byte download rate to be rather slow in these devices, down in the 100kBytes/second range. There is a CCLK speed bit which can be set in the bitstream which should allow a higher byte download rate (for all bytes *following* this bit...) This is well documented in the Xilinx data books and I expect it is not your problem. 2. Timing of the PROG* pin. There are 2 aspects of PROG* pin timing. First, how long should the pin remain LOW? Next, how long after the rising edge of PROG* should one wait before writing the first byte of the header? Neither of these times are specified in the Xilinx data book. Worse, there is no handshaking to allow the device to inform the "master" that it is OK to take PROG* high or write the first byte of the header. I had half expected the RDY/BUSY line to tell me when it was OK to write the first byte of the header, but it doesn't. I have used software delays on the order of milliseconds to drive both of these times, following some trial-and-error experimentation. If I am wrong on the lack of specification or lack of handshaking on this issue, I would be glad to be corrected! Note that the only time INIT can go low is at a frame boundary. If you see INIT going low at a point other than a frame boundary, start to suspect byte timing or PROG* timing. Good luck. -Steve GrossArticle: 6762
Steven Guccione wrote: > See http://www.vcc.com/ They have a Xilinx 6200-based board that has > full > Java support and does remote access pretty painlessly. Comes with > WebScope, a remote debug tool, also implemented in Java. > Also, I have Derek, a Java based tool for peering in at the 6200 when it sits on VCCs HotWorks card. Anyone is free to use that. Derek lives at http://www.eee.bham.ac.uk/James-RoxbyP/derek.htm PhilArticle: 6763
A while back ( 4 years?) intel was giving away its PLD software for its line of programmable logic. I was wondering if anyone still has the program set and if they could sent it to me. -- Anthony Marchini Electronic Design Engineer American Locker Security Systems www.americanlocker.comArticle: 6764
SALIX Technologies - Dan Simpkins <salix@clark.net> wrote in article <5or61p$26a@clarknet.clark.net>... > <deleted> > > So, my question is: Has anybody actually used a 3.3V Xilinx XC4000XL > part with true CMOS 5V drivers at the device inputs? > The fact that someone has done this without disaster is necessary but not sufficient proof that it is safe to do so. The latest XCELL (#25, 2nd Q/97) contains an encouraging statement: (p. 21): "The I/O structures of the XC4000XL FPGAs have been designed to tolerate being driven to a 5V rail by a low-impedance source. .. Power supply sequencing is not a problem; the inputs can be driven to 5V either before or after the 3.3 Vcc power is supplied without risking damage to the devices." (another quote for your timeline :). Still would be nice to see this on the data sheet, though. regards, tom tburgess@drao.nrc.caArticle: 6765
I'm looking for a board that would have a Xilinx or Altera FPGA(either soldered or socketed) with all the FPGA I/O pins brought out to a connector(s). A board like this would allow me to quickly protype designs without having to send out to a board-shop. Does anyone know of a company that makes a product like this? thanks, Steve Martindell s-martindell@ti.comArticle: 6766
a840272@# Replace this line with your news domain (Steve Martindell) writes: >I'm looking for a board that would have a Xilinx or Altera FPGA(either >soldered or socketed) with all the FPGA I/O pins brought out to a >connector(s). A board like this would allow me to quickly protype >designs without having to send out to a board-shop. Does anyone >know of a company that makes a product like this? Both companies make evaluation boards for a limited number of their devices. Altera just started making one for educational use that has a CPLD and one of their FLEX 10K devices on it. I've seen it, but haven't used it. I have used the Xilinx ones, though. They include a download cable that attaches to a PC. Cheers, Jake -- janovetz@coewl.cen.uiuc.edu| Once you have flown, you will walk the earth with University of Illinois | your eyes turned skyward, for there you have been, | there you long to return. -- da Vinci PP-ASEL | http://www.cen.uiuc.edu/~janovetz/index.htmlArticle: 6767
FS: CADKEY '97 (8.0)-100+ Available- Save $700.00 EACH!!! A new copy of Cadkey '97 is selling on the street for $1,195. To guarantee updates for the next year costs $350 more, bringing it up to $1,545 total! We have available over 100 NEW copies of Cadkey 6.0 on CD for DOS in unopened, shrinkwrapped boxes with all manuals & documentation; which can be upgraded to Cadkey '97, INCLUDING the year's worth of free updates, for the street price of $595. That's $950 LESS than the normal street price! Or, maybe Cadkey 6.0 has enough power for you with no upgrade at all! (Cadkey '97 is basically Cadkey 8.0) We're selling these for $250 each, so you save $700 over the normal street price of Cadkey '97! ($1,545 minus $595 upgrade price minus our price of $250 equals $700 savings!) If you would like more info on Cadkey 97, here's Cadkey 97's Web page: http://www.cadkey.com/cadkey/index.htm Thanks! Karl Kristianson DreamQuestArticle: 6768
Phani Putrevu (pputrevu@ececs.uc.edu) wrote: : Hi, : I was wondering if there are a class of circuits/systems that I should : not implement using FPGAs. If you can have the equivalent circuit for lot less (ie: an 8051, Z80, etc.) : I am told that floating point units dont perform well when implemented : using FPGAs. In fact a multiplier that I had implemented required 200ns : (16 bit). Present day CPUs have much better FPUs/ALUs. I'd say that's implementation dependent. Did you use a Wallace tree for your multiplier? --- #include <Standard.Disclaimer> about this post being a personal opinionArticle: 6769
In article <5os7pl$247$1@vixen.cso.uiuc.edu>, Jacob W Janovetz <janovetz@coewl.cen.uiuc.edu> wrote: >a840272@# Replace this line with your news domain (Steve Martindell) writes: >>I'm looking for a board that would have a Xilinx or Altera FPGA(either >>soldered or socketed) with all the FPGA I/O pins brought out to a >>connector(s). A board like this would allow me to quickly protype >>designs without having to send out to a board-shop. Does anyone >>know of a company that makes a product like this? > Both companies make evaluation boards for a limited number of their >devices. Altera just started making one for educational use that has >a CPLD and one of their FLEX 10K devices on it. I've seen it, but >haven't used it. I have used the Xilinx ones, though. They include >a download cable that attaches to a PC. For a number of projects, I've hand wired boards which contain a PLCC Xilinx and a timer circuit which allows the it to configured over a serial cable. I've found this to be very useful: you first configure the Xilinx and then a UART in the configuration can communicate with the computer over the serial cable. Of course the computer can be anything with an rs-232 port. If there's interest, I'll run off a bunch of these off with a prototyping area. I would like to do it, but I want to be able to sell enough of them to pay the board house bill. Basically, if I can sell at least 10 of them for $100 each I'll do it. A few other features might be: - able to power the FPGA with rs-232 signal lines as well as external power from a wall adaptor - led which indicates that the configuration has occured - can cause reconfiguration with change in DTR line status - Xilinx pins are on an IDC connector, which can be used for jumpers or cables as well as prototype wire-wraps. - cheap FPGA included - Socket for configuration prom and jumper option for master serial mode configuration. - Pads for a crystal and a socket for an oscillator can - I would include the board, software for PC and UNIX (including source) to download the configuration data, and a UART example in OrCAD, ActiveCAD and maybe Viewlogic formats (I have the OrCAD and Foundation devlopement systems, and 45 runs of Viewlogic left on my key :-) Should I use the 68 pin or 84 pin PLCC socket? How big should it be? It would also be useful to make an ISA bus Xilinx prototype card, but that interests me less at the moment. -- /* jhallen@world.std.com (192.74.137.5) */ /* Joseph H. Allen */ int a[1817];main(z,p,q,r){for(p=80;q+p-80;p-=2*a[p])for(z=9;z--;)q=3&(r=time(0) +r*57)/7,q=q?q-1?q-2?1-p%79?-1:0:p%79-77?1:0:p<1659?79:0:p>158?-79:0,q?!a[p+q*2 ]?a[p+=a[p+=q]=q]=q:0:0;for(;q++-1817;)printf(q%79?"%c":"%c\n"," #"[!a[q-1]]);}Article: 6770
nospam J.Szamosfalvi wrote: > > Phani Putrevu (pputrevu@ececs.uc.edu) wrote: > : Hi, > > : I was wondering if there are a class of circuits/systems that I should > : not implement using FPGAs. > > If you can have the equivalent circuit for lot less (ie: an 8051, Z80, etc.) > > : I am told that floating point units dont perform well when implemented > : using FPGAs. In fact a multiplier that I had implemented required 200ns > : (16 bit). Present day CPUs have much better FPUs/ALUs. > > I'd say that's implementation dependent. Did you use a Wallace tree > for your multiplier? Depending on the particular FPGA, there are better architectures than a wallace tree for multipliers that take advantage of the architecture. Remember, the routing in an FPGA is a limiting factor for speed. For example, a Xilinx 4K series -2 part can do 12 bit fixed point multiplies at better than 40 MHz with a four clock latency. A wallace tree design is considerably slower and uses more area because of the poor architectural fit. Floating point is normally a poor fit for FPGAs, as the hardware is considerably more complex than the equivalent fixed point hardware. Granted, a fixed point design requires the designer to do his homework up front, but in my opinion the resulting performance and density is worth it (and the results are potentially more accurate to boot). Generally speaking there are two limits to what can/should go in an FPGA. The lower limit is defined by the cost. If there is a cheaper solution that meets the required functionality and performance, it doesn't make sense to use the FPGA. An example of this is a design that will run on a conventional microprocessor/DSP processor at the desired data rate. The other end of the spectrum is defined by the clocking limits of the FPGA. Practically speaking, you are limited to clocks under about 100MHz for an FPGA design (certain carefully done designs can significantly exceed 100 MHz, but for general stuff that is an upper limit). Now, an FPGA being clocked at say 30-40MHz can outperform a microprocessor running at 200MHz, since the circuit can be designed to handle the entire function on every clock cycle instead of one basic function per instruction (ahh, the beauty of custom hardware). A third dimension that may come into play, is the somewhat limited number of gates available for a design. This can be alleviated somewhat by using multiple devices and properly partitioning the design. -Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email randraka@ids.net http://www.ids.net/~randrakaArticle: 6771
Steve Martindell wrote: > I'm looking for a board that would have a Xilinx or Altera FPGA(either > soldered or socketed) with all the FPGA I/O pins brought out to a > connector(s). A board like this would allow me to quickly protype > designs without having to send out to a board-shop. Does anyone > know of a company that makes a product like this? > > thanks, > Steve Martindell > s-martindell@ti.com See http://www.io.com/~guccione/HW_list.html for a list of commercial and university-based FPGA boards. Stephen -- Stephen D. Scott sds@cs.wustl.edu Department of Computer Science 535 Jolley Hall Washington University, Campus Box 1045 phone: (314) 935-4425 One Brookings Drive fax: (314) 935-7302 St. Louis, MO 63130-4899 URL: http://www.cs.wustl.edu/~sds/Article: 6772
Gianpaolo Scassellati wrote: >Hi there, > > I am learning VHDL but I have some troubles. In particular, I >need to know where I can find a online manual on the net, and if there >are any specific newsgroups on VHDL. > >Bye. > Gianpaolo http://server.vhdl.org/viuf/ is the home of the VHDL International Users' Forum and has lots of good information, including an extensive FAQ with a list of VHDL books. http://www.erols.com/aaps/x84lab/ has a VHDL tutorial, but I haven't tried it myself so I cannot vouch for it. Finally, try the newsgroup comp.lang.vhdl. Stephen -- Stephen D. Scott sds@cs.wustl.edu Department of Computer Science 535 Jolley Hall Washington University, Campus Box 1045 phone: (314) 935-4425 One Brookings Drive fax: (314) 935-7302 St. Louis, MO 63130-4899 URL: http://www.cs.wustl.edu/~sds/Article: 6773
Are you ready to look at pictures of Teens fucking and sucking? Then cum visit: http://www.sexy-girls.com All models represented on Sexy-Girls are 18 years of Age or Older.Article: 6774
In article <5os2ed$kk6@hammerhead.dadd.ti.com>, a840272@#.? writes >I'm looking for a board that would have a Xilinx or Altera FPGA(either >soldered or socketed) with all the FPGA I/O pins brought out to a >connector(s). A board like this would allow me to quickly protype >designs without having to send out to a board-shop. Does anyone >know of a company that makes a product like this? I'm working on a low-cost reconfigurable computing module, with a Xilinx XC6216 and four 1 MBit SRAMs. Some details are on my Web pages. Leon -- Leon Heller Amateur radio callsign: G1HSM Email: leon@lfheller.demon.co.uk http://www.lfheller.demon.co.uk Tel: +44 (0) 118 947 1424 (home) +44 (0) 1344 385556 (work)
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