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
Peter Alfke wrote: > > Rick, there really is no problem. Just look at their data sheet. They > use our nomenclature. > If I remember right, they use our Vpp pin differently, and they tend to > be slower than our parts, but that is no problem at 1 MHz configuration > rate ( but may be a worst-case problem at 8 MHz) > A long time ago, Atmel was also confused about the programming of the > reset polarity, but that must have been resolved long time ago. > > Enjoy their reprogrammability. ... and they need a faster Vcc ramp up. Gerhard -- on the air: DK4XP in the air: D-8551Article: 9126
Does anyone have a suggestion for a really good tect on FPGA for a beginner? I'm multistep process of going from code to a working device is not well explained in any software I have tried. Also I would like a book that explains the difference between using VHDL and other methods or creating a device. Steve -- * * * Author of Imagic and APE - The Atari Peripheral Emulator! * * * * * * Turn your 8-bit Atari into a powerhouse with APE! * * * * * * Ape Homepage: http://www.nacs.net/~classics * * * !! Request my *FOR SALE* LISTING OF CLASSIC VIDEO GAME STUFF !! 2000+ Lines !!Article: 9127
Princeton Information < www.princetoninformation.com > is seeking Ten (10) Senior Digital Design Engineers............ Digital FPGA Hardware Engineer -High Speed Digital Modem Technology Job Description: You will participate in the design and development of a flexible communications system breadboard utilizing FPGA's (ASIC) for Motorola's Celestri (tm) Modem Technology development program. These are specialized high-speed modems (Mbps) installed in satellites. Celestri is a multi-year, $13 billion project that will provide high-bandwidth packet switching - connections (e.g., internet) in outer SPACE!!! Celestial Data Communications for the 21st Century!!! BS/MS in Electrical Engineering +4 to 6 years of experience required. Familiarity with FPGA based design limitations and mitigation methods. Willingness to accept responsibility & technical challenge Ability to work independently and on a team Focus on customer satisfaction EMAIL your resume TODAY!!! ==> azjobs@usa.net Location: Phoenix/ Chandler, AZ Duration: Temp to Permanent Pay Rate: $55,000 => $85,0000 Also, Temp to Perm - "Try it you'll like it!" See our Website: < www.princetoninformation.com > Digital FPGA Hardware Designer - High Speed Digital Design - Modem Technology We have an opportunity for an individual who has done some complex Circuit Board/FPGA design to work at a place where cutting edge technology is the norm, and one of the very best design staffs in the country awaits. This position is for someone who has between 4-10 years of high performance custom circuit design under his/her belt. You will be working on some of the "neatest" projects you've ever seen, and will become a stellar hardware designer for your efforts. Some of the "buzz": We are looking for High Speed Digital Designers, Modem Technology development....... Experience with PLD's, FPGA's (ASIC), complex designs (nothing simple at this place), understands timings, etc... Not a person who still needs a lot of instruction, we are hoping to find an individual who can stand alone and bring a project in from scratch to production. Our client is located in Chandler, Arizona. Salary will be very nice, they're not cheap, as they're looking for the best we can bring in. Please E-mail your resume TODAY===> azjobs@usa.netArticle: 9128
E.E. Tools AllMax universal programmer (EPROMs, EEPROMs, serial EEPROMs, PALs, GALs, PLDs, Flash, micros, tests logic ICs, etc., etc.) with AllMax+ parallel port upgrade. Also has a separate socket on top for testing DRAM and SRAM modules. Works with your parallel port *or* with an ISA card (both interfaces included). Has a 48-pin DIP ZIF socket on top. Includes programmer, parallel port module, parallel port (Centronics) cable, PC card, ribbon cable, power supply, manual, and software. In excellent electrical and physical condition (works great, looks great). I'm moving, and need the money more than the universal programmer. Check out the features on the E.E. Tools website: www.eetools.com (keep in mind that this is the equivalent of the Allmax+, since it has the parallel port interface included). Asking price: $425 (includes shipping/C.O.D.)Article: 9129
My first impressions with 8.2 compared to 8.11 : - results are somewhat better in case you have fitting problems (my case is a 10K20) - it runs somewhat slower -- Koenraad SCHELFHOUT Switching Systems Division http://www.alcatel.com/ Microelectronics Department - VH14 _______________ ________________________________________\ /-___ \ / / Phone : (32/3) 240 89 93 \ ALCATEL / / Fax : (32/3) 240 99 47 \ / / mailto:ksch@sh.bel.alcatel.be \ / / _____________________________________________\ / /______ \ / / Francis Wellesplein, 1 v\/ B-2018 Antwerpen BelgiumArticle: 9130
Gerhard Hoffmann wrote: > Peter Alfke wrote: > > > > Rick, there really is no problem. Just look at their data sheet. > They use our nomenclature. > > If I remember right, they use our Vpp pin differently, and they tend > to > > be slower than our parts, but that is no problem at 1 MHz > configuration > > rate ( but may be a worst-case problem at 8 MHz) > > > ... and they need a faster Vcc ramp up. > > Gerhard > Let me explain this fundamental problem in more detail: During power-up and during reconfiguration, we strongly recommend that all INIT pins be interconnected and also be used to reset the SPROM. That takes care of any possible differences in the wake-up and internal "housecleaning" times of the FPGA(s). INIT goes High when the last of the FPGAs is ready, and the SPROM is being held reset as long as possible. But what if the SPROM power-on-reset is the longest in the system, longer than that of any FPGA ? Then the interconnect of the INIT pins doesn't help. The lead FPGA will start sending out CCLK, whether the SPROM has finished its power-on routine or not. That's what Gerhard was referring to. If Vcc rises slowly, the Atmel SPROM may have the highest Vcc-detect threshold in the system, and thus start its power-on-reset so late that the above mentioned disaster occurs. All such problems can, of course, be avoided by the use of a precision Vcc-detect circuit for <$1 from specialists like Maxim, Linear, Dallas, TI or Motorola. Peter Alfke, Xilinx ApplicationsArticle: 9131
I posted a recent MPEG video compression tutorial that I wrote to my website at http://members.aol.com/johnw39038/index.htm if you are interested. There's also a couple of QEX articles on DSP FIR filter implementations, and VHDL/FPGA chip design as related to amateur radio applications. Enjoy... jw ======================================================== John Wiseman Senior Researcher Hitachi America, Ltd., Digital Media Systems Laboratory Princeton, NJ Email: johnw@dmsl.hitachi.com Voice: (609)-520-0071 ext. 18 Fax: (609)-520-8953 Amateur Radio: KE3QG ========================================================Article: 9132
The JEDEC U and E fields are optional fields in a JEDEC file and are not part of the JEDEC-3C (JESD3-C) standard. The E field allows special feature fuses that do not affect the logic function of the device to be specified in JEDEC files. The E field reads left to right for the purpose of checksum calculations. In article <693cda$j1d$1@news.worldonline.nl>, Johan Kortenhoeven <jkorten@worldonline.nl> writes >Is there somebody who is familiar with the lines >starting with an "E" is the programming file of a mach 211 >device. >I am particular interested in the checksum calculations. > > > > -- William WhiteArticle: 9133
Hello, Sorry for the repost but nobody seems to know the answer to this simple question. I have been using Viewlogic 4 (DOS) and XACT6 PPR (DOS) for XC3000 designs. I have been attaching the above attrobutes to nets in the circuit, to force allocation to either Long Lines or to specify a maximum skew time, respectively. These are described in the Xilinx Viewdraw/LCA docs. I have seen some strange problems, never present in the old APR, and it has occurred to me that XACT6/PPR might be ignoring these net attributes. I know that the Xilinx preferred method today is timespecs, but this is a nice simple way to do it. Peter. Return address is invalid to help stop junk mail. E-mail replies to z80@digiXYZserve.com but remove the XYZ.Article: 9134
In article <34F0C991.F60D0BBA@now-online.com>, e.kobal@now-online.com wrote: >We are looking to implement signal arrival time detection using the >correlation method. Our design involves a steady stream of 8-bit >samples at a sampling rate of 20 Msps. Our pattern that we are >attempting to match is 40 samples wide. Now the problem we have arrived >at is the following: we need to perform a 40x8 bit matrix >multiplication. This is only the root of our design problem, since our Is this not a classic application of the "FIR" filter? A 40 tap FIR is quite possible to implement and much has been written on the subject. 20Msps should be easy. For information on the implementation of FIR filters (using "proper" coefficients) see for instance the Altera website (look for Application Note 73). A DSP solution will be difficult, you would be hard pressed to get a result in and out of the DSP in the 50nS you have available, let alone do a 40 tap FIR. >application is multichannel. We are not sure whether our design would >work best using a standard DSP, or whether we should use an FPGA to >allow for multiplication in parallel. Any information would be much >appreciated. > >Thank you, > >Erik Kobal, Cleveland State University >Article: 9135
Peter Alfke wrote in article <34F1A724.AFE62BD0@xilinx.com>...> >That's what Gerhard was referring to. If Vcc rises slowly, the Atmel >SPROM may have the highest Vcc-detect threshold in the system, and thus >start its power-on-reset so late that the above mentioned disaster >occurs. > >All such problems can, of course, be avoided by the use of a precision >Vcc-detect circuit for <$1 from specialists like Maxim, Linear, Dallas, >TI or Motorola. Disaster! Disaster ! The sky is falling in!!!! We have a specifically designed LOW POR threshold, very fast active from POR and additionally have a READY pin on our 512K and 1M devices to further simply system design. But you don't have to believe me - we will let you decide for yourself by happily shiping you a couple of FREE samples - for more details.... http://www.atmel.com/atmel/products/prod22.htm Martin Mason Atmel Corp.Article: 9136
Hi I'm working in Max+plus II of Altera and now I want change to Foundation of Xilinx. Somebody knows some program to make that ?? Felip Vicedo Roman Escuela Universitaria de Gandia Universidad Politécnica de Valencia e-mail: feviro@eug.upv.esArticle: 9137
Long ago someone sent a posting which included pesudorandom number generators. That person represented a model and then extended it to a Galois field. Do I remember correctly? If yes, would you please tell some links, books etc? UtkuArticle: 9138
This isn't a jobs newsgroup. The 'comp.lang' newsgroups aren't job newsgroups, either. I suggest that you download a current newsgroup list, and look for the word 'JOB'.Article: 9139
Very nice, well explained. Thanks John. Regards, Sam John Wiseman wrote: > I posted a recent MPEG video compression tutorial that I wrote > to my website at http://members.aol.com/johnw39038/index.htm > if you are interested. There's also a couple of QEX articles on > DSP FIR filter implementations, and VHDL/FPGA chip design > as related to amateur radio applications. Enjoy... > > jw > > ======================================================== > John Wiseman > Senior Researcher > Hitachi America, Ltd., Digital Media Systems Laboratory > Princeton, NJ > Email: johnw@dmsl.hitachi.com > Voice: (609)-520-0071 ext. 18 Fax: (609)-520-8953 > Amateur Radio: KE3QG > ========================================================Article: 9140
My company will shortly be respinning the prototype board for Xilinx FPGAs. We are requesting your help in choosing the appropriate feature set. Email your responses to arnaud@ecla.com 1. Board Form Factor Do you prefer (A) standalone board or (B) a board that plugs into a PC slot? 2. Bus Interface If PC size board, do you prefer (A) an ISA interface or (B) a PCI interface or (C) No interface, just power and ground. 3. Bus Interface Controller Do you require the board to incude: (A) Separate PCI bus controller (B) Separate ISA bus controller (C) the device should be able to drive the bus (D) No bus controller 4. Xilinx FPGA Should the device be (A) directly soldered to the board or (B) to a small module that's plugged into a connector on the board ? 5. If the device is on the board, which device would you prefer? 6. What package and pincount? 7. Should the board include a microcontroller? (A) Yes (B) No 8. What kind of connector should be used to bring out the IO? 9. How large should the wire-wrap area be? 10. The device will be able to be programmed through a serial prom. (A) Include a parallel Flash device? (B) Include a jtag (boundary scan) connector (C) Include a xchecker download/readback connector (D) If a micro is on the board, should the micro be able to program the device? 11. Other featgures or comments Thanks for you help. Alain Arnaud arnaud@ecla.comArticle: 9141
Nic wrote: > In article <34F0C991.F60D0BBA@now-online.com>, e.kobal@now-online.com wrote: > >We are looking to implement signal arrival time detection using the > >correlation method. Our design involves a steady stream of 8-bit > >samples at a sampling rate of 20 Msps. Our pattern that we are > >attempting to match is 40 samples wide. Now the problem we have arrived > >at is the following: we need to perform a 40x8 bit matrix > >multiplication. This is only the root of our design problem, since our > > Is this not a classic application of the "FIR" filter? A 40 tap FIR is quite > possible to implement and much has been written on the subject. 20Msps should > be easy. For information on the implementation of FIR filters > (using "proper" coefficients) see for instance the Altera website (look for > Application Note 73). > > A DSP solution will be difficult, you would be hard pressed to get a result > in and out of the DSP in the 50nS you have available, let alone do a 40 tap > FIR. > Well, _maybe_ it's an FIR problem. Signal arrival time detection is only partlysolved by using an FIR filter, since peak detection on the output only gives you the arrival time to the nearest sample. If you want subsample resolution, it's much easier to get the result by processing the data in the Fourier domain. Signal arrival type problems are very often a periodic process (it happens periodically), so you could use a DSP chip with appropriate data buffering hardware. The original poster is looking to find a 40 sample signal that arrives with an unknown periodic rate. If the rate is low enough, and the location uncertainty is small, then a DSP processor could do the job. On the other hand, an FIR correlator might be appropriate if the search space is too much for a DSP chip. I would submit that the original problem as posed is insufficiently specified to determine an optimal solution. Perhaps Mr Kobal can supply more information? For example, is the signal being sought a binary sequence, or multivalued? John McCluskey Lucent Technologies > >application is multichannel. We are not sure whether our design would > >work best using a standard DSP, or whether we should use an FPGA to > >allow for multiplication in parallel. Any information would be much > >appreciated. > > > >Thank you, > > > >Erik Kobal, Cleveland State University > >Article: 9142
Has anybody out there already done a completely digital PLL inside an FPGA ? I have to do an XC4028EX and one part of it shall be a programmable PLL with an input clock of about 300kHz and an output of about 10Mhz. But sometimes I need exactly 9.998 MHz, sometimes 10.011 and so on, as precise as possible. Any idea ? Holger Schuppan Neurotech@T-Online.deArticle: 9143
--------------0BE633ADE9983111DFB2A6B7 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Frank wrote: > Can anybody tell me what's the difference between BUFT and BUFE in Aldec > Foundation Series software? I use it to do design with Xilinx device. -- ************** Remove [ANTI_SPAM] to reply ************** BUFE is enabled when the E pin is high BUFT is tristated when the T pin is high. The only difference between the two is an inversion of the control pin. --------------0BE633ADE9983111DFB2A6B7 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit <HTML> Frank wrote: <BLOCKQUOTE TYPE=CITE>Can anybody tell me what's the difference between BUFT and BUFE in Aldec <BR>Foundation Series software? I use it to do design with Xilinx device.</BLOCKQUOTE> <PRE>-- ************** Remove [ANTI_SPAM] to reply **************</PRE> BUFE is enabled when the E pin is high <BR>BUFT is tristated when the T pin is high. <P>The only difference between the two is an inversion of the control pin.</HTML> --------------0BE633ADE9983111DFB2A6B7--Article: 9144
Erik Kobal <ekobal@now-online.com> wrote: >We are looking to implement signal arrival time detection using the >correlation method. Our design involves a steady stream of 8-bit >samples at a sampling rate of 20 Msps. Our pattern that we are >attempting to match is 40 samples wide. Now the problem we have arrived >at is the following: we need to perform a 40x8 bit matrix >multiplication. This is only the root of our design problem, since our >application is multichannel. We are not sure whether our design would >work best using a standard DSP, or whether we should use an FPGA to >allow for multiplication in parallel. Any information would be much >appreciated. >Thank you, >Erik Kobal, Cleveland State University You seem to be looking for a soft decision correlation . We use fpga's to do this . At that speed , I believe you are best to use fpga's. Some old app notes by trw (now Raytheon ) may still be available. Henri hgirard@nsicomm.comArticle: 9145
Erik Kobal <ekobal@now-online.com> wrote: >We are looking to implement signal arrival time detection using the >correlation method. Our design involves a steady stream of 8-bit >samples at a sampling rate of 20 Msps. Our pattern that we are >attempting to match is 40 samples wide. Now the problem we have arrived >at is the following: we need to perform a 40x8 bit matrix >multiplication. This is only the root of our design problem, since our >application is multichannel. We are not sure whether our design would >work best using a standard DSP, or whether we should use an FPGA to >allow for multiplication in parallel. Any information would be much >appreciated. >Thank you, >Erik Kobal, Cleveland State University Sounds like you are trying to do soft decision correlation . Some old app notes by trw (now Raytheon) described this. We use fpga's. You can email me for questions. H Girard hgirard@nsicomm.comArticle: 9146
H Girard wrote: > > Erik Kobal <ekobal@now-online.com> wrote: > > >We are looking to implement signal arrival time detection using the > >correlation method. Our design involves a steady stream of 8-bit > >samples at a sampling rate of 20 Msps. Our pattern that we are > >attempting to match is 40 samples wide. Now the problem we have arrived > >at is the following: we need to perform a 40x8 bit matrix > >multiplication. This is only the root of our design problem, since our > >application is multichannel. We are not sure whether our design would > >work best using a standard DSP, or whether we should use an FPGA to > >allow for multiplication in parallel. Any information would be much > >appreciated. > > >Thank you, > > >Erik Kobal, Cleveland State University > > Sounds like you are trying to do soft decision correlation . Some old > app notes by trw (now Raytheon) described this. We use fpga's. > You can email me for questions. > H Girard > hgirard@nsicomm.com H Girard Can you tell me what"s is app note numbers, and where can I get these notes. Jack LaiArticle: 9147
I just posted a note asking how to add pullups in Xilinx FPGAs. Well, I think I have answered my own question, however, I'd like to know your thoughts. The problem is that I could not place a pullup on an INPUT node. (Leonardo wouldn't let me) I am using direct port mapping to do this. However, if I redefine the signal as an "inout" it works fine. Although this isn't really what the signal is, I don't think there is much harm in doing this. Any thoughts? Cheers, Jake -- janovetz@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.ews.uiuc.edu/~janovetz/index.htmlArticle: 9148
In article <01bd40fa$242ef220$52912a9e@cabezon>, feviro@eug.upv.es says... > >Hi > >I'm working in Max+plus II of Altera and now I want change to Foundation of >Xilinx. >Somebody knows some program to make that ?? > > >Felip Vicedo Roman >Escuela Universitaria de Gandia >Universidad Politécnica de Valencia >e-mail: feviro@eug.upv.es The Exemplar synthesis tools allow you to specify an input technology and an output technology to do such translations. I've never done it, but it looks like it should be fairly painless. All disclaimers apply! Steve MitchellArticle: 9149
In article <6cvpke$k62$1@vixen.cso.uiuc.edu>, janovetz@ews.uiuc.edu says... > >I just posted a note asking how to add pullups in Xilinx FPGAs. >Well, I think I have answered my own question, however, I'd like >to know your thoughts. The problem is that I could not place >a pullup on an INPUT node. (Leonardo wouldn't let me) I am >using direct port mapping to do this. > >However, if I redefine the signal as an "inout" it works fine. >Although this isn't really what the signal is, I don't think >there is much harm in doing this. Any thoughts? > > Cheers, > Jake > >-- > janovetz@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.ews.uiuc.edu/~janovetz/index.html Check on the Xilinx web page for doing this. I recall having seen an app note on doing this within their help pages. Steve Mitchell
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