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
By being familiar with synthesis, I mean I'm a student and have had some coursework with the design and synthesis, but projects ended there. To be honest, they were more on the design aspect so I am familiar with the synthesis process but not too experienced with it. I do not use any gated clocks and all of my sequential logic runs of the same edge of the same 50 MHz clock except for the reset logic when the DCM is stabilizing. I have not set any OFFSET IN/OUT constraints, do you know where I can get more information about doing that? Thanks, Brandon >On 6/9/2010 8:01 AM, BrandonD wrote: >> Hi, >> >> I'm somewhat familiar with synthesis and Verilog but I am quite new to >> running the designs on FPGAs. I have a complex design of a processor that I >> am trying to get running on a Virtex 5 FPGA in a BEE3 module. The design >> synthesizes and goes through translate, map and par in Xilinx ISE 10.1 but >> it does not seem to run correctly when programmed on the FPGA. >> >> ISE says that all timing constraints have been met and the static timing >> report shows that it does too. What I am going to do now is look at the >> post-par simulation and see if there's a problem. Is there anything that >> maybe is a common mistake that I should also look into? >> >> I am using a DCM to turn the 100 MHz system clock into a 50 MHz clock. Any >> suggestions will be appreciated. >> >> Thanks, >> Brandon >> >> >> >> --------------------------------------- >> Posted through http://www.FPGARelated.com > >Anything other than single clock synchronous logic isn't represented in >your timing constraints; so anywhere that you're using a clock other >than your derived 50 is worth another look. This includes any logic >running off of a combinationally gated clock. You said you're used to >synthesis but not FPGAs; if that means ASICs then one thing to be aware >of is that clock nets are a lot more sacred in FPGAs. > >Unless you've specified your external OFFSET IN/OUT constraints properly >(this is rare), then your constraints don't properly cover your >relationships to external hardware. Worth another look. > >-- >Rob Gaddi, Highland Technology >Email address is currently out of order > --------------------------------------- Posted through http://www.FPGARelated.comArticle: 147976
Anssi Saari <as@sci.fi> wrote: >> Sudhir Singh <Sudhir.Singh@email.com> wrote: >>> I am just wondering if there are any standard ways of disabling an ip >>> core after an evaluation period of say 30 days. I am trying to provide >>> a potential customer a ip core but don't want them to continue using >>> it after the eval license period expires. The core will run on Xilinx >>> Spartan3 FPGAs. (then I wrote) >> Considering the ability to change the date on the computer, >> it is pretty hard to stop people from using something past >> a given date. Once the bit file is generated, it is pretty >> much impossible. > Well, at least the licensable Xilinx IP I've used included some kind > of timer in the bitfile itself if there's no license, so that the > block works for a few hours and then quits. Yes, I had forgotten about the ones that count clock cycles. But that doesn't help if you want a specific time or date. Even so, one could probably find a way to disable the counter. (I don't know if reset will reset the counter or not, otherwise one can always reload the device.) I was unsure from your original post if you meant the ability to generate bitfiles, or the ability to use them. If the device has internet access, you could use an NTP server to get the date/time. That would take a lot of logic, though! -- glenArticle: 147977
By being familiar with synthesis, I mean I'm a student and have had some coursework with the design and synthesis, but projects ended there. To be honest, they were more on the design aspect so I am familiar with the synthesis process but not too experienced with it. I do not use any gated clocks and all of my sequential logic runs of the same edge of the same 50 MHz clock except for the reset logic when the DCM is stabilizing. I have not set any OFFSET IN/OUT constraints, do you know where I can get more information about doing that? Thanks, Brandon >On 6/9/2010 8:01 AM, BrandonD wrote: >> Hi, >> >> I'm somewhat familiar with synthesis and Verilog but I am quite new to >> running the designs on FPGAs. I have a complex design of a processor that I >> am trying to get running on a Virtex 5 FPGA in a BEE3 module. The design >> synthesizes and goes through translate, map and par in Xilinx ISE 10.1 but >> it does not seem to run correctly when programmed on the FPGA. >> >> ISE says that all timing constraints have been met and the static timing >> report shows that it does too. What I am going to do now is look at the >> post-par simulation and see if there's a problem. Is there anything that >> maybe is a common mistake that I should also look into? >> >> I am using a DCM to turn the 100 MHz system clock into a 50 MHz clock. Any >> suggestions will be appreciated. >> >> Thanks, >> Brandon >> >> >> >> --------------------------------------- >> Posted through http://www.FPGARelated.com > >Anything other than single clock synchronous logic isn't represented in >your timing constraints; so anywhere that you're using a clock other >than your derived 50 is worth another look. This includes any logic >running off of a combinationally gated clock. You said you're used to >synthesis but not FPGAs; if that means ASICs then one thing to be aware >of is that clock nets are a lot more sacred in FPGAs. > >Unless you've specified your external OFFSET IN/OUT constraints properly >(this is rare), then your constraints don't properly cover your >relationships to external hardware. Worth another look. > >-- >Rob Gaddi, Highland Technology >Email address is currently out of order > --------------------------------------- Posted through http://www.FPGARelated.comArticle: 147978
On Tue, 08 Jun 2010 16:53:43 -0700, Sudhir Singh wrote: > Hi Guys, > > I am just wondering if there are any standard ways of disabling an ip > core after an evaluation period of say 30 days. I am trying to provide a > potential customer a ip core but don't want them to continue using it > after the eval license period expires. The core will run on Xilinx > Spartan3 FPGAs. > > Any suggestions will be much appreciated. > > Regards > Sudhir Xilinx doesn't have a mechanism yet for generating an IP license, they plan to have one that uses a new IEEE standard but that won't happen for a year. Altera does have tools that allow you to encrypt your source code and then generate a flexlm license that can restrict the type of FPGA it can go into, allow simulation only or just synthesis, and the time limits for the IP. However once you've generated a bit file it's good forever, there is no way to have a bit file that's time limited.Article: 147979
On Jun 9, 12:45=A0pm, "BrandonD" <BdOn003@n_o_s_p_a_m.n_o_s_p_a_m.gmail.com> wrote: > By being familiar with synthesis, I mean I'm a student and have had some > coursework with the design and synthesis, but projects ended there. To be > honest, they were more on the design aspect so I am familiar with the > synthesis process but not too experienced with it. > > I do not use any gated clocks and all of my sequential logic runs of the > same edge of the same 50 MHz clock except for the reset logic when the DC= M > is stabilizing. > > I have not set any OFFSET IN/OUT constraints, do you know where I can get > more information about doing that? > > Thanks, > Brandon > > > > > > >On 6/9/2010 8:01 AM, BrandonD wrote: > >> Hi, > > >> I'm somewhat familiar with synthesis and Verilog but I am quite new to > >> running the designs on FPGAs. I have a complex design of a processor > that I > >> am trying to get running on a Virtex 5 FPGA in a BEE3 module. The > design > >> synthesizes and goes through translate, map and par in Xilinx ISE 10.1 > but > >> it does not seem to run correctly when programmed on the FPGA. > > >> ISE says that all timing constraints have been met and the static > timing > >> report shows that it does too. What I am going to do now is look at th= e > >> post-par simulation and see if there's a problem. Is there anything > that > >> maybe is a common mistake that I should also look into? > > >> I am using a DCM to turn the 100 MHz system clock into a 50 MHz clock. > Any > >> suggestions will be appreciated. > > >> Thanks, > >> Brandon > > >> --------------------------------------- =A0 =A0 =A0 =A0 =A0 =A0 =A0 = =A0 > >> Posted throughhttp://www.FPGARelated.com > > >Anything other than single clock synchronous logic isn't represented in > >your timing constraints; so anywhere that you're using a clock other > >than your derived 50 is worth another look. =A0This includes any logic > >running off of a combinationally gated clock. =A0You said you're used to > >synthesis but not FPGAs; if that means ASICs then one thing to be aware > >of is that clock nets are a lot more sacred in FPGAs. > > >Unless you've specified your external OFFSET IN/OUT constraints properly > >(this is rare), then your constraints don't properly cover your > >relationships to external hardware. =A0Worth another look. > > >-- > >Rob Gaddi, Highland Technology > >Email address is currently out of order > > --------------------------------------- =A0 =A0 =A0 =A0 > Posted throughhttp://www.FPGARelated.com You may want to look at the "blogs" section of the Xilinx web forums. There is a 5-part piece on timing constraints that is quite useful when you're getting started. Also if you have severe problems, like it looks like nothing works at all, you could have a loop in the reset generation logic. Remember that the output of a DCM does not toggle until locked, so using it to release reset to the DCM will lock up the system because the reset prevents the DCM from locking. Regards, GaborArticle: 147980
On 6/9/2010 9:45 AM, BrandonD wrote: > By being familiar with synthesis, I mean I'm a student and have had some > coursework with the design and synthesis, but projects ended there. To be > honest, they were more on the design aspect so I am familiar with the > synthesis process but not too experienced with it. > > I do not use any gated clocks and all of my sequential logic runs of the > same edge of the same 50 MHz clock except for the reset logic when the DCM > is stabilizing. > > I have not set any OFFSET IN/OUT constraints, do you know where I can get > more information about doing that? > > Thanks, > Brandon > >> On 6/9/2010 8:01 AM, BrandonD wrote: >>> Hi, >>> >>> I'm somewhat familiar with synthesis and Verilog but I am quite new to >>> running the designs on FPGAs. I have a complex design of a processor > that I >>> am trying to get running on a Virtex 5 FPGA in a BEE3 module. The > design >>> synthesizes and goes through translate, map and par in Xilinx ISE 10.1 > but >>> it does not seem to run correctly when programmed on the FPGA. >>> >>> ISE says that all timing constraints have been met and the static > timing >>> report shows that it does too. What I am going to do now is look at the >>> post-par simulation and see if there's a problem. Is there anything > that >>> maybe is a common mistake that I should also look into? >>> >>> I am using a DCM to turn the 100 MHz system clock into a 50 MHz clock. > Any >>> suggestions will be appreciated. >>> >>> Thanks, >>> Brandon >>> >>> >>> >>> --------------------------------------- >>> Posted through http://www.FPGARelated.com >> >> Anything other than single clock synchronous logic isn't represented in >> your timing constraints; so anywhere that you're using a clock other >> than your derived 50 is worth another look. This includes any logic >> running off of a combinationally gated clock. You said you're used to >> synthesis but not FPGAs; if that means ASICs then one thing to be aware >> of is that clock nets are a lot more sacred in FPGAs. >> >> Unless you've specified your external OFFSET IN/OUT constraints properly >> (this is rare), then your constraints don't properly cover your >> relationships to external hardware. Worth another look. >> >> -- >> Rob Gaddi, Highland Technology >> Email address is currently out of order >> > > --------------------------------------- > Posted through http://www.FPGARelated.com I hope this is a stupid question but....it does simulate correctly, right? -- Rob Gaddi, Highland Technology Email address is currently out of orderArticle: 147981
On Jun 9, 2:37=A0pm, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote: > Anssi Saari <a...@sci.fi> wrote: > >> Sudhir Singh <Sudhir.Si...@email.com> wrote: > >>> I am just wondering if there are any standard ways of disabling an ip > >>> core after an evaluation period of say 30 days. I am trying to provid= e > >>> a potential customer a ip core =A0but don't want them to continue usi= ng > >>> it after the eval license period expires. The core will run on Xilinx > >>> Spartan3 FPGAs. > > (then I wrote) > > >> Considering the ability to change the date on the computer, > >> it is pretty hard to stop people from using something past > >> a given date. =A0 Once the bit file is generated, it is pretty > >> much impossible. > > Well, at least the licensable Xilinx IP I've used included some kind > > of timer in the bitfile itself if there's no license, so that the > > block works for a few hours and then quits. > > Yes, I had forgotten about the ones that count clock cycles. > But that doesn't help if you want a specific time or date. > > Even so, one could probably find a way to disable the counter. > (I don't know if reset will reset the counter or not, otherwise > one can always reload the device.) > > I was unsure from your original post if you meant the ability > to generate bitfiles, or the ability to use them. > > If the device has internet access, you could use an NTP server > to get the date/time. =A0That would take a lot of logic, though! > > -- glen Where I have seen evaluation versions of IP with hardware time limits, the clock cycles they count are internal to the core and generated with a ring oscillator to prevent reset without re-configuration. If your system can live with periodic reconfiguration, then obviously this method can be bypassed as well. Regards, GaborArticle: 147982
Gabor <gabor@alacron.com> wrote: (snip) > Where I have seen evaluation versions of IP with hardware time > limits, the clock cycles they count are internal to the core > and generated with a ring oscillator to prevent reset without > re-configuration. If your system can live with periodic > reconfiguration, then obviously this method can be bypassed > as well. Some devices naturally need a clock. I thought I remembered this test on an ethernet MAC, which would be pretty hard to use without a clock. On the other hand, how hard is it to find the carry chains in a design? If you can find the counter, it probably isn't so hard to disable it without doing anything else to the design. -- glenArticle: 147983
It does simulate behaviorally correctly. I forgot to mention that part. You were right about the loop of the reset signal back to the DCM. I did exactly that. I've fixed that problem and it still appears to be working incorrectly. I am going to try post-par simulation to see how that looks. I will read up on the constraints as well. >On 6/9/2010 9:45 AM, BrandonD wrote: >> By being familiar with synthesis, I mean I'm a student and have had some >> coursework with the design and synthesis, but projects ended there. To be >> honest, they were more on the design aspect so I am familiar with the >> synthesis process but not too experienced with it. >> >> I do not use any gated clocks and all of my sequential logic runs of the >> same edge of the same 50 MHz clock except for the reset logic when the DCM >> is stabilizing. >> >> I have not set any OFFSET IN/OUT constraints, do you know where I can get >> more information about doing that? >> >> Thanks, >> Brandon >> >>> On 6/9/2010 8:01 AM, BrandonD wrote: >>>> Hi, >>>> >>>> I'm somewhat familiar with synthesis and Verilog but I am quite new to >>>> running the designs on FPGAs. I have a complex design of a processor >> that I >>>> am trying to get running on a Virtex 5 FPGA in a BEE3 module. The >> design >>>> synthesizes and goes through translate, map and par in Xilinx ISE 10.1 >> but >>>> it does not seem to run correctly when programmed on the FPGA. >>>> >>>> ISE says that all timing constraints have been met and the static >> timing >>>> report shows that it does too. What I am going to do now is look at the >>>> post-par simulation and see if there's a problem. Is there anything >> that >>>> maybe is a common mistake that I should also look into? >>>> >>>> I am using a DCM to turn the 100 MHz system clock into a 50 MHz clock. >> Any >>>> suggestions will be appreciated. >>>> >>>> Thanks, >>>> Brandon >>>> >>>> >>>> >>>> --------------------------------------- >>>> Posted through http://www.FPGARelated.com >>> >>> Anything other than single clock synchronous logic isn't represented in >>> your timing constraints; so anywhere that you're using a clock other >>> than your derived 50 is worth another look. This includes any logic >>> running off of a combinationally gated clock. You said you're used to >>> synthesis but not FPGAs; if that means ASICs then one thing to be aware >>> of is that clock nets are a lot more sacred in FPGAs. >>> >>> Unless you've specified your external OFFSET IN/OUT constraints properly >>> (this is rare), then your constraints don't properly cover your >>> relationships to external hardware. Worth another look. >>> >>> -- >>> Rob Gaddi, Highland Technology >>> Email address is currently out of order >>> >> >> --------------------------------------- >> Posted through http://www.FPGARelated.com > >I hope this is a stupid question but....it does simulate correctly, right? > >-- >Rob Gaddi, Highland Technology >Email address is currently out of order > --------------------------------------- Posted through http://www.FPGARelated.comArticle: 147984
On Jun 10, 11:39=A0am, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote: > Gabor <ga...@alacron.com> wrote: > > (snip) > > > Where I have seen evaluation versions of IP with hardware time > > limits, the clock cycles they count are internal to the core > > and generated with a ring oscillator to prevent reset without > > re-configuration. =A0If your system can live with periodic > > reconfiguration, then obviously this method can be bypassed > > as well. > > Some devices naturally need a clock. =A0I thought I remembered > this test on an ethernet MAC, which would be pretty hard to use > without a clock. =A0 > > On the other hand, how hard is it to find the carry chains in > a design? =A0If you can find the counter, it probably isn't so > hard to disable it without doing anything else to the design. > > -- glen Hi Guys, Thanks for your replies. I think counting clock cycles will be a viable option for me. Cheers Sudhir From prenom.nom@gmail.com Thu Jun 10 00:26:20 2010 Path: unlimited.newshosting.com!s02-b57.iad!npeersf01.iad.highwinds-media.com!npeer02.iad.highwinds-media.com!news.highwinds-media.com!feed-me.highwinds-media.com!nntp.club.cc.cmu.edu!feeder.erje.net!newsfeed.straub-nv.de!proxad.net!feeder1-2.proxad.net!cleanfeed3-a.proxad.net!nnrp5-2.free.fr!not-for-mail Date: Thu, 10 Jun 2010 09:26:20 +0200 From: Matthieu Michon <prenom.nom@gmail.com> Newsgroups: comp.arch.fpga Subject: Re: Design passes synthesis and routing but fails on FPGA Message-Id: <20100610092620.1ca1e2c1.prenom.nom@gmail.com> References: <c_6dnR5Iae9KMZLRnZ2dnUVZ_jWdnZ2d@giganews.com> <mpadnbv9o_HnXJLRnZ2dnUVZ_v2dnZ2d@lmi.net> <oNWdnUOd8JXcWJLRnZ2dnUVZ_s-dnZ2d@giganews.com> <a76cf7b7-109a-4e74-b3f9-2d16ba8aa486@d8g2000yqf.googlegroups.com> X-Newsreader: Sylpheed 3.0.2 (GTK+ 2.10.14; i686-pc-mingw32) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 22 Organization: Guest of ProXad - France NNTP-Posting-Date: 10 Jun 2010 09:26:21 MEST NNTP-Posting-Host: 213.215.9.6 X-Trace: 1276154781 news-4.free.fr 20283 213.215.9.6:50071 X-Complaints-To: abuse@proxad.net Xref: unlimited.newshosting.com comp.arch.fpga:102229 X-Received-Date: Thu, 10 Jun 2010 07:26:21 UTC (s02-b57.iad) On Wed, 9 Jun 2010 15:05:23 -0700 (PDT) Gabor <gabor@alacron.com> wrote: (...) > Also if you have severe problems, like it looks like > nothing works at all, you could have a loop in the > reset generation logic. Remember that the output of > a DCM does not toggle until locked, so using it to > release reset to the DCM will lock up the system > because the reset prevents the DCM from locking. > > Regards, > Gabor To follow on with the DCM feature, performing an initial DCM reset with the timings in compliance with the user-guide and datasheet is a must (with special care when dealing with early Virtex-4 revisions, altough it is not the case here). I suggest to the OP that he should focus on verifying the obvious (pwr-supply levels, input clock waveform, pin-mapping, UCF file, JTAG connection, DCM lock and status outputs). -- Matthieu Michon <prenom.nom@gmail.com>Article: 147985
If you are convinced that your simulation of the design is correct then you really need to use Chipscope to see what is going on inside the device. Jon --------------------------------------- Posted through http://www.FPGARelated.comArticle: 147986
I am getting some warnings during post-par simulation that may be a clue. When I load my routed design in Modelsim, I am getting 14 warnings like this for different nets: Instance 'top.fab.\fs1/ras/reset_n_inv_shift1 ' - No solution possible for delayed timing check nets. Setting negative limit to zero. In the static timing report for the OFFSET IN BEFORE timing constraint (which I have not set so it must be determined by ISE) I have a total minimum clock path delay of -0.678 ns. For the OFFSET OUT AFTER constraint I have a maximum clock path delay of -0.256 ns. I wasn't sure if this meant something as the path for these delays are from the input clock buffer to the DCM. I don't believe the warnings in the simulation are normal but are negative delay values normal? Thanks, Brandon >If you are convinced that your simulation of the design is correct then you >really need to use Chipscope to see what is going on inside the device. > >Jon > >--------------------------------------- >Posted through http://www.FPGARelated.com > --------------------------------------- Posted through http://www.FPGARelated.comArticle: 147987
You shouldnt really need to do a post p&r sim. I have done designs running at > 300 MHz without a post p&r sim. As I said if you have access to Chipscope then you really need to use it. Jon --------------------------------------- Posted through http://www.FPGARelated.comArticle: 147988
I would like to have capabilities of Chipscope for Xilinx FPGA's are there any free alternatives to it? a.tArticle: 147989
have a look at this http://vmagic.sf.net and buid your own. We have done that using matlab here, but it's in german, I'm afraid: http://wwwhni.uni-paderborn.de/sct/projekte/hardware-in-the-loop-simulation-mit-raptor2000/hilde-offline-fpga-in-the-loop/Article: 147990
On Jun 10, 2:13=A0pm, apple <sys.app...@gmail.com> wrote: > I would like to have capabilities of Chipscope for Xilinx FPGA's > > are there any free alternatives to it? There is a poorman's FPGA debugging tool called "an I/O port". I have yet to learn to use chipscope or similar tools. I have a few I/Os available on my device which I tied to a header, often a 0.1" spaced header with 0.025" square pins (if I have room). I use a regular scope or logic analyzer. Since the number of I/Os is limited, I typically use a mux to select which signals come to the I/O header. Of course this has some limitations, mainly the PITA of often spinning a new build to get the signals into the mux I want. But that's not really different from chipscope. I guess they have a tool that routes the muxed signals on your existing design using unused resources, so in theory it doesn't disturb the design. But I've never had a problem where the design was sensitive to this. If you use good techniques for designing your FPGA and do a good job of simulating, I expect you will seldom need chipscope or similar tools. RickArticle: 147991
On Jun 10, 11:13=A0am, apple <sys.app...@gmail.com> wrote: > I would like to have capabilities of Chipscope for Xilinx FPGA's > > are there any free alternatives to it? It's actually pretty easy to write a logic analyzer core. Integrating it into a design isn't too bad if you instantiate it in your code. But ChipScope's Core Inserter feature is pretty great and is superior to instantiating the core. Writing host software to display it in a pretty manner is difficult if you don't know how to do it. Basically, ChipScope is one of Xilinx' shining lights, and is worth the cost. -aArticle: 147992
Hi, I'm working with Xilinx ISE 10.1 and I am having troubles with timing constraints. I've successfully implemented my design with a 20 ns cycle time and found that I needed to change something in the design. I make my changes and re-synthesize and implement design and I no longer meet the same timing constraint. Okay, I've changed the logic slightly so it's not going to be exactly the same, the static timing report says that I need a minimum cycle time of say 22 ns. So I only change the timing constraint to reflect a 25 ns and re-run the process and it fails again saying I need a 32 ns cycle time now. Okay, so I change the cycle time to 35 ns and then I need a 46 ns cycle time. I think this is understandable because with a relaxed constraint some optimizations or paths may not be chosen. When do you try it again with the same constraint and when do you relax the constraint? Thanks, Brandon --------------------------------------- Posted through http://www.FPGARelated.comArticle: 147993
On Jun 10, 8:49=A0pm, "BrandonD" <BdOn003@n_o_s_p_a_m.n_o_s_p_a_m.gmail.com> wrote: > Hi, > > I'm working with Xilinx ISE 10.1 and I am having troubles with timing > constraints. > > I've successfully implemented my design with a 20 ns cycle time and found > that I needed to change something in the design. > > I make my changes and re-synthesize and implement design and I no longer > meet the same timing constraint. Okay, I've changed the logic slightly so > it's not going to be exactly the same, the static timing report says that= I > need a minimum cycle time of say 22 ns. > > So I only change the timing constraint to reflect a 25 ns and re-run the > process and it fails again saying I need a 32 ns cycle time now. Okay, so= I > change the cycle time to 35 ns and then I need a 46 ns cycle time. > > I think this is understandable because with a relaxed constraint some > optimizations or paths may not be chosen. When do you try it again with t= he > same constraint and when do you relax the constraint? Actually, that doesn't make sense to me since the timing constraints are supposed to be used in routing. From what I have seen, synthesis, such as Synplify, is often driven by a variety of specs, but not necessarily the "constraints" that Xilinx lets you input. Under the Lattice tools I have to enter a separate timing constraint to have it sent to Synplify when I use the GUI. Even this is just one master clock speed and so can't have the detail to constrain the paths that have to be fast and relax the paths that don't need to run at full speed. There are also settings for "optimize for speed" or "optimize for power". I can't imagine what is going on that a path or paths can be de- optimized repeatedly to always be just out of reach of your timing constraint. Have you looked at the paths to see if it is the same path each time? If path A fails with a 20 ns constraint and path B fails with a 25 ns constraint what is path B doing under the 20 ns constraint? How does the tool make it meet 20 ns (or maybe just under 22 ns) and be slower otherwise? Repeat with path C under the 35 ns, etc. See if you can figure out what is changing in the result. There has to be something odd going on. RickArticle: 147994
On 6/11/2010 1:49 AM, BrandonD wrote: > Hi, > > I'm working with Xilinx ISE 10.1 and I am having troubles with timing > constraints. > > I've successfully implemented my design with a 20 ns cycle time and found > that I needed to change something in the design. > Hi Brandon, It strikes me that if you are using a modern FPGA and can't meet a 20ns timing constraint, you have bigger problems than Xilinx's crappy software. However, there are folks here who are queuing up to help. Tell us which device you are using and a bit about the portion of your design which fails timing. Once that rafter is out of the way, we can look at the straws. Cheers, Syms.Article: 147995
"BrandonD" <BdOn003@n_o_s_p_a_m.n_o_s_p_a_m.gmail.com> writes: > Hi, > > I'm working with Xilinx ISE 10.1 and I am having troubles with timing > constraints. > > I've successfully implemented my design with a 20 ns cycle time and found > that I needed to change something in the design. > > I make my changes and re-synthesize and implement design and I no longer > meet the same timing constraint. Okay, I've changed the logic slightly so > it's not going to be exactly the same, the static timing report says that I > need a minimum cycle time of say 22 ns. > > So I only change the timing constraint to reflect a 25 ns and re-run the > process and it fails again saying I need a 32 ns cycle time now. Okay, so I > change the cycle time to 35 ns and then I need a 46 ns cycle time. > > I think this is understandable because with a relaxed constraint some > optimizations or paths may not be chosen. When do you try it again with the > same constraint and when do you relax the constraint? That looks quite weird. BUT - I'm a bit confused -why would you relax the constraint? You set your constraint to match the physical system you are targetting. Unless you can change that system to match your new constraint, you can't just go changing constraints - are you planning to slow your clock down as well? Or have you overconstrained the design? Cheers, Martin -- martin.j.thompson@trw.com TRW Conekt - Consultancy in Engineering, Knowledge and Technology http://www.conekt.net/electronics.htmlArticle: 147996
hi all, i have a ddr2 controller that works on 4 burst mode. ddr2 dq width is 16. i must provide 2, 32 bit data to the controller before writing it to memory. As well as when reading from a memory location it gives out 2, 32 bit data out. my processor is a master AHB, has a 32 bit address and 32 bit data bus. How can i connect these 2 together? problem is ahb sends out 32 bit data but i have to write to 64 bit data to memory. also the burst is 4, so it will write to 2, 32 bit memory locations, so there would be a 32 bit memory location that would be written with redundant data(if i write 64 bit data to memory). thank you, randeel.Article: 147997
"apple" <sys.apple2@gmail.com> wrote in message news:20d4dfc1-6146-4ae7-a050-0343f93536d2@q39g2000prh.googlegroups.com... >I would like to have capabilities of Chipscope for Xilinx FPGA's > are there any free alternatives to it? Not free and not a direct alternative but I came up with a prototpype of this... http://www.1pin-interface.com/ ..for debugging my own designs. I found it invaluable for board and device debug so produced the 1 Pin Interface. It doesn't give you the low level logic analyser access that Chipscope does but it works at the next layer giving you easy access to internal status or control registers. All that's needed is 1 FPGA IO pin and a GND. Nial.Article: 147998
Answering "Is it possible to get consistent implementation results?". Yes, if your constraints are set up properly you should get reliable expected performance _every_ time you build the device (as long as the constraits are met). Nial.Article: 147999
You dont have to write all of the data into memory, you can use the data masks. Jon --------------------------------------- Posted through http://www.FPGARelated.com
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