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
On Mar 16, 12:08=A0pm, GaLaKtIkUs=99 <taileb.me...@gmail.com> wrote: > On Mar 15, 8:39=A0pm, d_s_klein <d_s_kl...@yahoo.com> wrote: > > > > > On Mar 13, 9:59=A0pm, GaLaKtIkUs=99 <taileb.me...@gmail.com> wrote: > > > > I googled for s g.723.1 codec ipcore for xilinx FPGA but didn't find > > > any :( > > > Please help! > > > > Mehdi > > > I Googled for "Next Week's Winning Lotto Numbers" and didn't find > > those either. > > > Here's a couple of links for you: > > <http://www.vhdl-online.de/tutorial/englisch/inhalt.htm> > > <http://en.wikipedia.org/wiki/G.723.1> > > > RK > > Thanks a lot! > I didn't know such a thing like wikipedia! Further I'll always take a > look there before asking stupid questions!!! > Waaaaaaaaaaaah ... VHDL is really cool!! I'll move away from > SystemVerilog. > > More seriously: Before asking I made some searches in major IP core > providers, opensource community, of course I took a look in wikipedia. > And I don't need courses of HDL since I have already an experience of > 8 years!! :) > One major reason why I'm looking for IPs is that I don't want to > beggin long administrative procedures of paying for G.723.1 patents. > The IP provider should have done this job!! :-) > > Any way I would like to thank you for the effort you made :-) > > -- > Mehdi I doubt you will find an IP Provide who has paid for the Patents. Typically the end user is responsible for Patent licensing. As an IP Provider, I can develop and offer an IP without paying for the patents. Best Regards, rudiArticle: 146401
rickman wrote: > You would think that. I just had to change some code to eliminate a > sync reset on a register to get rid of one level of LUTs in a Lattice > design which has a similar if not same FF structure with the dedicated > LSR input (Local Set/Reset). Synplify seems to not know how to use > that. Maybe this would get changed by the mapper, but I don't think > it does that. I was looking at it in the synthesis tool. Have you > tried an example? Does Synplify do a better job with the Xilinx parts > than they do with Lattice parts? I'm not convinced they do a good job > with the arithmetic units. It is more than once I've seen an adder > chain duplicated to get the carry out of the top! Hi Rick, I tried 5-input registered OR, like this: p_sync_or : process (clk) is begin if rising_edge(clk) then q <= a or b or c or d or e; end if; end process p_sync_or; In a Virtex4 (4 input LUTs) XST synthesis connects input 'e' to the set port of the register, and so does not require two LUTs to implement the gate. When it comes to Synplify I only have the actel edition ( only actel parts supported(?)) and I have hardly ever used it. I think actel parts only support async resets for their registers, in which case it is true as you say, that designing synchronous resets will generate extra logic in front of the registers. Maybe this is true for Lattice as well. In that case I would prefer using an async reset to avoid the extra logic and performance penalty. If the reset signal is synchronously deasserted, it can be constrained so that one is certain it will reach each register at a proper time. MagneArticle: 146402
Though it won't help with signed/unsigned issues, you may also want to look at the fixed point package. It automatically promotes result sizes to maintain accuracy for multiplication, addision and subtraction. You can use it for "integer" math by simply specifying your LSB index at 0. The idea is that intermediate result or operand resizing is not usually needed with fixed point, just a final resize prior to storage (like the implied resize that happens with integers). I really wish they had gone the extra step to make ufixed - ufixed = sfixed, but alas, that did not happen (not that it would be an issue in your problem). With that, we'd have 99% of the flexibility of integers (automatic signed and size promotion), with virtually unlimited data sizes, at reduced simulation performance (compared to integer, not signed/unsigned). AndyArticle: 146403
Hi. I was wondering if console output could be "locking" ISE and if redirecting to a file, or using flags for ISE tools to be silent could speed things up (xst, map, par). I use WinXP, btw. Cheers, hugoArticle: 146404
On Mar 16, 5:41=A0am, HS <hug...@gmail.com> wrote: > Hi. I was wondering if console output could be "locking" ISE and if > redirecting to a file, or using flags for ISE tools to be silent could > speed things up (xst, map, par). > > I use WinXP, btw. > > Cheers, > hugo Not enough to notice. ISE is very CPU intensive. RK.Article: 146405
On Mar 15, 9:08=A0pm, GaLaKtIkUs=99 <taileb.me...@gmail.com> wrote: > > And I don't need courses of HDL since I have already an experience of > 8 years!! :) > Mehdi Experience? Perhaps. Wisdom? No. RKArticle: 146406
d_s_klein wrote: > On Mar 16, 5:41 am, HS <hug...@gmail.com> wrote: >> Hi. I was wondering if console output could be "locking" ISE and if >> redirecting to a file, or using flags for ISE tools to be silent could >> speed things up (xst, map, par). >> >> I use WinXP, btw. >> >> Cheers, >> hugo > > Not enough to notice. ISE is very CPU intensive. It depends. :) I remember when they switched the GUI to QT (don't recall which release that was... ISE8 maybe?), they had a background process "cat.exe" that used up one entire CPU core. I suppose they used that to paste the console output into the GUI window... That one would've had me rolling on the floor laughing, hadn't it slowed down the entire process considerably on one-core machines. Fortunately they fixed that in the next SP, but still... Not that I use the GUI a lot, but for some quick testing of new releases it's handy sometimes. cu, Sean -- Replace "MONTH" with the three-letter abbreviation of the current month and the two-digit code for the current year (simple, eh?).Article: 146407
Hi, I want to buy books on CMOS digital circuit designs. Any advice on which is the best book on CMOS digital circuit design? I know the best textbook on software programming is a book written by MIT 4 professors "The Introduction to Algorithms". Knuth's "The Art of Computer Programming" 1-3 volumes are good, and I have read all of them, but they are at too low levels and cover too limited research area. I hope to buy the same books on CMOS digital circuit designs, even though many experts may have different points of views. Thank you. WengArticle: 146408
Can anyone tell me where to find out what Spartan6 and Virtex6 parts, packages, speed grades, and costs (small quantities) are or will be? I have a comparison charts from the Xilinx web site and I try to gleen availability from the Avnet website but this method seems like guess work. Thanks, Brad Smallridge AiVisionArticle: 146409
Brad, Xilinx provides all parts through authorized distributors. In North America, that is Avnet (as you have noted). I suggest you call them on the phone and request quotes and delivery. If they say 6 weeks, then it is as good as "in stock" but if it says 26 weeks that means the die are probably in die bank, and are not assembled into packages (yet). AustinArticle: 146410
Is AR#21774 the only official statement from Xilinx on this, or have I missed something? I want to mess about with non-standard load impedances. How do they maintain the common mode voltage? Do they string a potential divider between the outputs and servo the centre voltage to half-rail? Anyone know for sure? TIAArticle: 146411
On Mar 16, 2:04=A0am, Magne Munkejord <magnem...@yahoo.no> wrote: > rickman wrote: > > You would think that. =A0I just had to change some code to eliminate a > > sync reset on a register to get rid of one level of LUTs in a Lattice > > design which has a similar if not same FF structure with the dedicated > > LSR input (Local Set/Reset). =A0Synplify seems to not know how to use > > that. =A0Maybe this would get changed by the mapper, but I don't think > > it does that. =A0I was looking at it in the synthesis tool. =A0Have you > > tried an example? =A0Does Synplify do a better job with the Xilinx part= s > > than they do with Lattice parts? =A0I'm not convinced they do a good jo= b > > with the arithmetic units. =A0It is more than once I've seen an adder > > chain duplicated to get the carry out of the top! > > Hi Rick, > > I tried 5-input registered OR, like this: > > =A0 =A0 =A0p_sync_or : process (clk) is > =A0 =A0 =A0begin > =A0 =A0 =A0 =A0 =A0if rising_edge(clk) then > =A0 =A0 =A0 =A0 =A0 =A0 =A0q <=3D a or b or c or d or e; > =A0 =A0 =A0 =A0 =A0end if; > =A0 =A0 =A0end process p_sync_or; > > In a Virtex4 (4 input LUTs) XST synthesis connects input 'e' to the set > port of the register, and so does not require two LUTs to implement the > gate. > > When it comes to Synplify I only have the actel edition ( only actel > parts supported(?)) and I have hardly ever used it. > > I think actel parts only support async resets for their registers, in > which case it is true as you say, that designing synchronous resets will > generate extra logic in front of the registers. Maybe this is true for > Lattice as well. In that case I would prefer using an async reset to > avoid the extra logic and performance penalty. > > If the reset signal is synchronously deasserted, it can be constrained > so that one is certain it will reach each register at a proper time. > > Magne Magne, Your method is wrong ! You cannot connect 'e' to the set port of the register. It may compromise a new data which uses 'q'. Imagine the case: if 'e' is so earlier asserted that 'q' is asserted and changes a data 'New" which uses 'q' during setup or hold time around the next clock triggering edge.Next 'New' signal data may be invalid. p_sync_or : process (clk) is begin if rising_edge(clk) then q <=3D a or b or c or d or e; end if; end process p_sync_or; A : process(clk) begin if rising_edge(clk) then New <=3D a and q; end if; end process A; WengArticle: 146412
On Mar 12, 6:34=A0pm, John McCaskill <jhmccask...@gmail.com> wrote: > On Mar 12, 6:34=A0am, saras <saras_rajg...@yahoo.co.in> wrote: > > > hello all , I have triggered the chipscope on some error signal and is > > getting triggered on error.But i am unable to analyse the cause of the > > error. Is there any way we can sample the previous 20 or 50 samples > > before trigger so that i can see which are the signals causing the > > error ? Thanks in advance. > > In the trigger setup window, there is a subwindow named capture. =A0On > the right hand side of the capture window is a field named position > that defaults to 0. =A0Enter how many samples you want to see preceding > the trigger in that field. > > Regards, > > John McCaskillwww.FasterTechnology.com Thanks a lot. But in the trigger setup window, there is no subwindow named capture. All i have is " Enable trigger sequencer" , " Enable storage qualification" , "trigger width" , "match type", "max number of samples levels" , "samples depth" and then " Net connections". I'm using a licenesed Xilinx 10.1 version.Is there any problem with version or some other issue? Please help.Article: 146413
On Mar 16, 9:04=A0am, Andy <jonesa...@comcast.net> wrote: > Though it won't help with signed/unsigned issues, you may also want to > look at the fixed point package. It automatically promotes result > sizes to maintain accuracy for multiplication, addision and > subtraction. You can use it for "integer" math by simply specifying > your LSB index at 0. The idea is that intermediate result or operand > resizing is not usually needed with fixed point, just a final resize > prior to storage (like the implied resize that happens with integers). > > I really wish they had gone the extra step to make ufixed - ufixed =3D > sfixed, but alas, that did not happen (not that it would be an issue > in your problem). With that, we'd have 99% of the flexibility of > integers (automatic signed and size promotion), with virtually > unlimited data sizes, at reduced simulation performance (compared to > integer, not signed/unsigned). > Andy Actually, after converting the calculation to signed/unsigned types, it was still pretty groady, so I changed it back to integer and split it up. I need to optimize this design for size and I find that easier if I separate the arithmetic functions so I can more easily see how they are being implemented. I had originally used integer because it make the calculations easy, but doubted that this was the best way to express the calculations because of the mess of converting the inputs from signed/unsigned and back. As it turned out mixing signed and unsigned is still pretty messy. RickArticle: 146414
On Mar 16, 8:24=A0pm, Weng Tianxiang <wtx...@gmail.com> wrote: > On Mar 16, 2:04=A0am, Magne Munkejord <magnem...@yahoo.no> wrote: > > > > > > > > > rickman wrote: > > > You would think that. =A0I just had to change some code to eliminate = a > > > sync reset on a register to get rid of one level of LUTs in a Lattice > > > design which has a similar if not same FF structure with the dedicate= d > > > LSR input (Local Set/Reset). =A0Synplify seems to not know how to use > > > that. =A0Maybe this would get changed by the mapper, but I don't thin= k > > > it does that. =A0I was looking at it in the synthesis tool. =A0Have y= ou > > > tried an example? =A0Does Synplify do a better job with the Xilinx pa= rts > > > than they do with Lattice parts? =A0I'm not convinced they do a good = job > > > with the arithmetic units. =A0It is more than once I've seen an adder > > > chain duplicated to get the carry out of the top! > > > Hi Rick, > > > I tried 5-input registered OR, like this: > > > =A0 =A0 =A0p_sync_or : process (clk) is > > =A0 =A0 =A0begin > > =A0 =A0 =A0 =A0 =A0if rising_edge(clk) then > > =A0 =A0 =A0 =A0 =A0 =A0 =A0q <=3D a or b or c or d or e; > > =A0 =A0 =A0 =A0 =A0end if; > > =A0 =A0 =A0end process p_sync_or; > > > In a Virtex4 (4 input LUTs) XST synthesis connects input 'e' to the set > > port of the register, and so does not require two LUTs to implement the > > gate. > > > When it comes to Synplify I only have the actel edition ( only actel > > parts supported(?)) and I have hardly ever used it. > > > I think actel parts only support async resets for their registers, in > > which case it is true as you say, that designing synchronous resets wil= l > > generate extra logic in front of the registers. Maybe this is true for > > Lattice as well. In that case I would prefer using an async reset to > > avoid the extra logic and performance penalty. > > > If the reset signal is synchronously deasserted, it can be constrained > > so that one is certain it will reach each register at a proper time. > > > Magne > > Magne, > Your method is wrong ! > > You cannot connect 'e' to the set port of the register. It may > compromise a new data which uses 'q'. > > Imagine the case: if 'e' is so earlier asserted that 'q' is asserted > and changes a data 'New" which uses 'q' during setup or hold time > around the next clock triggering edge.Next 'New' signal data may be > invalid. > > p_sync_or : process (clk) is > =A0 =A0 =A0begin > =A0 =A0 =A0 =A0 =A0if rising_edge(clk) then > =A0 =A0 =A0 =A0 =A0 =A0 =A0q <=3D a or b or c or d or e; > =A0 =A0 =A0 =A0 =A0end if; > =A0 =A0 =A0end process p_sync_or; > > A : process(clk) > =A0 =A0 =A0begin > =A0 =A0 =A0 =A0 =A0if rising_edge(clk) then > =A0 =A0 =A0 =A0 =A0 =A0 =A0New <=3D a and q; > =A0 =A0 =A0 =A0 =A0end if; > =A0 =A0 =A0end process A; > > Weng I'm not clear on what your concern is. Perhaps you are thinging Magne is talking about e being connected to an async set? He is describing a FF with an async set which will do exactly the same thing as an OR of signal e with the rest of the inputs. RickArticle: 146415
On Tue, 16 Mar 2010 13:12:04 -0700, "Brad Smallridge" <bradsmallridge@dslextreme.com> wrote: > Can anyone tell me where to find out what >Spartan6 and Virtex6 parts, packages, speed >grades, and costs (small quantities) are or >will be? > I have a comparison charts from the Xilinx web >site and I try to gleen availability from the >Avnet website but this method seems like guess >work. > Thanks, > Brad Smallridge > AiVision > The parts are ready but the tools aren't. We're having ghastly timing problems and we aren't getting support. This is apparently chronic with S6. Plus, they pulled the plug on our distributor, NuHorizons, so we no longer have them putting pressure on Xilinx for support. I'd stick with S3 until they get all this fixed. JohnArticle: 146416
Weng Tianxiang wrote: > I want to buy books on CMOS digital circuit designs. Any advice on > which is the best book on CMOS digital circuit design? At least "Nanometer CMOS ICs, From Basics to ASICs" written by Harry Veendrick is quite nice overall book and is up to date with the technology. The only problem with the book is the price, which is quite high. --KimArticle: 146417
On Mar 17, 6:17=A0am, John Larkin <jjlar...@highNOTlandTHIStechnologyPART.com> wrote: > On Tue, 16 Mar 2010 13:12:04 -0700, "Brad Smallridge" > > <bradsmallri...@dslextreme.com> wrote: > > Can anyone tell me where to find out what > >Spartan6 and Virtex6 parts, packages, speed > >grades, and costs (small quantities) are or > >will be? > > I have a comparison charts from the Xilinx web > >site and I try to gleen availability from the > >Avnet website but this method seems like guess > >work. > > Thanks, > > Brad Smallridge > > AiVision > > The parts are ready but the tools aren't. We're having ghastly timing > problems and we aren't getting support. This is apparently chronic > with S6. Plus, they pulled the plug on our distributor, NuHorizons, so > we no longer have them putting pressure on Xilinx for support. > > I'd stick with S3 until they get all this fixed. > > John hi John it seems Xilinx is good on the road i predicted: I think some years ago I said that Xilinx products will be unusable because their software will become unusable when they hit 45nm technology. AnttiArticle: 146418
"John Larkin" <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote in message news:rnl0q5d1in9rnqhufimeij4jur878p750o@4ax.com... > On Tue, 16 Mar 2010 13:12:04 -0700, "Brad Smallridge" > <bradsmallridge@dslextreme.com> wrote: > >> Can anyone tell me where to find out what >>Spartan6 and Virtex6 parts, packages, speed >>grades, and costs (small quantities) are or >>will be? >> I have a comparison charts from the Xilinx web >>site and I try to gleen availability from the >>Avnet website but this method seems like guess >>work. >> Thanks, >> Brad Smallridge >> AiVision >> > > The parts are ready but the tools aren't. We're having ghastly timing > problems and we aren't getting support. There might be some fixes in ISE 11.5 which was release yesterday. From the website: "11.5 includes important updates and supports production devices for the Virtex-6 and Spartan-6 families. However, several workarounds may be required for Virtex-6 and some Spartan-6 customers using 11.5. Please review Answer Record 32147 before you upgrade to 11.5. These workarounds will be addressed in 12.1, scheduled for May 2010. Designers targeting families other than Virtex-6 or Spartan-6 do not need to install 11.5. To ensure that ISE Design Suite functions properly, it is essential that all 11.5 updates are applied together." Hans www.ht-lab.com This is apparently chronic > with S6. Plus, they pulled the plug on our distributor, NuHorizons, so > we no longer have them putting pressure on Xilinx for support. > > I'd stick with S3 until they get all this fixed. > > John > From prenom.nom@gmail.com Wed Mar 17 00:27:49 2010 Path: unlimited.newshosting.com!s03-b40!npeersf02.iad.highwinds-media.com!npeer03.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!nnrp17-1.free.fr!not-for-mail Date: Wed, 17 Mar 2010 09:27:49 +0100 From: Matthieu Michon <prenom.nom@gmail.com> Newsgroups: comp.arch.fpga Subject: Re: Question Rdging xilinx chipscope pro triggering Message-Id: <20100317092749.b1d0092c.prenom.nom@gmail.com> References: <18ffe316-e81d-4a95-bfac-2ae70ec5b2e3@s36g2000prf.googlegroups.com> <fe54e9b5-b2e2-45d1-88f0-93934200348d@d2g2000yqa.googlegroups.com> <0164d27f-7a6f-42a0-b020-ae95a060d6bb@m27g2000prl.googlegroups.com> X-Newsreader: Sylpheed 2.7.1 (GTK+ 2.10.14; i686-pc-mingw32) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Lines: 20 Organization: Guest of ProXad - France NNTP-Posting-Date: 17 Mar 2010 09:27:49 MET NNTP-Posting-Host: 213.215.9.6 X-Trace: 1268814469 news-4.free.fr 30353 213.215.9.6:56800 X-Complaints-To: abuse@proxad.net Xref: unlimited.newshosting.com comp.arch.fpga:99247 X-Received-Date: Wed, 17 Mar 2010 08:28:49 UTC (s03-b40) On Tue, 16 Mar 2010 20:54:46 -0700 (PDT) saras <saras_rajgiri@yahoo.co.in> wrote: (...) > But in the trigger setup window, there is no subwindow named capture. > All i have is " Enable trigger sequencer" , " Enable storage > qualification" , "trigger width" , "match type", "max number of > samples levels" , "samples depth" and then " Net connections". I'm > using a licenesed Xilinx 10.1 version.Is there any problem with > version or some other issue? Please help. Your are probably looking at the Core Inserter software. John was refering to the "Position" number field present in the ChipScope Pro __Analyzer__ software (the application used for capturing and displaying waveforms). Regards -- Matthieu Michon <prenom.nom@gmail.com>Article: 146419
jfh pisze: > Hi, > > The setup is the following : > - a stratix II GX hosts a Nios IIe > - one timer running continuously at a period of 1s, priority level 6, > the handler for this interrupt has some functions calls to usleep and > performs polling, this handler is made interruptible via the use of > the alt_irq_interruptible and alt_irq_noninterruptible pair of > functions > - one hw interrupt generates an interrupt every 500 us with a prioity > level 2 a uart running at 115 kbauds using the small driver > implementation (no interrupt) > - the main function of the C program is waiting forever for characters > to come in via the uart and when a command is received it processes it > and gets back to expecting a new command forever > > With this setup it seems that the uart does not work well. When I > pause the program after noticing it does not respond the program is > almost always located in the usleep function called from the low > priority interrupt handler. Is there anything in the setup described > that is bad design ? What am i missing ? All ideas are welcome ! > > Best regards, > > JF Hi, AFAIK interrupts in NIOS II e have software processing. So all interrupts are connected to same handler with soft processing. Change nios version to faster one or you can try add interrupt custom instruction to reduce interrupt latency. BR AdamArticle: 146420
rickman wrote: > On Mar 16, 8:24 pm, Weng Tianxiang <wtx...@gmail.com> wrote: >> On Mar 16, 2:04 am, Magne Munkejord <magnem...@yahoo.no> wrote: >> >> >> >> >> >> >> >>> rickman wrote: >>>> You would think that. I just had to change some code to eliminate a >>>> sync reset on a register to get rid of one level of LUTs in a Lattice >>>> design which has a similar if not same FF structure with the dedicated >>>> LSR input (Local Set/Reset). Synplify seems to not know how to use >>>> that. Maybe this would get changed by the mapper, but I don't think >>>> it does that. I was looking at it in the synthesis tool. Have you >>>> tried an example? Does Synplify do a better job with the Xilinx parts >>>> than they do with Lattice parts? I'm not convinced they do a good job >>>> with the arithmetic units. It is more than once I've seen an adder >>>> chain duplicated to get the carry out of the top! >>> Hi Rick, >>> I tried 5-input registered OR, like this: >>> p_sync_or : process (clk) is >>> begin >>> if rising_edge(clk) then >>> q <= a or b or c or d or e; >>> end if; >>> end process p_sync_or; >>> In a Virtex4 (4 input LUTs) XST synthesis connects input 'e' to the set >>> port of the register, and so does not require two LUTs to implement the >>> gate. >>> When it comes to Synplify I only have the actel edition ( only actel >>> parts supported(?)) and I have hardly ever used it. >>> I think actel parts only support async resets for their registers, in >>> which case it is true as you say, that designing synchronous resets will >>> generate extra logic in front of the registers. Maybe this is true for >>> Lattice as well. In that case I would prefer using an async reset to >>> avoid the extra logic and performance penalty. >>> If the reset signal is synchronously deasserted, it can be constrained >>> so that one is certain it will reach each register at a proper time. >>> Magne >> Magne, >> Your method is wrong ! >> >> You cannot connect 'e' to the set port of the register. It may >> compromise a new data which uses 'q'. It wasn't me, XST did it! >> >> Imagine the case: if 'e' is so earlier asserted that 'q' is asserted >> and changes a data 'New" which uses 'q' during setup or hold time >> around the next clock triggering edge.Next 'New' signal data may be >> invalid. >> >> p_sync_or : process (clk) is >> begin >> if rising_edge(clk) then >> q <= a or b or c or d or e; >> end if; >> end process p_sync_or; >> >> A : process(clk) >> begin >> if rising_edge(clk) then >> New <= a and q; >> end if; >> end process A; >> >> Weng Since the 'set' port of the register is synchronous the circuit will behave as your code above describes it. q will not change state before rising edge of clk even if e is asserted earlier. I would think the timing tools knows the required setup time for the synchronous set/reset ports, and cover their paths as any other synchronous elements. > > I'm not clear on what your concern is. Perhaps you are thinging Magne > is talking about e being connected to an async set? He is describing > a FF with an async set which will do exactly the same thing as an OR > of signal e with the rest of the inputs. > > Rick The second instance of 'async' I would correct to 'sync', but I guess that was a typo. MagneArticle: 146421
On Wed, 17 Mar 2010 00:13:52 -0700 (PDT), Antti <antti.lukats@googlemail.com> wrote: >On Mar 17, 6:17 am, John Larkin ><jjlar...@highNOTlandTHIStechnologyPART.com> wrote: >> On Tue, 16 Mar 2010 13:12:04 -0700, "Brad Smallridge" >> >> <bradsmallri...@dslextreme.com> wrote: >> > Can anyone tell me where to find out what >> >Spartan6 and Virtex6 parts, packages, speed >> >grades, and costs (small quantities) are or >> >will be? >> > I have a comparison charts from the Xilinx web >> >site and I try to gleen availability from the >> >Avnet website but this method seems like guess >> >work. >> > Thanks, >> > Brad Smallridge >> > AiVision >> >> The parts are ready but the tools aren't. We're having ghastly timing >> problems and we aren't getting support. This is apparently chronic >> with S6. Plus, they pulled the plug on our distributor, NuHorizons, so >> we no longer have them putting pressure on Xilinx for support. >> >> I'd stick with S3 until they get all this fixed. >> >> John > >hi John > >it seems Xilinx is good on the road i predicted: > >I think some years ago I said that Xilinx products will be unusable >because their software will become unusable when they hit 45nm >technology. > >Antti Their software has been spinning out of control for years. I've seen other big software projects like this. Once that pattern is established, it's nearly impossible to fix. What's special about 45 nm? Is it just the sheer increase in chip complexity that breaks the tools? The development software just sees CLBs. The worst problem we're having now is with the hard-core DRAM controller. The architecture is essentially unmanagable, and the tools throw timing errors even at low clock rates, or just refuse to compile. We've had cases open on this for months (we got very early S6/45 units) and the silence is deafening. I'm sure they'll patch it up eventually. The precise quantization of "eventually" will determine if we still have a customer and whether 2 man-years of work will be a total loss. JohnArticle: 146422
Looks like a phase controlled DCO. Maybe the frequency/phase d/dt fm effect can be used? It does look messy, modulus if its a power of 2 should be easy to remove by a (x downto y) subrange select. If modulus is n/(n-1) then consider MASH or bitstream delta sigma. OR use a fixed point overflow clock gating. Has anyone ever tried n/(n-2) via up/down clock gating of an n divider?? Cheers JackoArticle: 146423
On 16 Mar, 06:16, jfh <jeanfrancois.has...@gmail.com> wrote: > Hi, > > The setup is the following : > - a stratix II GX hosts a Nios IIe > - one timer running continuously at a period of 1s, priority level 6, > the handler for this interrupt has some functions calls to usleep and > performs polling, this handler is made interruptible via the use of > the alt_irq_interruptible and alt_irq_noninterruptible pair of > functions > - one hw interrupt generates an interrupt every 500 us with a prioity > level 2 a uart running at 115 kbauds using the small driver > implementation (no interrupt) > - the main function of the C program is waiting forever for characters > to come in via the uart and when a command is received it processes it > and gets back to expecting a new command forever > > With this setup it seems that the uart does not work well. When I > pause the program after noticing it does not respond the program is > almost always located in the usleep function called from the low > priority interrupt handler. Is there anything in the setup described > that is bad design ? What am i missing ? All ideas are welcome ! > > Best regards, > > JF maybe just have the one interrupt at 500us and every 2 interrupts do the 1s stuff. Maybe set a flag to return to sleep in the interrupt and do the actual sleep in the main loop. Cheers JackoArticle: 146424
Andrew, The common mode voltage is referenced to the band gap precise voltage reference, which is in every device since Virtex. The currents are set by design, and the output impedance is set by a resistance. LVDS has a 100 ohm transmit termination, as well as having a 100 ohm receive termination. If you don't use 100 ohms, it won't match! You can probably go anywhere from 75 ohms to 150 ohms, though, and the reflections will not be so bad that it affects the data link. It is a fairly easy exercise to add resistance matching networks, so that the device "sees" 100 ohms, and your transmission media "sees" what it likes. The matching networks add loss, so the distance you can travel is made less, but you can have a perfect match. Running some simulations using IBIS models will tell you what you need to know. And, no, we are not going to provide a schematic drawing of the LVDS output stage, as that is proprietary, and we have already provided everything you need (if you simulate the LVDS with the IBIS or encrytped hspice models that we provide). Austin
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