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
fpga00@gmail.com wrote: > hello, > > > I have a microblaze system with a uart and a timer (both produce > interrupt).The interrupts are handled by intc.Uart interrupt is the > highest priority followed by timer interrupt. > 1. can microblaze read only the required data from the uart at a > time.like say i want to get a frame from uart i.e data from one flag to > > another flag and then stop reading the bytes fro uart.When i tried > doing this i realised that I have to read the whole file that i am > transferring from the hyperterminal to the uart.but i dont want this to > > happen.any suggestions? The interrupt from uartlite is set when Receive FIFO goes from empty to non-empty. If you are not emptying the FIFO in your interrupt handler, you will not get any more receive interrupt. > 2. if the above behavior is not possible then i can send just one frame > > at a time to uart from the hyperterminal.if i do this the what happens > is that my timer is also running at the same time.when it expires it > interrupts the microblaze.But now if i send a frame to uart,it > interrupts microblaze and then i stop getting interrupts from the > timer.if i send another file (data) from hyperterminal to uart,it > interrupts the microblaze properly but the timer has stopped > working or otherwise its interrupt is not getting recognised. > I thought that probably when i am in the uart_int_handler probably my > timer interrupt was genearted and it was missed since i was in int. > routine.Is this a possibility.?is yes ,how can i avoid this?I want to > service the timer interrupt after i return fro the uart handler.my intc > > driver does not have XIN_SVC_ALL_ISRS_OPTION defined.so it does not > allow me to use it in the code. > Please share some information regarding this problem. > Thank you all. Timer interrupts and uart interrupts can of course be handled at the same time. The interrupt controller will keep track on what interrupts are active. You have to clear the interrupts serviced in your interrupt handler. EDK includes a default interrupt handler where you attach normal C function to handle each interrupt. More of this can be found in the documentation. Göran BilskiArticle: 84376
Thanks, I'll have a read through it. I did search for the answer to my problems via google - my apologies if i've asked a really simple question. Ta anyway :)Article: 84377
I have tried the following: To Assignment Name Value Enabled Pin_name1 Virtual Pin On Yes Pin_name1 Virtual Pin Clock clk_90_sys Yes clk_90_sys is the clock net name of the clock coming out of the Cyclone PLL. This clock is not routet to IOs, it is an internal clock. And yet I get the warning: #WARNING: Virtual pin clock "CLK_90_SYS" for virtual pin "Pin_name1" does not exist in the design - auto-selecting clock In the HELP of Quartus I read the following: CAUSE: You assigned the Virtual Pin Clock logic option to the specified pin. However this pin does not exist in the current design. When this condition occurs, Analysis & Synthesis automatically selects an appropriate virtual pin clock. ACTION: No action is required. To avoid receiving this message in the future, assign the Virtual Pin Clock assignment to an existing clock signal in the design. But "clk_90_sys" DOES exist. So what I am still doing wrong? Interesting thing: Some virtual pins do get an external clock coming into the FPGA. For these pins I do not get any warnings. Best regards Andr=E9Article: 84378
I would concur. We do exactly the same with a combination of licenses from these two vendors. ActiveHDL is very good at sorting file order out automatically and has in-built libraries that you don't need to compile to get going. The detachable waveform windows are very nice especially when you run multi-monitor machines like we do at Enterpoint. The cut and paste between these windows is also useful. In the last couple of years Mentor have improved the user interface of Modelsim and I hope they will continue the trend. I think they got a bit complacent. Five years ago I would have said Modelsim was the best but now I'm not so sure. Both these companies give out evaluation licenses so go try them out and see which you like best. John Adair Enterpoint Ltd. - Home of Broaddown2. The Ultimate Spartan3 Development Board. http://www.enterpoint.co.uk "Ray Andraka" <ray@andraka.com> wrote in message news:uxyie.11092$jj.3577@lakeread06... >I use a combination of Aldec and Modelsim PE. Aldec is more user friendly >and is really a complete design entry suite where Modelsim is a simulator. >I prefer Aldec for ease of use. Some customers insist on Modelsim, so I >have it. Modelsim is about the fussiest VHDL parser you'll find, and it is >absolutely true to the LRM, so it does serve as a decent VHDL code check to >make sure you aren't doing something that is slipping past your other tools >but might get you in trouble down the road. It is also useful to run on a >second machine while I'm, doing more design entry on Aldec. If I were to >buy just one, my choice would be Aldec for the more complete design entry >suite, the better customer support and the lower price. > > -- > --Ray Andraka, P.E. > President, the Andraka Consulting Group, Inc. > 401/884-7930 Fax 401/884-7950 > email ray@andraka.com http://www.andraka.com > "They that give up essential liberty to obtain a little temporary safety > deserve neither liberty nor safety." > -Benjamin Franklin, 1759 > >Article: 84379
Thanks Antti, Altera always talking about how successful the Jam Player is. I thought this is the right way to implement CPLD update in field using software. They publish the source code for Jam Byte-Code Player. But I'v never found a description of the Jam Byte-Code format. And the source code from Altera is also not well commented. If they do not really support the Jam player, they shouldn't promote it as an easy using, portable stuff. Regards DaniArticle: 84380
"dani" <user100@bluewin.ch> schrieb im Newsbeitrag news:1116405453.598651.58950@o13g2000cwo.googlegroups.com... > Thanks Antti, > > Altera always talking about how successful the Jam Player is. I thought > this is the right way to implement CPLD update in field using software. > They publish the source code for Jam Byte-Code Player. But I'v never > found a description of the Jam Byte-Code format. And the source code > from Altera is also not well commented. If they do not really support > the Jam player, they shouldn't promote it as an easy using, portable > stuff. > > Regards > > Dani > oh well they promote it :) just not updating the user resources and no support. there are several non altera uses of JAM as well, as example FS2 and Actel are using JAM, Xilinx tools can generate JAM (STAPL) but there is always a potential problem that the versions do have mismatch or something prevents the JAM useage. If you are working only with JAM files from single tool you should be ok though. Antti PS several FPGA vendors promote their JTAG stuff, Altera JAM, Xilinx XSVF, Lattice ispVM, etc.. all of them have problems. XSVF is changed with every new version of impact, the old player doesnt accept the new commands... everywhere about the same storyArticle: 84381
Hi, we recently bought the PCI Express Endpoint Core from Xilinx and implemented it into a Virtex 2 Pro 50 device. We have some experience with MGT connections and never had problems. But up to now we could not get the core running. We expect that the core will be detected by the BIOS of our PCI Express PC - but nothing happened. We could measure a stable reset pulse at boot-up of the PC. We expect that the signal trn_lnk_up_n will go high during link training and then goes low to indicate link up. But this signal stays low all the time. BTW: We heard that the Spread-Spectrum feature of some mainboards is critical, but as far we can measure there is no modulation of the 100MHz mainboard provided clock. Does anybody here have any experience with the usage of PCIexpress in FPGA ? Any suggestions welcome ! Regards AndreasArticle: 84382
"Andreas Loew" <fpgahardware@gmx.net> schrieb im Newsbeitrag news:ee8e58f.-1@webx.sUN8CHnE... > Hi, we recently bought the PCI Express Endpoint Core from Xilinx and implemented it into a Virtex 2 Pro 50 device. > > We have some experience with MGT connections and never had problems. > > But up to now we could not get the core running. We expect that the core will be detected by the BIOS of our PCI Express PC - but nothing happened. > > We could measure a stable reset pulse at boot-up of the PC. We expect that the signal trn_lnk_up_n will go high during link training and then goes low to indicate link up. But this signal stays low all the time. > > BTW: We heard that the Spread-Spectrum feature of some mainboards is critical, but as far we can measure there is no modulation of the 100MHz mainboard provided clock. > > Does anybody here have any experience with the usage of PCIexpress in FPGA ? Any suggestions welcome ! > > Regards Andreas both nital and dinigroup have working xilinx PCIe FPGA boards. try dini first see if they would be willing to help. AnttiArticle: 84383
Do you have any other information source which could be insteresting for me? You can just drop me the links which I find on the Web. Thanks DaniArticle: 84384
Gabor wrote: > amir.inti...@gmail.com wrote: > > Hello all, > > i am using the Spartan 3 (XC3s200). I have a clock signal > > coming into my FPGA expansion port and all i want to do is have the > > exact same clock signal going out of my FPGA, just delayed by about 3 > > micro seconds.I am using verilog and ISE 6.3. What is the best way to > > do this ???????. Thanks !!!!!!!!!!! > > Do you really mean 3 microseconds? That's a *LONG* time in a Spartan3. > What is your input clock frequency? Do you have another clock running > at a much higher frequency? > > If you really mean 3 microseconds, and you have a higher frequency > clock > I would suggest using a shift register. If you meant 3 nanoseconds, > you > need to use a DCM, preferably with external feedback, and use the phase > shift of the DCM to get your delay. hi, sorry, i should have mentioned the clock frequencies. Basically, the signal coming in is an "adcbusy" signal from an ADC converter. 1 - Busy 0 - not busy (take data). This positive cycle on the adcbusy signal lasts for about 1 micro second, followed by an 8 micro second negative cycle and so on.. This is the signal i want to reproduce on the output but with a slight delay. My Spartan3 is working at a woping 50 Mhz (20ns) however. I need to transmit the data to the PC in 1 micro second bursts, the same time as the adcbusy poitive cycle. Thats why i am trying to reproduce it at the output, but a delayed version. ....Cheers!!!!!Article: 84385
On Tue, 17 May 2005 17:55:45 -0700, Simon wrote: > B. Joshua Rosen wrote: >> On Sun, 15 May 2005 04:57:06 +0000, Ronald H. Nicholson Jr. wrote: >> >> Get yourself a Linux machine (x86 obviously, not Linux on PPC) to run your >> FPGA development environment. You can use the Mac as an X-Server, but >> thats as close as you are going to be able to get. It's inconceivable that >> the FPGA or CAE companies would add a third platform. > > Well, that's nice in theory, but certainly with the Wind/U toolkit it's > a no-go. Some X widgets (especially pulldown menus) are rendered in the > top-left corner of the screen, and the first choice is unavailable. In > other cases, the program simply crashes with a BadAccess error. That's curious because running the Xilinx tools with a Linux X-server works just fine, that's how I run it, I wonder what's broken in the OS-X implementation of X.Article: 84386
Hello, I'm trying to write a CORDIC macro for a polar transmitter FPGA design. I've used the parallel approach, but when I do a timing estimation the longest delay path is through the CORDIC routine, and limits the maximum clock rate of the whole design to about 40MHz. Other parts of the design on the same board need to run at much faster rate so I'm considering using a bit-serial version. As I understand it, the bit-parallel implementation has low latency and therefore high throughput, but because of the the word-wide shifts it clocks at a slower rate. Conversely, the bit-serial routine has a high latency and low throughput rate, but allows the board to run at a faster clock rate. Is this right? My question is: In the bit-serial implementation, you still need to perform the shift operation on the entire word to select the right bit to send to the bit-serial adder/subtractor, so how does this solve the problem of a slow clock rate due to the shift operation? Thanks, MeesArticle: 84387
randomdude@gmail.com wrote: > Thanks, I'll have a read through it. I did search for the answer to my > problems via google - my apologies if i've asked a really simple > question. > > Ta anyway :) Howdy Alan, Don't forget to search in groups.google.com... I found several discussions on strings (which probably aren't what you think they are), good VHDL books, and arrays - mostly in comp.lang.vhdl. To answer your "compiler" question, if the schematic tool generates moderately decent VHDL, then most synthesis tools (which is our term for compiler) will generate decent stuff. The last schematic tool I used was one made by Mentor (before it changed its name several times [the tool, not Mentor]). I was pretty happy with its output, but I'm sure it is way out of your price range. You can search in google.groups.com for more dicussion on this as well (in comp.arch.fpga). BTW, from the wording of your original question (especially the que(0)='a' part), it seems like you are approaching VHDL as a typical programming language. You know that it is not, right? When coding in an HDL (either VHDL or Verilog), you have to think in terms of the physical hardware you want (FF's, memories, counters, shift registers, and individual bits being OR'ed or AND'ed together). Have fun, MarcArticle: 84388
Hallo, i try to install debian (woody) for ppc on the xilinx ml310, but without success. It seems that the powerpc build in the fpga can't execute the binaries of the installer. Has someone tried to install a complete debian or suse linux on the ml310 board ? GreetingsArticle: 84389
One other old solution would be to put a x86 board inside, atleast thats what people did with earlier Macs, it was always a bit slower than regular PC but then the heat would be quite a extra burden. Don't know if Orange & the other guys are still in that business though.Article: 84390
I try to access the SDRAM in my program, and wait for enough time before writing data to SDRAM,but i find the value of SDRAM is FF,what should i do ? what is wrong?Article: 84391
"ARRON" <mlpei279@gmail.com> schrieb im Newsbeitrag news:ee8e5aa.-1@webx.sUN8CHnE... > I try to access the SDRAM in my program, and wait for enough time before writing data to SDRAM,but i find the value of SDRAM is FF,what should i do ? what is wrong? give more info in first place! what board what device what sdram controller IP core how its wired etc.. there could zilions of problems. the readout FF itself means nothing and is not enough to give you an answer AnttiArticle: 84392
Vaughn Betz wrote: > I definitely agree that STA provides 100% coverage, if you get all the > constraints correct. But there are many designs where the constraints are > incomplete, or may not be completely correct, and in those cases timing > simulation provides some level of back-up test. I agree a time sim can be a useful diagnostic when design rules go wrong. However, it is not an alternative to design rules. A timing sim covers function and timing too poorly to make a it an economical part of my standard development loop. > In designs with multicycles & cut paths set that I think there's value to > simulating to double-check you didn't cut or multicycle something you > shouldn't have. One of my design rules is no multicycle paths. I prefer pipelining. > Generally such constraints are set manually, and hence there > is the possibility of an error in entering them. True, but in a design composed of clocked processes, and synchronized inputs, fmax is often the only setting. It is impossible to verify synchronizers in simulation, so I don't try. > Of course, a timing > simulation is not guaranteed to catch a problem even if you did make a > mistake, but if you can afford the time, it may catch an issue. It might. When a design does not operate in the lab as it did in simulation either I violated a design rule or I need to add one. In this rare event, a timing sim is a useful tool. > Most designs I see use asynchronous clears, and don't set a recovery and > removal timing constraint to make sure the reset release occurs sufficiently > far from a clock edge for the design to come out of reset cleanly. That's a case of a design rule violation. I'd rather fix the design than see if it happens to work this time. > Timing simulation provides some level of back-up, although > it's definitely a poor substitute for a disciplined reset strategy and a > static timing analysis on it. I agree. -- Mike Treseler > >Article: 84393
Hi I have consulted the following link to initialize BRAM. http://www.fpga-faq.org/FAQ_Pa=ADges/0031_How_to_initialize_Blo=ADck_RAM.htm and it works fine for one BRAM module. Problem is when I use 2 RAMB4_S4 components (each address 9 to 0, data 3 to 0), thus totally 1024 addresses and 8 bit data width. Goal to initialize as follows address 0 : X"08" -- opcode address 1 : X"71" address 2 : X"34" address 3 : X"95" I am trying but MODELSIM simulation is showing unsatisfactory result so far ..... Could experienced one give comment for this? -------------------------------------------------------------------- .=2E.. -- RAMB4_S4 declared ONCE as below component RAMB4_S4 generic ( INIT_00 : BIT_VECTOR :=3D X"0000000000000000000000000000000000000000000000000000000000000000"; INIT_01 : BIT_VECTOR :=3D X"0000000000000000000000000000000000000000000000000000000000000000"; INIT_02 : BIT_VECTOR :=3D X"0000000000000000000000000000000000000000000000000000000000000000"; INIT_03 : BIT_VECTOR :=3D X"0000000000000000000000000000000000000000000000000000000000000000" ) .=2E.. -- 2 components of RAMB4_S4 are instantiated TWICE as below RAM0: RAMB4_S4 generic map( INIT_00 =3D> X"0000000000000000000000000000000000000000000000000000000000000000", INIT_01 =3D> X"0000000000000000000000000000000000000000000000000000000000000000", INIT_02 =3D> X"0000000000000000000000000000000000000000000000000000000000000000", INIT_03 =3D> X"0000000000000000000000000000000000000000000000000000000000000000" ) port map(WE=3D>we, EN=3D>en, RST=3D>rst, CLK=3D>clk, ADDR=3D>addr, DI=3D>di= (3 downto 0), DO=3D>do(3 downto 0)); RAM1: RAMB4_S4 generic map( INIT_00 =3D> X"0100000000000000000000000000000000000000000000000000000000000000", INIT_01 =3D> X"0000000000000000000000000000000000000000000000000000000000000000", INIT_02 =3D> X"0000000000000000000000000000000000000000000000000000000000000000", INIT_03 =3D> X"0000000000000000000000000000000000000000000000000000000000000000" ) port map(WE=3D>we, EN=3D>en, RST=3D>rst, CLK=3D>clk, ADDR=3D>addr, DI=3D>di= (7 downto 4), DO=3D>do(7 downto 4));Article: 84394
Hi, I want to thank you for this discussion. This highlights the criticity of static timing analyse, with a strong constraint coverage. The fact is that the design I'm working on requires a high level of safety, as mentionned Phil Hays. This means the cleanest design, well constrainted. But even with 100% cov. I will also use back annotated simulations, but more to cross checking a bug in the timing analyser tool/or a miss in the timing constraints as you mentionned through your experiences. Regards, Fred Cezilly Mike Treseler a écrit : > Vaughn Betz wrote: > >> I definitely agree that STA provides 100% coverage, if you get all the >> constraints correct. But there are many designs where the constraints >> are incomplete, or may not be completely correct, and in those cases >> timing simulation provides some level of back-up test. > > > I agree a time sim can be a useful diagnostic > when design rules go wrong. However, it is not > an alternative to design rules. > A timing sim covers function and timing too > poorly to make a it an economical part of my > standard development loop. > >> In designs with multicycles & cut paths set that I think there's value >> to simulating to double-check you didn't cut or multicycle something >> you shouldn't have. > > > One of my design rules is no multicycle paths. > I prefer pipelining. > >> Generally such constraints are set manually, and hence there is the >> possibility of an error in entering them. > > > True, but in a design composed of clocked processes, > and synchronized inputs, fmax is often the only setting. > It is impossible to verify synchronizers in simulation, so > I don't try. > >> Of course, a timing simulation is not guaranteed to catch a problem >> even if you did make a mistake, but if you can afford the time, it may >> catch an issue. > > > It might. When a design does not operate in the lab as > it did in simulation either I violated a design rule > or I need to add one. In this rare event, a timing > sim is a useful tool. > >> Most designs I see use asynchronous clears, and don't set a recovery >> and removal timing constraint to make sure the reset release occurs >> sufficiently far from a clock edge for the design to come out of reset >> cleanly. > > > That's a case of a design rule violation. > I'd rather fix the design than see > if it happens to work this time. > >> Timing simulation provides some level of back-up, although it's >> definitely a poor substitute for a disciplined reset strategy and a >> static timing analysis on it. > > > I agree. > > -- Mike Treseler > >> >>Article: 84395
Yeah, I'm trying very hard to visualise how the statements I make will map to hardware. Finding it kind of hard though, so I'm hoping some nice books will help me out. Ta for the searching-google-groups tip, btw! So are schematics compiled to VHDL which is then synthesised? I always assumed the HDL was 'compiled' into a schematic which was then mapped into the chip. I'm determined to master this fpga marlarkey.. even if it takes me a log time.. :) Thanks. -AlanArticle: 84396
I use a combination of Aldec and Modelsim PE. Aldec is more user friendly and is really a complete design entry suite where Modelsim is a simulator. I prefer Aldec for ease of use. Some customers insist on Modelsim, so I have it. Modelsim is about the fussiest VHDL parser you'll find, and it is absolutely true to the LRM, so it does serve as a decent VHDL code check to make sure you aren't doing something that is slipping past your other tools but might get you in trouble down the road. It is also useful to run on a second machine while I'm, doing more design entry on Aldec. If I were to buy just one, my choice would be Aldec for the more complete design entry suite, the better customer support and the lower price. -- --Ray Andraka, P.E. President, the Andraka Consulting Group, Inc. 401/884-7930 Fax 401/884-7950 email ray@andraka.com http://www.andraka.com "They that give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -Benjamin Franklin, 1759Article: 84397
Tommy Thorn wrote: > At $49 it starts looking good for small throw-away prototypes. It's > hard to tell, but it looks like the PCB has two unpopulated spots. SRAM? > I got one. There are two unpopulated spots labeled "CY7C1041CV33" which is a 4Mb (256Kx16) async SRAM. It came with the ByteBlaster II Parallel Port download cable, 9- and 25-pin serial cables, a 9V 500mA "wall-wart" supply and a CD-ROM with Quartus-II Web edition. At $49.00 I can guarantee Altera is giving these things away to win new design-ins. The manufacturer is "Axiom Manufacturing" at http://www.axman.com but you won't see this product on their website as it is private labeled for Future Electronics.Article: 84398
It doesn't do it because the signal would take longer to travel that path than if it's replicated, but also for more obvious reasons, and I exaplin... The signal at output of IOB register is not garanteed to be the signal that would get to the input, after it passed via the IO pad net. There's a lot of external interference that can occur, such as noise, .. Another example is if the output is sink-only (open-drain output). The IOB register may output a 1, while the IO pin is driven a zero (by an external driver sharing the same trace). This can even also occur with push-pull output, if there's (for some reason) contention on the IO pin. So, for those reasons, if a register need feedback to the FPGA fabric, but also require low output skew, then register replication is the only solution. To avoid warning, you could replicate the register in your source code, but this could make the code more obscure. You could always put a comment that specify that this extra register is for FPGA IOB register, the other for feedback.Article: 84399
JJ wrote: > One other old solution would be to put a x86 board inside, atleast > thats what people did with earlier Macs, it was always a bit slower > than regular PC but then the heat would be quite a extra burden. > > Don't know if Orange & the other guys are still in that business though. Not really cost-effective, what with throwaway $300 PCs! -a
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