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
Hi, I am using the Xilinx DDR3 controller (MIG). I need to trigger rewriting the mode registers of the DDR3 component during operation. The controller does not support this, so I'm working on a modification. The mode register write has to be performed after all read and write commands submitted so far have been processed (more specifically, after a complete row has been read or written). Any ideas on how to implement this? One possible approach is to wait until all bank machines are idle, trigger a refresh (so the operation won't be interrupted by the refresh timer), connect the address and control lines to my own logic instead of the memory controller and performing the mode register write manually (making sure to restore the connection in time for the next refresh operation). However, this approach seems a bit like a kludge to me. Also, I'm not sure how to avoid the ZQ calibration or a periodic read getting in the way. Any better ideas? Ideally, I would be able to enqueue a command to write the mode registers, just like I can enqueue read and write commands. I use "strict" ordering, so reordering should not be an issue. Martin -- Wise men don't need advice. Fools don't take it. - Benjamin FranklinArticle: 151376
Hi Last time I asked something about max array size. Now I have another problem and I think it has also to do with the memory. When I have a bigger program, sometimes it run how I want. But if I commented something out, the programm stops very early (the commented part are some thousand std_outs in a loop or something else, no data manipulation what influence my control flow). Sometimes it restarts at a specific point. It doesn't do what I want, but it does it allways the same way. So the problem seems to be reproducible, but arbitrarily. So what can the problem be? Unfortunately, I can't post a example, because the code is to large (but if you interessted I can send them via email). I need some ideas where I have to search for the problems. Thanks for every direct help or information where I can find some. Regards, TobiasArticle: 151377
Martin Herrmann wrote: > Hi, > > I am using the Xilinx DDR3 controller (MIG). I need to trigger rewriting > the mode registers of the DDR3 component during operation. The > controller does not support this, so I'm working on a modification. > > The mode register write has to be performed after all read and write > commands submitted so far have been processed (more specifically, after > a complete row has been read or written). Any ideas on how to implement > this? > > One possible approach is to wait until all bank machines are idle, > trigger a refresh (so the operation won't be interrupted by the refresh > timer), connect the address and control lines to my own logic instead of > the memory controller and performing the mode register write manually > (making sure to restore the connection in time for the next refresh > operation). > Don't even think about inserting your control logic at the chip interface. The timing won't work. > However, this approach seems a bit like a kludge to me. Also, I'm not > sure how to avoid the ZQ calibration or a periodic read getting in the > way. > > Any better ideas? Ideally, I would be able to enqueue a command to write > the mode registers, just like I can enqueue read and write commands. I > use "strict" ordering, so reordering should not be an issue. > > > Martin You talk about "The" DR3 controller, but really there's a different one for each device family. For the ones that don't use the hard MCB core you might find the right source module that schedules the refresh and other operations. The mode register write requires all banks to be precharged, as does the refresh command, so inserting the mode write just before starting a refresh would be a good place for it (or following a refresh after waiting Trfc). -- GaborArticle: 151378
FPGA Camp (http://www.fpgacentral.com/fpgacamp) is a conference, which brings engineers together to discuss FPGA, mainly NextGen FPGA technology, application, methodology, best practices and challenges. Also provide a location to meet other local FPGA designers to share their stories. Since its inception in Year 2009, FPGA Camp decided to stay vendor neutral. The attendance is completely FREE and so is putting up a booth. Due to the this approach we soon have been coined as an Open Source conference by Industry leaders like Eric Bogatin, Colin Warwick & Max Maxfield. REGISTER NOW http://bit.ly/fcjHCJ The event will focus on demonstrating key technologies available to bring processor inside the FPGAs. It will provide a glimpse of what to expect in the future, and how to use these great features for your next project. AGENDA (visit http://www.fpgacentral.com/fpgacamp for details) 4:00 PM Exhibit booths 4:25 PM Introductions 4:30 PM Tech Talk 1: On Die Instrumentation 5:05 PM Tech Talk 2: PCIe 3.0 case study 5:40 PM Tech Talk 3: Design Choices for Embedded Real-Time Control Systems 6:15 PM Dinner & Exhibit 6:45 PM Vendor Presentation 7:00 PM Panel: "State Of FPGAs - Current & Future" - Moderated by Dave Orecchio, CEO, Gaterocket 8:15 PM Closing Speakers & Moderator Dave Orecchio - GateRocket Gordon Hands - Lattice Chris Eddington - Synopsys Dave Bursky - Chip Design Magazine and more... Sponsors Lattice Semiconductor Altium Rhino Labs Agilent Technologies IEEE CNSV PLDA REGISTER NOW http://bit.ly/fcjHCJ (registration is FREE & Simple) Organized by: FPGA Cenral ( http://www.fpgacentral.com )Article: 151379
Gabor (2011-03-29): >> One possible approach is to wait until all bank machines are idle, >> trigger a refresh (so the operation won't be interrupted by the refresh >> timer), connect the address and control lines to my own logic instead of >> the memory controller and performing the mode register write manually > > Don't even think about inserting your control logic at the chip > interface. The timing won't work. I realize that the timing is quite challenging for the data lines, but isn't it easier for the address and control (RAS#, CAS#, WE#) lines? >> Any better ideas? Ideally, I would be able to enqueue a command to write >> the mode registers, just like I can enqueue read and write commands. I >> use "strict" ordering, so reordering should not be an issue. > > You talk about "The" DR3 controller, but really there's a different one > for each device family. Of course - my mistake. I'm using a Virtex6 device. > and other operations. The mode register write requires all banks > to be precharged, as does the refresh command, so inserting the > mode write just before starting a refresh would be a good place > for it (or following a refresh after waiting Trfc). That sounds like a good idea, I will have a look at the refresh logic. Martin -- Wise men don't need advice. Fools don't take it. - Benjamin FranklinArticle: 151380
"mah@k-space.org" <mah@n_o_s_p_a_m.k-space.org> writes: > HELLO > I want to use the below function to the xilinx system generator MCode > Block, but I faced some errors . how to solve this problem.The code > executed well in matlab but when I used MCode Block it gave me errors. > +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > %the function is: > +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > function s=dyn_focusing(RF,Lin) > prec = {xlSigned, 16, 0, xlRound, xlWrap}; > persistent m, m=xl_state(zeros(1,len),prec); > s=m; > len=512; > for k=1:len > if Lin(k)<=len > s(k)=RF(1,Lin(k)); > if Lin(k)>len > s(k)=0; > end > end > end > +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > where RF and Lin are (1X512) array > It might help us to help you if you tell us what errors you get! Cheers, Martin -- martin.j.thompson@trw.com TRW Conekt - Consultancy in Engineering, Knowledge and Technology http://www.conekt.co.uk/capabilities/39-electronic-hardwareArticle: 151381
On Mon, 28 Mar 2011 21:29:47 -0700, Mark wrote: > On Mar 29, 8:25Â am, Jim Granville <j.m.granvi...@gmail.com> wrote: >> On Mar 28, 11:01Â pm, Mark <markjsu...@gmail.com> wrote: >> >> > 3) Use 1-to-2 buffer (74xxx series) after the oscillator output and >> > split the clock to CPLD and the DAC. >> >> Â If you are jitter paranoid, then it is always best to avoid going >> thru complex logic. >> >> -jg > > Thanks jg. What do you suggest? Normal 74LS/74HC buffers or clock > buffers? Is there any advantage of using clock buffers over normal > buffers for digital audio? Neither 74LS nor 74HC. Possibly 74AC, or purpose built clock buffers. Fast clean edges (not HC) and full signal level switching (not LS) are important, but watch for edges so fast they propagate right through into the audio output. Also, run the buffer from its own power - at least a LC Pi filter to keep its supply quiet. (And I recommend at least the same, or separate regulators, to keep FPGA switching noise out of the main logic supply.) I used 74AC74 for re-clocking on an audio DAC. The effect on glitch energy from the (R-2R ladder DAC) was noticeable, and groups of listeners claimed to hear differences with different clock buffers (As I did). (I started with 74F74, when I switched to 74AC they stopped sending DACs back to the factory.) The US magazine "Stereophile" gave the result a "Class A" rating, for what that's worth. - BrianArticle: 151382
>Hi > >Last time I asked something about max array size. Now I have another >problem and I think it has also to do with the memory. > >When I have a bigger program, sometimes it run how I want. But if I >commented something out, the programm stops very early (the commented >part are some thousand std_outs in a loop or something else, no data >manipulation what influence my control flow). Sometimes it restarts at a >specific point. It doesn't do what I want, but it does it allways the >same way. So the problem seems to be reproducible, but arbitrarily. > >So what can the problem be? Unfortunately, I can't post a example, >because the code is to large (but if you interessted I can send them via >email). > >I need some ideas where I have to search for the problems. Thanks for >every direct help or information where I can find some. > >Regards, >Tobias > You might get a response if you post on the Xilinx forum at: http://forums.xilinx.com/t5/EDK-and-Platform-Studio/bd-p/EDK --------------------------------------- Posted through http://www.FPGARelated.comArticle: 151383
On 31.03.2011 10:34, RCIngham wrote: > > You might get a response if you post on the Xilinx forum at: > http://forums.xilinx.com/t5/EDK-and-Platform-Studio/bd-p/EDK > > > --------------------------------------- > Posted through http://www.FPGARelated.com I did it. I hope there's someone who can help. Here's the link: http://forums.xilinx.com/t5/Embedded-Processing/Microblaze-prorgam-behavior/td-p/140198Article: 151384
I've run into a bit of a problem. A camera module I would like to use has a MIPI CSI-2 interface. The SoC I want to connect to only has a parallel camera interface. How difficult would it be to use an FPGA in between? Or are there already MIPI to parallel converters? -- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... nico@nctdevpuntnl (punt=.) --------------------------------------------------------------Article: 151385
Hey guys, I work in the FPGA development kit sector, and I was just looking to see if you guys had any ideas regarding what would be your MUST HAVE features on an FPGA dev kit? Thanks in advance! AllenArticle: 151386
On 1 Apr., 07:43, allen <ayho...@gmail.com> wrote: > Hey guys, > > I work in the FPGA development kit sector, and I was just looking to > see if you guys had any ideas regarding what would be your MUST HAVE > features on an FPGA dev kit? > > Thanks in advance! > Allen Hi Allen, the only thing on which everyone will positively agree is an FPGA at least. (Oh, I forgot to mention the RS-232! :-) ) Everything else depends on the application, even the exact type of FPGA. You either end up with something that is so mediocre, that only beginners will use it if the price is low enough, but remains useless for professional designers, Or the board will be so overloaded and expensive, that noone can or will spend the money for it. Look for example at the boards offered on the Xilinx webpage. There you find a number of boards for each FPGA family. You almpost always find a Starter board, some DSP specialized board, some Video board, maybe a FPGA computing board and so on. Maybe you could explain a little more about the background and intention of your query. Have a nice synthesis EilertArticle: 151387
On 29 Mrz., 17:18, "cavalry" <Cavalrys@n_o_s_p_a_m.gmail.com> wrote: > Hey, > > I need to send data(1158 samples x 21 features) to Spartan 3e FPGA board > from the matlab workspace with rs232 and then do K-means classification o= n > that and then return the data (1158samples x 1 group(1-4)) from the FPGA > back into the matlab workspace. How do I do that? Kindly let me know if > there is some tutorial which is spartan 3e specific on this which can tel= l > me how to do it step by step. > > Thanks and Regards, > Cavalry > > --------------------------------------- =A0 =A0 =A0 =A0 > Posted throughhttp://www.FPGARelated.com Hi, search the forums at the Xilinx page for this. It's a common topic. One more hint if you are a beginner: Picoblaze and kcuart may be helpful for your approach, though it's a slow connection. Another approach would be using HW-Cosimulation with sysgen, which may work faster, but is expensive. Have a nice synthesis EilertArticle: 151388
On Apr 1, 7:43=A0am, allen <ayho...@gmail.com> wrote: > Hey guys, > > I work in the FPGA development kit sector, and I was just looking to > see if you guys had any ideas regarding what would be your MUST HAVE > features on an FPGA dev kit? > > Thanks in advance! > Allen It depends on the purpose of the board. For beginners and hobbyst "must have" features will be different. For professional users "must have" features will be different. But there may be some common features like rs232 or external memory.Article: 151389
Hi, A student board should contain : - Switches - Buttons - Seven Segments Displays (Multiplexed) - LCD or a port for a standard 16x2 LCD - Port for an optional graphical LCD (very nice) - RS232 - USB - Memory=20 - Minimal RGB driver - Optional port for more Clock sources - A/D converter - D/A converter - An add-on for motor driving (seperate board) - PS/2 interface A good idea for educational board is to that the FPGA could be changed e.g.= Altera or Xilinx. A decent board should have enough features so that a stu= dent can develop his graduate project on it so that students can spend more= time on developing the hardware. Professional board for computing purposes - Lots of Memory - Good Video driver - 1 or 2 A/D converter - 2 or 3 USB version 3 - PCI interface - Slot for more memory - 4 line LCD (debugging etc.) - Keyboard interface These are some ideas, Good luck for your development board Regards, JosephArticle: 151390
"allen" <ayhoung@gmail.com> wrote in message news:81b56b55-23ce-42a4-804a-f6cbc82955fc@q40g2000prh.googlegroups.com... > Hey guys, > > I work in the FPGA development kit sector, and I was just looking to > see if you guys had any ideas regarding what would be your MUST HAVE > features on an FPGA dev kit? > > Thanks in advance! > Allen I've never seen any point in general purpose development boards - I design a board for the job. Reference designs are useful. If I were going to attempt something with a big feasibility issue it might be worth buying a fancy dev board but it hasn't happened yet. Michael KellettArticle: 151391
On 3/31/2011 10:43 PM, allen wrote: > Hey guys, > > I work in the FPGA development kit sector, and I was just looking to > see if you guys had any ideas regarding what would be your MUST HAVE > features on an FPGA dev kit? > > Thanks in advance! > Allen Somewhere that I can actually solder down components. Plated through holes on 0.100" centers, with some grounds and VCCIOs interspersed. If I'm going to the dev kit it's either because a) I'm trying to get up to speed on a brand spanking new processor or b) I'm going off-datasheet and need to get measured results for things they won't spec. -- Rob Gaddi, Highland Technology Email address is currently out of orderArticle: 151392
I'm working on a Spartan 3 design and trying to figure out how, if at all, I can constrain what I'm looking for without resorting to hand placement and directed routing. I've got an analog pulse that feeds 4 comparators with different thresholds. This gives me 4 pulses of which each narrower one is contained entirely in time by each wider one, like the side view of the Towers of Hanoi experiment. The widest will be 3-4 ns, the narrowest will be narrow enough to not consistently be able to trip the internal flops. They'll be coming in at up to 80 MHz. Given that this is pretty fast for an S3, I've lit on the idea of using each pulse to clock a T-flop, then resynchronize the output of the T at 200 MHz into an XOR change detector; the classic pulse-toggle-pulse domain crosser but directly clocked rather than clock enabled. Code (untested) at the end of this message. So I've got a 2 ns event (rising edges only) and a 5 ns clock period. I need to make sure that all of my resynchronized pulses come out during clock cycle n or n+1. Which, so near as I can figure, means that I want to have no more than 3 ns of routing delay skew on these signals, from the pins through the Ts to the D input on the first synchronizer stage. Anyone have the foggiest how to bully the tools into pulling this off? Thanks, Rob -------- entity fast_rising_edge_det is generic ( SYNC_STAGES : positive := 2; REGISTER_OUTPUT : boolean := true ); -- Number of delay flip-flops port ( edge : in std_logic; -- Edge to be detected clk : in std_logic; -- Clock to synchronize the output to. pulse_out : out std_logic); -- Output pulse for the rising edge. end entity fast_rising_edge_det; architecture Behavioral of fast_rising_edge_det is signal toggle : std_logic := '0'; signal delay_stages : std_logic_vector(SYNC_STAGES-1 downto 0) := (others => ''0''); signal old_val : std_logic := '0'; signal change : std_logic; begin -- architecture Behavioral TFLOP : toggle <= not toggle when rising_edge(edge); DLINE : process (clk) is begin -- process if rising_edge(clk) then -- rising clock edge if (SYNC_STAGES = 1) then delay_stages(0) <= toggle; else delay_stages <= delay_stages(SYNC_STAGES-2 downto 0) & toggle; end if; end if; old_val <= delay_stages(delay_stages'high); end process; change <= old_val xor delay_stages(delay_stages'high); REG_OUT : if (REGISTER_OUTPUT) generate pulse_out <= change when rising_edge(clk); end generate REG_OUT; NOREG_OUT : if (not REGISTER_OUTPUT) generate pulse_out <= change; end generate NOREG_OUT; end architecture Behavioral; -------- -- Rob Gaddi, Highland Technology Email address is currently out of orderArticle: 151393
In my opinion: 1. FT245(or better FT2232 in FIFO mode - faster) for USB data transfer - this can be implemented easily either on PC side or FPGA side. 2. DDR or DDR2 memory. Better 2 separate chips if possible. Also easy to implement, because all vendors offer controllers for their FPGAs. 3. Ethernet. 4. IO expansions (well routed, not like Xilinx made on their S3E kit). These are 4 basic things. If I could find a PCI-E board + at least 2 LVDS RX and 2 TX outputs using SMA or other useful connectors in addition to everything I mentioned before - that would be perfect. There are plenty of boards for beginners, but nothing in the middle range to get more specific. The Cyclone IV GX starter kit would be just perfect, however it has no expansion IOs... Other PCI-E boards cost too much. "allen" <ayhoung@gmail.com> wrote in message news:81b56b55-23ce-42a4-804a-f6cbc82955fc@q40g2000prh.googlegroups.com... > Hey guys, > > I work in the FPGA development kit sector, and I was just looking to > see if you guys had any ideas regarding what would be your MUST HAVE > features on an FPGA dev kit? > > Thanks in advance! > AllenArticle: 151394
I am trying to make use of OC PCI Bridge, and it is working fine on the PCI Target side. But when i am trying to master the bus using my wishbone slave i was having a wrong behavior from the Wb slave of the bridge. When i try to read from the PC, the bridge should send a RTY until the data is ready.... but when i request the first time.. the slave assert the retry but never disassert, even after i disassert STB and CYC... i was looking at the verilog code at the pci_wb_slave.v but i cant find the reason for this behavior.... Any one ever had this problem? Any one have the contact with Miha Dolenc? I tryed sending an email to opencores email but i think it does not exist anymore.. Thank you!Article: 151395
On Thu, 31 Mar 2011 22:43:44 -0700, allen wrote: > Hey guys, > > I work in the FPGA development kit sector, and I was just looking to see > if you guys had any ideas regarding what would be your MUST HAVE > features on an FPGA dev kit? > > Thanks in advance! > Allen Not that I use dev kits, but if I did I would look for the following: - For an FPGA with transceivers, I would want two of them brought out to a pair of SFP+ sockets. The PCB routing should be of sufficient quality to allow this to work at 10Gb/s. The SFP+ control signals would also need to be connected to the FPGA (via level translators if the FPGA can't handle 3.3V hotplug signals). - Sites on the PCB for at least a few clock oscillators suitable for clocking the transceivers. These would be LVDS or PECL oscillators, not the CMOS ones that don't work so well above a few hundred MHz. One or two sites would be populated with oscillators that would allow the transceivers to do 1GbE and 10GbE. (The spare sites are to allow the user to solder in their own oscs to support other protocols.) - A USB to JTAG programming dongle built-in, probably using something like an FT2232H or FT232H. - Open source host driver software for the built-in JTAG programming hardware, with binaries available for at least WinXP, Win7 and major flavours of Linux. Regards, AllanArticle: 151396
Much delayed but will hopefully be actually be on show at ESC Silicon Valley in May our Raggedstone3 will have most of this for you. John Adair Enterpoint Ltd. On Apr 1, 9:33=A0pm, "scrts" <mailsoc@[remove@here]gmail.com> wrote: > In my opinion: > 1. FT245(or better FT2232 in FIFO mode - faster) for USB data transfer - > this can be implemented easily either on PC side or FPGA side. > 2. DDR or DDR2 memory. Better 2 separate chips if possible. Also easy to > implement, because all vendors offer controllers for their FPGAs. > 3. Ethernet. > 4. IO expansions (well routed, not like Xilinx made on their S3E kit). > > These are 4 basic things. > If I could find a PCI-E board + at least 2 LVDS RX and 2 TX outputs using > SMA or other useful connectors in addition to everything I mentioned > before - that would be perfect. > > There are plenty of boards for beginners, but nothing in the middle range= to > get more specific. The Cyclone IV GX starter kit would be just perfect, > however it has no expansion IOs... Other PCI-E boards cost too much. > > "allen" <ayho...@gmail.com> wrote in message > > news:81b56b55-23ce-42a4-804a-f6cbc82955fc@q40g2000prh.googlegroups.com... > > > Hey guys, > > > I work in the FPGA development kit sector, and I was just looking to > > see if you guys had any ideas regarding what would be your MUST HAVE > > features on an FPGA dev kit? > > > Thanks in advance! > > AllenArticle: 151397
I am trying to debug a Virtex 5 PCI Express core. When I insert it into a PC it is not detecting it. I have chipscope connected to the LTSSM state machine signals and it seems to be stuck in state 2. Is there any documentation from Xilinx that gives info about the states and the signals that are exchanged during the link training? Thanks Jon --------------------------------------- Posted through http://www.FPGARelated.comArticle: 151398
"Allan Herriman" <allanherriman@hotmail.com> a écrit : > - A USB to JTAG programming dongle built-in, probably using something > like an FT2232H or FT232H. > - Open source host driver software for the built-in JTAG programming > hardware, with binaries available for at least WinXP, Win7 and major > flavours of Linux. What are the advantages of using this chip compared to using the manufacturer USB cable ?Article: 151399
On Sat, 02 Apr 2011 13:41:58 +0200, PovTruffe wrote: > "Allan Herriman" <allanherriman@hotmail.com> a écrit : >> - A USB to JTAG programming dongle built-in, probably using something >> like an FT2232H or FT232H. >> - Open source host driver software for the built-in JTAG programming >> hardware, with binaries available for at least WinXP, Win7 and major >> flavours of Linux. > > > What are the advantages of using this chip compared to using the > manufacturer USB cable ? Nobody else in the office can borrow it if it's soldered to your board. Allan
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