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 Jul 28, 6:18=A0pm, Ben <leowy...@gmail.com> wrote: > I've read about daisy chaining fpgas with Platform flash, but is daisy > chaining fpgas possible with SPI flash? > The documents at Xilinx seems to give only examples of single fpga > configuration with SPI flash. no, you cant use multiply SPI flash in chained mode (you may configure more than one FPGA from one SPI flash) well, some newer FPGA's include multi SPI support, * Spartan-6 * Lattice-SC * Achronix Speedster all support multiply SPI flash in __parallel__ wiring (maybe some others too, i do not know all ) AnttiArticle: 142176
On Jul 28, 5:42=A0am, Mike Harrison <m...@whitewing.co.uk> wrote: > On Tue, 28 Jul 2009 10:50:48 +0200, Charles Gardiner <inva...@invalid.inv= alid> wrote: > >Hi Mike, > > >at what frequency is your SPI running. You can configure this in the > >*.lpf file or of course in the Design Planner. The actual frequency used > >is process/batch dependent since it uses an internal PLL. I think > >Lattice have specified an inaccuracy of 30%. Does it still fail if you > >set the load frequency to the lowest value? > > Default - 2.5MHz. > I've tried setting it faster and =A0do see the SPI rate increase once con= fig starts, so it is seeing > some of the datastream. > > >Other ideas: > >----------- > >I always put 10K pull-ups to 3.3V on the lines to and from the SPI > >flash. Have you tied any unused flash inputs? (e.g. WP/HOLD to 3.3V on > >Atmel parts) > > >You can also read a status register over the JTAG port with the Lattice > >USB thingy and ispVM. Is anything 'interesting' set in this if you read > >it after flash load fails? I've forgotten the exact feature/function in > >ispVM. I'll have mine hooked up later today so I'll take a look and post > >back. > > >Regards, > >Charles > > I've now tried retargetting the project to a LFEC6 and loaded it on a Lat= tice devboard with the same > result so AFAICS it looks like it's a software issue generating a bad CRC= on the bit file. If you have an earlier version of ispLever, you can try it to see if the CRC problem is a software bug. I tend to leave older versions of the tools around so I don't need to "upgrade" existing projects in order to make minor changes. All of my EC/ECP (not ECP2 or newer) were built using ispLever 6.x versions. Never seen this problem myself. Regards, GaborArticle: 142177
On Jul 28, 11:33=A0pm, "Antti.Luk...@googlemail.com" <Antti.Luk...@googlemail.com> wrote: > On Jul 28, 6:18=A0pm, Ben <leowy...@gmail.com> wrote: > > > I've read about daisy chaining fpgas with Platform flash, but is daisy > > chaining fpgas possible with SPI flash? > > The documents at Xilinx seems to give only examples of single fpga > > configuration with SPI flash. > > no, you cant use multiply SPI flash in chained mode > (you may configure more than one FPGA from one SPI flash) > > well, some newer FPGA's include multi SPI support, > > * Spartan-6 > * Lattice-SC > * Achronix Speedster > > all support multiply SPI flash in __parallel__ wiring > (maybe some others too, i do not know all ) > > Antti Actually, I was thinking of configuring multiple FPGAs from one SPI flash, but I don't see any examples in the Spartan6 configuration document. They give only the Platform flash example, which use pins like INIT_B and PROG_B instead of MOSI in SPI. Is the connection similar to the slave serial mode one using Platform flash, and do you know of any examples?Article: 142178
On Jul 28, 11:47=A0pm, Ben <leowy...@gmail.com> wrote: > On Jul 28, 11:33=A0pm, "Antti.Luk...@googlemail.com" > > > > <Antti.Luk...@googlemail.com> wrote: > > On Jul 28, 6:18=A0pm, Ben <leowy...@gmail.com> wrote: > > > > I've read about daisy chaining fpgas with Platform flash, but is dais= y > > > chaining fpgas possible with SPI flash? > > > The documents at Xilinx seems to give only examples of single fpga > > > configuration with SPI flash. > > > no, you cant use multiply SPI flash in chained mode > > (you may configure more than one FPGA from one SPI flash) > > > well, some newer FPGA's include multi SPI support, > > > * Spartan-6 > > * Lattice-SC > > * Achronix Speedster > > > all support multiply SPI flash in __parallel__ wiring > > (maybe some others too, i do not know all ) > > > Antti > > Actually, I was thinking of configuring multiple FPGAs from one SPI > flash, but I don't see any examples in the Spartan6 configuration > document. > They give only the Platform flash example, which use pins like INIT_B > and PROG_B instead of MOSI in SPI. > Is the connection similar to the slave serial mode one using Platform > flash, and do you know of any examples? Sorry, I mean master serial mode.Article: 142179
On Jul 28, 11:33=A0am, "Antti.Luk...@googlemail.com" <Antti.Luk...@googlemail.com> wrote: > On Jul 28, 6:18=A0pm, Ben <leowy...@gmail.com> wrote: > > > I've read about daisy chaining fpgas with Platform flash, but is daisy > > chaining fpgas possible with SPI flash? > > The documents at Xilinx seems to give only examples of single fpga > > configuration with SPI flash. > > no, you cant use multiply SPI flash in chained mode > (you may configure more than one FPGA from one SPI flash) > > well, some newer FPGA's include multi SPI support, > > * Spartan-6 > * Lattice-SC > * Achronix Speedster > > all support multiply SPI flash in __parallel__ wiring > (maybe some others too, i do not know all ) > > Antti I'm not sure that was what he was asking. If you have a single SPI flash big enough to hold all the bitstreams you can daisy chain the FPGA's (not the flash chips). One FPGA is in SPI master mode and the others in serial slave mode. Should be shown in the configuration UG.Article: 142180
On Jul 28, 11:49=A0pm, gabor <ga...@alacron.com> wrote: > On Jul 28, 11:33=A0am, "Antti.Luk...@googlemail.com" > > > > <Antti.Luk...@googlemail.com> wrote: > > On Jul 28, 6:18=A0pm, Ben <leowy...@gmail.com> wrote: > > > > I've read about daisy chaining fpgas with Platform flash, but is dais= y > > > chaining fpgas possible with SPI flash? > > > The documents at Xilinx seems to give only examples of single fpga > > > configuration with SPI flash. > > > no, you cant use multiply SPI flash in chained mode > > (you may configure more than one FPGA from one SPI flash) > > > well, some newer FPGA's include multi SPI support, > > > * Spartan-6 > > * Lattice-SC > > * Achronix Speedster > > > all support multiply SPI flash in __parallel__ wiring > > (maybe some others too, i do not know all ) > > > Antti > > I'm not sure that was what he was asking. =A0If you have a single > SPI flash big enough to hold all the bitstreams you can daisy > chain the FPGA's (not the flash chips). =A0One FPGA is in SPI master > mode and the others in serial slave mode. =A0Should be shown in the > configuration UG. Yes, I see this line "Another alternative is to use SPI mode for the first device. The daisy-chain data is still sent out through DOUT in SPI mode." Well, this is in the Serial Daisy Chain section and they use a Xilinx Platform Flash PROM.Article: 142181
On Tue, 28 Jul 2009 15:22:25 +0000, Allan Herriman wrote: > On Sun, 26 Jul 2009 10:28:03 -0700, Muzaffer Kal wrote: > >> On 26 Jul 2009 16:23:16 GMT, General Schvantzkoph >> <schvantzkoph@yahoo.com> wrote: >> >>>Has anyone benchmarked Core7 vs Core2 on NCverilog, Questa, Xilinx and >>>Altera FPGA tools? >> >> I'd also be very interested in Core7 vs Phenom II performance too (45nm >> AMD ie Phenom II 955 etc.) > > > We just got a new i7 machine for FPGA builds. It tested at just over > twice as fast as our high-end AMD box that was 2-3 years old. > > I will publish the results in this ng within a week or two, assuming I > can ever get the licensing for ISE 11.2 running on it. (Thanks Xilinx, > flexlm was a really good move.) > > Regards, > Allan What's the clock rate on each machine? iCore7 motherboards and processors are twice as expensive as Core2 motherboards and processors, and mainstream Core2s have a higher clock rate than iCore7s, so the clock rate normalized performance is what's important.Article: 142182
> After installing 11.1 I got another shock: projects are now not > backwards compatible with older versions of ISE. The project is > converted to the new format and gets a .xise extension. This will > significantly complicate collaboration as all users will need 11.1! It's much easier to work with the .xise because it is in text format. Users wanted this for a long time; finally the changes can be tracked in the source control software. I figured that ISE is still using the same .ise binary internally. When you open an .xise in the project navigator the first thing it does is converts it to the .ise format. I tried using that .ise from 11.1 in ISE 10.x and it worked (perhaps somebody could provide a case when it doesn't). EvgeniArticle: 142183
On Jul 28, 7:00=A0pm, Ben <leowy...@gmail.com> wrote: > On Jul 28, 11:49=A0pm, gabor <ga...@alacron.com> wrote: > > > > > > > On Jul 28, 11:33=A0am, "Antti.Luk...@googlemail.com" > > > <Antti.Luk...@googlemail.com> wrote: > > > On Jul 28, 6:18=A0pm, Ben <leowy...@gmail.com> wrote: > > > > > I've read about daisy chaining fpgas with Platform flash, but is da= isy > > > > chaining fpgas possible with SPI flash? > > > > The documents at Xilinx seems to give only examples of single fpga > > > > configuration with SPI flash. > > > > no, you cant use multiply SPI flash in chained mode > > > (you may configure more than one FPGA from one SPI flash) > > > > well, some newer FPGA's include multi SPI support, > > > > * Spartan-6 > > > * Lattice-SC > > > * Achronix Speedster > > > > all support multiply SPI flash in __parallel__ wiring > > > (maybe some others too, i do not know all ) > > > > Antti > > > I'm not sure that was what he was asking. =A0If you have a single > > SPI flash big enough to hold all the bitstreams you can daisy > > chain the FPGA's (not the flash chips). =A0One FPGA is in SPI master > > mode and the others in serial slave mode. =A0Should be shown in the > > configuration UG. > > Yes, I see this line "Another alternative is to use SPI mode for the > first device. The daisy-chain data is still sent out through DOUT in > SPI mode." > Well, this is in the Serial Daisy Chain section and they use a Xilinx > Platform Flash PROM.- Hide quoted text - > > - Show quoted text - Xilinx FPGA configuration controller has special function to BYPASS config data so you can stream it from the first device and use for any purpose you need (as config another FPGA as example) it is ir-relevant if the configuration comes from platform flash or spi flash or, you can use another approuch what i used with lattice FPGA as soon as FIRST FPGA is configured, it ROUTES the spi flash out to some downstream FPGA config interface and starts generating SPI clock, this "user config controller" is very small and doesnt take much resources AnttiArticle: 142184
On 28 Jul 2009 16:04:32 GMT General Schvantzkoph <schvantzkoph@yahoo.com> wrote: > On Tue, 28 Jul 2009 15:22:25 +0000, Allan Herriman wrote: > > > On Sun, 26 Jul 2009 10:28:03 -0700, Muzaffer Kal wrote: > > > >> On 26 Jul 2009 16:23:16 GMT, General Schvantzkoph > >> <schvantzkoph@yahoo.com> wrote: > >> > >>>Has anyone benchmarked Core7 vs Core2 on NCverilog, Questa, Xilinx > >>>and Altera FPGA tools? > >> > >> I'd also be very interested in Core7 vs Phenom II performance too > >> (45nm AMD ie Phenom II 955 etc.) > > > > > > We just got a new i7 machine for FPGA builds. It tested at just > > over twice as fast as our high-end AMD box that was 2-3 years old. > > > > I will publish the results in this ng within a week or two, > > assuming I can ever get the licensing for ISE 11.2 running on it. > > (Thanks Xilinx, flexlm was a really good move.) > > > > Regards, > > Allan > > What's the clock rate on each machine? iCore7 motherboards and > processors are twice as expensive as Core2 motherboards and > processors, and mainstream Core2s have a higher clock rate than > iCore7s, so the clock rate normalized performance is what's important. Clock rates is just an artificial number; I'd be more interested in performance normalized to non-recurring cost (purchase price) and operating cost (power consumption).Article: 142185
>Hi, i'm using ise 9.2i for my design, and the PAR runs successfully meeting >all my constraints with no timing errors, but the simulation reports >multiple warnings of hold/setup violation, what is the problem? > > > Hi, Please check whether the violated paths are constrained/not. if constrained then check if they were constrained as false paths/unrelated.Article: 142186
Hi! I'm having a small problem with simulations using Xilinx ISE 11.1. In my code there are some FSM which were declared like: process(clk, rst) begin if rst = '1' then state <= IDLE; elsif clk'event and clk='1' then case state is when IDLE => if start = '1' then state <= NEXT_STATE; end if; [...] With start being an entity port. Now, when running a behavior simulation, I get different results for this code when it's in the top level block (which receives its stimuli from the simulation testbench file) and when it's in an inner block of the design (which receives stimuli from other blocks). More specifically, the transition to NEXT_STATE happens in the same cycle start is raised to 1 in the top level block, but will only happen in the next cycle (when start already is 0) in inner blocks. I'm thinking, considering the real circuit will have delays, that the second behavior makes more sense. However, in the behavioral simulation, all combinational logic have zero delay, and so the changes in the signals happen in the same time - the least I expected was that the response of all circuits would be coherent. So, anyone got and explanation for this? Maybe a way to make all circuits behave the same way? Thank you! P.S.: I'm simulating using the ISim software included with ISE 11.1.Article: 142187
Watch out for delta delays on your clock between the top level and lower levels (i.e. from an assignment of one clock signal from another). The delta delay incurred by the clock will cause it not to see inputs on the same clock cycle in RTL simulation. AndyArticle: 142188
On Tue, 28 Jul 2009 09:42:08 -0700, Jason Zheng wrote: > On 28 Jul 2009 16:04:32 GMT > General Schvantzkoph <schvantzkoph@yahoo.com> wrote: > >> On Tue, 28 Jul 2009 15:22:25 +0000, Allan Herriman wrote: >> >> > On Sun, 26 Jul 2009 10:28:03 -0700, Muzaffer Kal wrote: >> > >> >> On 26 Jul 2009 16:23:16 GMT, General Schvantzkoph >> >> <schvantzkoph@yahoo.com> wrote: >> >> >> >>>Has anyone benchmarked Core7 vs Core2 on NCverilog, Questa, Xilinx >> >>>and Altera FPGA tools? >> >> >> >> I'd also be very interested in Core7 vs Phenom II performance too >> >> (45nm AMD ie Phenom II 955 etc.) >> > >> > >> > We just got a new i7 machine for FPGA builds. It tested at just over >> > twice as fast as our high-end AMD box that was 2-3 years old. >> > >> > I will publish the results in this ng within a week or two, assuming >> > I can ever get the licensing for ISE 11.2 running on it. (Thanks >> > Xilinx, flexlm was a really good move.) >> > >> > Regards, >> > Allan >> >> What's the clock rate on each machine? iCore7 motherboards and >> processors are twice as expensive as Core2 motherboards and processors, >> and mainstream Core2s have a higher clock rate than iCore7s, so the >> clock rate normalized performance is what's important. > > Clock rates is just an artificial number; I'd be more interested in > performance normalized to non-recurring cost (purchase price) and > operating cost (power consumption). Within an architecture CPU clock rate is the one thing that's directly proportional to performance. Memory performance has a small effect on performance, but in the age of multi-megabyte caches it's a distinctly second order effect. Price is something that you worry about at the time of purchase, the price that someone else paid for their system is completely irrelevant. What I want to know is the relative performance of Core2 vs iCore7 on a few important applications specifically NCverilog, which is where I spend 98% of my time, and the Xilinx and Altera FPGA tools. My current fastest machine is a 3GHz Core2 overclocked to 4GHz (I'm using a heatsink the size of Volkswagen). The motherboard for that system is $115 these days and the CPU is $168. A 2.66Ghz iCore7 is $280, a 3GHz iCore7 is $569 and an iCore7 motherboard is in the neighborhood of $250. If the clock for clock NCVerilog performance of an iCore7 is really 2X a Core2 then it makes sense to buy one assuming that I can get a similar overclock on the iCore7. If it's only 30% then it doesn't make sense because it's unlikely that a 2.66GHz iCore7 can be overclocked to the same clock as a 3GHz Core2.Article: 142189
On Jul 25, 11:36=A0pm, "MikeWhy" <boat042-nos...@yahoo.com> wrote: > "Andy Peters" <goo...@latke.net> wrote in message > > news:89bb1088-36f1-435c-8107-284367a3ac23@13g2000prl.googlegroups.com... > The project is in an svn repo. The great thing about this now-deleted > feature was that it boiled the ISE project file down to a couple of > tcl scripts which are scc-friendly. The only other things in the repo > are the sources and a .ucf. > > =3D=3D=3D=3D=3D=3D=3D=3D=3D > The <project>.xise file serves the need, without exporting or importing. True indeed, but if your project originated in the earlier version of the tools, it still needs to be recreated. I opened a WebCase, and the support person was able to figure out exactly how to use the tcl shell to import the previous version's script and make it into a proper .xise project file. > Now if only I can get it to stop polluting the project root with all thos= e > generated files. Yeah, well, you should be happy that you don't have to keep your HDL sources in the same directory as all of the project cruft! (How long did it take Xilinx to figure that out?) -a-Article: 142190
On Jul 27, 8:35=A0am, General Schvantzkoph <schvantzk...@yahoo.com> wrote: > On Fri, 24 Jul 2009 10:08:19 -0700, Andy Peters wrote: > > On Jul 24, 1:17=A0am, luudee <rudolf.usselm...@gmail.com> wrote: > >> On Jul 23, 12:36=A0am, Andy Peters <goo...@latke.net> wrote: > > >> > I installed the full-up Xilinx ISE 11.x tools on a spare machine so = I > >> > could give it a test-drive. (We have a site license.) > > >> ... > > >> > I opened a WebCase, which I suspect will be ignored like every other > >> > WebCase. > > >> > -a > > >> Well, Andy, I hate to say that, but it is your own fault. Perhaps you > >> are a new user ... > > > No, actually, I've being doing this for a long time -- remember XACT? > > >> Using anything but only the fundamental features of the tools is askin= g > >> for trouble. Even when just using the fundamental tools like "map > >> <option>" in a perl script you will see that they tend to rename and > >> change the way some switches work from release to release. > > > I realize that things change all the time, which is why I like to > > minimize my dependency on the tools. But unfortunately, that's not > > always possible. And this little archiving feature was one of Xilinx' > > better ideas, which is probably why they shitcanned it. > > I have one word for you, cvs. I would never use a vendor specific source > control system when there are so many first class open source choices > available. I prefer good old cvs because it's transparent, i.e. it > doesn't use any binary databases it just uses the regular file system and > a few text base files. It's also completely reliable across platforms, > all of my systems are Linux based while my partner uses cygwin on XP and > we have no problem sharing files. You could also use svn or git, I don't > like svn and I've never tried git but they are all free and packaged with > the major distros so you can make your own choice. Actually, as I've noted in other posts in this thread, I use Subversion. My point was that the old .ise file couldn't reasonably live in the repo, and that Xilinx added a feature called Project | Source Control | Import/Export boiled the binary ISE project file down to a couple of tcl scripts, which are revision-control friendly. It is unfortunate that Xilinx chose to put this feature in a menu item called "Source Control" when it really has nothing to do with scc. -aArticle: 142191
On Jul 28, 12:24=A0pm, "Antti.Luk...@googlemail.com" <Antti.Luk...@googlemail.com> wrote: > On Jul 28, 7:00=A0pm, Ben <leowy...@gmail.com> wrote: > > > > > On Jul 28, 11:49=A0pm, gabor <ga...@alacron.com> wrote: > > > > On Jul 28, 11:33=A0am, "Antti.Luk...@googlemail.com" > > > > <Antti.Luk...@googlemail.com> wrote: > > > > On Jul 28, 6:18=A0pm, Ben <leowy...@gmail.com> wrote: > > > > > > I've read about daisy chaining fpgas with Platform flash, but is = daisy > > > > > chaining fpgas possible with SPI flash? > > > > > The documents at Xilinx seems to give only examples of single fpg= a > > > > > configuration with SPI flash. > > > > > no, you cant use multiply SPI flash in chained mode > > > > (you may configure more than one FPGA from one SPI flash) > > > > > well, some newer FPGA's include multi SPI support, > > > > > * Spartan-6 > > > > * Lattice-SC > > > > * Achronix Speedster > > > > > all support multiply SPI flash in __parallel__ wiring > > > > (maybe some others too, i do not know all ) > > > > > Antti > > > > I'm not sure that was what he was asking. =A0If you have a single > > > SPI flash big enough to hold all the bitstreams you can daisy > > > chain the FPGA's (not the flash chips). =A0One FPGA is in SPI master > > > mode and the others in serial slave mode. =A0Should be shown in the > > > configuration UG. > > > Yes, I see this line "Another alternative is to use SPI mode for the > > first device. The daisy-chain data is still sent out through DOUT in > > SPI mode." > > Well, this is in the Serial Daisy Chain section and they use a Xilinx > > Platform Flash PROM.- Hide quoted text - > > > - Show quoted text - > > Xilinx FPGA configuration controller has special function to BYPASS > config data > so you can stream it from the first device and use for any purpose you > need > (as config another FPGA as example) > > it is ir-relevant if the configuration comes from platform flash or > spi flash > > or, you can use another approuch what i used with lattice FPGA > as soon as FIRST FPGA is configured, it ROUTES the spi flash > out to some downstream FPGA config interface and starts generating > SPI clock, this "user config controller" is very small and doesnt take > much resources > > Antti Actually at least for ECP2, SPI config of multiple parts is easy with Lattice. The first FPGA is in SPI master mode and the subsequent FPGA's are in serial slave mode. First FPGA CCLK is both SPI clock and downstream CCLK for serial config. The tricky part with lattice is figuring out how to merge the bitstreams into the single SPI PROM (you need the "Universal File Writer" for this). Regards, GaborArticle: 142192
On Jul 28, 12:44=A0pm, "arik" <ariff.a...@gmail.com> wrote: > >Hi, i'm using ise 9.2i for my design, and the PAR runs successfully > meeting > >all my constraints with no timing errors, but the simulation reports > >multiple warnings of hold/setup violation, what is the problem? > > Hi, > Please check whether the violated paths are constrained/not. > if constrained then check if they were constrained as false > paths/unrelated. Nine times out of ten, setup and hold violations stem from the testbench not meeting the constraints given to the place & route tools. It does no good to meet constraints if the external world doesn't meet them, too.Article: 142193
On Jul 28, 1:54=A0pm, Andy <jonesa...@comcast.net> wrote: > Watch out for delta delays on your clock between the top level and > lower levels (i.e. from an assignment of one clock signal from > another). The delta delay incurred by the clock will cause it not to > see inputs on the same clock cycle in RTL simulation. > > Andy Don't forget to have hold time in your testbench signals. You don't need delays, just a dependency on the clock edge. If you generate both the clock and the start signal together, you should expect the timing you're seeing. If start depends on clock'event then you should see the behavior of the lower level module.Article: 142194
> On the other hand, this particular item is something that I submitted > to the VHDL standards group a few years back and was accepted at that > time as a 'good' idea...maybe it made it into VHDL '08...if not theirs > always VHDL 201x. Kevin, Do you have a bug ID for it? I was looking through the bugzilla database and could not find it. JimArticle: 142195
Can anyone tell me what the following error messages mean please? ERROR:ProjectMgmt:357 - Problem updating Cs using the Transform Instance. ERROR:ConstraintSystem:170 - No constraint database file was specified. Thanks, RogArticle: 142196
On Tue, 28 Jul 2009 08:46:07 -0700 (PDT), gabor <gabor@alacron.com> wrote: >On Jul 28, 5:42 am, Mike Harrison <m...@whitewing.co.uk> wrote: >> On Tue, 28 Jul 2009 10:50:48 +0200, Charles Gardiner <inva...@invalid.invalid> wrote: >> >Hi Mike, >> >> >at what frequency is your SPI running. You can configure this in the >> >*.lpf file or of course in the Design Planner. The actual frequency used >> >is process/batch dependent since it uses an internal PLL. I think >> >Lattice have specified an inaccuracy of 30%. Does it still fail if you >> >set the load frequency to the lowest value? >> >> Default - 2.5MHz. >> I've tried setting it faster and do see the SPI rate increase once config starts, so it is seeing >> some of the datastream. >> >> >Other ideas: >> >----------- >> >I always put 10K pull-ups to 3.3V on the lines to and from the SPI >> >flash. Have you tied any unused flash inputs? (e.g. WP/HOLD to 3.3V on >> >Atmel parts) >> >> >You can also read a status register over the JTAG port with the Lattice >> >USB thingy and ispVM. Is anything 'interesting' set in this if you read >> >it after flash load fails? I've forgotten the exact feature/function in >> >ispVM. I'll have mine hooked up later today so I'll take a look and post >> >back. >> >> >Regards, >> >Charles >> >> I've now tried retargetting the project to a LFEC6 and loaded it on a Lattice devboard with the same >> result so AFAICS it looks like it's a software issue generating a bad CRC on the bit file. > >If you have an earlier version of ispLever, you can try it to see if >the >CRC problem is a software bug. I tend to leave older versions of the >tools around so I don't need to "upgrade" existing projects in order >to make minor changes. All of my EC/ECP (not ECP2 or newer) were >built >using ispLever 6.x versions. Never seen this problem myself. > >Regards, >Gabor Good idea, unfortunately 6.x does not appear to be available on the Lattice website - only the service packsArticle: 142197
http://i.cmpnet.com/eetimes/news/09/07/1565chart_pg18.gif JohnArticle: 142198
> http://i.cmpnet.com/eetimes/news/09/07/1565chart_pg18.gif Interesting, indeed. Cypress had viable products but I'm convinced that management was the problem. (I also remember they wanted a rather large premium for their CPLDs that were sometimes only marginally better than the competition's.) I don't know what Vantis's problem was, but at least after Lattive bought them they kept a few of the parts around. Intel doesn't have its heart in much of anything but their desktop CPUs -- they consistently bring out interesting products and then discontinue them just when they're starting to gain traction. You're pretty much a pure Xilinx man these days, aren't you, John?Article: 142199
> Hi all, > > I have a bunch of used FPGA boards available. They have been replaced > in a product by a similar board that has additional NAND Flash. > Here are the details of the board: > > * Altera Cyclone EP1C6Q240 FPGA > * 512 KB FLASH (for FPGA configuration and program code) > * 1 MB fast SRAM > * ByteBlasterMV port > * Watchdog with LED > * EPM7064 PLD to load FPGA from FLASH (on watchdog reset) > * Voltage regulator (1V5) > * Crystal clock (20 MHz) at the PLL input (up to 640 MHz internal) > * Serial interface (MAX3232) > * 56 general purpose IO pins > > Price is EUR 99,- > > See: http://www.jopdesign.com/cyclone/index.jsp > I forgot to say that all boards are fully tested. Furthermore, as beeing used without problem in the production for some time this served even as a reliability test ;-) Cheers, Martin
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