Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

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

Custom Search

Messages from 36725

Article: 36725
Subject: Re: Clock Divider or Multiplier ???
From: "Falk Brunner" <Falk.Brunner@gmx.de>
Date: Sat, 17 Nov 2001 20:09:58 +0100
Links: << >>  << T >>  << A >>
"Peter Alfke" <palfke@earthlink.net> schrieb im Newsbeitrag
news:3BF69776.A9E6B9A0@earthlink.net...
> Is 80 MHz a challenge these days?

;-))
Dont know,.
> That's 12.5 ns. Seems like an eternity for modern chips.

Almost. But also think about, not everyone is pushing >800 Mbit/s through
the inputs to build a terabit IP switch. There are also plenty of guys
around (including me) that eat humble pie (Thanks to dict.leo.org for the
phrase ;-) So why wasting time to make your circuit run @80 MHz when you
have a easy way (Virtex-II DCM) to generate a "real" 60 Mhz clock. I
wouldnt.

> We are presently haggling over shaving an extra 100 picoseconds off the
input
> set-up time...

And this is very appreciated.

--
MfG
Falk



Article: 36726
Subject: Xilinx Student edition F1.5 and Webpack compatible ?
From: "Rob Finch" <robfinch@sympatico.ca>
Date: Sat, 17 Nov 2001 16:05:23 -0500
Links: << >>  << T >>  << A >>
Is Xilinx Student edition F1.5 compatible with Webpack 4.1 ? I'm trying to
update an older design using Student Edition F1.5 and am getting strange
results. I've installed Webpack 4.1 since I last used the student edition. I
tried my original design in F1.5 and it's now telling me it's 4x (1800CLB's)
too big for the part an XC4010XL It used to fit no problem. If I try it in
Webpack 4.1 the result is only 400 LUTs (= about 200 CLBs).

Thanks
Rob




Article: 36727
Subject: Does anybody knows where have a free(open hardware) FPGA PCI Development board whith PCB data.....?
From: tt889@163.net (apple88888)
Date: 17 Nov 2001 14:00:57 -0800
Links: << >>  << T >>  << A >>
It may like Spartan-II PCI develop board(DS-PCI32S-KIT2) or something like this.

remail me :  tt889@163.net

Article: 36728
Subject: Q: XILINX binary .bit file header - ?
From: "Alex Sherstuk" <sherstuk@iname.com>
Date: Sat, 17 Nov 2001 22:43:01 GMT
Links: << >>  << T >>  << A >>
Dear colleagues,

   I am trying to use XILINX .bit-file as an input for a self-made FPGA
configuration load program.

The question is:
- What is contained in the header of the .bit-file, before actual
configuration data?
- How to skip the header?

Yes, I know, that I can use ASCII .rbt file as an input for my program (and
actually do that).
But .rbt files are very large, compared to .bit files. It is very
unconvenient to distribute several megabytes of firmware codes ...

Of course, I can invent my own binary format, derived from .rbt-files. But
why to do that, if there is a standard .bit-file?

Thanks,
   Alex Sherstuk




Article: 36729
Subject: Re: Q: XILINX binary .bit file header - ?
From: Lasse Langwadt Christensen <langwadt@ieee.org>
Date: Sat, 17 Nov 2001 23:52:03 GMT
Links: << >>  << T >>  << A >>
Alex Sherstuk wrote:
> 
> Dear colleagues,
> 
>    I am trying to use XILINX .bit-file as an input for a self-made FPGA
> configuration load program.
> 
> The question is:
> - What is contained in the header of the .bit-file, before actual
> configuration data?
> - How to skip the header?
> 
> Yes, I know, that I can use ASCII .rbt file as an input for my program (and
> actually do that).
> But .rbt files are very large, compared to .bit files. It is very
> unconvenient to distribute several megabytes of firmware codes ...
> 
> Of course, I can invent my own binary format, derived from .rbt-files. But
> why to do that, if there is a standard .bit-file?
> 
> Thanks,
>    Alex Sherstuk

 looking at the rbt and the bit files it looks like the firste 72 bytes
of 
the .bit file is header the rest of it looks the same as the .rbt file

-Lasse
-- Lasse Langwadt Christensen, 
-- A Dane in Phoenix, Arizona

Article: 36730
Subject: Re: Q: XILINX binary .bit file header - ?
From: Philip Freidin <philip@fliptronics.com>
Date: Sun, 18 Nov 2001 01:22:27 GMT
Links: << >>  << T >>  << A >>
The answer is in the FAQ:

     http://www.fpga-faq.com/FAQ_Pages/0026_Tell_me_about_bit_files.htm

Philip


On Sat, 17 Nov 2001 22:43:01 GMT, "Alex Sherstuk" <sherstuk@iname.com> wrote:
>Dear colleagues,
>
>   I am trying to use XILINX .bit-file as an input for a self-made FPGA
>configuration load program.
>
>The question is:
>- What is contained in the header of the .bit-file, before actual
>configuration data?
>- How to skip the header?
>
>Yes, I know, that I can use ASCII .rbt file as an input for my program (and
>actually do that).
>But .rbt files are very large, compared to .bit files. It is very
>unconvenient to distribute several megabytes of firmware codes ...
>
>Of course, I can invent my own binary format, derived from .rbt-files. But
>why to do that, if there is a standard .bit-file?
>
>Thanks,
>   Alex Sherstuk
>
>

Philip Freidin
Fliptronics

Article: 36731
Subject: Re: Q: XILINX binary .bit file header - ?
From: khtsoi@cse.cuhk.edu.hk
Date: 18 Nov 2001 01:33:16 GMT
Links: << >>  << T >>  << A >>

I have written a program on Linux to download the bit file
to XCV100E through the Xilinx Parallel Cabel III. You may
be intreset in that.

http://www.cse.cuhk.edu.hk/~khtsoi/project/Xilinux/index.htm

May people in this news group help me to build that. But
it's now using raw inb(), outb() and must be setuid as root.
If you (or anyone) can modify it to have nicer security,
please let me know. Thanks!

---- Brittle

Article: 36732
Subject: Re: CASE vs. IF statements
From: Tom Verbeure <tom.verbeure@verizon.no_spam.net>
Date: Sun, 18 Nov 2001 06:39:16 GMT
Links: << >>  << T >>  << A >>

You could use an array of arrays. Multi-dimensional types are not supported 
by some synthesis tools (eg. Synopsys DC), but an array of arrays usually 
is.
So you would get:

        type t_Array is array(0 to something) of integer;
        type t_ArrayOfArray is array(0 to 9) of t_Array;

        output_sig <= myArray(cntr2)(cntr);

BTW, while I think it should be ok, I am not 100% sure that the last line 
is syntactically correct. If not, you can do:

        v_ActArray : t_Array;

        v_ActArray := myArray(cntr2);
        output_sig <= v_ActArray(cntr);


To answer the original question: I am a bit surprised that the case 
statement gives worse results than the IF-ELSE case. I don't really see who 
the IF-ELSE case can have less fanout than the CASE case.
And the IF-ELSE case will likely have a higher number of levels of logic.

Tom


Sul Weh wrote:

> Yes, but if I have *10* roms, each each 10 integers in them then it would
> be if (cntr2 = 0) then
> output_sig <= array_of_int0(cntr)
> 
> so I have 10 arrays of 10 integers
> "Russell Shaw" <rjshaw@iprimus.com.au> wrote in message
> news:3BF06C0F.32B632D2@iprimus.com.au...
>>
>>
>> Sul Weh wrote:
>> >
>> > What is the advantages/disadvantages of using a CASE statement or an if
>> > statement?
>> > For example, if I have 10 arrays of 10 integers (array0, array1,...)
>> > and
> I
>> > want to sequentially output each array I can either use a CASE statment
> or a
>> > 10 separate IF statements.
>> > e.g. if (counter = 0) then ... end if
>> > if (counter = 1) then ... end if
>> >
>> > What I have seen is the clk is slower in a CASE statement than an IF
>> > statement because the counter has a huge fanout.
>> >
>> > any advice or comments?
>>
>> I'd do it this way (less code): Declare an array of 10 items with 10-bit
>> integers, and initialize them. The synthesizer should see this
>> hard-coding,
>> and make a small rom. Now all you need to do is have a counter index the
>> pointer of this array:  output_sig<=array_of_int(cntr);


Article: 36733
Subject: WebPACK 4.1 under Win95
From: pbleyer@embedded.cl (Pablo Bleyer Kocik)
Date: 17 Nov 2001 22:54:39 -0800
Links: << >>  << T >>  << A >>
Hi!

 We recently updated WebPACK 3.3 to the latest 4.1 version. The nice
thing is that now, in our Win95 machines, WebPACK fails to start. The
pthread shared library needs some threading functions not present in
the Win95 Kernel32 library (TryEnterCriticalSection, for example -- it
seems the libpthread.dll library that Xilinx is using is not very
robust).

 Has anybody solved this problem? We can't switch to other Windoze
versions in these machines because we have other software/hardware
that depends on it...

 Best wishes.

Article: 36734
Subject: Re: WebPACK 4.1 under Win95
From: Simon Gornall <simon@gornall.net>
Date: Sun, 18 Nov 2001 07:23:40 +0000
Links: << >>  << T >>  << A >>
Pablo Bleyer Kocik wrote:
> 
> Hi!
> 
>  We recently updated WebPACK 3.3 to the latest 4.1 version. The nice
> thing is that now, in our Win95 machines, WebPACK fails to start. The
> pthread shared library needs some threading functions not present in
> the Win95 Kernel32 library (TryEnterCriticalSection, for example -- it
> seems the libpthread.dll library that Xilinx is using is not very
> robust).
> 
>  Has anybody solved this problem? We can't switch to other Windoze
> versions in these machines because we have other software/hardware
> that depends on it...

It says in the release notes that 4.1 will not work under Win95. Win95
and POSIX-compliance (pthreads) are after all rareley mentioned in the
same sentence, unless there's a NOT! in there as well :-)

ATB,
	Simon.

Article: 36735
Subject: fft instantiation in foundation tools
From: gopalakr@eng.usf.edu (Mouli)
Date: 18 Nov 2001 01:32:11 -0800
Links: << >>  << T >>  << A >>
Hi all :

I am trying to instantiate a fft16 core (created using coregen) in a
bigger project. It starts off ok and then gives the following warnings
:
<===================================================>

WARNING:NgdBuild:454 - logical net 'FFT16/done' has no load
WARNING:NgdBuild:454 - logical net 'FFT16/ovflo' has no load
....
<===================================================>

120 of them !!!!

Then it carries on to mapping which it does with no errors/warnings. 

Then it goes on to routing and then gives the following warnings ...

<===================================================>

WARNING:Route:49 - The signal "FFT16/dfly/xmul/pi_add/COUT16_NET_O"
has no
loads so was not routed. 
WARNING:Route:49 - The signal
"FFT16/dfly_i_2comp/TWOS_COMP_SUB/COUT_NET_16"
has no loads so was not routed. 
<===================================================>

Quite surprisingly - it then says ..

<===================================================>

All signals are completely routed.

Total REAL time to PAR completion: 43 secs 
Total CPU time to PAR completion: 42 secs 

Placement: Completed - No errors found.
Routing: Completed - No errors found.

PAR done.

<===================================================>

Then it goes on to timing .. goes pretty smoothly until when it says

<===================================================>

WARNING:Anno:197 - NGDAnno found 8 physical component(s) for which 100
percent
   back-annotation is not possible. Simulation models for these
components will
   be constructed from the NCD. Rerun NGDAnno with the -report option
for
   additional details, including any net and instance names which are
lost in
   this process.

<===================================================>

and then aborts with a fatal error :

FATAL_ERROR:NetListWriters:basnewrite.c:709:1.5.6.1 - No block bus
info found
   for bus d(3:0). Process will terminate.  To resolve this error,
please
   consult the Answers Database and other online resources at
   http://support.xilinx.com

<===================================================>



And ofcourse support.xilinx.com didnt have the answer :(( and that is
why i am writing here. If someone could help me out , it would be
great !

I am using Foundation tools 3.11. I have included the vff16.edn to the
project. I am instantiating the core in my vhd file the exact same
way, the coregen manual tells me to .. I dont quite understand why
this error comes up.

Would really appreciate it, if you could also copy replies to 
cgopalak@csee.usf.edu

Thanks in advance. ...

--Mouli

Article: 36736
Subject: Re: WebPACK 4.1 under Win95
From: "Leon Heller" <leon_heller@hotmail.com>
Date: Sun, 18 Nov 2001 10:55:43 -0000
Links: << >>  << T >>  << A >>

"Pablo Bleyer Kocik" <pbleyer@embedded.cl> wrote in message
news:d26ba1a8.0111172254.75e620d5@posting.google.com...
> Hi!
>
>  We recently updated WebPACK 3.3 to the latest 4.1 version. The nice
> thing is that now, in our Win95 machines, WebPACK fails to start. The
> pthread shared library needs some threading functions not present in
> the Win95 Kernel32 library (TryEnterCriticalSection, for example -- it
> seems the libpthread.dll library that Xilinx is using is not very
> robust).
>
>  Has anybody solved this problem? We can't switch to other Windoze
> versions in these machines because we have other software/hardware
> that depends on it...


You should be able to upgrade to Win98 without any problems, I'd have
thought. The new Webpack work fine with that, of course.

Leon
--
Leon Heller, G1HSM leon_heller@hotmail.con
http://www.geocities.com/leon_heller
Low-cost Altera Flex design kit: http://www.leonheller.com





Article: 36737
Subject: Xilinx Fpga Editor support for Virtex 2...does it exist in 3.x? How about 4.x?
From: "Austin Franklin" <austin@da22rkroom.com>
Date: Sun, 18 Nov 2001 12:44:23 -0500
Links: << >>  << T >>  << A >>
I thought I had a full installation of the latest 3.x tools...and brought up
the FPGA Editor...but it doesn't seem to allow me to select any of the
Virtex 2 parts...  Does anyone know if this is supposed to be available in
FPGA Editor with the 3.x tools?  If not, then is it available with the 4.x
tools?




Article: 36738
Subject: Re: Decoupling capacitors on Virtex II
From: Rick Filipkiewicz <rick@algor.co.uk>
Date: Sun, 18 Nov 2001 18:55:31 +0000
Links: << >>  << T >>  << A >>


Austin Lesea wrote:

> More to consider than just bypassing....
>
>  http://www.xilinx.com/xlnx/xil_prodcat_product.jsp?title=si_pcbcheck
>
> Austin
>

Nice checklist although the "simulation" rqeuirements presuppose access to a SPICE
engine. Is there a freeware/opensource/public-domain implemetation ?

Only one quibble on TI (= TextIntegrity): Are you really recommending milliFarad
(e.g. 4.7mF) caps ? Perhaps its a misprint for 4.7MF :-).



Article: 36739
Subject: Re: Clock Divider or Multiplier ???
From: Rick Filipkiewicz <rick@algor.co.uk>
Date: Sun, 18 Nov 2001 19:04:50 +0000
Links: << >>  << T >>  << A >>


Falk Brunner wrote:

> "Peter Alfke" <palfke@earthlink.net> schrieb im Newsbeitrag
> news:3BF69776.A9E6B9A0@earthlink.net...
> > Is 80 MHz a challenge these days?
>
>
> > We are presently haggling over shaving an extra 100 picoseconds off the
> input
> > set-up time...
>
> And this is very appreciated.
>
> --
> MfG
> Falk

Even more appreciated would be some extra -ve hold time. Sometime soon I've
got to deal with a chip that has a 0.5nsec -ve min Tco. This is unloaded so
the 10pF of trace + input capacitance should get me within the Virtex-E's =
-0.4 BUT ....

PALs & GALs had fairly large -ve hold requirements that were quoted in the
early data sheets but then the manufacturers stopped giving this info &
reverted to type in only saying 0. AFAIK Xilinx are the only ones willing to
stick their necks out these days - at least for FPGA's IOB FFs.



Article: 36740
Subject: Re: 'Timing' simulation in ModelSIM
From: Rick Filipkiewicz <rick@algor.co.uk>
Date: Sun, 18 Nov 2001 19:14:21 +0000
Links: << >>  << T >>  << A >>


Austin Franklin wrote:

> > Hey there,
> >     It seems I have another challange to overcome in attempting 'timing'
> > simulations in ModelSIM.
>
> Kris,
>
> Why are you doing timing simulations in the first place?
>
> Austin

Post P&R simulation is the only way to find NGDBUILD/MAP/PAR bugs [BITGEN bugs
= you are basically stuffed]. Particularly in the early days of MX.Y there
were a lot of issues with timing contraints. Even as recently as 2.1i you
couldn't separate the 2 parts of a BlockRAM if they were on different clocks -
lead to some interesting problems.

Found a new 2.1iSP6 MAP bug just the other day when re-building an old release
with the same toolset as the original build. Needed post PAR simulation to
find it.




Article: 36741
Subject: Re: Xilinx Fpga Editor support for Virtex 2...does it exist in 3.x? How about 4.x?
From: "Austin Franklin" <austin@da22rkroom.com>
Date: Sun, 18 Nov 2001 15:16:06 -0500
Links: << >>  << T >>  << A >>
I loaded 4, and the XC2V3000 FF1152 is supported in FPGA Editor...  Still
curious about if any version of 3 supports the XC2V3000 in FPGA Editor.

"Austin Franklin" <austin@da22rkroom.com> wrote in message
news:tvfsrtqmn2hk52@corp.supernews.com...
> I thought I had a full installation of the latest 3.x tools...and brought
up
> the FPGA Editor...but it doesn't seem to allow me to select any of the
> Virtex 2 parts...  Does anyone know if this is supposed to be available in
> FPGA Editor with the 3.x tools?  If not, then is it available with the 4.x
> tools?
>
>
>



Article: 36742
Subject: WebPACK 4.1 under Win95 : solved
From: pbleyer@embedded.cl (Pablo Bleyer Kocik)
Date: 18 Nov 2001 15:27:32 -0800
Links: << >>  << T >>  << A >>
Ok, I solved the problem -- at least the technical aspects of it. I
replaced the libpthread.dll library that comes with Xilinx WebPACK
(origin unknown) with GNU's pthread-win32 library (pthreadGCE.dll -
GCC + exceptions compiled natively with Mingw). And voilą, I have been
using 4.1 smoothly since yesterday -- however I have not programmed
any devices yet.

 The pthread-win32 library (http://sources.redhat.com/pthreads-win32/)
is released under the LGPL. I don't know which *legal* consequences
these may have under Xilinx WebPACK ISE legal terms. However... should
I suggest that Xilinx uses this well supported & tested library
instead of the one it is using now?

 Ideas? Suggestions? Flames? Software guys from Xilinx here? (or worse
-- Xilinx lawyers ;^)

 Cheers!

Article: 36743
Subject: jtag programming xilinx xc9572 cpld
From: Daniel Nilsson <e9danne@etek.chalmers.se>
Date: Mon, 19 Nov 2001 03:53:04 +0100
Links: << >>  << T >>  << A >>
Hi.
I've managed to read some sort of IDCODE out of my XC9572 pc44 cpld, and
I wonder if someone using webpack or similar software could tell me a
valid 32-bit IDCODE from such a chip, so I have something to compare
with.
The reason I'm asking this is that I dont have microsoft windows
installed.

/Daniel Nilsson M.Sc.EE student

Article: 36744
Subject: Re: jtag programming xilinx xc9572 cpld
From: "Jim" <johnsonw10@hotmail.com>
Date: Mon, 19 Nov 2001 04:07:55 GMT
Links: << >>  << T >>  << A >>
Based on xc9572_pc44.bsd, it is

attribute IDCODE_REGISTER of xc9572_pc44 : entity is
    "XXXX" &        -- version
    "1001010100000100" &    -- part number
    "00001001001" &     -- manufacturer's id
    "1";            -- required by standard

Hope this helps,
Jim

"Daniel Nilsson" <e9danne@etek.chalmers.se> wrote in message
news:3BF87410.438BF1F7@etek.chalmers.se...
> Hi.
> I've managed to read some sort of IDCODE out of my XC9572 pc44 cpld, and
> I wonder if someone using webpack or similar software could tell me a
> valid 32-bit IDCODE from such a chip, so I have something to compare
> with.
> The reason I'm asking this is that I dont have microsoft windows
> installed.
>
> /Daniel Nilsson M.Sc.EE student



Article: 36745
Subject: Problems of Amplify about Modular Design
From: hitajian <hitajian@21cn.com>
Date: Mon, 19 Nov 2001 12:40:52 +0800
Links: << >>  << T >>  << A >>
Now I'm studying Modular Design using Amplify of Synplicity. However,
now I encountered some problems. I have thought them for about three
days. But, till now, I cannot fix them yet. Therefore, I query them on
Internet. Could you kindly help me on these problems? 
 
1. I study this design flow following the tutorial guide of Amplify.
However, after I completed the design budget in Design leader project,
I was informed to "Use the top-level .edf and .ncf files to run
Initial Design Budgeting in the place-and-route tool". At this point,
I was confused. There are all abstract sub-modules in the EDIF
netlist, which were treated by Amplify as black box. How could
ngdbuild expand the design? 
 
2. When I try to run the .edf and .ncf files in the place-and-route
tool (I use the Design Compiler in ISE 4.1, Service Pack II), the tool
reported the error as follows: The entry: E:\..\.edf contains spaces.
Currently spaces are not supported. But the .edf file in this format
(I mean the file contain spaces) is outputed by Amplify. If such a
format could not be supported by the P&R tool, why does Amplify output
such a file? 
 
3. In Amplify, how can I expand a bus when I would like to do some
constraints by pesudo port region. In the example design in Amplify, a
bus has been expanded into bits. The bus port in bit is draged&droped
from the RTL view to the pesudo port region. But, I cannot find a way
to expand the bus in my design. How can I make this? 
 
4. After I complete the design of team leader, I need distribute the
source code, SCOPE files and physical constraint file to the team
members to begin the implementation of their respective P&R regions.
At this point, I wonder in what way should I distribute these files,
by sending a copy of my whole project to each member or in other way?
Judging from the specification of the guide, I regard I should forward
a copy of my whole project to every member, right? If it's right,
every member just cares the design that is allcated to himself or
herself and doesn't care others although his project file includes
other P&R region information. Is it right? 
 
Thanks 


Article: 36746
Subject: Virtex-II Pin-Incompatibility
From: Jonas Weiss <jweiss@kontronmedical.ch>
Date: Mon, 19 Nov 2001 10:49:32 +0100
Links: << >>  << T >>  << A >>
I'm working on a design using either two XC2V1000 (FG456 package) or
XC2V3000 (FG676 package). As the board is part of a prototype system we
wanted to start with the small devices and if necessary up grade to the
3000s later. First we discovered that roughly half of the LVDS channels
can not be used because the differential pairs on one package do not
correspond to the pairs on the other package (what an inconvenience).
Running a place and route yielded that of the pins I thought would
match, there are still pairs with inversed polarity between the
packages...which was confirmed by a look at the data sheet. :-(
Namely:
3000                    1000

M22->N                K20->P
M21->P                K19->N

N21->P                L19->N
N22->N                L20->P

L21->P                 J19->N
L22->N                 J20->P

I remember having seen a nice Xilinx brochure with a transparent print
of one package which fits perfectly over the drawing of the next bigger
package...how misleading.

If anybody knows of any other hidden obstacles I might struggle over,
any comments will be appreciated very much.

Thanks

Jonas


Article: 36747
Subject: Re: How can I solve the "clock" warning of synplify.
From: phil <phil1807@genion.de>
Date: Mon, 19 Nov 2001 02:15:07 -0800
Links: << >>  << T >>  << A >>
It seems that you have connected combinational logic to the clock input of a flop (you have implemented a not recommended "gated clock" structure which can cause problems because of skew on clock signals). Synplify provides you with that warning message because it assumes that any signal which is connected to the clock input of a flop is a "real"clock signal. Synplify is not able to calculate the frequency of that gated clock signal and, therefore, assumes that the frequency of the incoming clock signal is also applied to all the flopswhich are connected to the gated clock.   

To anwer your questions:

Yes, that warning message is important because it provides you with information on potential problems with your design. You can resolve it by either resolving the gated clock structure of your design or applying a clock constraint on the flops driven by the gated clock

Article: 36748
Subject: how to limit the fanout in APEX20K400E
From: shengyu_shen@hotmail.com (ssy)
Date: 19 Nov 2001 02:20:09 -0800
Links: << >>  << T >>  << A >>
Hi everyone

I use a sinplify pro to synthesis, and the target device is
APEX20K400E, but the sinplify pro do not allow me to specify the max
fanout, and after the synthesis, some net(not clock) have fanout as
large as 500.

how to deal with this?

thank you

Article: 36749
Subject: Re: FPGA synthesis
From: Madhura <madhura@controlnet.co.in>
Date: Mon, 19 Nov 2001 02:45:31 -0800
Links: << >>  << T >>  << A >>
The design is being targetted to VirtexE family. For the gate level simulation, I require VHDL netlist. If I generate that through dc_shell, the INIT generic for the LUT's are not being associated with the LUT's. So it gives warning during the compilation. So the netlist has to be given to FC2 to have these INIT parameters. Now those parameters are associated with the LUT's. But now the warning is given for the RAM's..RAMB4_S*.. The xnf files for these can not be opened for analyze.
can u suggest some solution for this?



Site Home   Archive Home   FAQ Home   How to search the Archive   How to Navigate the Archive   
Compare FPGA features and resources   

Threads starting:
1994JulAugSepOctNovDec1994
1995JanFebMarAprMayJunJulAugSepOctNovDec1995
1996JanFebMarAprMayJunJulAugSepOctNovDec1996
1997JanFebMarAprMayJunJulAugSepOctNovDec1997
1998JanFebMarAprMayJunJulAugSepOctNovDec1998
1999JanFebMarAprMayJunJulAugSepOctNovDec1999
2000JanFebMarAprMayJunJulAugSepOctNovDec2000
2001JanFebMarAprMayJunJulAugSepOctNovDec2001
2002JanFebMarAprMayJunJulAugSepOctNovDec2002
2003JanFebMarAprMayJunJulAugSepOctNovDec2003
2004JanFebMarAprMayJunJulAugSepOctNovDec2004
2005JanFebMarAprMayJunJulAugSepOctNovDec2005
2006JanFebMarAprMayJunJulAugSepOctNovDec2006
2007JanFebMarAprMayJunJulAugSepOctNovDec2007
2008JanFebMarAprMayJunJulAugSepOctNovDec2008
2009JanFebMarAprMayJunJulAugSepOctNovDec2009
2010JanFebMarAprMayJunJulAugSepOctNovDec2010
2011JanFebMarAprMayJunJulAugSepOctNovDec2011
2012JanFebMarAprMayJunJulAugSepOctNovDec2012
2013JanFebMarAprMayJunJulAugSepOctNovDec2013
2014JanFebMarAprMayJunJulAugSepOctNovDec2014
2015JanFebMarAprMayJunJulAugSepOctNovDec2015
2016JanFebMarAprMayJunJulAugSepOctNovDec2016
2017JanFebMarAprMayJunJulAugSepOctNovDec2017
2018JanFebMarAprMayJunJulAugSepOctNovDec2018
2019JanFebMarAprMayJunJulAugSepOctNovDec2019
2020JanFebMarAprMay2020

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

Custom Search