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 25900

Article: 25900
Subject: Re: Xilinx Student Edition 2.1i with "Digital Design:Principles and
From: Anna Acevedo <acevedo@xilinx.com>
Date: Mon, 25 Sep 2000 12:16:15 -0700
Links: << >>  << T >>  << A >>
Yes, the XSE 2.1i in the Wakerly's Digital Design: Principles and Practices
book is the same
as the standalone Xilinx Student Edition 2.1i by Prentice Hall

The Xilinx Student Edition software is available in several Prentice Hall
books
For a list of XDS (Xilinx Design Series) books visit
http://vig.prenhall.com/
and do a catalog search, key work Xilinx

Anna



Guest Internet User wrote:

> I was in Irvine ScitechBooks, today, looking for Xilinx Student Edition
> 2.1i.  (The website www.scitechbooks.com doesn't show it as stocked, but
> I've learned from the retail store workers that the website and store
> inventory are separate entities.)  I came across Wakerly's "Digital
> Design:Principles and Practices" (third edition, ISBN: 0137691912)
>
> The book was bundled with Xilinx Student Edition 2.1i.  I was surprised
> to see 2.1i bundled with a book published back in 1999, but the pricetag
> sticker showed a date code of '09/00' (meaning Scitech
> received/processed it this month.)  The book was shrinkwrapped so I
> couldn't even open it, but the back of the book had a blue cardboard
> envelope package labeled "Xilinx Student Edition 2.1."
>
> Does anyone know if this is the same software as the standalone Xilinx
> Student Edition 2.1i by Prentice Hall?  Seems like the standalone
> package is hard to find.

--
*****************************
Anna M. Acevedo
Xilinx University Program
2100 Logic Drive
San Jose, CA 95124
PH: (408) 879-5338
FAX: (408) 879-4780

Email: anna.acevedo@xilinx.com
http://www.xilinx.com/programs/univ.htm
*****************************


Article: 25901
Subject: Re: Dual monitor display possible with modelsim on a PC?
From: "Keith R. Williams" <krw@btv.ibm.com>
Date: Mon, 25 Sep 2000 15:39:01 -0400
Links: << >>  << T >>  << A >>


Nial Stewart wrote:
> 
> I'm about to order bits to build a PC for FPGA simulation/
> compilation/PAR.
> 
> I'm going for a 750M Althlone Thunderbird on an Asus A7V motherboard
> with
> 256M memory (to start), a 15G IBM ATA100 hard drive with a
> Yamaha 8x8x24 CD-WR for project documentation.

> Any comments?
 
I get along "fine" with a PII/333 with a PIII-650 in the "background"
for PAR. Synplify and ModelSim work fine on a PII/333 w/192K.  In fact
Synplify is almost acceptable on my P120 - zero L2 laptop.  However,
Foundation 3.1i sucks a system to its knees for hours sometimes (even
with 640MB of RAM) so I have a separate system for PAR.  I put the
"background" system in the lab with and also use it for debug,
simulators, EPROM programmer, and such.  

> I've been thinking about buying a Matrox G400 video card with
> dual output as I was thinking that it would be very handy to
> be able to view a modelsim wave window on one monitor, with
> all the other modelsim windows on another one.
> 
> Has anyone tried this? I know that windows98 supports dual
> monitors, but does the application software have to support
> them too, or does it jut create more desktop on the second one?

This system has two Matrox Mystique Business/220s for dual screen in a
"KeithKit" with the above PII/333.  I have a 20" on the left and a 19"
to the right (well, they were available ;-).  The setup essentially
gives me one 2560x1024 (2 x 1280x1024)desktop under NT4. It works very
nicely.

There is some strangeness with messages wanting to go to the center
(across the split between the monitors) and windows that refuse to
remember where I put them last, but it is a very nice setup. Running
simulations on one screen while compiling or looking at schematics (with
Synplify) on the other is very nice! Cut-n-paste to get entities and
mappings right is really nice too.

----
  Keith
Sender: eric@ruckus.brouhaha.com
Article: 25902
Subject: Difference between Foundation Base and Foundation ISE Base Express?
From: Eric Smith <eric-no-spam-for-me@brouhaha.com>
Date: 25 Sep 2000 15:44:51 -0700
Links: << >>  << T >>  << A >>
Does anyone have a simple explanation of the differences between
Foundation Base ($1090) and Foundation ISE Base Express ($695)?
They both seem to support the same devices (except for the XC3000 and
XC5200 series)?  Does Foundation Base not support VHDL?

And how do those packages compare to the student edition 2.1i, which
Amazon lists for $55?  I can't even *find* a supported device list
for the student edition.

The Xilinx marketing people must have worked long hours and legal
holidays to make sure that it's difficult to get useful information from
their web site.  They reference comparison charts that are "currently
not available."

I want to do VHDL design using Spartan and Spartan II parts, and I'd
like to get started as cheaply as possible.
Article: 25903
Subject: Re: Difference between Foundation Base and Foundation ISE Base Express?
From: rickman <spamgoeshere4@yahoo.com>
Date: Mon, 25 Sep 2000 20:25:35 -0400
Links: << >>  << T >>  << A >>
I don't know what is in the Student Edition. They keep that under pretty
close wraps. But here are the two pages the tell you what is in the ISE
and the Foundation packages...

http://www.xilinx.com/products/foundise.htm
http://www.xilinx.com/products/found.htm

If you can get Xilinx marketing to focus on providing data rather than
the glossy info-deprived web pages and literature you will be a better
man than I. Xilinx has done this for years. I remember opening the first
Dynabook searchable data book only to find an incredible multi-layered
graphic that took over a minute to draw each time I started it up or
came back to the main page. 

They get beat up about things like this and correct them. But nobody
does anything about the guy that keeps making new problems. So they add
a lot of silly, useless and time wasting stuff to the install script or
to the web pages or to anything else they think they can use to
brainwash you into using their parts... instead of just giving you the
information that will let you see how their part fits perfectly into
your design. 

Can you tell that I am having a bad day?



Eric Smith wrote:
> 
> Does anyone have a simple explanation of the differences between
> Foundation Base ($1090) and Foundation ISE Base Express ($695)?
> They both seem to support the same devices (except for the XC3000 and
> XC5200 series)?  Does Foundation Base not support VHDL?
> 
> And how do those packages compare to the student edition 2.1i, which
> Amazon lists for $55?  I can't even *find* a supported device list
> for the student edition.
> 
> The Xilinx marketing people must have worked long hours and legal
> holidays to make sure that it's difficult to get useful information from
> their web site.  They reference comparison charts that are "currently
> not available."
> 
> I want to do VHDL design using Spartan and Spartan II parts, and I'd
> like to get started as cheaply as possible.

-- 

Rick "rickman" Collins

rick.collins@XYarius.com

Ignore the reply address. To email me use the above address with the XY
removed.



Arius - A Signal Processing Solutions Company
Specializing in DSP and FPGA design

Arius
4 King Ave
Frederick, MD 21701-3110
301-682-7772 Voice
301-682-7666 FAX

Internet URL http://www.arius.com
Article: 25904
Subject: Re: Announce: Free HC11 CPU Core
From: "yaohan" <yaohan@tm.net.my>
Date: Tue, 26 Sep 2000 08:27:02 +0800
Links: << >>  << T >>  << A >>
Hi,
May i know how to get the free core that U  mention?
thank you
regards,
chew
"Scott Thibault" <thibault@gmvhdl.com> wrote in message
news:8qfo69$rf4$1@slb7.atl.mindspring.net...
> Hi Ray,
>
> VHDLStudio is only required to run the included HC11 debugger.  The core
is
> in standard VHDL which can be simulated and synthesized with any standard
> tool.
>
> Regards,
> --Scott Thibault
> Green Mountain
> Computing Systems, Inc.
>
> "S. Ramirez" <sramirez@deleet.cfl.rr.com> wrote in message
> news:ETIy5.20208$W45.2292373@typhoon.tampabay.rr.com...
> > Ray,
> >      I agree that this is a substantial amount of effort, but the site
> does
> > say that you need VHDL Studio.  I do not know why it is needed, but this
> > could explain why the core is free.  This could be an example of bait
and
> > switch.  Depending on the price of VHDL Studio, this could be a "gift
> > horse," too, as you say.
> > -Simon Ramriez, Consultant
> >  Synchronous Design, Inc.
> >
> >
> >
> > "Ray Andraka" <ray@andraka.com> wrote in message
> > news:39CB52F5.9DDCB49C@andraka.com...
> > > Evan, his website tells all.  He's apparently sythesized it to Altera
> > 20K100 and
> > > Virtex400E, both of which ran at about 31-32 MHz.  One could probably
> > double the
> > > speed and cut the size by a considerable amount by tailoring the
design
> to
> > a
> > > specific part, but hey this ain't bad and who's to look a gift horse
in
> > the
> > > mouth?
> > >
> > > Nice piece of work, and a substantial amount of effort to give away.
> > >
> > >
> > > eml@riverside-machines.com.NOSPAM wrote:
> > > >
> > > > On Thu, 21 Sep 2000 10:17:19 -0400, "Scott Thibault"
> > > > <thibault@gmvhdl.com> wrote:
> > > >
> > > > >We have a synthesizable HC11 CPU core available for download.  See
> > > > >http://www.gmvhdl.com/hc11core.html for details.
> > > > >
> > > > >--Scott Thibault
> > > > >Green Mountain
> > > > >Computing Systems, Inc.
> > > > >http://www.gmvhdl.com
> > > >
> > > > Very nice. Have you done an FPGA implementation? Do you know how
> > > > big/fast it is?
> > > >
> > > > Evan
> > >
> > > --
> > > -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  or http://www.fpga-guru.com
> > >
> >
> >
>
>


Article: 25905
Subject: Point to point core
From: "yaohan" <yaohan@tm.net.my>
Date: Tue, 26 Sep 2000 08:29:10 +0800
Links: << >>  << T >>  << A >>
Hi,
Do anyone have resources or related material regarding the PPP point to
point
core development in fpga using vhdl.

thanx.

regards,
Hew


Article: 25906
Subject: Re: Announce: Free HC11 CPU Core
From: "S. Ramirez" <sramirez@deleet.cfl.rr.com>
Date: Tue, 26 Sep 2000 01:32:57 GMT
Links: << >>  << T >>  << A >>
Scott,
     Is there a list of errata for this core?
-Simon Ramirez, Consultant
 Synchronous Design, Inc.


"Scott Thibault" <thibault@gmvhdl.com> wrote in message
news:8qd59o$iom$1@slb3.atl.mindspring.net...
> We have a synthesizable HC11 CPU core available for download.  See
> http://www.gmvhdl.com/hc11core.html for details.
>
> --Scott Thibault
> Green Mountain
> Computing Systems, Inc.
> http://www.gmvhdl.com
>
>
>
>


Article: 25907
Subject: Looking for ASIC,FPGA Designers
From: "Barry Schneider" <barry61s@worldnet.att.net>
Date: Tue, 26 Sep 2000 02:25:46 GMT
Links: << >>  << T >>  << A >>
I am presently working at a ASIC consulting company and am extremely over
worked.  We need help and will pay well.  We have a great office and have
very flexible hours.   We are looking for Verilog and/or VHDL experience.
Synthesis and/or Mixed Signal a plus. If you are interested in a Good Job
e-mail me at barry61s@att.net.

  Hope to hear from you.

                        Sincerely,
                                        Barry


PS: We have needs in:       Commack, Long Island New York,
                                         Hazlet, New Jersey
                                         Bethlehem, Pennsylvania.





Article: 25908
Subject: FEC in FPGAs?
From: shahzad2512@my-deja.com
Date: Tue, 26 Sep 2000 03:31:36 GMT
Links: << >>  << T >>  << A >>
Forward error correction (FEC) can be realized in ASSP, for example
AHA, in ASICs and in FPGAs. Does anyone know What are the key
advantages of realizing Viterbi Decoder and Reed Solomon Decoder in
FPGAs, in terms of speed and cost. Of course flexibility is the
inherent feature of FPGAs. Thanks in advance.
Regards,
Shahzad


Sent via Deja.com http://www.deja.com/
Before you buy.
Article: 25909
Subject: Re: Using the xilinx "pull-up to 5V" in VHDL
From: korthner@hotmail.nospam.com (K. Orthner)
Date: 26 Sep 2000 04:54:59 GMT
Links: << >>  << T >>  << A >>
Hi, Marc.

You can do this:

output_pin <= 'Z' when desired_output = '1' else '0';

This casuses the output pin to mimic Open-Collector behavior, and the pull-
up resistor that you install external to the chip is responsible for your 
TTL '1' (5v).

The place-and-route tools will know that this uses a tri-state output 
buffer.

HTH,
Kent.

--

Marc.Battyani@fractalconcept.com (Marc Battyani) wrote in
>I've seen on the Xilinx web an application note (vtt002.pdf) where they
>use a pull-up resistor to 5V and use the input to put the output in 3
>state when it has reached a 3.3V 1 level. (not very clear description.
>i't better explained in the AN)
>
>How can I use this in VHDL? I use this output as a 3-state bus.
>Should I instanciate an OBUFT gate or is there a more clever way to do
>this? 
>
>Thanks
>
>Marc Battyani
>
>
>

Article: 25910
Subject: clock skew,,,,,,,DLL instantiation
From: sivakumar1974@my-deja.com
Date: Tue, 26 Sep 2000 05:22:38 GMT
Links: << >>  << T >>  << A >>
Hi all,  in my design, i am using Xilinx virtex 1000E and design operates at
two clock frequencies (100Mhz one side and 62.5 Mhz on the other side). The
design contains more than 16 FIFO (async.FIFO).

 The design doesn't meet the timing requirements(10ns). So i instantiated
DLLs(CLKDLLHF).Then also the clock skew is more.  i have some doubts,  1.How
to reduce the RAM setup time timing violations and how to control the clock
skew.  2.What is the use of DUTY_CYCLE_CORRECTION constraint and how to use
it.  3.What is the use of NODELAY constraint and how to use it properly.Will
it improve the DLL performance?.  4.In the FIFO the write clock frequency
(100MHz) is more than the read clock frequency(62.5MHz).How to set delay
constaint for this.

could anybody suggests me?

thanks in advance,
sivakumar.s




Sent via Deja.com http://www.deja.com/
Before you buy.
Article: 25911
Subject: Re: Xilinx Student Edition 2.1i with "Digital Design:Principles and
From: Guest Internet User <INVALID@INVALID.NET>
Date: Mon, 25 Sep 2000 22:53:20 -0700
Links: << >>  << T >>  << A >>
Thank you!  I don't suppose Xilinx could also update their webpage, to
announce the availability of XSE 2.1i?  The webpage still lists 1.5.

And after double checking the book title "Digital Design:Principles and
Practices", I'd like to add a correction.  The full title is Digital
Design:Principles and Practices, Updated Edition" (3rd edition)  

The original third-edition comes with XSE 1.5, apparently.

> Yes, the XSE 2.1i in the Wakerly's Digital Design: Principles and Practices
> book is the same
> as the standalone Xilinx Student Edition 2.1i by Prentice Hall
> 
> The Xilinx Student Edition software is available in several Prentice Hall
> books
> For a list of XDS (Xilinx Design Series) books visit
> http://vig.prenhall.com/
> and do a catalog search, key work Xilinx
Article: 25912
Subject: Re: Using the xilinx "pull-up to 5V" in VHDL
From: Steven <steven.yeung@xilinx.com>
Date: Mon, 25 Sep 2000 23:31:38 -0700
Links: << >>  << T >>  << A >>
What you need to do is to use OBUFT to create a open-drian output. When you need to drive a logic 0, the output buffer will drive it low, if you need a logic 1, the output buffer will be tri-state and the external pull-up will do the job. In your vhdl code, you can do:

output <= '0' when (oe = '0') else 'z';

So, when oe is 0, output will be 0, and vice versa.
Article: 25913
Subject: Re: Using the xilinx "pull-up to 5V" in VHDL
From: "Marc Battyani" <Marc.Battyani@fractalconcept.com>
Date: Tue, 26 Sep 2000 09:00:42 +0200
Links: << >>  << T >>  << A >>

"K. Orthner" <korthner@hotmail.nospam.com> wrote in message
news:8FBB82E6Fkorthnerhotmailcom@158.202.232.7...
...
> You can do this:
>
> output_pin <= 'Z' when desired_output = '1' else '0';
>
> This casuses the output pin to mimic Open-Collector behavior, and the
pull-
> up resistor that you install external to the chip is responsible for your
> TTL '1' (5v).

Thanks. This will indeed mimic an open drain output, but the Xilinx trick is
to force the output to "1" and then tristate it so that the rise time is
much shortened compared to only tristating it.
Doing this reduce the rise time from 20ns to 3ns.

It's what I would like to do in VHDL.

Marc Battyani


Article: 25914
Subject: Re: Using the xilinx "pull-up to 5V" in VHDL
From: "Marc Battyani" <Marc.Battyani@fractalconcept.com>
Date: Tue, 26 Sep 2000 09:03:11 +0200
Links: << >>  << T >>  << A >>

"Steven" <steven.yeung@xilinx.com> wrote in message
news:ee6e195.0@WebX.sUN8CHnE...
> What you need to do is to use OBUFT to create a open-drian output. When
you need to drive a logic 0, the output buffer will drive it low, if you
need a logic 1, the output buffer will be tri-state and the external pull-up
will do the job. In your vhdl code, you can do:
>
> output <= '0' when (oe = '0') else 'z';

Thanks, but the xilinx trick is to use a feedback from the output to force
the output to 1 before going to tristate. (see my previous post).

Marc Battyani


Article: 25915
Subject: Re: Using the xilinx "pull-up to 5V" in VHDL
From: Klaus Falser <kfalser@durst.it>
Date: Tue, 26 Sep 2000 07:52:22 GMT
Links: << >>  << T >>  << A >>
In article <8qpibm$bf3$1@reader1.fr.uu.net>,
  "Marc Battyani" <Marc.Battyani@fractalconcept.com> wrote:
>
> "K. Orthner" <korthner@hotmail.nospam.com> wrote in message
> news:8FBB82E6Fkorthnerhotmailcom@158.202.232.7...
> ...
> > You can do this:
> >
> > output_pin <= 'Z' when desired_output = '1' else '0';
> >
> > This casuses the output pin to mimic Open-Collector behavior, and the
> pull-
> > up resistor that you install external to the chip is responsible for
your
> > TTL '1' (5v).
>
> Thanks. This will indeed mimic an open drain output, but the Xilinx
trick is
> to force the output to "1" and then tristate it so that the rise time is
> much shortened compared to only tristating it.
> Doing this reduce the rise time from 20ns to 3ns.
>
> It's what I would like to do in VHDL.
>
> Marc Battyani
>
>
Maybe this slightly modified statement does what you want :

output_pin <= 'Z' when desired_output = '1' else desired_output;

This assumes that threestating is slower than the gate delay
(which usually holds true).

--
Klaus Falser
Durst Phototechnik AG
I-39042 Brixen


Sent via Deja.com http://www.deja.com/
Before you buy.
Article: 25916
Subject: Re: Category : why is 0?
From: chsw <chen.songwei@mail.zte.com.cn>
Date: Tue, 26 Sep 2000 00:59:15 -0700
Links: << >>  << T >>  << A >>
Hi: Now, i am doing "16bits converting 64bits fifo" of four channels ,my application is to write a lot of data to a fifo first by polling ,and then read it from the fifo by polling, my design is: the input data is 16 bits width,and use four 16 bits latches and 2 bits counter,the data is sent into 64 bits fifo when counter==2'b11.And then i will transfer it at the top level,if i write and read only a certain fifo ,the data of being read is correct.However,when i switch channel,the reading first data is 0 from the switched channel.Obviously ,the "0" of 64 bits is excrescent,but the back data is correct,that is ,it reads a "0" more.the phenomena happens only switching channel. why?
Article: 25917
Subject: Coregen
From: Christophe Heyert <heyertc@rsd.bel.alcatel.be>
Date: Tue, 26 Sep 2000 10:01:24 +0200
Links: << >>  << T >>  << A >>

A very strange thing happened to me yesterday. I'm using a RAM block
generated by the Coregenerator in
my structural design.
The first time I didn't have any problems i my behavioral simulation,
but the second time the following error appeared.

Internal ERROR : Access violation
                 Occurred in architecture BEHAVIORAL of entity
C_MEM_BLOCK_V1_0 (which is a single port RAM block).

Where does this problem come from?


christophe
Article: 25918
Subject: Re: Using the xilinx "pull-up to 5V" in VHDL
From: "Marc Battyani" <Marc.Battyani@fractalconcept.com>
Date: Tue, 26 Sep 2000 10:08:40 +0200
Links: << >>  << T >>  << A >>

"Klaus Falser" <kfalser@durst.it> wrote in message
news:8qpkjh$rbk$1@nnrp1.deja.com...
> In article <8qpibm$bf3$1@reader1.fr.uu.net>,
>   "Marc Battyani" <Marc.Battyani@fractalconcept.com> wrote:
> > ... the Xilinx  trick is
> > to force the output to "1" and then tristate it so that the rise time is
> > much shortened compared to only tristating it.
> > Doing this reduce the rise time from 20ns to 3ns.
> >
...
> Maybe this slightly modified statement does what you want :
>
> output_pin <= 'Z' when desired_output = '1' else desired_output;
>
> This assumes that threestating is slower than the gate delay
> (which usually holds true).

I think it could be more like :
output_pin <= 'Z' when desired_output = '1' and output_pin = '1' else
desired_output.

But the second output_pin should be the actual external signal after the
output buffer not the internal one...
may be I will have to instanciate low level gates. (OBUFT...)

Marc Battyani




Article: 25919
Subject: Re: Using the xilinx "pull-up to 5V" in VHDL
From: korthner@hotmail.nospam.com (K. Orthner)
Date: 26 Sep 2000 08:42:48 GMT
Links: << >>  << T >>  << A >>
Marc,

My apologies.

I should have read the appnote before I responded.

I beleve that this is what you're looking for.

I synthesized it with Synopsys FPGA Express, and it gave what the appnote 
described.


pin <= '0' when desired_output = '0' else
       '1' when pin = '0' else
       'Z';

-Kent

Article: 25920
Subject: Re: Using the xilinx "pull-up to 5V" in VHDL
From: "Marc Battyani" <Marc.Battyani@fractalconcept.com>
Date: Tue, 26 Sep 2000 11:21:05 +0200
Links: << >>  << T >>  << A >>

"K. Orthner" <korthner@hotmail.nospam.com> wrote in message
news:8FBBBD919korthnerhotmailcom@158.202.232.7...
> Marc,
>
> My apologies.
>
> I should have read the appnote before I responded.
>
> I beleve that this is what you're looking for.
>
> I synthesized it with Synopsys FPGA Express, and it gave what the appnote
> described.
>
>
> pin <= '0' when desired_output = '0' else
>        '1' when pin = '0' else
>        'Z';

Thanks! Exactly what I needed.

Marc Battyani



Article: 25921
Subject: Global clock buffers and secondary clock buffers.
From: Richard Meester <rme@quest-innovations.com>
Date: Tue, 26 Sep 2000 14:27:55 +0200
Links: << >>  << T >>  << A >>
Hello All,

I have some trouble regarding primary global and secondary global clock
lines.

I have a design that generates a clock from two inputs. This clock is
then used to clock other ff's. The problem now is that there are only 4
global clock lines with low skew in the spartan II (which i am using)
and i need aprox 8. The 4 global clock lines are used for other
purposes.  There are however 24 secondary global lines which also could
do the job.

Now my design is not yet finished, but when i implement/synthesize part
of the design, it inferres BUFG (1 of the primary global clocks) instead
of just a IBUF, and placing the pin on the top or bottom row of the
device. I looked at the xilinx site, nothing usefull found. I have the
constraints setup for these lines with USELOWSKEWLINES. I also tried by
hardcoded instantiating an IBUF to the clock line, but it gets replaced.

Anybody some suggestions????

Richard

--
Quest Innovations
tel: +31 (0) 227 604046
http://www.quest-innovations.com


Article: 25922
Subject: Re: Global clock buffers and secondary clock buffers.
From: "Simon" <simonb@tile.demon.co.cuthis.uk>
Date: Tue, 26 Sep 2000 14:35:53 +0100
Links: << >>  << T >>  << A >>
There are _no_ secondary clock nets on the Virtex/E/SpartanII.  This was
an unfortunate choice of wording in (early?) Virtex data.

You should be able to stop your synthesiser inserting BUFGs in your
secondary nets.  Synplify does this via the 'syn_noclockbuf' attribute.

You can get good low-skew routing off the top/bottom pins to adjacent
columns, particularly if you keep within the span of a 'hex' line.
Try playing around in the Editor to see what is possible.


Richard Meester wrote in message
<39D0964B.50AD8AD0@quest-innovations.com>...
>Hello All,
>
>I have some trouble regarding primary global and secondary global clock
>lines.
>
>I have a design that generates a clock from two inputs. This clock is
>then used to clock other ff's. The problem now is that there are only 4
>global clock lines with low skew in the spartan II (which i am using)
>and i need aprox 8. The 4 global clock lines are used for other
>purposes.  There are however 24 secondary global lines which also could
>do the job.
>
>Now my design is not yet finished, but when i implement/synthesize part
>of the design, it inferres BUFG (1 of the primary global clocks) instead
>of just a IBUF, and placing the pin on the top or bottom row of the
>device. I looked at the xilinx site, nothing usefull found. I have the
>constraints setup for these lines with USELOWSKEWLINES. I also tried by
>hardcoded instantiating an IBUF to the clock line, but it gets replaced.
>
>Anybody some suggestions????
>
>Richard
>
>--
>Quest Innovations
>tel: +31 (0) 227 604046
>http://www.quest-innovations.com
>
>


Article: 25923
Subject: Re: Paranoid...
From: Johan Petersson <Johan.Petersson@uab.ericsson.se>
Date: Tue, 26 Sep 2000 16:54:27 +0200
Links: << >>  << T >>  << A >>
Gary Watson wrote:
> 
> I'm doing a board with a Xilinx Spartan II (XC2S150) and have provided
> places for both a XC17S150XL and a XC18V01S20 config prom.  Am I being
> overly paranoid?

Hi Gary,

Well, maybe not... It all depends on how careful you have been with your
architecture and area/speed calculations. But I wouldn't take space for
more
prom's than the ones I plan to use. On the other hand, you never know
when the
customer changes his mind about product requirements :)

Regards,
Johan
 
 
> Gary Watson
> gary2@nexsan.com
> Nexsan Technologies Ltd.
> Derby DE21 7BF  ENGLAND
> http://www.nexsan.com
Article: 25924
Subject: Re: Global clock buffers and secondary clock buffers.
From: SteVe <Ste@spammerssuck.it>
Date: Tue, 26 Sep 2000 17:04:32 +0200
Links: << >>  << T >>  << A >>
Richard Meester wrote:

> Hello All,
>
> I have some trouble regarding primary global and secondary global clock
> lines.
> [...]
> Now my design is not yet finished, but when i implement/synthesize part
> of the design, it inferres BUFG (1 of the primary global clocks) instead
> of just a IBUF, and placing the pin on the top or bottom row of the
> device. I looked at the xilinx site, nothing usefull found. I have the
> constraints setup for these lines with USELOWSKEWLINES. I also tried by
> hardcoded instantiating an IBUF to the clock line, but it gets replaced.
>
> Anybody some suggestions????

I had a similar trouble a while ago. I was using a Virtex 1000 and I needed
a clock pin to be located on a general I/O pad. The workaround I used was to
manually edit the edif netlist generated from Synopsys DC and to replace the
BUFGP component instantiation with an IBUF. Then I read this netlist with
the Xilinx tool and all worked fine!

Hope this will help.

--
SteVe




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