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 14800

Article: 14800
Subject: Re: Digital PLL
From: Wiggo Olufsen <wiggo.olufsen@online.no>
Date: Wed, 17 Feb 1999 23:25:16 +0100
Links: << >>  << T >>  << A >>
Antonio,

I found the following document on the Actel site:
http://www.actel.com/appnotes/s04_18.pdf

Wiggo.

schaltung@hotmail.com wrote:

> Hi everyone,
>
> I want to have some sort of Digital PLL implemented in an XILINX FPGA. Does
> anyone know of any Literature, Core or information of this type of PLLs?
>
> Regards
> Antonio Moreno
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own



--
+---------------------------------------------------------------+
| Wiggo Olufsen                                                 |
| Cypress Software AS           Phone : +47-73-52 46 59         |
| P.O.Box 2668                  Fax   : +47-73-52 46 80         |
| N-7001 TRONDHEIM              E-mail: wiggo.olufsen@online.no |
| NORWAY                                                        |
+---------------------------------------------------------------+


Article: 14801
Subject: Re: "Altera FreeCore Library" back on the web
From: carlhermann.schlehaus@t-online.de (Carlhermann Schlehaus)
Date: Wed, 17 Feb 1999 23:38:40 +0100
Links: << >>  << T >>  << A >>
Hi,

Paul Baxter <paje@nospam.globalnet.co.uk> schrieb in Nachricht
7af83k$rvs$1@newnews.global.net.uk...///
>Thanks Rune for the efforts you put in over the last few years with the
>freecore library.
>
>Sad to see that Altera couldn't make you an enticing offer to stay with
>their product range!
>
>Paul Baxter
>

Let me tell you about our experiences with ALTERA:
we are working with ALTERA FPGA (FLEX8K, 6K) and two weeks ago I tried to
get a new software version.
When we bought our MAX+PLUSII we also made a maintenance contract. That way
we got several updates...
When the software maintenace ran out, we decided to keep the software as it,
so we are working with
MAX+PLUSII ver 8.0 at the moment.
Now forced to change to 6K or a great 10K in our designs we have to upgrade
and ...

ALTERA does not honour if you have an old version. Buying a new software
(and a new dongle) costs as a "upgrade" :-(
ALTERA takes $2000 for one year, so after one year you can freeze your
software version or pay another $2000     :-(

This is no good politics, as this forces you to pay year by year to get
newest devices support.
Thik this will force some guys to change to XILINX ?

Bye, Carlhermann Schlehaus







Article: 14802
Subject: Re: Flex6016 config. problem.
From: carlhermann.schlehaus@t-online.de (Carlhermann Schlehaus)
Date: Wed, 17 Feb 1999 23:43:34 +0100
Links: << >>  << T >>  << A >>
Hi,

Philip Freidin <fliptron@netcom.com> schrieb in Nachricht
fliptronF7AEJy.6Jo@netcom.com...///
>Check the signal integrity on your configuration clock. In particular,
>look at both the rising and falling edge with a 300MHz scope (or better.
>a 100MHz or 200MHz scope can miss clock problems that the FPGA/EPROM can
>react to). Check at both ends of the trace, and check BOTH edges.
>Remember to keep the scope ground wire as short as possible (less than 2
>inches would be a good goal.) You are looking for "U" turns in the edges
>of the clock during the transitions. These can be only a few nanoseconds
>wide, and maybe only a few 100 mV high.
>If you see anything like this, you have a termination problem that you
>must remedy.
>
>Philip

Are the 6K FPGA such sensitve to the signals @ DCLK ? Well I've just started
a transition from an EPF81500 (which get'S to small in LCells and to
expensive
in money) to an EPF6016A. Hope it will work without such problems.
Furthermore, how and with what values should the Configurations lines be
terminated
to ensure correct function of FPGA ?

Thanx, Carlhermann Schlehaus



Article: 14803
Subject: "Altera FreeCore Library" back on the web
From: "Rune Baeverrud" <fpga@iname.com>
Date: Wed, 17 Feb 1999 15:32:27 -0800
Links: << >>  << T >>  << A >>
Hi All,

Until quite recently - I worked as a Field Applications Engineer with one of
the distributors of Altera programmable logic. During that time - I created
the "Altera FreeCore Library" - a web site of free information and logic
cores written specifically for Altera CPLD's.

Recently - I changed my job - and I'm now working with the Norwegian
distributor of Xilinx programmable logic, BIT Elektronikk.

It is obvious that I cannot actively support Altera anymore. Instead - I
created freecore.com - my own personal web site - where I will share some of
my programmable logic experiences with you. Focus will be on VHDL, Verilog,
Design Tools and Architectures.

Once donated to the Public Domain - and with many contributions from
excellent designers throughout the world - it is not my right to completely
remove the "Altera FreeCore Library" from the web. You will find a "frozen"
version of it here - but there is no support for it. If you have questions -
don't email me - I cannot help you. Sorry.

Website is at http://freecore.com

Regards,
Rune Baeverrud
rune@freecore.com







Article: 14804
Subject: Announcing HDL Conference
From: "HDL Conference" <info@vhdl.org>
Date: Wed, 17 Feb 1999 16:56:47 -0700
Links: << >>  << T >>  << A >>
The International HDL Conference & Exhibition

                                     April 6th - 9th

                Santa Clara Convention Center

                                   Santa Clara, CA.


Advanced and introductory topics on  Verilog and VHDL.design, and Innovative
HDL Techniques for System-on-Chip (SoC) design.   This conference and
exhibition will feature papers, tutorials, panel sessions, keynote speakers,
and 50 EDA exhibitors.  Some of the topics covered include Intellectual
Property, Timing Issues, Synthesis, Verification, and Language Specifics.


Registering for the conference automatically registers you for our big prize
raffle, where 10 lucky winners will each take home a Canon-Advantix camera
(Retail $200).


Stop by the web site for more information and on-line registration.

                http://www.hdlcon.org


Mark your calendars and sign-up now (Limited seating capacity).


Sign up in advance for the full conference $300 (or $400 at the door).

==================


Article: 14805
Subject: virtex vs apex ?
From: muzok@nospam.pacbell.net (muzo)
Date: 17 Feb 1999 19:26:52 PST
Links: << >>  << T >>  << A >>
hi,
has anyone done any comparative analysis of these two news families ? Would you
like to share ?

muzo

Verilog, ASIC/FPGA and NT Driver Development Consulting (remove nospam from email)
Article: 14806
Subject: Re: Free circuit design
From: Geir Harris Hedemark <geir@dod.no>
Date: 18 Feb 1999 08:28:31 +0100
Links: << >>  << T >>  << A >>
Sander Vesik <sander@haldjas.folklore.ee> writes:
> Can you actually name these tools? Yes, I specificly mean those written
> by Linux people as opposed to say "magic".

FreeHDL is one, although only in its alpha stages.

Geir
  
Article: 14807
Subject: Re: Xilinx Spartan and pin-locking
From: murray@pa.dec.com (Hal Murray)
Date: 18 Feb 1999 07:37:37 GMT
Links: << >>  << T >>  << A >>
One of the main advantages of FPGAs is that you can fix a bug
or make a minor change to the design and ship new bits to the field.
That requires pinlocking.  So I assume that any software that doesn't
work well with locked pins is brain damaged, aka useless.

An alternate way of looking at things is that floorplanning can
extend to the pins too.  I often pick pins after carefully considering
the local routing in order to make sure the final result will be
fast enough.  This sort of thing can be very important with chips like
the Xilinx 3000 series which have direct connections that save a
significant fraction of the total cycle time over misc local routing.

Am I out to lunch to be thinking that way?

-- 
These are my opinions, not necessarily my employers.
Article: 14808
Subject: Re: "Altera FreeCore Library" back on the web
From: Achim Gratz <gratz@ite.inf.tu-dresden.de>
Date: 18 Feb 1999 08:54:47 +0100
Links: << >>  << T >>  << A >>
"Rune Baeverrud" <fpga@iname.com> writes:

> Once donated to the Public Domain - and with many contributions from
> excellent designers throughout the world - it is not my right to completely
> remove the "Altera FreeCore Library" from the web. You will find a "frozen"
> version of it here - but there is no support for it. If you have questions -
> don't email me - I cannot help you. Sorry.
> 
> Website is at http://freecore.com

Are there any plans of extending the site to other architectures?


Achim Gratz.

--+<[ It's the small pleasures that make life so miserable. ]>+--
WWW:    http://www.inf.tu-dresden.de/~ag7/{english/}
E-Mail: gratz@ite.inf.tu-dresden.de
Phone:  +49 351 463 - 8325
Article: 14809
Subject: Re: Free circuit design
From: Thomas Reinemann <thomas.reinemann@mb.uni-magdeburg.de>
Date: Thu, 18 Feb 1999 09:06:11 +0100
Links: << >>  << T >>  << A >>
Geir Harris Hedemark wrote:
> 
> Sander Vesik <sander@haldjas.folklore.ee> writes:
> > Can you actually name these tools? Yes, I specificly mean those written
> > by Linux people as opposed to say "magic".
> 
> FreeHDL is one, although only in its alpha stages.

No, there are still least Electric an Alliance, have a look at the FAQ posted in
this group.


Bye Tom!
Article: 14810
Subject: Re: Free circuit design
From: Sander Vesik <sander@haldjas.folklore.ee>
Date: 18 Feb 1999 10:20:52 GMT
Links: << >>  << T >>  << A >>
In comp.arch Thomas Reinemann <thomas.reinemann@mb.uni-magdeburg.de> wrote:
> Geir Harris Hedemark wrote:
> > 
> > Sander Vesik <sander@haldjas.folklore.ee> writes:
> > > Can you actually name these tools? Yes, I specificly mean those written
> > > by Linux people as opposed to say "magic".
> > 
> > FreeHDL is one, although only in its alpha stages.

> No, there are still least Electric an Alliance, have a look at the FAQ posted in
> this group.

But Alliance counts just as magic. 

> Bye Tom!

-- 
	Sander

	There is no love, no good, no happiness and no future -
	all these are just illusions.
Article: 14811
Subject: Re: Digital PLL
From: Jeff Streznetcky <jeff.streznetcky@lmco.com>
Date: Thu, 18 Feb 1999 07:33:52 -0500
Links: << >>  << T >>  << A >>
Look at the Xilinx Virtex series of FPGAs.

For info on the Virtex devices see:  http://www.xilinx.com/partinfo/virtex.pdf

For info on the DLLs in the Virtex devices see:
http://www.xilinx.com/xapp/xapp132.pdf


-Jeff


schaltung@hotmail.com wrote:

> Hi everyone,
>
> I want to have some sort of Digital PLL implemented in an XILINX FPGA. Does
> anyone know of any Literature, Core or information of this type of PLLs?
>
> Regards
> Antonio Moreno
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own

Article: 14812
Subject: four signals into array?
From: jamie morken <foster@uvic.ca>
Date: Thu, 18 Feb 1999 05:23:39 -0800
Links: << >>  << T >>  << A >>
Hi all,

I have some VHDL code (below) that I would like to "put" into an array R

so that I can select the register by typing R[ra]  where ra is 0 for
REGA, 1 for REGB, 2 for REGC and 3 for REGD.  Any ideas?

--current and next state of the four 8-bit general purpose registers
SIGNAL curr_REGA, next_REGA: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
A
SIGNAL curr_REGB, next_REGB: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
B
SIGNAL curr_REGC, next_REGC: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
C
SIGNAL curr_REGD, next_REGD: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
D

Thanks for your time

Jamie Morken




Article: 14813
Subject: Re: Synplify resource usage report for Virtex devices
From: brian@shapes.demon.co.uk (Brian Drummond)
Date: Thu, 18 Feb 1999 14:39:44 GMT
Links: << >>  << T >>  << A >>
On Wed, 17 Feb 1999 13:24:35 +0000, Eduardo Augusto Bezerra
<E.A.Bezerra@sussex.ac.uk> wrote:

>
>Thank you for the detailed explanation. I know I have a long way
>ahead. I think my big mistake was to try to use the technology
>independent feature of VHDL. Just in case, I'll save the present
>version in a safe place, and I hope in few years I can synthesize
>it properly.
>
>I'm going to make some changes in my design in order to turn it
>a device specific one.
>
I believe that code written to allow ram inferencing will also
successfully synthesise to individual FF's when you move it to an ASIC
process. It is just an array of storage locations, expressed in a way
that the synthesis tool will recognise as a special case when targetting
FPGA. There is nothing Xilinx-specific about the way it is expressed (or
Altera- specific, or for that matter Leonardo or Synplify-specific)
about the code involved. So your modified solution should still be
portable to ASIC later on.

- Brian

Article: 14814
Subject: Re: Synplify resource usage report for Virtex devices
From: Andrew Brown <andrewbr@nortelnetworks.com>
Date: Thu, 18 Feb 1999 15:29:00 +0000
Links: << >>  << T >>  << A >>
ram inference (like synplify can do) is not really suitable for ASIC (as
yet) - no RESET !

Brian Drummond wrote:

>
> I believe that code written to allow ram inferencing will also
> successfully synthesise to individual FF's when you move it to an ASIC
> process. It is just an array of storage locations, expressed in a way
> that the synthesis tool will recognise as a special case when targetting
> FPGA. There is nothing Xilinx-specific about the way it is expressed (or
> Altera- specific, or for that matter Leonardo or Synplify-specific)
> about the code involved. So your modified solution should still be
> portable to ASIC later on.
>
> - Brian

Article: 14815
Subject: Re: four signals into array?
From: Brian Dam Pedersen <brian@kom.auc.dk>
Date: Thu, 18 Feb 1999 16:37:23 GMT
Links: << >>  << T >>  << A >>
jamie morken wrote:
> 
> Hi all,
> 
> I have some VHDL code (below) that I would like to "put" into an array R
> 
> so that I can select the register by typing R[ra]  where ra is 0 for
> REGA, 1 for REGB, 2 for REGC and 3 for REGD.  Any ideas?
> 
> --current and next state of the four 8-bit general purpose registers
> SIGNAL curr_REGA, next_REGA: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
> A
> SIGNAL curr_REGB, next_REGB: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
> B
> SIGNAL curr_REGC, next_REGC: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
> C
> SIGNAL curr_REGD, next_REGD: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
> D
> 
> Thanks for your time

Declare your register file like this:

SUBTYPE byte IS STD_LOGIC_VECTOR(7 downto 0);

TYPE regs IS ARRAY (3 downto 0) OF byte;

SIGNAL reginstance : regs;

Now you can access reginstance(0) and so on. If you need individual bits, you
have to put your register into a temporary variable before accessing it eg:

tmp:=reginstance(0);

nibble<=tmp(7 downto 4);

Hope this helps
--                                               

Brian Pedersen, DSP Student                    _/     _/_/_/  _/_/_/  _/
Applied Signal Processing and Implementation  _/_/   _/       _/   _/ _/
Department of Communication Technology       _/  _/   _/_/_/  _/_/_/  _/
Aalborg University, Denmark                 _/_/_/_/       _/ _/      _/
URL: http://www.danbbs.dk/~kibria/brian/   _/      _/ _/_/_/  _/      _/
Article: 14816
Subject: Re: Orcad Express Plus vs Foundation Express
From: "Wenwei Qiao" <WWQiao@SIGNTECH.COM>
Date: Thu, 18 Feb 1999 17:12:15 GMT
Links: << >>  << T >>  << A >>
>Pawel Michocki wrote:
>
>Hi Pawel,
>
>I started a project last year using Orcad Express 7.2. I don't think
>they had a "Plus" version back then. After taking the 5 day training
>course I convinced myself that I should try using VHDL for my design. I
>did some work is a small section of the design and decided to proceed
>with VHDL. Once I had much of my design entered and implemented, the
>VHDL tools started behaving badly. I had numerous problems with crashes
>(due to constructs in my code that the compiler or simulator didn't
>like). The support was not adequate. I finally abandoned Orcad for
>Xilinx Foundation. I get much better support and although the Foundation
>Express compiler has its problems, I always seem to be able to get my
>work done.
>

Exactly same experience here.  We wasted a lot of time in trying to get
OrCAD v7.x work well with VHDL design.  Finally, gave up and upgraded Xilinx
software from Alliance to Foundation.

The tech supports from Xilinx and OrCAD are the same situations with their
products.


>I have received my copy of Orcad Express Plus v9.0, but have not even
>put it on my machine. After posting my story a few times in this
>newsgroup I even received an email from someone at Orcad asking me to
>give the new version a try. But when I wrote back asking for specifics
>about which problems had been fixed, I never received a reply. So the
>Orcad box still sits on my shelf (or is it on the floor beside the trash
>can?). Unfortunately I can't, at this point, fully evaluate the SW if I
>wanted to. We converted our Alliance license to a Foundation license and
>I no longer have the libraries for Orcad (??unless they come on the
>Orcad CD??).
>

We got OrCAD Express V9.0 only, not Plus version.  I just tried it for one
day and never used it again.  Now I only use V7.2 for schematic drawing.

Another story, we bought OrCAD Layout plus together with Express.  Almost 6
months later, we had a guy with 20+ year PCB layout experience joined us.
He tried OrCAD Layout for a few months and finally switched to ACCEL.


>If you are using VHDL for your design, I would say to drop Orcad like a
>hot potato(e). If you have had different results, please let me know.
>
>P.S. I once asked through this newsgroup, for anyone using Orcad doing
>FPGA work to let me know of their success. I got no replies. I repeat
>the request.
>
>***************************************************************************
**************
>If anyone out there is using Orcad successfully for FPGA design, please
>post your results.
>***************************************************************************
**************
>
>
>--
>
>Rick Collins
>
>redsp@XYusa.net
>
>remove the XY to email me.


Article: 14817
Subject: Re: "Altera FreeCore Library" back on the web
From: "Andy Peters" <apeters@noao.edu.NOSPAM>
Date: Thu, 18 Feb 1999 10:43:14 -0700
Links: << >>  << T >>  << A >>
Carlhermann Schlehaus wrote in message <7afhg8$h0d$1@news03.btx.dtag.de>...

>This is no good politics, as this forces you to pay year by year to get
>newest devices support.
>Thik this will force some guys to change to XILINX ?


Xilinx is no different.  No Virtex or Spartan support for the XACT 6.x.x
tools!  Gotta get M1.x

And the new M1.x tools don't support the (obsolete) XC73xx chips (we have
several of those in a design I have to maintain).  So you need to maintain
two versions of the tools; one for the old chips and one for the new.


-- andy
------------------------------------------
Andy Peters
Sr. Electrical Engineer
National Optical Astronomy Observatories
950 N Cherry Ave
Tucson, AZ 85719
apeters@noao.edu

Don't waste apostrophes!  The plural of the acronym for "personal computers"
is PCs, NOT PC's.



Article: 14818
Subject: edge-triggered registers on Xilinx 4000e.
From: Zhen Luo <zhenluo@ee.princeton.edu>
Date: Thu, 18 Feb 1999 13:11:28 -0500
Links: << >>  << T >>  << A >>
Hello,

I am trying to map a design to some XC4010e-3 FPGAs, in this design, I
need some registers to be triggered at the clock's rising edge and some
registers to be triggered at the clock's falling edge. I am not sure if
Xilinx 4000e FPGAs provide registers that can be triggered at falling
edge of the clock. Further more, I am using some counters by logiblox
and it seems I can hardly modify them to let them be triggered at the
falling edge. Any suggestions? 

Thanks!

-- Zhen
Article: 14819
Subject: Packing, Placement and Routing Tools for Academic FPGA Research
From: vaughn@eecg.utoronto.ca (Vaughn Betz)
Date: 18 Feb 99 18:20:33 GMT
Links: << >>  << T >>  << A >>
Hi,

New versions (Version 4.22) of Versatile Place and Route (VPR) and VPack

(Version 2.09) are now on my web page, at
http://www.eecg.toronto.edu/~vaughn/vpr/vpr.html.

VPR is a placement and routing tool for FPGA architecture research,
while VPack is a logic block packing tool for FPGA architecture
research.
Both are available free for non-profit, non-commercial use (such as
academic research).

This new version has some very significant enhancements over the last
release version (Version 3.99).  Some of the biggest ones are:

- Support for routing architectures with wires that span multiple logic
  blocks (i.e. general segmented architectures)

- Timing-driven routing

- A delay estimator and path-based timing analyzer tell you how fast a
circuit
  will run.

- A detailed model of an FPGA's routing area is built in, and lets you
  evaluate the area of different architectures.

If you have any questions, contact me at vaughn@rtrack.com.

Vaughn Betz


Article: 14820
Subject: Re: Xilinx Spartan and pin-locking
From: fliptron@netcom.com (Philip Freidin)
Date: Thu, 18 Feb 1999 18:37:56 GMT
Links: << >>  << T >>  << A >>
Hal, depending on the time of day, and your level of hunger, you may or 
may not be out to lunch. As for pin locking, and floorplanning, you are 
right on target.

The people at FPGA vendors that recommend letting their sw pick the 
pinouts are clueless as to the issues that face a system designer, 
primarily because they have never done a real design with these products 
(or any other) in their life. While it is true that the tools may for a 
single instance select pins that are more appropriate than totally 
randomly selected pins by the user, any designer that is willing to think 
about the design's data and control flow for half an hour or more, is able 
to come up with a better pin floorplan than the tools can. 

Designers though are somewhat to blame for the situation, as the message 
that many deliver to the FPGA vendor is that they want a single button 
implementation flow. They dont want to spend any time learning what 
the products can do, or how they might adjust their design style to suit 
what the silicon has to offer. Telling a designer that the tools will 
pick a pinout for them is easier than telling them that they should read 
the design-style 101 section of the product data sheet, and take some 
responsibility for the implementation flow. 

Philip Freidin


In article <7agg01$43c@src-news.pa.dec.com> murray@pa.dec.com (Hal Murray) writes:
>One of the main advantages of FPGAs is that you can fix a bug
>or make a minor change to the design and ship new bits to the field.
>That requires pinlocking.  So I assume that any software that doesn't
>work well with locked pins is brain damaged, aka useless.
>
>An alternate way of looking at things is that floorplanning can
>extend to the pins too.  I often pick pins after carefully considering
>the local routing in order to make sure the final result will be
>fast enough.  This sort of thing can be very important with chips like
>the Xilinx 3000 series which have direct connections that save a
>significant fraction of the total cycle time over misc local routing.
>
>Am I out to lunch to be thinking that way?
>
>-- 
>These are my opinions, not necessarily my employers.


Article: 14821
Subject: Re: P&R times for Altera10K200E and Virtex
From: fliptron@netcom.com (Philip Freidin)
Date: Thu, 18 Feb 1999 18:47:51 GMT
Links: << >>  << T >>  << A >>
Place and route times are totally dependent on your design style. While 
it is not uncommon to see postings in this news group from people 
complaining that their HDL based designs are taking tens of hours to 
place and route at 70% utilization in a device like a 4028XL, There are 
others of us here that have designs that are at 95% utilization of 
devices like 4062XL that place and route in 1.5 hours. 

The biggest difference in route times is effected by floorplanning your 
design. In my opinion, this is far easier to do with a schematic based 
flow than an HDL one, but is also dependent on learning what these 
devices are capable of, and adjusting the design to leverage these 
capabilities.

Philip Freidin

In article <7af9cb$lhv$1@nnrp1.dejanews.com> edwinpark@my-dejanews.com writes:
>Does anyone have any P&R times for these two devices.  Also, could you post
>some info about the design (% utilization, # of registers, # I/Os, computer
>used to P&R, etc).
>I am going to start a design that needs many iterations and am very worried
>about P&R times.
>-Edwin


Article: 14822
Subject: Re: Flex6016 config. problem.
From: fliptron@netcom.com (Philip Freidin)
Date: Thu, 18 Feb 1999 18:56:24 GMT
Links: << >>  << T >>  << A >>
I do not believe the Altera 6000 parts are any more sensitive to clock
quality than any other vendor's FPGAs. The issue is that all these parts
have very fast logic in them, even in areas where it is not needed like
the configuration logic. So even if you are configuring at 1MHz, a 1nS
glitch on the clock line can cause double clocking, and make configuration
fail. Often, configuration support is left till the end of the project,
and not given much attention, and because it is relatively slow compared
to the rest of the design, clock integrity is often overlooked. 

With regard to how to terminate the clock signals, there is nothing
special about how this should be done for FPGAs, that is different for any
other clock signal. Your termination should be such that the rising and
falling edges are monotonic at their destinations, and they arrive at the
right time. This will depend on how you design the PCB, and not the FPGA. 

In article <7afhg8$h0d$2@news03.btx.dtag.de> carlhermann.schlehaus@t-online.de (Carlhermann Schlehaus) writes:
>Philip Freidin <fliptron@netcom.com> schrieb in Nachricht
>>  ... me on signal integrity of config clock signal ...
>>Philip
>
>Are the 6K FPGA such sensitve to the signals @ DCLK ? Well I've just started
>a transition from an EPF81500 (which get'S to small in LCells and to
>expensive
>in money) to an EPF6016A. Hope it will work without such problems.
>Furthermore, how and with what values should the Configurations lines be
>terminated
>to ensure correct function of FPGA ?
>
>Thanx, Carlhermann Schlehaus
>
>
>


Article: 14823
Subject: Re: edge-triggered registers on Xilinx 4000e.
From: "Alex V. Sherstuk" <sherstuk@iname.com>
Date: Thu, 18 Feb 1999 22:04:55 +0300
Links: << >>  << T >>  << A >>


XILINX XC4000 CLB's can be configured to be triggered in any way: rising
edge/falling edge.
There are two kinds of D-flip-flop in XILINX schematic library:
   FD
   FD_1
If you are working not with a single trigger, but with a register - do not
hesitate to place INV
element on the clock circuit. Clever XILINX tools will absorb this element -
it will just create proper
kind of trigger configuration.

Regards,
   Alex Sherstuk
   AMSD Company

Zhen Luo wrote in message <36CC57D0.41C6@ee.princeton.edu>...
>Hello,
>
>I am trying to map a design to some XC4010e-3 FPGAs, in this design, I
>need some registers to be triggered at the clock's rising edge and some
>registers to be triggered at the clock's falling edge. I am not sure if
>Xilinx 4000e FPGAs provide registers that can be triggered at falling
>edge of the clock. Further more, I am using some counters by logiblox
>and it seems I can hardly modify them to let them be triggered at the
>falling edge. Any suggestions?
>
>Thanks!
>
>-- Zhen


Article: 14824
Subject: Re: four signals into array?
From: Lasse Langwadt Christensen <fuz@kom.auc.dk>
Date: Thu, 18 Feb 1999 19:42:43 GMT
Links: << >>  << T >>  << A >>
Brian Dam Pedersen wrote:
> 
> jamie morken wrote:
> >
> > Hi all,
> >
> > I have some VHDL code (below) that I would like to "put" into an array R
> >
> > so that I can select the register by typing R[ra]  where ra is 0 for
> > REGA, 1 for REGB, 2 for REGC and 3 for REGD.  Any ideas?
> >
> > --current and next state of the four 8-bit general purpose registers
> > SIGNAL curr_REGA, next_REGA: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
> > A
> > SIGNAL curr_REGB, next_REGB: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
> > B
> > SIGNAL curr_REGC, next_REGC: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
> > C
> > SIGNAL curr_REGD, next_REGD: STD_LOGIC_VECTOR (7 DOWNTO 0); -- register
> > D
> >
> > Thanks for your time
> 
> Declare your register file like this:
> 
> SUBTYPE byte IS STD_LOGIC_VECTOR(7 downto 0);
> 
> TYPE regs IS ARRAY (3 downto 0) OF byte;
> 
> SIGNAL reginstance : regs;
> 
> Now you can access reginstance(0) and so on. If you need individual bits, you
> have to put your register into a temporary variable before accessing it eg:
> 
> tmp:=reginstance(0);
> 
> nibble<=tmp(7 downto 4);
> 
> Hope this helps
> --

AFAIR you can actually do this instead  

 nibble <= reginstance(0)(7 downto 4); 


--L2C                 
--___--_-_-_-____--_-_--__---_-_--__---_-_-_-__--_----
Lasse Langwadt Christensen, MSEE (to be in 1999) 
Aalborg University, Department of communication tech.    
Applied Signal Processing and Implementation (ASPI)      
http://www.kom.auc.dk/~fuz , mailto:langwadt@ieee.org


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