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
2017JanFebMarApr2017

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 30200

Article: 30200
Subject: Re: speech
From: "Victor Schutte" <victors@mweb.co.za>
Date: Wed, 28 Mar 2001 08:54:25 +0200
Links: << >>  << T >>  << A >>
Suggestion: Try and use a PIC with either a CODEC  or even a cheaper Delta
modulator. With Delta modulation you will achieve excellent results with
100kbit/s.  Speech quality will also be acceptable with bit rates lower than
64kbit/s.  With PICs running at 5mips you should be able to perform at least
limited speech processing.

Can't you use a simple resistor/capacitor network? Some of the best
solutions sometimes cost a couple of cents.

I would agree with the other people that a FPGA is too expensive for simple
speech tasks.


"catherina" <floreq10@hotmail.com> wrote in message
news:99g823$1gb$1@news.qub.ac.uk...
> hello,
>
> a simple question, how many bits should be allocated to cover the natural
> speech signal dynamic ?
>
> I have never seen an  implementation related to speech processing on FPGA.
> Is that  because there is no need for Mhz speed ?
>
>
>
>
>



Article: 30201
Subject: Re: Powerup problems with XC9500XL
From: kfalser@REMOVETHISdurst.it (Klaus Falser)
Date: Wed, 28 Mar 2001 06:59:19 GMT
Links: << >>  << T >>  << A >>
On Tue, 27 Mar 2001 17:58:30 +0100, "david garnett"
<dave.garnett@metapurple.co.uk> wrote:

>I've run into power up problems with the XC95288XL that are similar to those
>encountered with fpga's. The power up ramp is definitely not monotonic and
>has three or four quite large current spikes - my current probe says >1.5A !
>There are no numbers that I can find in the datasheet about powerup - only
>an exhortation to provide a smoothly rising ramp. The end result is a chip
>that just hangs up.
>
>So, since since some of you must have found this problem and got round it,
>what did you do ?
>
>I'm thinking along the lines of hanging a large low esr cap or two off the
>5V rail and using a low dropout regulator with shutdown (LT1528) for the 3.3
>V rail. So, turn on the +5, wait 200 ms and then turn on the LDO.
>
>Any comments/ideas ?
>
>regards
>Dave
>
>--
>"Don't worry - I've read all about this sort of thing in books !"
>Dave Garnett                                         Metapurple Limited
>dave.garnett@metapurple.co.uk
>
>
>

We are using the LM1117 3.3V LDO (only 800 mA) for powering 2
XC95288XL from the 5V line. No delay is used.
Until now we had no problems, so I'm surprised about your 1.5A spikes.

Best regards 
Falser Klaus
R&D Electronics Department
Company	: Durst Phototechnik AG
	Vittorio Veneto Str. 59
	I-39042 Brixen
Voice	: +0472/810235
	: +0472/810111
FAX	: +0472/830980
Email	: kfalser@IHATESPAMdurst.it 

Article: 30202
Subject: Re: speech
From: Michael Strothjohann <strothjohann@rheinahrcampus.de>
Date: Wed, 28 Mar 2001 08:15:50 +0100
Links: << >>  << T >>  << A >>

catherina schrieb:

> I have never seen an  implementation related to speech processing on FPGA.
> Is that  because there is no need for Mhz speed ?

In real: FPGAs are not for free.
Xilinx and Altera like You have to pay for them.
So you will see them only in advanced ( no low cost )
speech signal processing.
bye
michael strothjohann



Article: 30203
Subject: Re: Xilinx FPGA Config file sizes.
From: Petter Gustad <spam@gustad.com>
Date: 28 Mar 2001 09:46:19 +0200
Links: << >>  << T >>  << A >>
"Dave Brown" <dbrown@novatel.ca> writes:

> for a given device? And finally, how big would a hex file be for an
> XCV1000E?

I'm using XC18V04's for the XCV1000E and I generate two MCS files. One
is 1441932 and the other is 822760 bytes. It takes almost one hour to
load and program these files on a DataIO 3980 programmer...

Petter

-- 
________________________________________________________________________
Petter Gustad       8'h2B | (~8'h2B) - Hamlet      http://www.gustad.com
#include <stdio.h>/* compile/run this program to get my email address */
int main(void) {printf ("petter\100gustad\056com\nmy opinions only\n");}

Article: 30204
Subject: Xilinx par -m
From: Petter Gustad <spam@gustad.com>
Date: 28 Mar 2001 10:07:33 +0200
Links: << >>  << T >>  << A >>

I'm trying to speed up my par jobs using multiple hosts. I have some
older sparc's which have two cpu's. I tried to list the host twice in
the nodenames file, but it seems like par is "smart" enough to detect
this and only start one par process on the machine. Any suggestions?

If only Xilinx could provide their software under Linux, then I could
run par on my Linux clusters :-(

Petter
-- 
________________________________________________________________________
Petter Gustad       8'h2B | (~8'h2B) - Hamlet      http://www.gustad.com
#include <stdio.h>/* compile/run this program to get my email address */
int main(void) {printf ("petter\100gustad\056com\nmy opinions only\n");}

Article: 30205
Subject: Re: No inputs on XC9536XL
From: Kolja Sulimma <kolja@em.informatik.uni-frankfurt.de>
Date: Wed, 28 Mar 2001 11:16:22 +0200
Links: << >>  << T >>  << A >>
Hello!

Most of the answers I get are related to one of the following:
- How I could get the signal to the correct logic level without using the CPLD (As
Peter is suggestion below)
- How to avoid having the inputs removed by the software (There were no inputs in the
first place).

The best response so far was the "out <= inputX or '1'". Unfortunatly it does not
work, as inputX is optimized away
early in the process, as it is redundant.

My current solution is the following: "out <= inputX" by using an input that has an
external pullup.
This solution is very unsatisfactory, because I might not allways have an handy Input
like this around..
So here is the import question again:

-----------
How can I implement the following VHDL code in the  XC95xx family without using bogus
extra inputs:
out <= '1';
-----------

One application for this, is to program the frequency of an ICS525-01 via JTAG.
Of course I can replace the CPLD with a dip switch and pullup, as Peter suggests
below.
And I also can replace my Spartan-II by a PLX PCI bridge, and I can replace my Virtex
by
TI DSP, and wow! I suddently got rid off all my Foundation problems :-)

Another application is to set the outputs of an CPLD that is allready soldered but not
used yet to a known level.

Overall this error of the foundation software falls in the category "Errors that
really should be warnnigs".

Kolja

Peter Alfke wrote:

> Yes, since the outputs are 3-stated ( with a week pull-up, I assume ), you can
> generate a High by connecting 10 kilohm pull-ups on the pins in question. Once the
> chip is active, the driver will easily override these external resistors.
>
> Peter Alfke, Xilinx Applications


Article: 30206
Subject: Please help a poor student with virtexe
From: "Frederic Darre" <darre@irit.fr>
Date: Wed, 28 Mar 2001 11:19:13 +0200
Links: << >>  << T >>  << A >>
Hello I am a student who work on the xilinx virtexe XCV100EPQ240,
when I use the project manager I have the following error message

_________________________________________________________________________
JTAG Programmer Started 2001/03/28 11:13:19
Cable Hardware:  1
Cable Firmware:  101
Cable FPGA:      c
Cable Algorithm: 100
Loading Boundary-Scan Description Language (BSDL) file
'C:/Xilinx/virtexe/data/xcv100e_pq240.bsd'.....completed successfully.
Checking boundary-scan chain integrity...ERROR:JTag - Boundary-scan chain
test failed at bit position '3' on instance 'toto(Device1)'.
 A problem may exist in the hardware configuration.
 Check that the cable, scan chain, and power connections are intact,
 that the specified scan chain configuration matches the actual hardware,
and
 that the power supply is adequate and delivering the correct voltage.
ERROR:JTag - Boundary scan chain has been improperly specified.  Please
check your configuration and re-enter the boundary-scan chain information.
Boundary-scan chain validated unsuccessfully.
ERROR:JTag - : The boundary-scan chain has not been declared correctly.
 Verify the syntax and correctness of the device BSDL files, correct the
files,
 reset the cable and retry this command.
ERROR:JTag - : The boundary-scan chain has not been declared correctly.
 Verify the syntax and correctness of the device BSDL files, correct the
files,
 reset the cable and retry this command.
_________________________________________________________________________

Someone could help me?
I dont know if my connectic with the multilinx is correct (on the jtag port)
or if my schema is right with the good pin number!



Article: 30207
Subject: Re: Xilinx par -m
From: =?iso-8859-1?Q?J=F6rg?= Ritter <ritter@informatik.uni-halle.de>
Date: Wed, 28 Mar 2001 12:59:36 +0200
Links: << >>  << T >>  << A >>
Hallo Petter,
who about a simple shell script, which replace the par command ?

#!/bin/csh

...../xilinx/bin/sol/par $*
...../xilinx/bin/sol/par $*

I don't know exactly, if the computed results of the two p&r jobs are
recognized.


> I'm trying to speed up my par jobs using multiple hosts. I have some
> older sparc's which have two cpu's. I tried to list the host twice in
> the nodenames file, but it seems like par is "smart" enough to detect
> this and only start one par process on the machine. Any suggestions?
>

> If only Xilinx could provide their software under Linux, then I could
> run par on my Linux clusters :-(

take the NT software and try the emulation software wine.

ciao
Joerg


Article: 30208
Subject: Re: speech
From: Ray Andraka <ray@andraka.com>
Date: Wed, 28 Mar 2001 13:25:22 GMT
Links: << >>  << T >>  << A >>
It depends on how much processing you are doing.  Audio is slow enough that you
can easily do bit serial for many channels through the same hardware, and even
then you can get it into a SpartanII-50 with room to spare at a cost of less
than $17 in small quantities.  A studio mixer might be an application where an
FPGA makes sense (Metalithic did something in this area about 10 years ago with
FPGAs).  It may also make sense where you have other functions to take care of
that you can integrate into the FPGA.

There are several barriers to using the FPGA in the average audio project: 1)
The horsepower needed to do the task is available in cheaper microporcessors, so
the more expensive FPGA is not warranted until you are doing enough processing
to push you into a more expensive DSP processor.  2) FPGA talent with the
systems know-how to keep the design small (ie bit serial, channel multiplexing)
is hard to come by.  3) Many DSPs are already integrated with decent ADCs and
DACs, so the integration with the processor may be higher, 4) the DSP processor
is often needed for control/front panel applications on a part time basis
anyway.

Michael Strothjohann wrote:
> 
> catherina schrieb:
> 
> > I have never seen an  implementation related to speech processing on FPGA.
> > Is that  because there is no need for Mhz speed ?
> 
> In real: FPGAs are not for free.
> Xilinx and Altera like You have to pay for them.
> So you will see them only in advanced ( no low cost )
> speech signal processing.
> bye
> michael strothjohann

-- 
-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

Article: 30209
Subject: P4 vs Athlon
From: "Phil Martin" <phil@martphil.demon.co.uk>
Date: Wed, 28 Mar 2001 14:40:30 +0100
Links: << >>  << T >>  << A >>
Dear All,

Does anyone have any comparison info on the P4 vs Athlon, running the
typical ModelSim, Leonardo, Xilinx P&R tools?

Is DDR for Athlon worth the price difference?

On large designs, my simulations can take a couple of hours, so a few % can
make a significant difference.

Thanks,

Phil Martin
Convergence Media Solutions Ltd.





Article: 30210
Subject: Problem with Virtex XCV1000BG560-4 ES
From: Ronald Hecht <ronald.hecht@gmx.de>
Date: Wed, 28 Mar 2001 16:34:52 +0200
Links: << >>  << T >>  << A >>
Hi,

we have a problem with programming a Virtex XCV1000BG540-4 ES
(engineering sample) using JTAG.

The JTAG Programmer gets a wrong Device ID and does not download the
bitstream. What can we do?

Where are the differences between the engineering sample and a normal
device? Are there differences between both devices at download via JTAG
and serial or parallel download? Do you know the Device IDs?

Can anybody help me?
Thanks

Article: 30211
Subject: Re: P4 vs Athlon
From: Muzaffer Kal <muzaffer@dspia.com>
Date: Wed, 28 Mar 2001 15:15:24 GMT
Links: << >>  << T >>  << A >>
On Wed, 28 Mar 2001 14:40:30 +0100, "Phil Martin"
<phil@martphil.demon.co.uk> wrote:

>Dear All,
>
>Does anyone have any comparison info on the P4 vs Athlon, running the
>typical ModelSim, Leonardo, Xilinx P&R tools?
>
In normal integer apps Athlon is much faster than a P4 even when the
comparison is 1.2 GHz Athlon to a 1.5 GHz P4 and now you can buy a
1.33 GHz Athlon. 

>Is DDR for Athlon worth the price difference?
>
Athlon large cache hides most of the %100 bandwidth increase; the
difference between PC133 and PC266 is around %7 but this might be
enough for you from your description.

>On large designs, my simulations can take a couple of hours, so a few % can
>make a significant difference.

Muzaffer

FPGA DSP Consulting
http://www.dspia.com

Article: 30212
Subject: Re: PCI-X core
From: Chris Dunlap <cdunlap@xilinx.com>
Date: Wed, 28 Mar 2001 09:12:26 -0700
Links: << >>  << T >>  << A >>

--------------F1BF8BF186D2E19DABB77D31
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I can only comment on the Xilinx PCI-X core which is available right now at
66Mhz.   The 133Mhz version of this core will be ready soon and yes . . . .

   * Virtex-II device can operate PCI-X at 133Mhz.
   * The core PCI-X code is encrypted
   * A testbench is included which can very easily be modified to do some
     powerful testing.  (there are dummy targets 32bit and 64bit and the
     stimulus file will simulate either single data phase transactions or
     burst transactions with some slight modification.
   * The interface signals are very straightforward.  The Core decodes BAR
     hits and informs the user application.  The Core will also arbitrate
     for bus access as well as verify bus mode on startup.  The core uses
     just a few signals to interface the PCI bus with the user application.
     The user application is responsible for sourcing and sinking of data.

Chris


Chris Briggs wrote:

> I'm looking for recommendations for a PCI-X core. I'm working on a
> design which will be initially implemented in an FPGA. We're targeting a
> Virtex-II 3000. For initial development, we'll be using PCI @ 33 MHz,
> then we plan to switch to PCI-X @ 66 MHz (or faster, if possible).
> Ultimately, we want to harden the design into an ASIC and have PCI-X @
> 133 MHz. This is an embedded application, so once we switch to PCI-X, we
> won't actually need to worry
> about (plain) PCI anymore, and don't expect to need PCI @ 66 MHz.
>
> I know about and am looking into the following cores:
> 1. Xilinx LogiCORE PCI-X
> 2. inSilicon PCI-X
> 3. DCM Technologies Corex-V10 and V20
> 4. Synopsys DesignWare DW_PCIX
> 5. Compaq X-caliber (free reference implementation)
>
> Does anyone have any experience with any of these, or others, that they
> can share? I realize that most detailed info on these comes with an NDA
> attached.
>
> My main questions are these:
> 1. How fast can I do PCI-X in a Virtex-II?
> 2. Does the core include unencrypted RTL source? (looking for Verilog)
> 3. Does it come with a testbench? (And is it any good?)
> 4. What does the interface to the application logic look like?
>
> Any info is appreciated. TIA.
>
> -cb

--------------F1BF8BF186D2E19DABB77D31
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
I can only comment on the Xilinx PCI-X core which is available right now
at 66Mhz.&nbsp;&nbsp; The 133Mhz version of this core will be ready soon
and yes . . . .
<ul>
<li>
Virtex-II device can operate PCI-X at 133Mhz.</li>

<li>
The core PCI-X code is encrypted</li>

<li>
A testbench is included which can very easily be modified to do some powerful
testing.&nbsp; (there are dummy targets 32bit and 64bit and the stimulus
file will simulate either single data phase transactions or burst transactions
with some slight modification.</li>

<li>
The interface signals are very straightforward.&nbsp; The Core decodes
BAR hits and informs the user application.&nbsp; The Core will also arbitrate
for bus access as well as verify bus mode on startup.&nbsp; The core uses
just a few signals to interface the PCI bus with the user application.&nbsp;
The user application is responsible for sourcing and sinking of data.</li>
</ul>
Chris
<br>&nbsp;
<p>Chris Briggs wrote:
<blockquote TYPE=CITE>I'm looking for recommendations for a PCI-X core.
I'm working on a
<br>design which will be initially implemented in an FPGA. We're targeting
a
<br>Virtex-II 3000. For initial development, we'll be using PCI @ 33 MHz,
<br>then we plan to switch to PCI-X @ 66 MHz (or faster, if possible).
<br>Ultimately, we want to harden the design into an ASIC and have PCI-X
@
<br>133 MHz. This is an embedded application, so once we switch to PCI-X,
we
<br>won't actually need to worry
<br>about (plain) PCI anymore, and don't expect to need PCI @ 66 MHz.
<p>I know about and am looking into the following cores:
<br>1. Xilinx LogiCORE PCI-X
<br>2. inSilicon PCI-X
<br>3. DCM Technologies Corex-V10 and V20
<br>4. Synopsys DesignWare DW_PCIX
<br>5. Compaq X-caliber (free reference implementation)
<p>Does anyone have any experience with any of these, or others, that they
<br>can share? I realize that most detailed info on these comes with an
NDA
<br>attached.
<p>My main questions are these:
<br>1. How fast can I do PCI-X in a Virtex-II?
<br>2. Does the core include unencrypted RTL source? (looking for Verilog)
<br>3. Does it come with a testbench? (And is it any good?)
<br>4. What does the interface to the application logic look like?
<p>Any info is appreciated. TIA.
<p>-cb</blockquote>
</html>

--------------F1BF8BF186D2E19DABB77D31--


Article: 30213
Subject: problem when printing from Xilinx FPGA editor
From: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
Date: Wed, 28 Mar 2001 19:43:41 +0200
Links: << >>  << T >>  << A >>
Hi,

I got a strange problem when trying to print a chip image from the
Xilinx FPGA editor (xilinx foundation 3.2i). After I set up the printer
correctly and press the print button Windows reports that the program
will be terminated because of errors. Does anybody know this problem and
have a solution ?

Would be nice :-)

Matthias

-- 
-------------------------------------------------
\ Matthias Fuchs                                 \
 \ esd electronic system design Gmbh              \
  \ Vahrenwalder Straße 205                        \
   \ D-30165 Hannover                               \
    \ email: matthias.fuchs@esd-electronics.com      \
     \ phone: +49-511-37298-0                         \
      \ fax:   +49-511-37298-68                        \
       --------------------------------------------------

Article: 30214
Subject: Re: Xilinx par -m
From: Brant Soudan <brants@xilinx.com>
Date: Wed, 28 Mar 2001 10:20:02 -0800
Links: << >>  << T >>  << A >>
petter,

try using upper cases:

foo
Foo
FOO
FOo

this should work.  if not, add aliases for the machine.

foo
foo1
foo2
foo3
.
.

brant


Petter Gustad wrote:

> I'm trying to speed up my par jobs using multiple hosts. I have some
> older sparc's which have two cpu's. I tried to list the host twice in
> the nodenames file, but it seems like par is "smart" enough to detect
> this and only start one par process on the machine. Any suggestions?
>
> If only Xilinx could provide their software under Linux, then I could
> run par on my Linux clusters :-(
>
> Petter
> --
> ________________________________________________________________________
> Petter Gustad       8'h2B | (~8'h2B) - Hamlet      http://www.gustad.com
> #include <stdio.h>/* compile/run this program to get my email address */
> int main(void) {printf ("petter\100gustad\056com\nmy opinions only\n");}


Article: 30215
Subject: Re: Pinout tables
From: "Simon Bacon" <simonb@tile.demon.co.cut_this.uk>
Date: Wed, 28 Mar 2001 19:40:55 +0100
Links: << >>  << T >>  << A >>

"Phil Hays" <spampostmaster@home.com> wrote...

> I would like to have a file for each type of device, with the Pin number (such
> as AB22), the physical name (PIN2), the adjacent CLB (X1Y14), the bank number
> and a description of the pin.  A sample line, not from a real part, might look
> like:
>
> PAD20      X1Y14       AB22      1       IO_LO1P/VREF
>
> Currently, for a design going into Virtex E, I am cutting and pasting out of
the
> .pdf pinout table to build a .ucf file. Starting with the .pdf pinout table I
> get the bank and the pin name, but not the pad or the nearest CLB, unless I
care
> to edit them in by hand.
>
> Starting a file like this, I can easily edit this into ucf format, easily sort
> and search by pad or by CLB location as well as pin name.  This would make
> creating a pinout much easier and less error prone.  My question, does this
sort
> of file exist for Virtex E?  If so, how can I get a copy?  If not, why not?

We have all been asking for this for ages.  Even a simpler version would
be useful.

Does anyone out there have any leverage with Xilinx?





Article: 30216
Subject: Wanted: test vector generation software
From: tom_systek@msn.com
Date: 28 Mar 2001 19:58:16 GMT
Links: << >>  << T >>  << A >>

Does anyone know of test vector generation free- or shareware?

Thanks.




 -----  Posted via NewsOne.Net: Free (anonymous) Usenet News via the Web  -----
  http://newsone.net/ -- Free reading and anonymous posting to 60,000+ groups
   NewsOne.Net prohibits users from posting spam.  If this or other posts
made through NewsOne.Net violate posting guidelines, email abuse@newsone.net

Article: 30217
Subject: JTAG Chain problem and Altera -- has anyone seen this before?
From: "Emanuel Machado" <emachado@bu.edu>
Date: Wed, 28 Mar 2001 15:05:36 -0500
Links: << >>  << T >>  << A >>
Hi,

I have this 9Ux400mm board that has one JTAG chain covering 32 devices. 12
of them are Altera devices. the oethes are TI and XILINX. Configuration
information (typically Altera POF and SOF files) is sent out from a computer
through the Byte-Blaster. All JTAG lines are buffered initially by a little
PLD (already programmed), and in addition, TCK and TMS are buffered with the
help of a TI '245, so that each TCK line only serves 4 or 5 loads.

The JTAG chain has been properly configured and tested, and it DOES work
(i.e. I can configure all the devices) -- most of the time. What is strange
is that when it does not work, the following happens: If I press "Detect
JTAG Chain Info" button, the correct information is displayed, every time,
with no problem. However, if I go and try to program the device, it claims
that the hardware does not match the chain information. Now remember, most
of the times, it DOES work, so the solution offered in altera's solution
database is not adequate.

My feeling is that there may be a problem in a line, probably TCK. Maybe the
IDCODE of the devices is not being reported correctly some of the time. With
that in mind, I downloaded the 2nd version of IDCODE.ZIP, a little jam file
that tests the chain and reads the idcode of the devices involved. I run it,
and found that it NEVER fails, not even once. So the IDCODE and other chain
characteristics are always being reported correctly.

So what else is being checked at programming time that I am having problems
with?

Any ideas on how to solve the puzzle?

Emanuel
emachado@bu.edu



Article: 30218
Subject: Re: Pinout tables
From: Vikram Pasham <Vikram.Pasham@xilinx.com>
Date: Wed, 28 Mar 2001 12:10:59 -0800
Links: << >>  << T >>  << A >>
Phil,

Yes, we do have package pinout tables with this info.
These pinout tables are available in  text format for different packages for all
Virtex devices.

The package files has pinout in the following format

Physical   Row/Col-RAM      Package       VCCO       IO           Output LVDS
Name       Location                Name           Bank        Type
Asynchronous


Call the hotline to get these package pinout in text format or let me know the
exact device /package and I will email it to you.


-Vikram
Xilinx Applications


Phil Hays wrote:

> I would like to have a file for each type of device, with the Pin number (such
> as AB22), the physical name (PIN2), the adjacent CLB (X1Y14), the bank number
> and a description of the pin.  A sample line, not from a real part, might look
> like:
>
> PAD20      X1Y14       AB22      1       IO_LO1P/VREF
>
> Currently, for a design going into Virtex E, I am cutting and pasting out of the
> .pdf pinout table to build a .ucf file. Starting with the .pdf pinout table I
> get the bank and the pin name, but not the pad or the nearest CLB, unless I care
> to edit them in by hand.
>
> Starting a file like this, I can easily edit this into ucf format, easily sort
> and search by pad or by CLB location as well as pin name.  This would make
> creating a pinout much easier and less error prone.  My question, does this sort
> of file exist for Virtex E?  If so, how can I get a copy?  If not, why not?
>
> --
> Phil Hays


Article: 30219
Subject: Re: Pinout tables
From: Vikram Pasham <Vikram.Pasham@xilinx.com>
Date: Wed, 28 Mar 2001 12:17:21 -0800
Links: << >>  << T >>  << A >>
The  package file format did not show up properly in my earlier mail.

Physical   Row/Col-RAM Package  VCCO   IO                 Output LVDS
Name       Location    Name     Bank   Type               Asynchronous

-Vikram


Vikram Pasham wrote:

> Phil,
>
> Yes, we do have package pinout tables with this info.
> These pinout tables are available in  text format for different packages for all
> Virtex devices.
>
> The package files has pinout in the following format
>
> Physical   Row/Col-RAM      Package       VCCO       IO           Output LVDS
> Name       Location                Name           Bank        Type
> Asynchronous
>
> Call the hotline to get these package pinout in text format or let me know the
> exact device /package and I will email it to you.
>
> -Vikram
> Xilinx Applications
>
> Phil Hays wrote:
>
> > I would like to have a file for each type of device, with the Pin number (such
> > as AB22), the physical name (PIN2), the adjacent CLB (X1Y14), the bank number
> > and a description of the pin.  A sample line, not from a real part, might look
> > like:
> >
> > PAD20      X1Y14       AB22      1       IO_LO1P/VREF
> >
> > Currently, for a design going into Virtex E, I am cutting and pasting out of the
> > .pdf pinout table to build a .ucf file. Starting with the .pdf pinout table I
> > get the bank and the pin name, but not the pad or the nearest CLB, unless I care
> > to edit them in by hand.
> >
> > Starting a file like this, I can easily edit this into ucf format, easily sort
> > and search by pad or by CLB location as well as pin name.  This would make
> > creating a pinout much easier and less error prone.  My question, does this sort
> > of file exist for Virtex E?  If so, how can I get a copy?  If not, why not?
> >
> > --
> > Phil Hays


Article: 30220
Subject: VHDL question
From: Jean-Marie Bussat <JMBussat@lbl.gov>
Date: Wed, 28 Mar 2001 12:45:06 -0800
Links: << >>  << T >>  << A >>

Hi VHDL wizards,

I'm using foundation 2.1i to program some xilinx XC4005E. I'm quite
new to vhdl FPGA programming (up to now, I was not targeting to real
devices and I was in the "everything is perfect" theoretical world).

I have a piece of code but the synthetizer doesn't react as I would
expect at first glance. I have to following simple code (rxready and
autosync are two input ports and datamode and syncerr are output
ports):

process(rxready, autosync)
begin
	if rxready='0' then
		syncerr <= '1';
		if autosync='1' then   
			datamode <= '0';
		else
			datamode <= '1';
		end if;	
	else
		syncerr <= '0';
		datamode <= '1';
	end if;
end process;

When I synthetize it with Fpga Express I got a warning saying
that no net is attached to port autosync. I dont understand
why and if I modify the code the following way ( I believe that
the functionnality is the same ):

process(rxready, autosync)
begin
	if rxready='0' and autosync='1' then    
		datamode <= '0';
		syncerr  <= '1';
	else
		syncerr <= not rxready;
		datamode <= '1';
	end if;
end process;

Then, the synthetizer doesn't report the warning anymore.

Does anybody has an idea about what I'm doing wrong ?

Many thanks in advance.

Jean-Marie

-- 
  _______________________________________
 |                                       | 
 | Jean-Marie Bussat                     |
 | Lawrence Berkeley National Laboratory |
 | 1 Cyclotron Road - MS 50A-6134        |
 | BERKELEY, CA 94720 - USA              |
 | Email: JMBussat@lbl.gov               |
 | Phone: (510)-486-5687                 |
 | Fax:   (510)-486-5977                 |
 |_______________________________________|

Article: 30221
Subject: Re: Pinout tables
From: "Simon Bacon" <simonb@tile.demon.co.cut_this.uk>
Date: Wed, 28 Mar 2001 23:43:53 +0100
Links: << >>  << T >>  << A >>
A pointer to the web address would be terrific.

http://www.xilinx.com/support/.... <g>

"Vikram Pasham" <Vikram.Pasham@xilinx.com> wrote

> The  package file format did not show up properly in my earlier mail.
>
> Physical   Row/Col-RAM Package  VCCO   IO                 Output LVDS
> Name       Location    Name     Bank   Type               Asynchronous
>
<snip>

> > Call the hotline to get these package pinout in text format or let me know
the
> > exact device /package and I will email it to you.
> >
> > -Vikram
> > Xilinx Applications




Article: 30222
Subject: Re: Pinout tables
From: Rick Filipkiewicz <rick@algor.co.uk>
Date: Thu, 29 Mar 2001 00:28:06 +0100
Links: << >>  << T >>  << A >>


Vikram Pasham wrote:

> The  package file format did not show up properly in my earlier mail.
>
> Physical   Row/Col-RAM Package  VCCO   IO                 Output LVDS
> Name       Location    Name     Bank   Type               Asynchronous
>
> -Vikram
>
> Vikram Pasham wrote:

If you have them why don't you just put them on the Web ?


Article: 30223
Subject: Re: Please help a poor student with virtexe
From: Vikram Pasham <Vikram.Pasham@xilinx.com>
Date: Wed, 28 Mar 2001 15:30:57 -0800
Links: << >>  << T >>  << A >>
Frederic,

Try to see if you can read Device id using JTAG programmer. Also make sure all
the JTAG pins are
connected properly . Refer sol#8255 for some JTAG debugging tips.

http://support.xilinx.com/xlnx/xil_ans_display.jsp?iLanguageID=1&iCountryID=1&getPagePath=8255



-Vikram
Xilinx Applications


Frederic Darre wrote:

> Hello I am a student who work on the xilinx virtexe XCV100EPQ240,
> when I use the project manager I have the following error message
>
> _________________________________________________________________________
> JTAG Programmer Started 2001/03/28 11:13:19
> Cable Hardware:  1
> Cable Firmware:  101
> Cable FPGA:      c
> Cable Algorithm: 100
> Loading Boundary-Scan Description Language (BSDL) file
> 'C:/Xilinx/virtexe/data/xcv100e_pq240.bsd'.....completed successfully.
> Checking boundary-scan chain integrity...ERROR:JTag - Boundary-scan chain
> test failed at bit position '3' on instance 'toto(Device1)'.
>  A problem may exist in the hardware configuration.
>  Check that the cable, scan chain, and power connections are intact,
>  that the specified scan chain configuration matches the actual hardware,
> and
>  that the power supply is adequate and delivering the correct voltage.
> ERROR:JTag - Boundary scan chain has been improperly specified.  Please
> check your configuration and re-enter the boundary-scan chain information.
> Boundary-scan chain validated unsuccessfully.
> ERROR:JTag - : The boundary-scan chain has not been declared correctly.
>  Verify the syntax and correctness of the device BSDL files, correct the
> files,
>  reset the cable and retry this command.
> ERROR:JTag - : The boundary-scan chain has not been declared correctly.
>  Verify the syntax and correctness of the device BSDL files, correct the
> files,
>  reset the cable and retry this command.
> _________________________________________________________________________
>
> Someone could help me?
> I dont know if my connectic with the multilinx is correct (on the jtag port)
> or if my schema is right with the good pin number!


Article: 30224
Subject: Re: PCI-X core
From: Rick Filipkiewicz <rick@algor.co.uk>
Date: Thu, 29 Mar 2001 00:56:56 +0100
Links: << >>  << T >>  << A >>


Chris Briggs wrote:

> I'm looking for recommendations for a PCI-X core. I'm working on a
> design which will be initially implemented in an FPGA. We're targeting a
> Virtex-II 3000. For initial development, we'll be using PCI @ 33 MHz,
> then we plan to switch to PCI-X @ 66 MHz (or faster, if possible).
> Ultimately, we want to harden the design into an ASIC and have PCI-X @
> 133 MHz. This is an embedded application, so once we switch to PCI-X, we
> won't actually need to worry
> about (plain) PCI anymore, and don't expect to need PCI @ 66 MHz.
>
> I know about and am looking into the following cores:
> 1. Xilinx LogiCORE PCI-X
> 2. inSilicon PCI-X
> 3. DCM Technologies Corex-V10 and V20
> 4. Synopsys DesignWare DW_PCIX
> 5. Compaq X-caliber (free reference implementation)
>
> Does anyone have any experience with any of these, or others, that they
> can share? I realize that most detailed info on these comes with an NDA
> attached.
>
> My main questions are these:
> 1. How fast can I do PCI-X in a Virtex-II?
> 2. Does the core include unencrypted RTL source? (looking for Verilog)
> 3. Does it come with a testbench? (And is it any good?)
> 4. What does the interface to the application logic look like?
>
> Any info is appreciated. TIA.
>
> -cb

Could someone enlighten me ? Is PCI-X just a clock speed uprated 66MHz PCI
or are there any fundamental differences ? As far as I can see the only
significant change is to allow split cycles.




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
2017JanFebMarApr2017

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