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 3250

Article: 3250
Subject: ANNOUNCE: New Products and Prices
From: William Billowitch <wdb@vhdl.com>
Date: Fri, 03 May 1996 09:11:48 -0400
Links: << >>  << T >>  << A >>
We are announcing two new products. Please visit our site
http://www.vhdl.com  for more details.

New: Sledgehammer-4 source code editor for C, C++, ABEL, VHDL, Verilog
New: Dragster VHDL model generation tool

--
Sincerely,

-------------------------------------------------------------
William Billowitch                   e-mail: wdb@vhdl.com
The VHDL Technology Group            Web:    http://www.vhdl.com
100 Brodhead Road, Suite 140         Phone : 610-882-3130
Bethlehem, PA 18017                  Fax   : 610-882-3133


Article: 3251
Subject: Re: Simple Xilinx board
From: Kaj Norman Nielsen <knn@inet.ats.dk>
Date: 3 May 1996 13:23:22 GMT
Links: << >>  << T >>  << A >>
ian@PROBLEM_WITH_INEWS_DOMAIN_FILE (Ian McCrum STAFF) wrote:
>We are designing a simple lowcost prot-typing board that takes 4 84 pin
>Xilinx FPGAs and provides a wirewrap area and some leds and switches. We
>want to daisychain these in such a way that the standard Xilinix software
>will download into them. we want to use a simple parallel cable, no
>expensive Xchecker. I recall that the pinouts were in the very old Xilinx
>manuals, and maybe the circuit of their old prototype boards.
>
>Can anyone help supply me with the pinouts? I suppose the daisychain
>programming uses slave serial mode, as listed in the current manuals.
>I do realise that using the parallel cable is not as powerful as an Xcehcker
>but there is a hint in the modern docs that it is possible.
>
>Anyone else wish to contribute their hardware scematics if they have done
>anything similiar?
>
>I'll post all data on my web... within a few months Thanks in anticiaption
>
>--
>Regards from  Ian McCrum, Lecturer in Digital Systems and ECAD   
>Email:  <IJ.MCCRUM@ulst.ac.uk>   WEB:  http://www.eej.ulst.ac.uk/index.html
>POST-MAIL: School of Electrical & Mechanical Engineering 
>University of Ulster at Jordanstown,   Northern Ireland, BT37 0QB
>Tel: +44 1232 366364 or at home          +44 1247 882889
>Fax: +44 1232 362804 or preferably       +44 1247 882894
>------------------------------ ends ------------------------------------ 8-}


Hi . I'm afraid i cant give you any help at the moment but Im very 
interested in your project as I am about to do the same on my own in the 
autumn.

I'm lecturing in digital systems and for the moment I'm doing the same 
kind of breadboard for an AMD-445 IC.


My AMD breadboard is designed from the AMD-manuals and there has not been 
any problems yet( it hasn't been tested  :-)).

I'm having done a PCB'layout at the moment if you are interested .


Maybe we can exchange  design, PCB-layout etc at a later time ..

Good luck

Best regards  Kaj



Article: 3252
Subject: Re: Simple Xilinx board
From: Kaj Norman Nielsen <knn@inet.ats.dk>
Date: 3 May 1996 13:23:58 GMT
Links: << >>  << T >>  << A >>
ian@PROBLEM_WITH_INEWS_DOMAIN_FILE (Ian McCrum STAFF) wrote:
>We are designing a simple lowcost prot-typing board that takes 4 84 pin
>Xilinx FPGAs and provides a wirewrap area and some leds and switches. We
>want to daisychain these in such a way that the standard Xilinix software
>will download into them. we want to use a simple parallel cable, no
>expensive Xchecker. I recall that the pinouts were in the very old Xilinx
>manuals, and maybe the circuit of their old prototype boards.
>
>Can anyone help supply me with the pinouts? I suppose the daisychain
>programming uses slave serial mode, as listed in the current manuals.
>I do realise that using the parallel cable is not as powerful as an Xcehcker
>but there is a hint in the modern docs that it is possible.
>
>Anyone else wish to contribute their hardware scematics if they have done
>anything similiar?
>
>I'll post all data on my web... within a few months Thanks in anticiaption
>
>--
>Regards from  Ian McCrum, Lecturer in Digital Systems and ECAD   
>Email:  <IJ.MCCRUM@ulst.ac.uk>   WEB:  http://www.eej.ulst.ac.uk/index.html
>POST-MAIL: School of Electrical & Mechanical Engineering 
>University of Ulster at Jordanstown,   Northern Ireland, BT37 0QB
>Tel: +44 1232 366364 or at home          +44 1247 882889
>Fax: +44 1232 362804 or preferably       +44 1247 882894
>------------------------------ ends ------------------------------------ 8-}


Hi . I'm afraid i cant give you any help at the moment but Im very 
interested in your project as I am about to do the same on my own in the 
autumn.

I'm lecturing in digital systems and for the moment I'm doing the same 
kind of breadboard for an AMD-445 IC.


My AMD breadboard is designed from the AMD-manuals and there has not been 
any problems yet( it hasn't been tested  :-)).

I'm having done a PCB'layout at the moment if you are interested .


Maybe we can exchange  design, PCB-layout etc at a later time ..

Good luck

Best regards  Kaj



Article: 3253
Subject: Re: Simple Xilinx board
From: Kaj Norman Nielsen <knn@inet.ats.dk>
Date: 3 May 1996 13:24:54 GMT
Links: << >>  << T >>  << A >>
ian@PROBLEM_WITH_INEWS_DOMAIN_FILE (Ian McCrum STAFF) wrote:
>We are designing a simple lowcost prot-typing board that takes 4 84 pin
>Xilinx FPGAs and provides a wirewrap area and some leds and switches. We
>want to daisychain these in such a way that the standard Xilinix software
>will download into them. we want to use a simple parallel cable, no
>expensive Xchecker. I recall that the pinouts were in the very old Xilinx
>manuals, and maybe the circuit of their old prototype boards.
>
>Can anyone help supply me with the pinouts? I suppose the daisychain
>programming uses slave serial mode, as listed in the current manuals.
>I do realise that using the parallel cable is not as powerful as an Xcehcker
>but there is a hint in the modern docs that it is possible.
>
>Anyone else wish to contribute their hardware scematics if they have done
>anything similiar?
>
>I'll post all data on my web... within a few months Thanks in anticiaption
>
>--
>Regards from  Ian McCrum, Lecturer in Digital Systems and ECAD   
>Email:  <IJ.MCCRUM@ulst.ac.uk>   WEB:  http://www.eej.ulst.ac.uk/index.html
>POST-MAIL: School of Electrical & Mechanical Engineering 
>University of Ulster at Jordanstown,   Northern Ireland, BT37 0QB
>Tel: +44 1232 366364 or at home          +44 1247 882889
>Fax: +44 1232 362804 or preferably       +44 1247 882894
>------------------------------ ends ------------------------------------ 8-}


Hi . I'm afraid i cant give you any help at the moment but Im very 
interested in your project as I am about to do the same on my own in the 
autumn.

I'm lecturing in digital systems and for the moment I'm doing the same 
kind of breadboard for an AMD-445 IC.


My AMD breadboard is designed from the AMD-manuals and there has not been 
any problems yet( it hasn't been tested  :-)).

I'm having done a PCB'layout at the moment if you are interested .


Maybe we can exchange  design, PCB-layout etc at a later time ..

Good luck

Best regards  Kaj



Article: 3254
Subject: Re: Simple Xilinx board
From: Kaj Norman Nielsen <knn@inet.ats.dk>
Date: 3 May 1996 13:26:33 GMT
Links: << >>  << T >>  << A >>
ian@PROBLEM_WITH_INEWS_DOMAIN_FILE (Ian McCrum STAFF) wrote:
>We are designing a simple lowcost prot-typing board that takes 4 84 pin
>Xilinx FPGAs and provides a wirewrap area and some leds and switches. We
>want to daisychain these in such a way that the standard Xilinix software
>will download into them. we want to use a simple parallel cable, no
>expensive Xchecker. I recall that the pinouts were in the very old Xilinx
>manuals, and maybe the circuit of their old prototype boards.
>
>Can anyone help supply me with the pinouts? I suppose the daisychain
>programming uses slave serial mode, as listed in the current manuals.
>I do realise that using the parallel cable is not as powerful as an Xcehcker
>but there is a hint in the modern docs that it is possible.
>
>Anyone else wish to contribute their hardware scematics if they have done
>anything similiar?
>
>I'll post all data on my web... within a few months Thanks in anticiaption
>
>--
>Regards from  Ian McCrum, Lecturer in Digital Systems and ECAD   
>Email:  <IJ.MCCRUM@ulst.ac.uk>   WEB:  http://www.eej.ulst.ac.uk/index.html
>POST-MAIL: School of Electrical & Mechanical Engineering 
>University of Ulster at Jordanstown,   Northern Ireland, BT37 0QB
>Tel: +44 1232 366364 or at home          +44 1247 882889
>Fax: +44 1232 362804 or preferably       +44 1247 882894
>------------------------------ ends ------------------------------------ 8-}


Hi . I'm afraid i cant give you any help at the moment but Im very 
interested in your project as I am about to do the same on my own in the 
autumn.

I'm lecturing in digital systems and for the moment I'm doing the same 
kind of breadboard for an AMD-445 IC.


My AMD breadboard is designed from the AMD-manuals and there has not been 
any problems yet( it hasn't been tested  :-)).

I'm having done a PCB'layout at the moment if you are interested .


Maybe we can exchange  design, PCB-layout etc at a later time ..

Good luck

Best regards  Kaj



Article: 3255
Subject: Re: Simple Xilinx board
From: ft63@dial.pipex.com (Peter)
Date: Fri, 03 May 1996 14:58:24 GMT
Links: << >>  << T >>  << A >>


>We are designing a simple lowcost prot-typing board that takes 4 84 pin
>Xilinx FPGAs and provides a wirewrap area and some leds and switches. We
>want to daisychain these in such a way that the standard Xilinix software
>will download into them. we want to use a simple parallel cable, no
>expensive Xchecker. I recall that the pinouts were in the very old Xilinx
>manuals, and maybe the circuit of their old prototype boards.

The easiest way will be to load the first one in parallel, and load
the other three from the first one's DOUT & CCLK signals.

On the PC's parallel port, tie the BUSY line into a permanently ready
state (on your board), and then when you output a byte to LPT1 it
should get strobed into the first FPGA. You need to make sure you have
a certain min gap between consecutive bytes, but that should be OK
because of the time it takes to run through the LPT BIOS code etc.

Then just use MAKEPROM to make a hex file, and you need a little
program which converts this into a binary image which you can "COPY /b
filename lpt1" to the board. This program needs to allocate a RAM
block, fill it with FFs, then load the hex image into it according to
the address in each hex record. These programs are sometimes used to
drive EPROM programmers. A pity MAKEPROM cannot (I think) generate
binary images directly.

I cannot see why this should not work. In fact I remember seeing an
article somewhere, with example code.

Peter.


Article: 3256
Subject: Re: FPGA leaders - Who are they? Xilinx, Altera, Actel?
From: jcooley@world.std.com (John Cooley)
Date: Fri, 3 May 1996 19:42:59 GMT
Links: << >>  << T >>  << A >>
Peter Alfke <peter@xilinx.com> wrote:
>Here are the calendar 1995 revenue numbers for the biggest players in the
>Programmable Logic Market:
>
>Xilinx $ 520 M
>Altera $ 402 M
>AMD $ 285 M
>Lattice $ 185 M
>Actel $ 106 M
>Cypress $ 76 M.

Peter, do you know QuickLogic's sales numbers for this time period?  Because
they get all their silicon from Cypress, I'd be tempted to lump the two
companies together for comparing market sizes.  (A combined Cypress-QuickLogic
might even be bigger than Actel or possibly even Lattice -- don't know for
sure -- just guessing.)
                           - John Cooley
                             Part Time EDA Consumer Advocate
                             Full Time ASIC, FPGA & EDA Design Consultant

===========================================================================
 Trapped trying to figure out a Synopsys bug?  Want to hear how 3713 other
 users dealt with it ?  Then join the E-Mail Synopsys Users Group (ESNUG)!
 
      !!!     "It's not a BUG,               jcooley@world.std.com
     /o o\  /  it's a FEATURE!"                 (508) 429-4357
    (  >  )
     \ - /     - John Cooley, EDA & ASIC Design Consultant in Synopsys,
     _] [_         Verilog, VHDL and numerous Design Methodologies.

     Holliston Poor Farm, P.O. Box 6222, Holliston, MA  01746-6222
   Legal Disclaimer: "As always, anything said here is only opinion."


Article: 3257
Subject: Re: Simple Xilinx board (and cool application)
From: jhallen@world.std.com (Joseph H Allen)
Date: Sat, 4 May 1996 07:23:18 GMT
Links: << >>  << T >>  << A >>
You guys should think about using the serial port instead:

Use a MAX232 chip for converting rs-232 to ttl levels.  This is a convenient
device because it has its own +/- 9V power supply (requiring four external
capacitors), and can work off of 5V only.  These chips cost $2 plus $2 more
for the capacitors.  It may be worth while to use the MAX233 instead- it has
built in capacitors, but costs about $5.

To make the PC's serial port software happy, you need to connect CTS to RTS
and DTR to DCD and DSR.  You should receive DTR and send-data with the
MAX232, and you should transmit receive-data with the MAX232.  Also some
serial port cards have a jumper so that you can get 5V on the
otherwise-useless ring-detect line (or you can cut and jumper your serial
port board to do this- 5V on the ring detect line when not harm modems).

You want to be able to reprogram the xilinx, so when DTR drops the PROG/DONE
line should be pulled down.  The MAX-232 chip inverts the signals so you can
do this with an NPN transistor (like 2n3904) and a resistor.  Connect the
collector to the PROG/DONE line.  Connect the emitter to ground and connect
the base though a 1K resistor to DTR from the max232.

Send data will have the configuration data, so it should be connected to DIN
on the xilinx.  To generate CCLK, use a 74LS221 dual one-shot (costs about
$0.60).  Use one of the one-shots (triggered from a low going edge) to delay
to the middle of an asynchronous data frame (about 500uS for 9600 baud). 
Feed this to the second one shot, which will make the actual low-going cclk
pulse (about 1uS).  Thus the start bit of each serial frame will trigger the
one-shots so that a clock pulse appears when the data bits are stable on the
DIN line.  For each bit to be sent to the xilinx you have to transmit a 0xFF
(for a 1) or a 0x00 (for a 0) (remember data on rs-232 is inverted, but the
max-232 fixes it).

You probably also want to connect the receive data line to one of the xilinx
pins- this way your xilinx 'program' can communicate bidirectionally with
the PC after configuration (if you implement a serial port in the xilinx,
which is easy to do).  Also it's trivial to cascade several xilinx's
together.

I've made tiny little boards containing this circuit with a XC2064 (yes a
2064! $1 each surplus for these) and a DB9 serial port connector.

I use these to convert credit card machines (which also are used for
Medicaid verification and a bunch of other things) so that I can remotely
type on the keypad and read the vacuum flourescent screen (this hardware
solution is much much eaiser than hacking the protocol on the phone line
(which tends to change often anyway)).  The xilinx is better suited for this
because a microcontroller may not be able to keep up with keypad's scanning
lines.  Anyway, I sell them and a small program to automate methadone
treatment companys Medicaid verification process for $2500 each. Which is
well worth the money because it usually eliminates an employee- each patient
has to be verified every day, 100 patients can take 5 hours to do.  Small
companies can not get an automated program from the state (or it costs too
much).
 
The same board can be easily reconfigured for other keypad/displays.  I've
also used it to make a laser camera (an expensive device which prints MRI
films) remote controllable (for an image archiving system).  You can get a
factory upgrade for the laser camera so that it can be remote controllable
(using their proprietary interface), but for this particular brand it costs
$25,000 (about the same as the entire archiving system).
-- 
/*  jhallen@world.std.com (192.74.137.5) */               /* Joseph H. Allen */
int a[1817];main(z,p,q,r){for(p=80;q+p-80;p-=2*a[p])for(z=9;z--;)q=3&(r=time(0)
+r*57)/7,q=q?q-1?q-2?1-p%79?-1:0:p%79-77?1:0:p<1659?79:0:p>158?-79:0,q?!a[p+q*2
]?a[p+=a[p+=q]=q]=q:0:0;for(;q++-1817;)printf(q%79?"%c":"%c\n"," #"[!a[q-1]]);}


Article: 3258
Subject: Re: Simple Xilinx board
From: Scott Kroeger <scott.kroeger@mei.com>
Date: Sat, 04 May 1996 18:33:10 -0500
Links: << >>  << T >>  << A >>
Ian McCrum STAFF wrote:
> 
> We are designing a simple lowcost prot-typing board that takes 4 84 pin
> Xilinx FPGAs and provides a wirewrap area and some leds and switches. We
> want to daisychain these in such a way that the standard Xilinix software
> will download into them. we want to use a simple parallel cable, no
> expensive Xchecker. I recall that the pinouts were in the very old Xilinx
> manuals, and maybe the circuit of their old prototype boards.
> 
> Can anyone help supply me with the pinouts? I suppose the daisychain
> programming uses slave serial mode, as listed in the current manuals.
> I do realise that using the parallel cable is not as powerful as an Xcehcker
> but there is a hint in the modern docs that it is possible.
> 
> Anyone else wish to contribute their hardware scematics if they have done
> anything similiar?
> 
> I'll post all data on my web... within a few months Thanks in anticiaption


Hello Ian,

I hacked the following program together years ago when I first started 
playing with Xilinx parts and "C".  It is compatible with the Xilinx 
demo boards and accepts an intel hex bitfile.  I just hacked in a crude 
wiring diagram which I believe is correct.  I didn't document it when I 
wrote it (bad dog!) and had to reverse engineer it just now.  The 
program uses four PC parallel port pins to drive three signal lines 
(DIn, CClk and *Prog/Done) to a Xilinx part in slave serial mode.  You 
can hack the -r switch code to generate a reset from another port pin 
instead of prompting to push the demo board reset button.  A single 
Schottky diode (any kind will do) is required to isolate the parallel 
port pin that drives *Prog when it goes high so that the state of Done 
can be read by the PC on the SelectIn pin.

I have schematics for the demo board, but not in electronic form.  They 
may be copyrighted by Xilinx.  Peter Alfke of Xilinx often drops in on 
this newsgroup and may be able to help.  If you have no luck, get back 
to me and I'll see what I can do.

Have fun,
Scott

Program follows (originally compiled with QuickC):

/*
   A simple program to download intel hex format Xilinx bitmap files to
   Xilinx devices in slave serial mode from the PC parallel port (LPT1).
   
   Scott Kroeger Copyright 1989

  Wiring:

	LPT Port Pin	Port Name	Xilinx Pin
	2				D0----------DIn
	3				D1----------CClk
	6				D4----|<----*Prog/Done
	13				SelectIn----*Prog/Done
					
	Note: -|<- = Schottky diode.

*/

#include <stdio.h>
#include <ctype.h>
#include <fcntl.h>
#include <conio.h>

main(argc,argv)

char *argv[];
int argc;

{
	FILE *in;
	int doReset;

	printf("Xilinx downloader version 1.0\n" );

	if (argc < 2) {
		printf("correct usage is: download infile\n");
		exit(1);
	}

	in = fopen (argv[argc-1],"rt");
	if(in == NULL){
		printf("can't open input file: %s\n",argv[argc-1]);
		exit (1);
	}

	doReset=1;
	if( strstr(argv[1],"-r") ) doReset = 0;

	download(in,doReset);

	fclose(in);
	return (0);
}

download(in,doReset)
FILE *in;
int doReset;
{
	register int c,byteCount;
	outp(0x378,0x03); /* assert prog */
	if(doReset){
		printf("Reset the device and press a key when ready.\n");
		while (kbhit()==0);
	}
	outp(0x378,0x13); /* release prog */
	if (inp(0x379) & 0x10) {
		printf("Done did not stay low, loading stopped.\n");
	}
	getc(in);   /* discard first record */
	while ((c=getc(in)) != EOF){
		if(c == ':') {
			if(byteCount = GetByte(in)){
				downloadRecord(byteCount,in);
			}
		}
	}
	if (inp(0x379) & 0x10)
		printf("Load Done.\n");
	else
		printf("Done did not go high, bad load.\n");
}

downloadRecord(byteCount,in)
FILE *in;
int byteCount;
{
	int a;
	GetByte(in);    /* discard load address and record type */
	GetByte(in);
	GetByte(in);

	for(;byteCount>0;byteCount--){
		SendByte(GetByte(in));
	}
}

GetByte(in)
FILE *in;
{
	int     b,temp;
	temp = getc(in);
	b=toupper(temp)-'0';
	if (b > 9) b-=7;
	temp=b<<4;

	b = getc(in);
	b=toupper(b)-'0';
	if (b > 9) b-=7;
	b=b+temp;
	return(b);
}

SendByte(byte)
int byte;
{
	int bit;

	for (bit=0;bit<=7;bit++) {
		if(byte & 1) {
			outp(0x378,0x11); /* Data = 1, CClk = 0 */
			outp(0x378,0x11); /* waste time to avoid overruns */
			outp(0x378,0x13); /* Data = 1, CClk = 1 */
			outp(0x378,0x13);
		}
		else {
			outp(0x378,0x10); /* Data = 0, CClk = 0 */
			outp(0x378,0x10);
			outp(0x378,0x12); /* Data = 0, CClk = 1 */
			outp(0x378,0x12);
		}
		byte = byte >> 1;
	}
}


Article: 3259
Subject: FS Actel development system
From: dwp@deltanet.com
Date: 5 May 1996 01:02:35 GMT
Links: << >>  << T >>  << A >>
Not sure if this is the appropriate place, but I have an Actel 
development system that I don't expect to be using soon and 
would like to sell it.  Includes software and device programmer.
Both for the IBM PC platform.

As I seldom visit this group please email any interest.

dwp@deltanet.com

thanks, dennis



Article: 3260
Subject: Re: Simple Xilinx board
From: daveb@iinet.net.au (Dave)
Date: Sun, 05 May 1996 22:37:56 GMT
Links: << >>  << T >>  << A >>
ft63@dial.pipex.com (Peter) wrote:



>>We are designing a simple lowcost prot-typing board that takes 4 84 pin
>>Xilinx FPGAs and provides a wirewrap area and some leds and switches. We
>>want to daisychain these in such a way that the standard Xilinix software
>>will download into them. we want to use a simple parallel cable, no
>>expensive Xchecker. I recall that the pinouts were in the very old Xilinx
>>manuals, and maybe the circuit of their old prototype boards.

>The easiest way will be to load the first one in parallel, and load
>the other three from the first one's DOUT & CCLK signals.

>On the PC's parallel port, tie the BUSY line into a permanently ready
>state (on your board), and then when you output a byte to LPT1 it
>should get strobed into the first FPGA. You need to make sure you have
>a certain min gap between consecutive bytes, but that should be OK
>because of the time it takes to run through the LPT BIOS code etc.

>Then just use MAKEPROM to make a hex file, and you need a little
>program which converts this into a binary image which you can "COPY /b
>filename lpt1" to the board. This program needs to allocate a RAM
>block, fill it with FFs, then load the hex image into it according to
>the address in each hex record. These programs are sometimes used to
>drive EPROM programmers. A pity MAKEPROM cannot (I think) generate
>binary images directly.

>I cannot see why this should not work. In fact I remember seeing an
>article somewhere, with example code.

>Peter.

 If you are using XC4000 parts (don't know about later ones), you can
do a neat remote configuration using the IEEE 1149 TAP port. This can
be connected to the PC's parallel port (just use a couple of data
lines), and generate a TAP protocol to configure it.
 I have built a number of devices using this method, commonly
comprising an embedded CPU, memory and FPGA as a specialised
peripheral. With the software (and the real configuration) in a flash
ROM, this method allows the PC to "take over" the target hardware, and
turn it into an in-system flash ROM programmer. This is then used to
install the working software. The method will work with a factory-new
ROM in the system (ie no boot code is needed). The boards are arranged
so that on power-up, if the programming cable is connected, the CPU is
locked off the bus, and the FPGA awaits remote configuration. Else,
the FPGA self-configures (in master parallel mode, from the ROM), then
takes Reset off the CPU to run the system.
 One tip: parallel port lines often have awful bounce on them.
Generate the TAP clock on-board, using a R-S latch controlled by TWO
parallel data lines, one to set and one to reset. Now the bounce is
eliminated, similar to debouncing a Form-C switch contact. This avoids
double-clocking the TAP logic, which is fatal.



--  Dave Brooks
PGP public key: finger  daveb@opera.iinet.net.au
                servers daveb@iinet.net.au
    fingerprint 20 8F 95 22 96 D6 1C 0B  3D 4D C3 D4 50 A1 C4 34



Article: 3261
Subject: plpl2.3 device def. format?
From: stilti@leland.Stanford.EDU (Timothy Scott Stilson)
Date: 6 May 1996 01:12:46 GMT
Links: << >>  << T >>  << A >>

Does anybody have information on the plpl2.3 (not plpl2.2) device
definition file format (the files named "p<device>")?  I was able
to glean most of the format from the 2.2 source files, but it looks
like the support for OR arrays was added in v2.3 (the lines in the
files that begin with ">"), and I don't know if my interpretation
of the format for these is correct.

I'm trying to hack up a definition file for a GAL6002.  Ithink I
can do it, basing it on the p22v10, pls168 (for the or array)
and the p29m16 (for the dual-feedback macrocells) files, but
I'd like to verify a few points in the syntax.

Maybe someone has already done the GAL6002?


--Tim Stilson
stilti@ccrma.stanford.edu


Article: 3262
Subject: Re: FPGA leaders - Who are they? Xilinx, Altera, Actel?
From: "f.j.koons" <fjk@aloft.att.com>
Date: Mon, 06 May 1996 16:12:21 -0400
Links: << >>  << T >>  << A >>
My two cents worth:

"FPGA - The field-programmable gate array (FPGA) 
 was developed as a programmable alternative to 
 the masked gate array, another popular semicustom 
 ASIC device. An FPGA uses a channeled programmable 
 interconnect matrix to join blocks of configurable 
 logic. Unlike the CPLD, an FPGA's electrical 
 characteristics are not fixed, so qualities 
 such as line length and number of fuses are not 
 known until routing. Although this makes for some 
 unpredictability of timing, the trade-off is the 
 FPGA's increased logic complexity and flexibility."

So....segmenting the FPGA portion of the programmable 
logic market by this definition yields:

"Company     95 Sales     94 Sales      95 MS      94 MS
 ---------------------------------------------------------
 Xilinx        $504.5       $315.1      70.4%      67.3%
 Actel         $108.4        $76.0      15.1%      16.2%
 AT&T Micro     $77.8        $44.9      10.9%       9.6%
 QuickLogic     $17.1         $7.5       2.4%       1.6%

 Table 6-22: FPGA vendors' sales (in millions) and market shares,
 1995, and comparision to 1994."

 - definition and data from: "PLD Market Report 1995", 
                             by Rhondalee Rohleder and
                             published by Pace Technologies

Altera, AMD, Lattice, and Cypress account for most of their revenue 
from the sales of simple and complex PLDs. Lucent Technologies 
(formerly AT&T Microelectronics) accounts for all of its revenue from
the sales of SRAM-based FPGAs (ATT3000 and ORCA-Series).

======================================================
Frederick Koons     Strategic Marketing Manager, FPGAs         
Lucent Technologies          E-Mail: fjk@aloft.att.com
====================================================== 


> Peter Alfke wrote:
> 
> In article <4lqme8$2iq@news.tiac.net>, Eric Ryherd <eric@vautomation.com> wrote:
> 
> > Does anyone have a reference on the annual number of FPGAs shipped
> > (or even revenue) by the top 3 FPGA companies.
> >
> > I assume the order is
> > 1) Xilinx,
> > 2) Altera,
> > 3) Actel.
> >
> Here are the calendar 1995 revenue numbers for the biggest players in the
> Programmable Logic Market:
> 
> Xilinx $ 520 M
> Altera $ 402 M
> AMD $ 285 M
> Lattice $ 185 M
> Actel $ 106 M
> Cypress $ 76 M.
> 
> Xilinx sales are mainly SRAM-based FPGAs, plus some CPLDs and some
> Antifuse-based FPGAs.
> 
> Altera does not call any of its devices "FPGA", and most of their sales
> are CPLDs in EPROM and EEPROM technology.
> AMD is 100% CPLDs and PALs.
> Lattice is CPLDs and GALs.
> Actel is 100% antifuse-based FPGAs.
> Cypress is mostly CPLDs.
> 
> I hope nobody is confused or offended by this explanation of the alphabet-soup.
> 
> FPGA stands for Field-Programmable Gate Array,
> CPLD stands for Complex Programmable Logic Device
> PAL stands for Programmable Array Logic, trademarked by AMD
> GAL stands for Generic Array logic, trademarked by Lattice.
> EPROM stands for Erasable Programmable Read-Only Memory, and
> EEPROM stands for Electrically Erasable Programmable Read-Only Memory. Wow!
> 
> Peter Alfke, Xilinx Applications


Article: 3263
Subject: On FPGAs as PC coprocessors [rererepost]
From: jangr@microsoft.com (Jan Gray)
Date: 6 May 1996 22:12:11 GMT
Links: << >>  << T >>  << A >>
[I've had many problems posting this message.
Netcom corrupted several attempts, and two repostings
from Microsoft's Usenet site never made it out the door.
Apologies if you've already read this or are tired of the reposts,
but I really wanted to get the message out in its entirety.]


One of the "way out speculation" questions asked at FCCM 96 (IEEE 
Symposium on FPGAs for Custom Computing Machines) was "when, if ever, 
will an FPGA coprocessor ship on every PC motherboard?"

Ignoring the daunting language and interface standards issues, and just 
looking at current hardware approaches to FPGA "coprocessors", the 
answer must be "not any time soon".

Consider that today's power user's CPU (such as a 200 MHz Pentium Pro 
or a 400 MHz Alpha 21164A) *running out of L1 cache* issues peak three 
or four instructions per 2.5-5 ns clock.  Also consider that the 
current "low latency high bandwidth" approach to FPGA coprocessor 
integration is to hang the FPGA on the PCI bus.

In this scenario, the kinds of quasi-general purpose computing problems 
that an FPGA coprocessor can usefully assist with are quite limited.  
Issuing a write and then a read back operation to the FPGA could easily 
take 10 PCI bus cycles (300 ns), assuming no PCI bus contention.  In 
that time (assuming hand crafted code (less effort than a FCCM)) a 
Pentium Pro could issue as many as 180 instructions; the Alpha, 480 
64-bit instructions. Future versions of such processors will soon be 
doing VIS- or MMX-like limited bytewise 8+-way SIMD parallelism, and 
eventually superscalar versions of same.  Such designs might issue 
between 500 (*) and 4000 (**) hand coded packed byte operations while 
that single 300 ns FPGA write/read is still in progress.


((At peak speeds like 400e6 clock/s * 4 issue/clock * 8 byte ops/issue, 
e.g. 12e9 byte ops/s, you have to agree that superscalar micros with 
bytewise SIMD are going to displace many FPGA applications on PC and 
workstation platforms.))


So as long as FPGAs are attached on relatively glacially slow I/O buses 
-- including 32-bit 33 MHz PCI -- it seems unlikely they will be of 
much use in general purpose PC processor acceleration.  Sure, for 
applications such as cryptography, image and signal processing, they 
might be a win (***), given a semi-autonomous problem which either fits 
in the FPGA and local storage, or which can employ DMA to stream data 
into or through the FPGA without much CPU intervention or management.

Of course, the PCI ASIC crowd has the same latency problems, but they 
donít share FCCM aspirations of accelerating general purpose computing, 
rather they focus on the same aforementioned special purpose 
applications.


Five times better latency and four times better bandwidth could be 
achieved if FPGA vendors invent a way to directly connect their parts 
to the Pentium Pro external bus, as a peer of the memory/bus 
controller.  A custom, dedicated Pentium Pro interface would probably 
be required, since FPGA configurable logic would be too slow and 
electrically incompatible.

This could be a good volume business, and not quite the moving target 
it might appear -- I expect the PPro external bus to be just as 
ubiquitous and as long lived as have been the 486 and Pentium buses.  
Someone could make a plug in card which sits in the PPro ZIF socket and 
which hosts a PPro and its FPGA(s).

Alternately, the FPGA coprocessor could be attached on the new advanced 
graphics memory port, or whatever it is to be called, that will be 
available in future Intel memory/PCI controller chipsets.

One might argue that Xilinx made a big mistake in not offering a 
version of the XC6200 with a dedicated 66 MHz Pentium external bus 
interface -- after all it is by far the most popular and most supported 
processor interface for the most lucrative general computing market.

If any vendor does pursue this idea, I would appreciate a couple of 
sample parts. :-)

--
(*) 500 op in 300 ns: forthcoming 200 MHz PPro with MMX: 60 clocks x 1 
8-byte MMX insn/clock

(**) 4000 op in 300 ns: hypothetical 400 MHz Alpha with each integer 
unit enhanced for bytewise SIMD: 120 clocks x 4 8-byte insns/clock

(***) "win": much cheaper/faster than simply adding a second processor
--

Acknowledgements: this posting is a spin-off of a discussion with Mark 
Shand, and the "way out speculative" question was suggested by Mike 
Butts.

Jan Gray
Redmond, WA
-- 
The opinions expressed in this message, and any opinions
I might express about FPGAs or the computer industry,
are my own personal views and, of course, do not reflect
the official views of Microsoft Corporation.
Also, I acknowledge the trademarks of the various products named.





Article: 3264
Subject: FS: Altera EPX780QC132-15
From: davilla@4pi.com (Scott D. Davilla)
Date: Tue, 7 May 1996 01:37:51 GMT
Links: << >>  << T >>  << A >>
For Sale: 72 Altera EPX780QC132-15 devices.
        In factory trays sealed in moisture barrier film, bag seal date
1/24/96, will sell 24 pieces, 48 pieces, or all 72.  Asking $40/piece or
best offer, terms are COD.  We are phasing our Nubus product into PCI which
uses a larger FPGA.

        Thanks,



-----------------------------------------------------------------------
Scott D. Davilla                              Phone: 919 489-1757 (tel)
4pi Analysis, Inc.                            Fax:   919 489-1487 (fax)
3500 Westgate Drive, Suite 403                email: davilla@4pi.com
Durham, North Carolina  27707-2534            web: http://www.4pi.com





Article: 3265
Subject: Re: so little posts about PCI :(
From: jcurtis@atl.mindspring.com (John W. Curtis)
Date: Tue, 07 May 1996 02:11:19 GMT
Links: << >>  << T >>  << A >>
It is a tough thing to implement.  Been there.

Xilinx has a lengthly info packet on the subject.  Quiet helpful.  you
can probably get it from your distributor.


tw38966@vub.ac.be (Rafiki Kim Hofmans) wrote:


>Anyone else busy with implementing PCI interfaces in FPGA's ?
>Or am I the only newsreader busy with it at the moment ? :(

>Regards,

>Kim 

>--


>==============================================================================

>			************************************
>			*	Hofmans Kim 		   *	
>  		       	*				   *
>			*	tw38966@vub.ac.be	   *
>			*	khofmans@info.vub.ac.be	   *
>			*                                  *
>			*	Brouwerijstraat 62         *
>			*	1630 Linkebeek             *
>			*	Belgium 		   *
>			*				   *
>			*	32-2-3771012		   *
>			*				   *
>			************************************





Article: 3266
Subject: Re: SILAGE
From: nachterg@imec.be (Lode Nachtergaele)
Date: Tue, 7 May 1996 06:48:40 GMT
Links: << >>  << T >>  << A >>
Hi,

you can find a public domain simulator called S2C in

ftp://ftp.imec.be/pub/vsdm/SW-distrib/s2c/s2c.4.13.tar.Z

In that distribution, documentation on the Silage language
is included.

Hope this helps,

Lode

In article <DqtI8L.DKI@tumlis.lis.e-technik.tu-muenchen.de>, markusr@tumlis.lis.e-technik.tu-muenchen.de (Markus Rettinger) writes:
|> Hi everybody,
|> 
|> I have a short question:
|> 
|> Does anybody know some good books regardig silage?
|> 
|> I use a power estimation tool which is a free program available at Berkeley.
|> It uses silage at the beginning of the optimization.
|> 
|> Thanks a lot in advance.
|> 
|> Best regards
|> 
|> Markus Rettinger
|> 
|> ---------------------------------------------------------------------
|> -   Lehrstuhl fuer Integrierte Schaltungen      |                   -
|> -   Techn. Universitaet Muenchen                |                   -
|> -   Dipl.-Ing Markus Rettinger                  |                   -
|> -   Arcisstr. 21                                |                   -
|> -   80290 Muenchen                              |                   -
|> ------------------------------------------------|                   -
|> -   Tel: +49 89 289  225 15                     | neue Telefonnr.   -
|> -   Fax: +49 89 289  283 23                     | seit 01.05.1996   -
|> ------------------------------------------------|                   -
|> -   E-Mail:                                     |                   -
|> -   M_Rettinger@lis.e-technik.tu-muenchen.de    |                   -
|> ---------------------------------------------------------------------
|> 
|> 

-- 
================================================================================
Lode Nachtergaele             
IMEC V.Z.W.                       
Kapeldreef 75                                    Touch the ground        
3001 Heverlee                                     while carving !
Belgium                                        
E-mail: nachterg@imec.be   
WEB   : http://www.imec.be/
Phone : +32 (0)16 28.15.12   
Fax   : +32 (0)16 28.15.15


Article: 3267
Subject: Re: Simple Xilinx board
From: wolf@aur.alcatel.com (William J. Wolf)
Date: 7 May 1996 12:39:07 GMT
Links: << >>  << T >>  << A >>
ian@PROBLEM_WITH_INEWS_DOMAIN_FILE (Ian McCrum STAFF) writes:
> we want to use a simple parallel cable, no expensive Xchecker. 

Xchecker comes with the software, so I don't follow. 

I think you get one per license, so unless people hoard them 
this should suffice.  How much is an extra Xchecker cable?


---
- Bill Wolf, Raleigh NC
- My opinions, NOT my employer's




Article: 3268
Subject: Re: so little posts about PCI :(
From: wen-king@myri.com (Wen-King Su)
Date: 7 May 1996 06:27:05 -0700
Links: << >>  << T >>  << A >>
tw38966@vub.ac.be (Rafiki Kim Hofmans) wrote:

>Anyone else busy with implementing PCI interfaces in FPGA's ?
>Or am I the only newsreader busy with it at the moment ? :(

I have just finished with a PCI master/slave interface in Lucent Orca.
It supports full bandwidth 33MHz arbitrary length burst in both directions
and as master or slave. 

It can be done, but you can't hope to do every operation using the minimum
number of clock cycles.  Careful pre-assignment of I/O signals to pads is
critical, but unfortunately there is no easy to learn how to place them
without first going over the process a few times yourself.  This is the
third circuit board I cut, and still there will be things I would like
to change should there be another board made. 

You may not be able to place the signals in a way that is optimal for
wiring out to the PCI edge connector.  You need to watch out for places
where you absolutely cannot pad with extra clock cycles to allow for
internal operation.  Architectural restrictions can make things impossible
if signals were not placed carefully.  If your design is meant only as a
stop gap measure before you go to gate array in production boards, you
can also cut a few corners as long as it do not affect the normal operation
of the bus.
Article: 3269
Subject: Re: MAX+plusII LPMs, Synthesis Options & AHDL Design Style
From: Alfred Fuchs <alfred.fuchs@banyan.siemens.co.at>
Date: Tue, 07 May 1996 12:15:55 -0500
Links: << >>  << T >>  << A >>
Andre Klindworth wrote:
> 
> ...
> This posting seeks feedback on the library of parametrized modules (LPM),
> AHDL design style and logic synthesis options within MAX+plusII version 6.01.
> ...
> 
> Within a larger design project, I had to design a 16-bit binary up-/down
> counter with synchronous and asynchronous reset for an Altera MAX-7000 CPLD.
> ...
> Since we just got our hands on MAX+plusII version 6.01, I decided to give
> a try to the new LPM-modules and instantiated an 8-bit LPM_COUNTER with the
> following source code:
> ...
> The results were disappointing: 17 LCs in a EPM7032 !
> > ...
> Next, I tried the following straightforward source code:
> ...
> > The results, as reported in the .rpt file, where quite good: 8 LCs as expected,
> but - what the heck? - 4 sharable expanders? What are those used for. So I took
> a closer look to the report file and found the following equation for bit 5:
> > ...
> For the higher order bits (bits 6 and 7), DFFEs and sharable expanders
> (for value0 & ... & value5  and  !value0 & ... & !value5) were used.
> ...
> So I tried FAST, NORMAL, WYSIWYG, switching of this and that advanced option,
> turning XOR synthesis on and off, but I didn't arrive at the optimal design.
> I finally typed in the logic equation for each bit and compiled with WYSIWYG -
> and got the optimum design at last. This hand-made counter is reported to run
> at 62 MHz and uses 8 LCs (16 LCs for the 16-bit version).
> 
> Finally I tried the old-style macrofunction 8COUNT and this also result
> in an 8 LCs mapping with no expanders.
> 

Hi Andre,

nice to meet people, who are examining the SW behind a beautiful front-end.
I can confirm your experience. The LPMs are not the best choice so far (except for multiply and add 
functions of course). I did not yet check Version 6.1, but expect improvements no earlier than in V. 
6.2.
It would be interesting to know, if the Altera people did the LPM-subdesigns themselves or 
outsourced them. I also wonder if anyone knows the approaches they take in synthesis, because their 
tool is so much faster than you would expect when you read research papers on that problems. 
Possibly they trade synthesis speed versus design speed (??). 

You did not mention parallel expanders; even if synthesis is not optimal, turning on this option 
should improve performance.
And be aware that a "used" shareable expander can also be a product term that could alternatively be 
used as a sh. exp (in MAX 7k and 9k), but is actually used to feed the OR. Real sh. exp. have an 
equation named X... = ... in the equations section of the report file.

With advanced synthesis options I typically turn OFF "subfactor extraction" and "reduce logic". 
Especially "reduce logic" is problematic: I found that it can increase design size, it can crash the 
logic synthesizer and in some earlier versions it simply led to wrong synthesis results in FLEX8k 
devices.

With LPMs there are unexpected advantages too: You can assign a clique to a LPM_DFF, whereas you 
cannot do it with a DFF connected to a bus (the "primitive array").

By the way: There is a PLL-macro provided by Altera (in the LPM-directory) that is inherently 
unreliable, because it uses combinatorial feedback (My customer proved that it is unreliable !...)
And you must be careful with a Gray-counter macro (I think it is named gray16.tdf) on their BBS, 
because it tends to give you a wrong MSB.


Did you get feedback from Altera to your questions?

Alfred


Article: 3270
Subject: Re: On FPGAs as PC coprocessors
From: Alfred Fuchs <alfred.fuchs@banyan.siemens.co.at>
Date: Tue, 07 May 1996 13:11:14 -0500
Links: << >>  << T >>  << A >>
Scott Kroeger wrote:
> 
> The FPGA belongs on the CPU die.  Once there it gets "free" use of much of the CPU's
> facilities (registers, cache, memory interface etc).  This should allow the FPGA to
> achieve performance improvements over the CPU for operations too wierd to be
> contemplated in the normal instruction set.  For example, think how easy it would be to
> rewire an ALU to perform CRC.
> 
> I expect (hope) to see FPGAs on low end embedded chips as well.  What else are we going
> to use all that silicon for!
> 
> Cheers,
> Scott

Three notes:

a. FPGA vendors could increase the credibility of the reconfigurable 
coprocessor concept, if they tried to speed up their own SW with such a 
board. It need not be fully application independent.
It is my impression that Xilinx intentionally delays the 6200 (targeted at 
such applications), because they do know, that there is no market for it.

b. There was a project funded by the EC where a FPGA (CAM-based from Oxford?) 
was included on a processor from SGS-Thomson (I believe). As far as I know 
Thomson stopped the work after delivering first silicon (wasn't it in the 
EETimes?). I also was told that TI investigated FPGA in a DSP. Did they 
publish their experiences?

c. The most promising way: Ask Intel to dug a third cave into the PPro 
package and specify an interface ... :-)



Alfred


Article: 3271
Subject: Please help with CRC hardware implementation - crcgen.zip (0/1)
From: peter@kksystems.com
Date: Tue, 07 May 1996 20:18:56 GMT
Links: << >>  << T >>  << A >>
Hi All,

I am working on an ASIC, and I need a hardware implementation of the
standard CCITT CRC algorithm: x^16+x^12+x^5+1 as used in floppy disk
controllers.

There are lots of software versions of this, mostly table-driven, but
none are in the form comprising a 16-bit shift reg, with some XOR
gates around it, which I know are all that is necessary. I have access
only to 1 bit of the data at any time, whereas all the s/ware
implementations load in at least 1 byte at a time.

Attached is a zipfile containing a postscript file of my present
circuit. (I hope no-one objects to this "binary" post; it is only 6k.)
This very nearly works, but only if the input data is all zeroes! So
there must be something wrong with the way I am feeding in the input
data. Note that my init of 0xe295 (instead of 0xffff) is deliberate.

Any help is appreciated. I have already done the usual web search, not
to mention lots of books.

Pete


Article: 3272
Subject: Please help with CRC hardware implementation - crcgen.zip (1/1)
From: peter@kksystems.com
Date: Tue, 07 May 1996 20:18:57 GMT
Links: << >>  << T >>  << A >>
begin 644 crcgen.zip
M4$L#!!0``@`(`$NHIR"#IE:(<AD``(-S```)````0U)#1T5.+E!3Q5U+C]S(
MD;X+T'_@'@3(6(EBOICD`#Z,-5K#F+$]8WD-+[P^M+I+K8*K']M=FI$L^+\O
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MKC/=P]WR2,`-_0#_F>6_L,A0DF!CEUJ>&:FM.MG=7LDG?[\\RON%?7]W>TQN
MTM]*XY^+RJO/<(?=[M/E!RK_5)?4;O*S`_TW4E\>IOM<R+]W-Q?_V)V^]O%+
M%>OD\OYNOY@4_+J'NV/Z\_CA[I<3`?CX87&5Y/<7UU+TK/MD7GPV+S[9%Y_M
MBT_NQ6?WXI-_\=DO5N]*W+)JR^]+?C7JOY>'N\<=Q/CE87_?%3M^.SMPK-[I
MP[M%U"W_H^LFDD5*8E_=7^UN'_?'SPQ]*-_7=0*E'W;[ZP_'NA0C1!9&VQ77
M5_N?N_1F4F@F,9K016X^'CKRS$J8]JO[_>7C\:%[UA'QX>YP]=C=W>Z2UBVT
M6ZBY^H[=Q=55-W7[=/5G5'CWOMO?7%SONOUM]^[S<??8_;J[W.T7^^OGH/%J
M^M7J:+G2(D"^/!@ZHMLUW=^03\%(9;>[Z\+\'6V^E`#>+P]'S_"PN[CZL/N$
M%X'6"NH.+>`NJ]K<Y;96K>S=X;YI>'*9TO#LOJSB_].&A[<RK&76FYC_G_:Z
M:;K>U=^4_&HN-[=H[3XM3_)`'\<!`SVS6-N9*R]L^VB;=\]N\*N_AO!5__'A
M[G!W_?3)TR<AQG[H;.BG[N;8(3>9L;?=X=B9J>%`\Y`U,S?T9NCN/SY]\NKU
MW<>'_>ZA>\_H9.<7Q??_?/KD^7>_^\.ONGGH9MN'SOC0N^Z80F!Q8`;7CYV9
M76^(MH/M8W<@^>Q3D(`8R"2UJW1:+F&=Z3V1UC/IHC_Z,`^A>QE[9T;KQU08
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M_"<3H&J\M;WOXI"J[)C*/=98G)9GE37F3,*/:F0:^E`CRLJ0KFH,Y5ACBKUS
MT/3$U'P0/:5F2)7#O2!]$/?'Y<6_G5*G;:TQ8G,5Q5E4(+&YQICM9HTMIL.\
M56/IV>U6C:E*SDZIW7`0SDB+H#4NIJ`%.=$RJ.%%$7*@#XI_KJO9$RW\NX@2
M=G_"WD]CN2;75>V1EL\G04HJ9@1]<Q))7:GVE&[6;WE2E?3W\;RG1NGY]]]W
M;__K[9_?_/YM]\.?O^O&5^'5/,*K&*SMIVYRO4UO8FEGK>OGSI1/R^L_O?[M
M&_RV!Y-:KBDN\OSJ^O3=7#RE,C\/R[](EWJ9$]8AM3>9MKX/3#XM+H,;4E.&
MM`U,'H:4008[];[0<:$/ZO7=XJ5<7_]P_/8/W^&S#`X,A^7?HPBBTP]CY]1.
M^RDY(WKV[&;LE-(>DB.=WK*JY0\VM39$F\#!@F0F!IN3F>`B)3-3&"F9"7%6
MDYGDT#H3QNB76XB]'^-L4\XMLIDQCI3-3%84GL]FDO-AFHSST^)\[L,T3]&[
MXCPU97,7!FRA;8O,'%/5H!SH(M](:-R0$DJT/)/0A"'F)"3$83NAL9#O36'T
MZ-S%,2RR.J%)OE972D(SV%,7F!:G<XQ#,/4%3F0TQK"4IGPW6>LX3@':R$1'
M*P(+Y1';P%57IBP^!FC7$ZW)N?]*CL',[)%6Y>PE.)QMR=(UQJV6;%7"!"?:
M%`\E.8HVI##*V5##)N7#JDSLUR1'H`R?VN<_46X]I?9%IBYF2LT`I@-(R]2$
MY)":<-V"G$DM&R6#0&MR[K^2>PMU`O9$B^\RE[-KG:B9=(EIWL@@&B4SC243
MO4D=RI'GJ0FVT+*4]7#;K9I)ICZP).CY3_C.&)>^&57-0!9"R`%=U0S*L6:8
M;I''E$@B6DAK<NZ_DL]^M2=:9!Q<SJYUJF:62PQQJV9J)>.F`G:J&6)S5;C0
MCRV;:X;9;M9,,G6B9@:HF2FF_I6HF&DN728@9;6@%&J%*190+*:&`)IM^V:\
M3#/W8^IKHSG1HDZXG%WJ1)VD*X2-*JEUYJ&/:X401Q4PQ:6TY@A^9K=5&<G0
M\KKX\4]O?OCCMYCW>&_2`(U-XRDE[XGPV`$;\40''QELO,S[5)>:G.Q!CG21
M0YDQ24N3FP`).929`1HI*6=E1(MJXSZ)EHT@YAUHSZZE7C^TG0,;3,KH0,Y]
M:?Z)KGK\B2,Y\Z7Z1UIV?GQ,S0/*F2_5/]+2?D;/Z5_N2_-/]$E\N2_5/]+R
MM8(ZPY@A>C1-?(CX$G*L$V$OX@\Q)SG35>V1%OX1<Y1S7YI_HJ4]8$YRYDOU
MC[2P1\Q)SGRI_I&6]H`YRKDOS3_1)_'GOE3_2&=[[]-7S)C0V]2>939UY`Y'
M:&QBRP8<>.:V6VU:,AW)]N3X1[HW.V,@FI-];56I]K0\JX];GE2E.M>=YWX,
M&YYTI=I3BARWY4E5JCVE&#9;GE2E.BM);].PX4E7JCQ!#&[=DZZD1P&$Z[`1
M!962'@4;GE0E/0K.>]*5]"C8\*0JZ5&PX4E5TJ/@O"==28^"\YX:I>>O?_@>
MDYL0(',;<4@+&[<0H96!Y`-H9T7RPLJP(:SDK(QH-[+!8R['QE':\S*BA;V0
M^_;^>!G1TI[)\>,@[7D9T<)>R.'C5-FS,J*E/9?#Q[&R9V5$2WLFQX^SM.=E
M^/&OY*R,:.%?V",MZX_7/]"CF7CRP\JPKBLY*R,:QJ<U.=:UM.=E6)>5G)5A
M74DY+\.ZJ.2L#+&NY*P,\9%R7H;X5G)61K1X?F&/M)#SZQ,MY?S^D99R_OQ(
M"SG'CV@IY_@C+>]?;<9`WVVT=962WOZ>]Z0KZ=^$#4^JDOZ=VO"D*NG?SO.>
M="7]>[[A2572OR[G/>E*>A1`J^$WHJ!2TJ/@O"==28^"#4^JDAX%&YY4)3T*
MSGO2E?0HV/"D*NE1<-Z3KE1["C&-1VYXJI5\2#/@:Y>)V-Q'"FFY1\/F+A.S
MW>PR)5/;=)G8Y"W1<>;#HFQ8C&@IQZ%.`TD5T@.?(^/C)T1G>S.FE6#(IL%(
M8,DZ#3>FD>J*)=NL7-CRX,]_^_O__"L-S[LT1(8J:<'!VS<_?(N"`+G'E-;9
MP$H$4D]7<@,-33[_#9:F6&2ZWS+=4LKS"L@EB/9C"Q3*D=;D?/PYTEBC"5-1
M3T"1]SEE/8GU,$XS5E`0GL4*6?19X*W8[(3;"G@%5G$@K+[CI7XBK%XSM%T8
M<);X^7]\QZL'+P&U,*T#Q'*V"P9N<0(*2#G7A5*8ZF**<@`X!$#=MC-9O$PS
MAX@G<Z3E2!27LTN='2`^/\]5Z\!`;YGE(HYFM6!(N.9H3HO9?=T`\3K!Q?,X
MR'V(=K'!'<1(*E)6*^.(L_C&I*E]5$Y1B9[]0%,.0U(C5SR6C2N72%;(HL_E
MP=.X4<5E%]Q2B628=1]'0XN2OF.%RSU9'L<#/(>W%-TYCF%V&_U#&,]L!DK&
M,;[5&&!(RT@F.8#&=>4,$P4;T)J<^Z_D,,-$]DC+8.9R=JWS,U#GH[E1PEFD
M$L^9I1#&*:>&I2#FME\Y`U4'->MP$"W"EF!#.=*:G%=1"6QHI-<01>\EL*&1
M5@(;&MK5"MD<RM@JUVP);69[LI&N8YM*Z^"&1KH-;FRD672;@4U]5^$-T]$4
M?D!7X8URQ([IRJEM##FD-3GW7\EA:IOLD9:C\ES.KG5^ZGLCO&LEG+Y>PYO8
M',\PU]VP.;R9[5=.?5?AS?O+1,OP1=@PO)'6Y+R*<GC;(13UM-H"O>?P3LN1
MC1**=DAK))D5L"6>IV%]@PJ;G7!;+;RA\IOPQM(JO"T$JA+><2K?BJ,<7T`,
MD188T7H86G-HR^MQ9NU-P3"XHI[00.\%0S^O%<(QA"AE5L!FT*R-:S45MF#(
M;!4,K7/PY9084FF-84CHM1A:.Y9&"9L(RQ;V5NN6IO4+A72U;@GEB-W4?N%P
ML2V^MDAK<NZ_DL-=DCW2<N*,R]FU#F?7UVZL6ZJ5<)GONFZ)V+Q0"=8$-VQ>
MM\1LOW*)<-U$L"$QHF7X(FPH1UJ3\RK*X>T&(P(5O9?PGN-Z,7;K;O`EMURL
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M7]._$=ZUDHNQ=$E2?1*;XSEZWC_/;`YO9KL9WLFTZ;_P"0FB9?C&M7]#M";G
M553">XZ\BTW>2WC/3FUI_3#P7@^Q)9ZCX8TYL=D)M]7"&[8>-N&-I75XSR-T
MV9OP'N>2P!WE!!Q@1+3`R-LUB29:DP.&1&<,O0^\-T?>,X9IS8HV%.%]Y*DU
ML1DT;V8^UD%LP=!'-6DD#-/]30V&5%IAF.YV5C#T)I9&Z2@G67$P$VDQF(EK
M+4&,I")E:SK+0"<,#JU#EN0Y#W3"X)`RT`E#/*L5LGED$X:#*JX,<S++4X-#
M]2@G%M:#G#`XU`YRPN!0&>/D\]N`#=%R'!A7=2.T2&MROH(\PX=]M`)$]D[P
M81^M?5CL9S$K8,O`\#2L0]&%S4ZX[<D^6C-.C*45A-A'4\:)H8]68<C6'1#M
MV]UC)$=:D_,MG05#Z*.M:)#WC"'TT10,,:=:K8#-H&&GK&8+ALSV9!^MQI!*
M:PRAC]9BB'VT"D.V]H-HB1%NG$0YTIJ<;[+,&&)'8$6#O&<,H2/0/B\F\\6*
MV`):#&LU%38[X;8G.P(-AEA:88@=`05#Z`BP^8K`^@'5_E7Z/N-"%OB.B#7M
M)`?HN*[,\W&I.=*:G/NOY-@!1GNBYZ:?P/V?7=9.*?[Y=>V-$N;R965[9FDQ
M.R;^#4O+V;GM5_8#JIVP?&44T;[=?TIRZBV,YZNH1#?T`]8X)>\4W=@/4*([
M6/YI(S:',R;^-5NB.UAUAD_T`^KHIM(ZNJ$?T$8W]@-8=(^L&U!%=USWN"-=
M13?*$;K8[H?'U!PC#FE-SOU7<FCZR)[HN>DF</_GHQLS_(WHKI4PE5^CF]@<
MSI#W-VR.;F;[E=V`.KK9ND&B?;N[F>2QG8M6JJA$-W0#UC@E[SFZH1O01B*F
M\L6*V!+.D/?7;';";4]V`YKHQM(ZNJ$;H$0W=`-8=$=V)(&,;KZDDV@!'6;\
M)+=M^D9E`"W1&5KL':P@D7>"%GL'"K0^\J2.V(PE=@=JMD#KHYJNB-Y!#2V5
M5M!B[Z"%%GL'*[2TMZC[=3=\>F/G0!E?FAOSGO4@(/''%Q=)L:H$3W8!*9)"
MBIT&6'*"I%QQPG8$$5U6G$"^OZXXP?0_+S'!OH'@UO4FS/+$>A/(];7E)M`W
MJ)>;8#>B6FTR0VM5+3;!XT'D6I,)#PN9RU18M2B'[2\D6LJQWX"+=I`6$/)-
M2T1G>\SX"X34`2BK=+![(-F"&K<]M6@'LGT%1>P>-(MVL"M1+]K!/4?UHAV:
M%N!`\J,EB!9`\2VT1$LY=AX`2*(ED&SW%M$%R+`NIEJ[#!DYZB-(=@62V9X`
M$E-^#4CH(]1`4G^B`M+B1J\*2-2M@)R&$G%$2Z#8P3=$2SGV(!!(I"60;!L;
MT=D><_\")'4%"G+849!L08[;G@(2\GX%2.PH-$!BIZ(&$DZ/:(`$70DD/VR%
M:`$43E&1W,7FU:=DU>`B]FE=0B?W]J$]T@7(L`;XFJ)FY"@GE>P*9!#+\S0@
M,<74@/23>%U%5EL!Z=S$7V("$G4K(-FI+D1+H%C'B&@IQ[P(@41:`LDV-A*=
M[3&C*4!2@E.0P_1'L@4Y;GL*2,AF-"#G4;RN(H&J@9Q&$;W?,ET))&8Q^*!$
M"Z!X#DZTE&,6!$`2+8#D.W2)+D#Z-<#7W"<C1\F.9%<@O6@R-2`Q=U&`Q&2G
M!I(2HPI([[RR0A1U-2!QSH+H.'`@V)P&T5*.@\V03I`])`S5MDVT1SK;(QC(
M%B#).B-7LVA;E#-[&DA448`,T>'Q<P+(,.>CP#B0JRX'LI2R5Q,?E&@!%)_[
M)%K*<>8#QVS17@#)][0278"$UW,%$MF,'+[+-5N`9+9G7VT%2'RU:R#Q=:V!
MQ%>[!I)T)9#P@:`'15H"Q=9($"WE.$.*0**]!))M+B8ZV^,'HP!);$$.OBXU
MFY'CMF<_-AJ0'A^S`M+B&U`!"1^;!DC4%4!BRH(/2K0`BDYW0OD\\%>3KZJ@
MG@+8"R#Y+F^B"Y"#+T%6TI^"'.8[-5N`9+9GTQ\-R&$04<;3GQI(3']J($E7
M`@E)-#THTA(HMK"/:"G'%5<()-I+(-EV>Z*S/2;5!4AB"W)6-)G$9N2X[=F$
M7`$2$_(&2/P`U$!2LUX!21\+#B1VZ_!!B19`\07`1$LY+O\#(,E>`LG./2"Z
M``G=O!5(9#-RV">LV0(DLSW;1=2`',2G0G01:R`-->L22-*50,(R7P(*:0D4
MVQ9`M)3C,F$$$NT%D/P`"J*S/2[P+4`26Y!SHLDD-B/';<]ND%&`Q-7`#9#X
M`:B!A`TR#9#TL>!`PE`-/N<T-]]L&.9!Z3SQUY)M)P`,I[GY7O,S1(@N$,*X
MS0HALADS&..IN`(@LSPWXJ/A-X@O!!_QJ>&#:]?HT0="M(OM48%B[4LS)";W
M/+>;M*2\'0^2>[+;80XI;WOO<D]ZVRF5\K:O)?=DMUT((:_#H]JQV[R!E;QI
MZJH=Q<TWI9(W'^]J1W&3)57R)AVM=HPW>7^1EZ,#H5O6>SPZ\-]QU#6DX>;T
MQ?#B/$]QRN!?7M/P[@Q3X33F.PZPACXUAC?'CG'I1/!T%GCAY.0%R@)N'0$U
MPR9N1N@D@R5W*;N94,:[G**.R7IU).L0RYBF8LP&GF4%HI@[TISSH5?MUKDO
MS3\?*#N_K!CT_<:RXDI)W\-PWI.NI.Y<.^^HUMD.SN`"'N/Y%<&)9\+FX`PN
M?79R<"*7@[-P2G`&GZ8J28T?L0XQA9;<I3A6%H\+QI-@J^.&9SQ.>$X-U)P/
M+9ZJ8VWQ_<JT8W(\RC;YC(46+X>XON/''5<GV%H[LY-1^3'(ZGWC,`(\*Q^>
MD$TGP;(BI)E31UK(>1GOM*MR[#]*>U;&^ZJ:G-^J=OL\CSN_HPWT_<:.MDI)
M7S)QWI.NI$]/;WA2E3!3@HX*R[UDH\;%MMW&@F5<53/G?0W9YJ&<^]+\\T[?
M03E[F?M2_;/>]_DV->E/<:--K93TPVHW/*E*>NM\WI.NY+SM(YU:3S3\7D9U
M@CW*D9YMLQ'&6EQ&![3QO+TY??+T^C,:-GKUY.DT;#*$P3HSPYXV:P:?SE4^
M_3L:0_CJW]&`NQW&.$X33"L[MQ3;6)P[W/4_8)(ZM,CXD%I$E`-=Y%LG3X<A
MY4A@69T\[>9Q_2D-.&,!3HNV8SA[\G2<T^$4WO5ACF:>PN(\G3(Q&QM\??)T
M\@6NW#Q;?EV\K)G4"TPAI=JVMV9TXSRJ%U!B060GT!WS8_IXL$Y:E9K#"AVD
MY6H@93U$M9IG7;^&-%Y*E9^T9[<GS>L3.BPLX#F_V*=1\K">ORSVR2RM[O&P
MGK]A:74/M]T\H2.9AN9''?#G<D8X')']=,Y!M@4CKFQ>VP5-SGT5!$<X"HO;
MGY0C+0_[IB[)M";N<EN(LEY>;NL`G]Q>D^,UN:XFIU/XY6'DK(R?V']J/5A:
M`#_#\_IP^HB:2@FKN6P+R2SM`\&8:%C:!\)MOS)$ZFWM:@"?W]FB*]6>$K9X
MNKL]^8G2E9K?:9*-"AR3[=-WF_W<TR*=`XQ1I%V/Z;@+X+S%)5*SA6,X!`>:
MAZR9N;,8)L,!=4V"\/$#_#;FP\7^L'M(/Y;UOU!+`0(4`!0``@`(`$NHIR"#
MIE:(<AD``(-S```)``````````$`(`````````!#4D-'14XN4%-02P4&````
/``$``0`W````F1D`````
`
end


Article: 3273
Subject: Is XC7336 the least expensive CPLD?
From: gf0570 <fang@signus.utah.edu>
Date: 7 May 1996 23:43:14 GMT
Links: << >>  << T >>  << A >>
Hi,
    Is XC7336 the cheapest CPLD one can find in the CPLD marcket?
What's the best price in quantity? Who should I order from?
    Thanks in advance for any information in this regard.

-GF-



Article: 3274
Subject: Implementation of a ROM
From: safiri@engn.uwindsor.ca (Safiri H.,13310,1100,g)
Date: Wed, 8 May 1996 06:25:40 GMT
Links: << >>  << T >>  << A >>

I am trying to implement a design consist of a 256*8 size ROM on a Xilinx FPGA. The rest 
of the circuit consists of a simple combinational and sequential logic.  I use Synopsys
FPGA compiler and the the design discription is in VHDL.  I have used both "Memgen"
program and the behavioural discription to generate the ROM.  The behavioural discription
almost always(with different contents for the ROM) gives a more optimum design in terms
of resources used.  Still when the content of the ROM is more random a large part of the resources
are taken by the ROM and in many cases PPR tool can not succesfully place the circuit
in one FPGA,(this very much depends on the content of the ROM).

Can anybody give me some other solutions for the design for the ROM. Again ROM is
256*8 bits, and I am using Xilinx 4000series(4003, 4005).




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