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 1400

Article: 1400
Subject: Re: Any one working on Cypress PLD's ?
From: pngai@mv.us.adobe.com (Phil Ngai)
Date: Thu, 15 Jun 1995 07:44:27 GMT
Links: << >>  << T >>  << A >>
In article <3rlhvl$1ge@marina.cinenet.net> kirani@cinenet.net (kayvon irani) writes:
>	I am wondering if any one out there has tried out the Cypress's
>	FLASH series ? They're supposed to perform same as or better than
>	Altera parts but supposedly, they are much easier to re-route once
>	the pins are fixed. BTW, any comment on AMD suit against them ?

Altera has a broad selection of parts so it's hard to tell what you
mean by Altera parts.

Their 7000 family has a deep secret which is never discussed in the
data book: the interconnect matrix is quite sparse, the price they pay
for their speed. That is probably the source of the re-routing concern
you mention.

Their so-called Flash Logic parts (so-called because only one member of
the family is actually built with Flash technology and it's not really
available right now) is what they euphemistically call "feature-rich".
This translates as EXPENSIVE. Even if money and availability is not a
concern, the fact is that a universal interconnect matrix is not the
final answer to re-routing problems. When you have product-term sharing
between adjacent macrocells like in the Altera Flash Logic and the
Cypress 370 series, a change in product term usage COULD result in a
locally starved product term shortage preventing fitting.

(note: I think the 370 macrocells can only borrow from neighboring
macrocells, I haven't convinced myself for sure of this from their
literature yet)

The best solution is that offered by Lattice and AMD: input/output
switch matrixes. IOSMs allow you to avoid the problem of neighboring
macrocells competing for product terms. There is a speed penalty but
for many applications the speed is adequate and the flexibility is
wonderful. (I think Lattice could use more internal routing, based on
the studies I have done, but the Mach 445 will swallow almost anything
reasonable.)


-- 
 Should the 1st amendment apply to the Internet? Ask Senator Feinstein.



Article: 1401
Subject: Understanding Lattice equations
From: eddie@merlion.singnet.com.sg (q)
Date: 15 Jun 1995 10:21:59 GMT
Links: << >>  << T >>  << A >>
I am going through some LATTICE eqns., and trying to understand the 
logic, without having the Lattice manual. Could somebody using Lattice 
explain to me the foll. queries :

I only have the information about the foll. macrofunctions from the 
Lattice System Macro Library. The macrofunctions are 2/4/8 bit down 
counters CBD42, CBD44, CBD44_1, CBD44_2, CBD48, CBD48_1, CBD48_2, 
CBD48_3, CBD48_4.

I could not understand the 'underscore' deviations of the respective 
counters. Moreover for CBD48_4(CAO,[Q0..Q7],CAI,EN), there is no clock. 
How can this be a down-counter ?

The eqns. that I have use two successive lines as below :

CBD48_2(pc4,pc5,[pc0..pc3],pd4,pd5,1,clk27,0,pcao,1,0);
CBD48_3(pc6,pc7,[pc0..pc5],pd6,pd7,1,clk27,0,pcao,1,0);

How can the output signals [pc0..pc3] be driven twice ? 


Thanks & Regards,


Article: 1402
Subject: Unified Library for Xilinx
From: COMTECH ELECTRONICS <comtech@freenet.ufl.edu>
Date: Thu, 15 Jun 1995 09:53:12 -0400
Links: << >>  << T >>  << A >>
Has anyone built the Unified Library macros for Xilinx 5.0 or greater 
using FutureNet, and you be willing to share them? Or, would anyone want 
to be part of a team to construct these macros? 

Regards,

Bob Bachus
904-373-6785 (voice)
comtech@freenet.ufl.edu



Article: 1403
Subject: Need Help with Altera FLEX programing.
From: Joerg Wittenberger <joerg.wittenberger@inf.tu-dresden.de>
Date: 15 Jun 1995 20:15:03 +0200
Links: << >>  << T >>  << A >>
Hi all,

this will be a longer article about some experiences I got from
programing Altera FLEX logic using Max-II software and in the second
part some problems I can't solve at all.

What I have is a RIPP 10 board from Altera with a FLEX 8452
controlling the ISA bus interface and two FLEX 81188. Additionaly
there are 512Kx8 RAM (accessiable from the 81188's).

For programing I have the Max-II 5.2.

I should eventually mention, that I'm not a FPGA guru. Instead I'm
coming from computer science making my way educating myself... Since a
couple of weeks I try to create a small module without much luck.

*				- I -

** When I compiled and simulated the first designs I created, I
figured that the "global logic synthesis style"s I can choose don't
behave as their names suggest. There is "normal", "fast" and
"wysiwyg". But it turned out, that fast seems to be only a little bit
faster than wysiwyg and normal seems to make the best speed. But both,
fast and normal do some things I can't understand. Therfore the result
has lots of glitches. Especially when I use state machines.

Is there any explanation for this?

** With the former version of the software (Max-II 5.1) I tried to
figure why some designs didn't work. Finaly I installed the update and
those designs happend to work. At least something about state machines
is affected, they don't enter unused states whithout any visible
reason anymore. Are there other problems known like this? Are there
experiences which constructs to use and which one should avoid?

** There seems to be no real correlation between the complexity of a
equation and the time nessesary to compute it. Are the any rules what
I can expect?

Well, obviewsly the fact whether the result is supposed to drive a pin
or is only used by the next equation is important. But also these
timings vary about as much their average value is. Normal?

** I would guess, that the following two AHDL excerpts (A, B) should
compile into the same code. But the simulation I get from is sometimes
quite different (in terms of the timing, not the logic).

  A:

  IF ena THEN
    bus[N..0] = reg[N..0];
  ELSE
    bus[N..0] = GND;
  ENd IF;

  B:

  bus[N..0] = reg[N..0] AND ena;

Does someone has an idea what will happen? (Unfortunatly I'm not
really able to reproduce those results, this seems to happen by
chance.)

** Is there any rule how to write state machines? Should I better use
Moore or Mealy automata? While the notation of those machines is
really convinient I never know what they'll do. See below.


*				- II -

The designs I'm fighting at the moment had a state machine like this:

  sm		: MACHINE WITH STATES (res, wal, wah, wvl, rvl);

and later a logic table like this:

  TABLE
    sm,	 /iow, /ior => sm, (dd, dga, dgd, glbl_drv);
    res, VCC,  VCC  => res, B"0000"; -- What should we do? Wait.
    res, VCC,  GND  => wal, B"0000";
    res, GND,  VCC  => wal, B"0000";
    wal, VCC,  GND  => rvl, B"0101";
    wal, GND,  VCC  => wvl, B"0101";
    rvl, VCC,  GND  => rvl, B"1001"; -- Wait until host is away
    wvl, GND,  VCC  => wvl, B"0011"; -- dto.
  END TABLE;

The glbl_drv signal is immediatly connected to a pin, while the other
signals driven from this table (dd, dga, dgd) are used in other
equations.

This was almost fine except that I realized, that the time of one
clock cycle the dgd signal is active is too short. Therefore I changed
the above declaration like this:

  sm		: MACHINE WITH STATES (res, wal0. wal, wah, wvl, rvl);

and:

  TABLE
    sm,	 /iow, /ior => sm, (dd, dga, dgd, glbl_drv);
    res, VCC,  VCC  => res, B"0000"; -- What should we do? Wait.
    res, VCC,  GND  => wal0, B"0000";
    res, GND,  VCC  => wal0, B"0000";
    wal0,  x,  x    => wal, B"0101";
    wal, VCC,  GND  => rvl, B"0101";
    wal, GND,  VCC  => wvl, B"0101";
    rvl, VCC,  GND  => rvl, B"1001"; -- Wait until host is away
    wvl, GND,  VCC  => wvl, B"0011"; -- dto.
  END TABLE;

I expected the glbl_drv signal to be active from about 8-12ns or so
after sm enters the wal0 state until reset. (As it became active about
the same time after entering the wal state in the former design.)

Unfortunatly it takes about 25ns until active now. Also there is a
"glitch" of about 8.5ns one clock cycle after it became active for the
first time. Moreover the signals derived from dga and dgd are no
longer stable.


Is there any half the way general rule I am missing? As long as each
and every small change to the design requires a almost complete
rewrite I guess there must be something wrong with the way I go.


Any help appreciated!!!

Thanks for listening

/Jerry
--

-----------------------------------------------------------------------------
Joerg Wittenberger | email: joerg.wittenberger@inf.tu-dresden.de
Rietzstr. 32b      | 
01139 Dresden      | http://www.inf.tu-dresden.de/~jw6
Germany            | PGP: D4 B2 DA AE C3 02 50 9C 45 3E AD 99 C1 1A 8E F8




Article: 1404
Subject: Re: Need Help with Altera FLEX programing.
From: trev@ss11.wg.icl.co.uk (Trevor Hall)
Date: Fri, 16 Jun 1995 07:00:10 GMT
Links: << >>  << T >>  << A >>
Joerg Wittenberger <joerg.wittenberger@inf.tu-dresden.de> writes :-

>** I would guess, that the following two AHDL excerpts (A, B) should
>compile into the same code. But the simulation I get from is sometimes
>quite different (in terms of the timing, not the logic).
>
> A:
>
>  IF ena THEN
>   bus[N..0] = reg[N..0];
>  ELSE
>    bus[N..0] = GND;
>  ENd IF;
>
>  B:
>
>  bus[N..0] = reg[N..0] AND ena;


Yes it will compile into the same code. The timing however is dependant on the routing
resources used, although in the FLEX I would only expect small (+/- 1ns) deviations
with the above.
Are you using global clocking ? Product term clocking of the above registers may 
account for changes in timing. 


Stuff about state machines deleted.

Observation 1)
Are you using a clocked state machine, i.e.
sm		: MACHINE WITH STATES (res, wal, wah, wvl, rvl);

sm.clk = clockinput;

An unclocked state machine may explain your glitches.

Observation 2)
You may wish to make your glbl_drv output synchronous. This will stop any state 
machine decode glitches going to the 'outside world'.

Observation 3)
Again, have you told the compiler to use Global Clock, rather than product term
clocking.

Cheers,
T.H.



Article: 1405
Subject: Orbit Semiconductor
From: cks@ccd.harris.com (Chris Shipman)
Date: 16 Jun 1995 10:49:09 -0400
Links: << >>  << T >>  << A >>
Has anyone used the Orbit Semiconductor FPGA netlist conversion
service?  They call their process "Encore".

Chris



Article: 1406
Subject: summary: 80x51 in FPGA anyone?
From: granville@decus.org.nz
Date: 17 Jun 95 08:32:23 +1200
Links: << >>  << T >>  << A >>
80x51 in FPGA ?
-  results 
I have lots of email saying 'let me know if you find something'
and some asking about the 6502 quoted, but alas, no 51 core yet.. :-)
( might be just too tough still .. )

The design ex of 6502 was from xilinx, pge 11-18 of their 'breakthru 95'
road show manual.
 uses x8108 new device ( too big / complex for older parts  :-)?
 vhdl source
 fits 96% OF 8106
  FILLS 81% OF a1 15k gate cpld ( compeditor ?)
  fills 197% of 10k a1 fpga ( compeditor ?)
 4000 lsi logic gates
 no speed spec mentioned...

If anyone hits a 80x51, let me know...
jim



Article: 1407
Subject: test
From: chibane@alpha.fdu.edu (Cherif Chibane)
Date: Sat, 17 Jun 1995 02:19:35 GMT
Links: << >>  << T >>  << A >>
Plesae ignore. Test



Article: 1408
Subject: VHDL vs. Verilog happened at SNUG not IVC
From: jcooley@world.std.com (John Cooley)
Date: Sat, 17 Jun 1995 15:37:35 GMT
Links: << >>  << T >>  << A >>
rose@src.honeywell.com (Fred Rose) writes:
> One data point which may help shed some light on the mind set of the two
> HDL conference organizers, and why they may not see eye to eye.  The
> contest in question at the Verilog conference was the design of a counter.
> The design contest at a recent VHDL conference was the Denver Airport
> baggage handling system.

Kirk Anderson <kirka@ctron.com> wrote:
>AMEN !

Kirk, again I'd like to clarify, this design contest wasn't held at *any*
Verilog conference -- it was held at the Synopsys Users Group Meeting (which
is more of a synthesis oriented conference.)  As I told Fred, if you wish to
catagorize the Synopsys meeting from historical company tastes, this meeting
would perhaps be seen as more VHDL than Verilog biased from a Synopsys, Inc.
point of view because they only sell a VHDL simulator.  How the customers
see it is another story.  

Also, I'd like to stress that this design contest was not crafted to be a
referendum on whether VHDL or Verilog is a better language for serious
hardware designers to use.  The fact that 89% of the Verilog designers
completed while 100% of the VHDL design *didn't* complete it was a surprize
to an awful lot of us!

I know that this may seem like I'm nitpicking; but I just want to try to keep
the facts in this contest straight in the public's mind when this contest is
being discussed.
                           - 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 3443 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: 1409
Subject: MACH110 Uk distributer ?
From: dp11@unix.york.ac.uk (D Plunkett)
Date: 18 Jun 1995 14:55:31 GMT
Links: << >>  << T >>  << A >>
Hello

Does anyone know the email address or phone number for the uk distributer of
the MACH110 series? 

Does anyone know anyone cheap (free) development software?

Thanks

Dominic



Article: 1410
Subject: Re: Understanding Lattice equations
From: Gush Bhumbra <GUSH@mgnelect.co.uk>
Date: 18 Jun 1995 23:33:07 +0100
Links: << >>  << T >>  << A >>
In article <3rp1k7$t4a@lantana.singnet.com.sg>
           eddie@merlion.singnet.com.sg "q" writes:

> I am going through some LATTICE eqns., and trying to understand the 
> logic, without having the Lattice manual. Could somebody using Lattice 
> explain to me the foll. queries :
> 
> I only have the information about the foll. macrofunctions from the 
> Lattice System Macro Library. The macrofunctions are 2/4/8 bit down 
> counters CBD42, CBD44, CBD44_1, CBD44_2, CBD48, CBD48_1, CBD48_2, 
> CBD48_3, CBD48_4.
> 
> I could not understand the 'underscore' deviations of the respective 
> counters. Moreover for CBD48_4(CAO,[Q0..Q7],CAI,EN), there is no clock. 
> How can this be a down-counter ?

The underscore macrofunctions are called sub-macros and are what would
appear in each General Logic Block (GLB) if you placed them manually.
If you place the parent named macro into one GLB then you need not worry
about the underscore ones as they are placed automatically (certainly in
the pDS software). The submacros contain the actual equations for
a particular GLB.

The CBD48_4(..) only generates the combinatorial CAO carry out and hence
needs no clock.

> 
> The eqns. that I have use two successive lines as below :
> 
> CBD48_2(pc4,pc5,[pc0..pc3],pd4,pd5,1,clk27,0,pcao,1,0);
> CBD48_3(pc6,pc7,[pc0..pc5],pd6,pd7,1,clk27,0,pcao,1,0);
> 
> How can the output signals [pc0..pc3] be driven twice ? 
> 

The grouping [pc0..pc5] does not imply that all are outputs of that GLB.
Each 1000 series Lattice GLB only has four normal outputs and so
couldn't possibly generate six registered outputs. The first four
[pc0..pc3] are generated in the CBD48_1 sub-macro. In this case the group
[pc0..pcn] comprises inputs to the GLB.

> 
> Thanks & Regards,
> 

Anytime.

--
/-----------------------------------------------------------------------\
| Gush Bhumbra                          |           Gush@mgnelect.co.uk |
| MGN Electronics, 23 Garendon Way,     |       Tel:   +44 116 232 2472 |
| Groby, Leicestershire, LE6 0YR, UK.   |       Fax:   +44 116 232 2535 |
|***********************************************************************|
|    When I was young I wanted to be a farmer .....                     |
|               .. so I could be outstanding in my own field.           |
\-----------------------------------------------------------------------/


Article: 1411
Subject: Re: MACH110 Uk distributer ?
From: Leon Heller <Leon@lfheller.demon.co.uk>
Date: 18 Jun 1995 23:44:03 +0100
Links: << >>  << T >>  << A >>
In article <3s1ep3$4qb@mailer.york.ac.uk>
           dp11@unix.york.ac.uk "D Plunkett" writes:

> Does anyone know the email address or phone number for the uk distributer of
> the MACH110 series? 
> 
> Does anyone know anyone cheap (free) development software?

Kudos Thame Ltd. has an advert. in the current Electronics Weekly for
the MACH devices. Their number is (01734) 351010. They have a BBS on
(01734) 352942.

Leon
-- 
Leon Heller, G1HSM                | "Do not adjust your mind, there is
E-mail leon@lfheller.demon.co.uk  |  a fault in reality": on a wall
Phone: +44 (0)1734 266679         |  many years ago in Oxford.


Article: 1412
Subject: Who was the winner on latest PREP benchmarks?
From: kirani@cinenet.net (kayvon irani)
Date: 19 Jun 1995 03:51:11 GMT
Links: << >>  << T >>  << A >>
	Hi every one!

	Any one out there has regular subscription to PREP newsletter? If so,

	for the benefit of all of us could some one let us know which FPGA 

	and CPLD fared the best in terms of capacity and speed?

 	I remember last time, Altera CPLDs came slighty ahead of Cypress's 

	CPLDs and on FPGA catagory Quick-logic fared the best ahead of Actel 

	and others. BTW, do you guys think that PREP has become too political?

	Some FPGA vendors think so.

From:	Kayvon Irani
	Lear Astronics Corp.
	3400 Airport Ave.
	Santa Monica, Ca 90405
	(310)915-6000 Ext. 3696
	(310)915-8369 Fax



Article: 1413
Subject: altera mail adress ?
From: tw38966@vub.ac.be (SH.RYU KIM HOFMANS)
Date: 19 Jun 1995 11:56:30 GMT
Links: << >>  << T >>  << A >>

Can someone tell me if Altera has a technical support email adress ?

Thanx in advance !!!



Article: 1414
Subject: Re: Any company for conversion FPGA to ASIC?
From: peter.sels@student.kuleuven.ac.be (Pete)
Date: 19 Jun 1995 14:50:49 GMT
Links: << >>  << T >>  << A >>
There is one service in Leuven (Belgium(Europe)) called EASICS that
does these conversions.
If you are interested I can give you their address.

Peter Sels


Article: 1415
Subject: Re: altera mail adress ?
From: q (Pete)
Date: 19 Jun 1995 15:05:21 GMT
Links: << >>  << T >>  << A >>
In article <3s3olf$62r@rc1.vub.ac.be>, tw38966@vub.ac.be (SH.RYU KIM HOFMANS) says:
>
>
>Can someone tell me if Altera has a technical support email adress ?
>
>Thanx in advance !!!
>
try at http://www.altera.com/
you 'll find it if it exists...





Peter Sels


Article: 1416
Subject: VHDL synthesis in ViewLogic.
From: cheng@news.cs.columbia.edu (Fu-Chiung Cheng)
Date: 19 Jun 1995 11:18:57 -0400
Links: << >>  << T >>  << A >>

Hi,

	I am using Viewlogic to synthesize very simple circuits using VHDL.
I managed to analyze the vhld programs( by using VHDL vhdl.vhd) and do the 
simulation (by using viewsim). Howerer when I used the technology mapping
command it give me core dump with different reasons.

Case I:
======================================================
$ vhdldes

VHDLDes - V2.2; Workview 4.1.3 062292, 6000 Series
c Copyright 1985,1992 by Viewlogic Systems, Inc.
1: VHDLDes> tech x4000
 -- reading library file /proj/iso/Powerview/standard/x4000.sml
Processing Viewlogic X4000 Library, Version - 2/22/93
Bus error(coredump)
$
======================================================

Case II:
======================================================
$ vhdldes

VHDLDes - V2.2; Workview 4.1.3 062292, 6000 Series
c Copyright 1985,1992 by Viewlogic Systems, Inc.
1: VHDLDes> tech lsi10k
 -- reading library file /proj/iso/Powerview/standard/lsi10k.sml
Processing Viewlogic LSI10K Library, Version - 3/3/93
Memory fault(coredump)
$
======================================================

	lsi10k.sml and x4000.sml files are in /proj/iso/Powerview/standard/.

	Any help or suggections will be appreciated?


						-John


Article: 1417
Subject: ASIC TEST, etc. courses from UC Berkeley
From: course@garnet.berkeley.edu ()
Date: 19 Jun 1995 21:20:05 GMT
Links: << >>  << T >>  << A >>
UC BERKELEY EXTENSION ANNOUNCES

4 Summer Short Courses at the San Francisco Airport

1.  "TESTING ASICS, BGAS, KNOWN GOOD DIE (KGD) AND
    MULTICHIP MODULES"  August 2-4, 1995  (2.1 ceu)

    Topics covered: dynamic simulation at CAE, scan testing       
    testing laminates, environmental stress screening (ESS)

    Instructor:  Robert Hanson, M.S.E.E., AmeriCom Services,
    a test and manufacturing consulting company.  Mr. Hanson
    has extensive experience designing test hardware and 
    operation/test software.

2.  "SURFACE MOUNT ASSEMBLY AND FINE PITCH"  August 8-9, 1995
    (1.4 ceu)

    Topics covered:  introduction to SMT/FPT, SMT/FPT 
    components, SMT substrates, types of SMT/FPT assemblies,
    design for manufacturability, SMT process details, 
    typical defects and inspection, rework/repair, starting
    an SMT operation.

    Instructor:  Charles Hutchins, Ph.D., an independent
    consultant recognized worldwide for his experience in SMT.
    He has been President of the Surface Mount Technology 
    Association, and is the author of 30 technical papers
    and the textbook "Understanding and Using Surface Mount
    and Fine Pitch Technology."

3.  "BALL GRID ARRAY (BGA)/FLIP CHIP AND CHIP ON BOARD (COB)
    TECHNOLOGIES"  August 10-11, 1995 (1.4 ceu)

    Topics covered:  background, package types, properties
    and characteristics, chip attachment and interconnection,
    interconnection materials, printed wiring board design
    and specification, second level assembly, process control,
    reliability, future technology directions.

    Instructor:  Charles E. Bauer, Ph.D., Managing Director
    of TechLead Corporation, an engineering and management
    services company.  He has more than 17 years experience
    in electronics packaging, interconnection and assembly
    from printed wiring boards, ceramic hybrids and IC 
    metallization to multichip modules, micropackaging,
    smart cards and most recently PCMCIA design and assembly.
   
4.  "MULTICHIP MODULE (MCM) DESIGN"  August 14-16 (2.1 ceu)

    Topics covered;  introduction, materials, resistor design,
    thick film, thin film, MCM technology, CAD, thermal 
    management, assembly processes, screening techniques.

    Instructor,  Al Krum, M.S.E.E., a manager at Hughes
    Aircraft where he has more than 20 years experience
    in design, test and manufacturing of microelectronic
    packaging, including hybrids and multichip modules.
    He is the author of numerous papers in the field, and
    holds 2 patents.

For a brochure describing these courses in detail please
contact us as follows:

e-mail to:     course@garnet.berkeley.edu
fax to:        510-643-8683 (att:  Engineering)
write to:      Continuing Education in Engineering
               UC Berkeley Extension
               2223 Fulton St.
               Berkeley, CA 94720

please specify "microelectronic packaging and test courses"




Article: 1418
Subject: Re: Low cost ISA board
From: jhseng@xmission.com (Jeffrey L. Hutchings)
Date: Tue, 20 Jun 95 03:19:35 GMT
Links: << >>  << T >>  << A >>
In article <D9zCMG.ILx@undergrad.math.uwaterloo.ca>,
   pdgray@undergrad.math.uwaterloo.ca (Peter D. Gray) wrote:
>In article <1995Jun9.034024.4769@super.org>,
>Steve Casselman <sc@vcc.com> wrote:
>>We are designing a low cost ISA board for reconfigurable
>>computing/prototype development and would like to have  
>>everyones thoughts on the subject like:
>>
>>Cost, functionality, programmablity, prototype area,
>>mezzanine busses, external connectors, development software,
>>driver software and type of projects you might want to do
>>with such a card.
>>
>>This will help us design a product more in tune with what
>>you all might need.

Hmmm...  Since the definition of the board is somewhat nebulous, I can't say 
defintively what features it should have.  However, I can volunteer some of 
mmy own experience in designing such boards and the applications that run on 
them.

First, I feel it would be really nice to have a very low-cost board out there. 
 Just enough for people to whet there appetite on the technology.  Something 
similar to TI's $99.00 TMS320C5x development kit.  After all, the more 
individuals we can entice into using reconfigurable hardware - the better.  I 
also realize this may not be easy, FPGAs are NOT cheap.

Second, the ISA bus isn't a bad choice if your target market is low-end 
developers on PCs.  In fact, it is an excellent choice for the hobbiest 
market.  A market I feel has been much ignored.  Obviously it will limit the 
boards capabilites, but for an introductory platform, that's probablt not a 
problem.  I would reccommend against the PCI bus if this is an introductory 
low-cost board.  The PCI bus can add a lot of complexity that would be 
unecessary at this level.  My guess is that someone's first venture into 
reconfigurable hardware is not going to be video-rate computing.

Third, functionality should be constrained enough that the novice won't be 
overwhelmed.  We all know that just learning FPGA design is hard enough.  Give 
them a basic platform that will allow them to learn and experiment cheaply and 
efficiently.

Fourth, if by programmability you mean how many reconfigurable resources there 
should be, I'd say one.  Simply put one FPGA on the board with some memory and 
I/O.  This way the user can develop simple uP and controllers that use the ISA 
interface or an exteral interface without contending with multiple FPGAs on 
the board (bus contention, fires, smoke, etc.)

A proptotype area would be very useful.  In fact, I would like to see a great 
deal of space available for such things.  For example, one could place a CODEC 
on the board with some basic analog circuitry and have a simple audio 
development platform.  Or, the more courageous user could place a DSP or other 
coprocesor on the board.  Of course, some I/O from the FPGA should be 
dedicated for this purpose.

In terms of external connectors, it would be nice if I could hook a logic 
analyzer up easily to any busses I have on the board.  In addition, expansion 
connectors for memory or whatever would be very nice.  If you provice a good 
prototype area, the user can add his own connectors as well.

Develpment software is very important.  We all know that one of the biggest 
problems facing the Reconfigurable Hardware area is a lack of tools.  At least 
provide routines the user can use to read/write the board.  Reconfigure the 
FPGA, end perhaps even a sample application could be provided.

As far as projects go.  Individuals with backgrounds in other areas will no 
doubt want to compare reconfigurable solutions to the solutions they have used 
in the past.  For example, a DSP engineer would like to be able to develop 
filters in the FPGA and compare them to DSPs.  Audio projects, control, data 
acquisition and stored program uProcessors are all nice introductory projects 
that could be handled by such a board.

My goal for such a board would be to help expose more people to reconfigurable 
hardware.  Unfortunately, for this to work well, the board will have to be 
cheap, easy to use, simple, and come with some good documentation and support. 
 This will be very difficult to do and maintainn profitability.

Regards,

Jeff Hutchings


Article: 1419
Subject: Re: Low cost CPLD/FPGA tools
From: Uwe Kremmin <100114.2166@CompuServe.COM>
Date: 20 Jun 1995 11:40:18 GMT
Links: << >>  << T >>  << A >>
In case you have access to CompuServe and would like to use 
AMD's MACHs:

Go MAGNA,  section "Design & Elektronik". There you find MACHXL 
2.0 including docu and download softare for Windows 
for FREEEEEE!

Uwe Kremmin
AMD Munich
uwe.kremmin@amd.com


Article: 1420
Subject: Help with Viewlogic
From: an222663@anon.penet.fi
Date: Tue, 20 Jun 1995 13:48:15 UTC
Links: << >>  << T >>  << A >>
Hello, 
      I want to simulate the whole circuite I has designed using Orcad or
      Viewlogic. I have common chips with Xilinx devices. I would like to
      do a "timing" simulation, but I'm not sure if this is possible.

	       Could anybody help me?

				Thanks in advance.

					     Q. 
----------------------------------------------------------------------------
To find out more about the anon service, send mail to help@anon.penet.fi.
If you reply to this message, your message WILL be *automatically* anonymized
and you are allocated an anon id. Read the help file to prevent this.
Please report any problems, inappropriate use etc. to admin@anon.penet.fi.


Article: 1421
Subject: Find One Paper of D.F.Wong
From: eea80593@maddux.EE.NCTU.edu.tw (Chuang Hsien-Ho)
Date: 20 Jun 1995 17:09:24 GMT
Links: << >>  << T >>  << A >>
Hi:

I want to find this paper:
"Circuit Clustering for Delay Minimization Under Area and Pin Constraints"
        H. Yang and D.F. Wong
EDTC '95(European Design & Test Conference)

I can't find this proceeding in local library. Anyone or the author could
send me the ps file or text file or anything helpful?
Thanks!

--
===============================
Hsien-Ho Chuang 
eea80593@yankees.ee.nctu.edu.tw
===============================


Article: 1422
Subject: Re: Low cost ISA board
From: wirthlim@fpga.ee.byu.edu (Michael J. Wirthlin)
Date: 20 Jun 1995 12:48:07 -0600
Links: << >>  << T >>  << A >>

|> >In article <1995Jun9.034024.4769@super.org>,
|> >Steve Casselman <sc@vcc.com> wrote:
|> >>We are designing a low cost ISA board for reconfigurable
|> >>computing/prototype development and would like to have  
|> >>everyones thoughts on the subject like:
|> >>

|> 
|> First, I feel it would be really nice to have a very low-cost board out there. 
|>  Just enough for people to whet there appetite on the technology.  Something 
|> similar to TI's $99.00 TMS320C5x development kit.  After all, the more 
|> individuals we can entice into using reconfigurable hardware - the better.  I 
|> also realize this may not be easy, FPGAs are NOT cheap.

It seems to me that there are plenty of boards out there already with one or
only a couple FPGAs. As has been said earlier in this newsgroup, what we really
need is low cost tools. There has been some discussion on the one-FPGA board
made by XESS corporation - for a low fee (I don't remember, but it is under
$150), you get a board with an FPGA *and* free access to some tools. The tools
may not provide you with all the fancy features of synthesis, partitioning,
etc., but they are enough to get a design running. What makes such a board so
attractive is not the board itself, but the tools that come at no additional
cost.

Boards are a dime-a-dozen, but decent and *low-cost* development tools are 
an "arm and a leg and then some". I think these discussions should follow the
theme "what types of features should we put in a low-cost tool". Provide almost
any kind of board with good and inexpensive tools, not the fanciest low-cost
board with tools costing over $1k.

On a slightly different subject, how many people would be interested in a >$200
FPGA based board over some of the commercially available DSP boards? If I had a
choice between a single FPGA (small) based board with no software and a DSP board
($89) with a stereo codec, memory, compiler, windows interface, etc., there 
is no question - I would take the DSP board. Am I alone?

- Mike
-- 
Michael J. Wirthlin
Brigham Young University - Electrical Engineering Department
Reconfigurable Logic Laboratory (801) 378-7206


Article: 1423
Subject: Re: Understanding Lattice equations
From: iisakkil@gamma.hut.fi (Mika Iisakkila)
Date: 21 Jun 1995 15:06:58 GMT
Links: << >>  << T >>  << A >>
eddie@merlion.singnet.com.sg (q) writes:
>I could not understand the 'underscore' deviations of the respective 
>counters. Moreover for CBD48_4(CAO,[Q0..Q7],CAI,EN), there is no clock. 
>How can this be a down-counter ?

The XXX_N macros are submacros of the actual function block, which is
too complex to fit into a single cell. The real thing is CBD48. The
software links the submacros for you, and you only have to care about
the signals given in the CBD48 declaration. 
--
Segmented Memory Helps Structure Software


Article: 1424
Subject: Re: altera mail adress ?
From: crm182c@bmers245.bnr.ca (Hing-Fai Lee)
Date: 21 Jun 1995 17:02:23 GMT
Links: << >>  << T >>  << A >>
In article <3s43nh$qsp@chaos.kulnet.kuleuven.ac.be>,
Pete <peter.sels@student.kuleuven.ac.be> wrote:
>In article <3s3olf$62r@rc1.vub.ac.be>, tw38966@vub.ac.be (SH.RYU KIM HOFMANS) says:
>>
>>
>>Can someone tell me if Altera has a technical support email adress ?
>>
>>Thanx in advance !!!
>>
Try sos@altera.com. If you can't wait, call 1-800-800-EPLD
-------------------------------------------------------------
- Name:  Hing-Fai Lee   Address: Bell-Northern Research     -
- Phone: 613-765-2097            Dept 1D15, Mail Stop 170   -
- FAX:   613-763-2108            P.O. Box 3511, Station C   -
- Email: crm182c@bnr.ca          Ottawa, Canada, K1Y 4H7    -
-------------------------------------------------------------
- Disclaimer: I do not speak for BNR nor BNR speaks for me. -
-------------------------------------------------------------




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