OK Xilinx users, it's time I was let in on the joke...

Mark McDougall wrote:

> And worst of all - _you_ lot, telling me that Xilinx actually works, and
> that you _can_ use it for more than flashing LEDs on the Spartan starter
> kit. And I was gullible enough to believe you! :O :(
> 
> I've seen the light. You _cannot_ convince me that it is possible to
> produce a commercial product in silicon using these tools. Period.
Umm, we have been developing stuff for some time for mostly in-house use
here at Washington University, mostly using Virtex 2 stuff here.  I also 
have been doing some very UN-challenging units on both CPLDs and Spartan 
2E FPGAs at Pico Systems, my "night job".  You can see some of the 
products I have developed at 
http://pico-systems.com/oscrc4/catalog/index.php?cPath=3
for CNC motion control.  Most of the boards on that page have either a
CPLD or an FPGA in them.

I have mostly moved development (at both sites) over to Linux and iSE 
10.1.  I do have an annoying problem where you can't print schematics 
without some fooling around, but otherwise it seems to work well.  I 
only use schematics now on some interface-type CPLDs, so that's no big
deal.  I still find their modelsim simulator cumbersome to use, but I 
live with it.  I understand they are phasing out the timing diagram 
entry for sim stimulus, and I really prefer that to writing the test 
bench in words.

Now, I have to admit, most of this stuff is NOT pushing the chips to the
ultimate.  I am using 40 MHz clocks on the Spartan 2E, for instance!
I do have a CPLD design wehre I used every FF and every macrocell, and 
any time I want to change something it is a great hassle, but it does 
get it routed.

I have sold about 180 of the motion controller boards, so it is real 
production to ME, at least!

Jon

On Nov 7, 4:54=A0am, Curt Johnson <curt.john...@dicombox.net> wrote:
> If it is any solace, the Mentor schematic capture and PCB routing tools
> are much worse. In addition to continuous license issues, crashes,
> version compatibility problems, and file corruption, they produce
> unreliable output. I now have to manually check every trace on a 10
> layer PCB with multiple BGA packages ever since we discovered that the
> tools can randomly delete nets from a fully routed board without warning.

Err, but a random-deleted net is actually easy to find, as you usually
have more than one copy (as in, in the SCH, or even in a copy of the
PCB ) ?

Which Mentor flows/versions are you using ?

On Nov 6, 8:01=A0pm, Oscar Almer <o.al...@gmail.com> wrote:
> On Fri, 06 Nov 2009 17:14:21 +1100
>
>
>
> Mark McDougall <ma...@vl.com.au> wrote:
> > I've had it up to the eyeballs with Xilinx tools now. I'm seriously
> > ready to go postal in the lobby of Xilinx HQ. I don't expect
> > perfection but this really is beyond a joke.
>
> > Can someone please put me out of my misery, and finally admit that you
> > have _all_ been having me on for the past few years now! :O ...that
> > it has all been an elaborate hoax instigated by someone I offended in
> > a past life. ...that a team of engineers has been working for years
> > to produce an IDE that crashes randomly, and steadfastly refuses to
> > launch tools on Tuesday mornings and Friday afternoons? ...working
> > for years on a synthesizer that removes random bits of logic, or sits
> > spinning in an infinite loop on an entity that works in another
> > project? ...that corrupts my project file bi-monthly.
>
> > And worst of all - _you_ lot, telling me that Xilinx actually works,
> > and that you _can_ use it for more than flashing LEDs on the Spartan
> > starter kit. And I was gullible enough to believe you! :O :(
>
> > I've seen the light. You _cannot_ convince me that it is possible to
> > produce a commercial product in silicon using these tools. Period.
>
> I, at least, gave up on the ise wrapper about a year and a half ago -
> it just didn't do it, anymore. Instead I drive the flow (xst, map, par,
> etc) from a single makefile, and a short one at that - maybe 20 lines.
>
> The things I know im missing out on is 1. the pretty XML reports and 2.
> COREgen etc hook-ins. The former I can survive without, as its
> easier to grep through plain text reports anyway, and the latter I
> typically only need to run once anyway, at which point I suffer ise
> long enough to move the generated files somewhere sensible.
>
> Just my experience.
>
> //Oscar


I used to do the same. And I still use the scripts when I just need
to recompile the entire thing because of an RTL change in my code.

But, since ISE 11, I have started to create MPD files and use EDK
to put together the system. The GUI is now (for me at least) very
stable, and reasonable easy to use. (This is all on linux).

I am even starting to support the EDK plugin for most of our IP
cores, as it seems to be actually working pretty well now.

Regards,
rudi

Mark McDougall <markm@vl.com.au> wrote:

>I've had it up to the eyeballs with Xilinx tools now. I'm seriously ready
>to go postal in the lobby of Xilinx HQ. I don't expect perfection but this
>really is beyond a joke.
>
>Can someone please put me out of my misery, and finally admit that you
>have _all_ been having me on for the past few years now! :O ...that it has
>all been an elaborate hoax instigated by someone I offended in a past
>life. ...that a team of engineers has been working for years to produce an
>IDE that crashes randomly, and steadfastly refuses to launch tools on
>Tuesday mornings and Friday afternoons? ...working for years on a
>synthesizer that removes random bits of logic, or sits spinning in an
>infinite loop on an entity that works in another project? ...that corrupts
>my project file bi-monthly.
>
>And worst of all - _you_ lot, telling me that Xilinx actually works, and
>that you _can_ use it for more than flashing LEDs on the Spartan starter
>kit. And I was gullible enough to believe you! :O :(
>
>I've seen the light. You _cannot_ convince me that it is possible to
>produce a commercial product in silicon using these tools. Period.

Its about time someone writes a decent Eclipse plugin / makefile
generator.

-- 
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
                     "If it doesn't fit, use a bigger hammer!"
--------------------------------------------------------------

On Fri, 06 Nov 2009 07:54:47 -0800, Curt Johnson
<curt.johnson@dicombox.net> wrote:

|If it is any solace, the Mentor schematic capture and PCB routing tools 
|are much worse. In addition to continuous license issues, crashes, 
|version compatibility problems, and file corruption, they produce 
|unreliable output. I now have to manually check every trace on a 10 
|layer PCB with multiple BGA packages ever since we discovered that the 
|tools can randomly delete nets from a fully routed board without warning.
|=============

Mentor was an expensive joke ten years ago and I am not so surprised
that it has not changed. I am glad I am retired and don't have to use
their tools. 

james

On Nov 6, 1:32=A0pm, Curt Johnson <curt.john...@dicombox.net> wrote:
> rickman wrote:
> > The reason for using commercial tools is supposed to be the great
> > support and the lack of serious bugs... =A0so if you don't like your
> > expensive layout tools, why use them? =A0I used FreePCB on my last
> > project and found it very suitable. =A0I am sure there are some things
> > that you need to do manually that expensive tools might do
> > automatically, but hand checking all the nets is not one of them.
>
> > Not trying to be smart, this is a serious question.
>
> > Rick
>
> Good question. Mostly inertia.
>
> If FreePCB will import our libraries and designs, I'll check it out in a
> minute.
>
> Maintenance is an issue. We are in the medical racket. Product lifetimes
> can be a dozen years or more. We are still shipping product with
> XC3142's in them. When parts go obsolete, we often need to quick turn a
> board to keep shipping product. That means either we keep the tools
> around for the duration, or find something that can read the files.
>
> Most of my new designs are more or less 50% original and 50% cut and
> pasted from previous designs. It saves a bit of time not having to
> recreate parts and redraw schematics.
>
> I've used Pads, Tango, Calay, Cadroid, Orcad, and a few other packages
> over the years. Some were better than others, but all of them had
> problems of one sort or another. I've kind of got to the point where if
> my PC doesn't catch fire after I install a new version, I consider it a
> victory.
>
> Curt

FreePCB is not the sun and moon of PCB layout.  It won't import any
other layouts or libraries I'm pretty sure.  It is a pretty good
tool.  Do *any* layout tools import other layouts?  There have been
efforts over the years to establish a common format for schematic and
layout tools and it has always met with failure because none of the
tool vendors want to let a user switch easily.  That alone is reason
enough for me to want to never use a commercial tool again.

Rick

On Nov 9, 4:51=A0pm, rickman <gnu...@gmail.com> wrote:
> =A0Do *any* layout tools import other layouts? =A0

 Yes, Mentor have Layout translators from Protel(Altium), Cadstar,
OrCAD, and also Schematic Translators from the same.

 They work quite well, and are good for harvesting web resource,
and things like reference designs.

 Last time I checked, Analog Devices and SiLabs used Mentor's
PADS for their designs, but others may use OrCAD Schematic.

DXF is also a useful interchange format, often overlooked.

 I remember creating a good decal for an Open-Frame relay, via
the DXF file supplied.
 In fact, the DXF file had vital info that was missing off the
drawing !!


-jg

On Sun, 8 Nov 2009 20:34:15 -0800 (PST), -jg <jim.granville@gmail.com> wrote:

>On Nov 9, 4:51&#4294967295;pm, rickman <gnu...@gmail.com> wrote:
>> &#4294967295;Do *any* layout tools import other layouts? &#4294967295;
>
> Yes, Mentor have Layout translators from Protel(Altium), Cadstar,
>OrCAD, and also Schematic Translators from the same.
>
> They work quite well, and are good for harvesting web resource,
>and things like reference designs.
>
> Last time I checked, Analog Devices and SiLabs used Mentor's
>PADS for their designs, but others may use OrCAD Schematic.
>
>DXF is also a useful interchange format, often overlooked.

But probably one of the least standard 'standards' out there, at least in terms of implementation...
I find that an imported/exported DXF rarely resembles the original in all respects - even really
basic stuff like the scale can be completely wrong!

On Nov 8, 11:34=A0pm, -jg <jim.granvi...@gmail.com> wrote:
> On Nov 9, 4:51=A0pm, rickman <gnu...@gmail.com> wrote:
>
> > =A0Do *any* layout tools import other layouts? =A0
>
> =A0Yes, Mentor have Layout translators from Protel(Altium), Cadstar,
> OrCAD, and also Schematic Translators from the same.
>
> =A0They work quite well, and are good for harvesting web resource,
> and things like reference designs.
>
> =A0Last time I checked, Analog Devices and SiLabs used Mentor's
> PADS for their designs, but others may use OrCAD Schematic.
>
> DXF is also a useful interchange format, often overlooked.
>
> =A0I remember creating a good decal for an Open-Frame relay, via
> the DXF file supplied.
> =A0In fact, the DXF file had vital info that was missing off the
> drawing !!

Curt was asking about drawings and libraries.  Do the converters also
convert user built libraries?  A lot of people have significant
libraries which represent a lot of work although, if the info is
captured in a way that is easily read, I think  most footprints are
pretty easy to recreate.  It is coming up with the dimensions that is
the hard part I think.  I know FreePCB makes it *very* easy to draw
footprints, it is just figuring out how to design them that is the
hard part.  It would only be a weeks work to convert 100 footprints if
you had the dimensions readily available.  Schmatic symbols are a
different thing because each schematic package has its own quirks
about how they are made I think.

FreePCB does not include schematics at all.  Many people use TinyCAD
for that, but since layout is not closely tied to schematic capture,
it doesn't matter a lot what you use.

Rick

malcolm wrote:

> Err, but a random-deleted net is actually easy to find, as you usually
> have more than one copy (as in, in the SCH, or even in a copy of the
> PCB ) ?
> 
> Which Mentor flows/versions are you using ?

If I have plots from an earlier revision, I compare them with the new 
revision one page at a time on a light box. Every difference has to be 
traced to ensure that it is an intended change.

On a 10 layer PCB with 4 BGAs, about a dozen smaller packages and a 
couple of hundred discretes, it takes me two or three days.

Not difficult, but extremely tedious.

Viewdraw & Pads. I'm using Viewdraw (or DxDesigner I think they call it 
now) version 2006.1. I havn't been able to get the 2007 release to read 
in any of our current projects.

The random net deleting feature has been in place since before I started 
here in 1993; but before BGAs, we could always rework the boards.

Curt