[SI-LIST] Re: Which tool is the best
- From: Ravinder.Ajmani@xxxxxxxx
- To: si-list@xxxxxxxxxxxxx
- Date: Wed, 14 Apr 2004 14:47:30 -0700
I got in to Signal Integrity work around mid-nineties. At that time, we
had a Cadence SigNoise license. The software version had many bugs and I
got frustrated after trying it for a few times. The set up was really
cumbersome, as our logic symbols were not designed with many SigNoise
parameters. I looked at other tools, such as XTK (then owned by
ViewLogic), Interconnectix (then an independent company), UniCad, Pacific
Numerix, as well as Cadence BoardQuest. I really liked XTK and
Interconnectix, but found that both had steep learning curve. Since SI
was only a small part of my total work, I could not justify spending weeks
learning a tool, which I may best use once in a month or two months.
It was then I came across a demo version of HyperLynx (which was also
independent then), and developed an instant liking for it because of its
ease of use. Within minutes, I was able to start simulation on a board. I
bought a node locked license of HyperSuite (including BoardSim and
LineSim), and have been using them ever since. Later when HyperLynx came
out with Crosstalk module, I was one of their Beta testers. The software
has grown in features and performance, and can no longer be considered a
low-end tool. Its ease of use is just superb, and LineSim is a great tool
for pre-layout simulations. That said, I have a few gripes about it.
1) The simulation step size is selected by the program (what Kellee calls
"expert system automation"). This sometimes causes problem when
simulating a driver with very fast slew rate. One driver I use in our
design often causes high-frequency oscillations. I had taken up this
issue with the support people but could not get a resolution mainly
because there is no user control on step size.
2) If one is using resistor packs as series terminators for a data bus,
then the input and output pins for resistors always get scrambled. Thus
you will end up simulating two different nets, one from the driver to
resistor, and a totally different net from resistor to the receiver.
3) The EMI module can not simulate differential nets.
4) Batch mode requires that models be attached to the ICs being simulated.
When an IBIS model is assigned to an IC, the default mode is receiver.
One has to select every pin individually to change the mode to driver.
Perhaps there is a simpler way that I am not aware of.
Currently, I am also using Cadence SPECCTRAQuest a lot for my simulation
work. This tool has improved a lot from the SigNoise days. Since we use
Cadence front-end and back-end tools, there is no need for translation.
SPECCTRAQuest has definitely more features than HyperLynx, and a much
better waveform display. Differential signal simulation is also very
good, and I find batch simulation very useful. My main issue with
SPECCTRAQuest is that unlike HyperLynx, I can't assign a single buffer to
one pin of an ASIC. I have to create a model for the complete ASIC before
I can run simulation. I can't export a net to SigXplorer unless I have
models assigned to drivers and receivers. By the way, SigXplorer is not
as flexible as HyperLynx LineSim, since I can't build my own stackup. Also
at times during crosstalk simulation, simulator shows no crosstalk whereas
I can clearly see parallel nets at close spacing. These same nets produce
significant crosstalk, when simulated on HyperLynx.
Overall, I find HyperLynx and SPECCTRAQuest as a good complementary pair
of tools, since each one has its strong points as well as weaknesses.
Regards, Ravinder
Server PCB Development
Hitachi Global Storage Technologies
Email: Ravinder.Ajmani@xxxxxxxx
Dan Bostan <dbostan@xxxxxxxxx>
Sent by: si-list-bounce@xxxxxxxxxxxxx
04/14/2004 11:42 AM
Please respond to dbostan
To: kelleecrisafulli@xxxxxxxxxxx, si-list@xxxxxxxxxxxxx
cc:
From: si-list-bounce@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: Which tool is the best
A bird told me that the "new" Hyperlynx incorporates
now the XTK/Quad engine, which Mentor got from
Viewlogic/Innoveda, which was very, very good.
As you may see on the Mentor's web site, the XTK is
downplayed nowadays, and Hyperlynx is heavily
promoted, for good reasons...
/dan
--- Kellee Crisafulli <kelleecrisafulli@xxxxxxxxxxx>
wrote:
> Hello all,
>
> I often hear comments like the one below which seems
> to be based on here-say
> heard from a sales person as opposed to real user
> feedback on the current
> version of the tool:
>
> " I would have said that Hyperlynx is a low-end tool
> as well, although I
> understand that it's come
> a long way in the last few years"
>
> For example to say the Hyperlynx Si tools are
> low-end may have been true in
> 1991
> but saying that today is like saying Harley Davidson
> makes low end motorcycles.
>
> Many features in Hyperlynx are "best-in-class" for
> example I feel the
> cross-talk feature may be more
> advanced than any other tool in the market because
> of it's ability to
> automatically
> identify the net's that need to be included in the
> simulation. Most other
> tools in the market require
> the user to "guess" which nets needs need to be
> included by specifying a
> distance around the signal
> being analyzed. Even the best SI engineers cannot
> "guess" correctly all
> the time. Having a tool with the
> expert system built in to determine the required
> nets is very powerful.
>
> All SI tools have dozen's of settings that determine
> the details of how the
> simulation will be created.
> Many people forget about these settings when they
> use the tools. As a
> result some times the results are not as accurate as
> desired. HyperLynx
> has expert-system automation for many of these
> settings resulting in much
> more accurate
> simulations by people that do not use the tools
> every day.
>
> Since being acquired by Pads then Innoveda, and now
> Mentor the HyperLynx
> tool has been improved each time it was acquired by
> choosing some of the
> best technologies available within the new parent
> company.
>
> I don't believe there is "best tool" in the market
> some tools have better
> batch mode features
> some have better user interfaces, some have complex
> automation which
> safe-guards the user (like HyperLynx).
> Some tools are only 2D or 2.5D while others are full
> 3D analyzers.
> A 2D tool uses a 2 dimensional field analyzer to
> create the simulation. A
> full 3D tool (when enabled for 3D)
> will use a full 3D analysis (generally VERY SLOW)
> and very accurate.
> Still other tools have features for power supply or
> RF analysis that are
> best in class.
> And to complicate matters even more the tools are
> constantly evolving.
>
> 1) I urge people to refrain from comments about
> tools based on 3 to 10 year
> old information received from
> a competing sales person.
> If you have concrete information based on your
> own work with a tool
> than by all means share it.
> 2) A "best tool" is the one that reliably get's your
> job done in a timely
> manor.
> 3) If only HyperLynx had that nice Harley Davidson
> rumble driving out of
> the parking lot it would be really great :)
>
> My bias (for those who do not know me):
> This is is a biased comment from one of the author's
> of the HyperLynx tool
> and a founder of the IBIS group.
>
> best regards,
> Kellee Crisafulli
>
> At 12:37 AM 4/14/2004, you wrote:
> >I would have said that Hyperlynx is a low-end
> >tool as well, although I understand that it's come
> >a long way in the last few years.
> >
> >I'm most familiar with SpecctraQuest, and I highly
> >recommend it. It's my impression that ICX is
> comparable
> >in every way.
> >
> >SpecctraQuest is particularly recommended if you
> >are using Allegro for board design, since there is
> no
> >translation procedure required for moving the board
> >database into SpecctraQuest in that case.
> >
> >
> >On Apr 13, 2004, at 10:51 PM, Dan Bostan wrote:
> >
> > > Hyperlynx, ICX and Spectraquest.
> > > Protel is a low end tool, good only for small
> PCB
> > > designs.
> > > /dan
> > >
> > > --- Tayyab Jamil <tayyab@xxxxxxxxxxxxxx> wrote:
> > >> Hello,
> > >>
> > >> Can any one seggest me the good and reliable
> tools
> > >> for signal integrity
> > >> simulations? Also how do you people rate PROTEL
> 99
> > >> SE for PCB designing and
> > >> signal integrity simulations?
> > >>
> > >> Tayyab Jamil Qureshi
> > >>
> > >> Sr. Design Engineer
> > >> CARE Pvt. Ltd.
> > >> 19 Attaturk Avenu, G-5/1
> > >> Islamabad, Pakistan
> > >> Phone: +92 51 2874794
> > >> +92 51 2874115
> > >> Fax: +92 51 2874614
> > >> Cell: +92 300 9506986
> > >> email: tayyab@xxxxxxxxxxxxxx
>
>
>
------------------------------------------------------------------
> To unsubscribe from si-list:
> si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in
> the Subject field
>
> or to administer your membership from a web page, go
> to:
> //www.freelists.org/webpage/si-list
>
> For help:
> si-list-request@xxxxxxxxxxxxx with 'help' in the
> Subject field
>
> List FAQ wiki page is located at:
>
> http://si-list.org/wiki/wiki.pl?Si-List_FAQ
>
> List technical documents are available at:
> http://www.si-list.org
>
> List archives are viewable at:
> //www.freelists.org/archives/si-list
> or at our remote archives:
> http://groups.yahoo.com/group/si-list/messages
> Old (prior to June 6, 2001) list archives are
> viewable at:
> http://www.qsl.net/wb6tpu
>
>
__________________________________
Do you Yahoo!?
Yahoo! Tax Center - File online by April 15th
http://taxes.yahoo.com/filing.html
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
