cell libraries and place and route

[Frank E. Gennari]

On Fri, 10 Jun 2005 19:37:30 -0700, alino717 wrote:

Hi,

I am try to make glass conductive with SnCl2 but it require lots of
time. I use the pieces of 7-3 inch have this structure.

| _____________________ |
| == == |
| _____________________ | <----------- A pot or iron
| == == |
|_______________________|
Means that placing pieces of glass in pot by placing iorn nuts between
them to saparate them. I have to remain placed the pot in fire for
about 5 to 6 hours. Thats is only problem of time.
Please guide me any other mathod to make glass conductive in the way
that the visibility of glass should not be lossed. You can also guide
about another site.


Thanks in advance,

Asad

Ah, conductive, transparent glass. First you need something conductive,
like gold. But it also has to be transparent. I can sell you an 8oz bottle
of my special "invisible gold" for only $599.95. Please send me a check or
money order. When you receive the bottle it may look empty, but don't
worry, it's there - just invisible.


Sorry, I just had to say that! Really - this is a newsgroup for Cadence
users (CAD tools for layout, simulation, etc.). We generally don't work
with conductive glass here. I don't know which newsgroup you want, maybe
material sciences or chemistry or something (if you're serious about this).

Frank

 

[I. Bitchslapped U]

Time to replace all those Gieger Counter batteries...

Asad wrote:

Hi Dear,


I have follow your link
http://www.teralab.co.uk/Exper­iments/Experiments_Menu.htm. I am also
performing the all process that is explained in this link.
I do not blow the vapour through between the glass plates in my
experiment. Could you please explain what is effact of blowing the
vapour between the glass plates.


I am using tin chloride as I can find it easily. But tin oxide or zinc
oxide will
work very batter I think. You can also guide me to make tin oxide or
zinc oxide.
It would be very batter if I succeed in making the tin oxide or zinc
oxide


Thanks in advance,


Asad

 

[Guenther Sohler]

Hallo sunkey,

I use this code with cadence IC/5.1.41 and it works for me

ciwBannerMenus=hiGetBannerMenus(window(1))
if((member('ciwNLTools ciwBannerMenus) == nil)
then
ciwBannerMenus=reverse(cons('ciwNLTools reverse(ciwBannerMenus)))
hiReplaceAllBannerMenus(window(1) ciwBannerMenus )


'ciwNLTools is the pulldown menu created


have a nice day

 

[Andrew Beckett]

On 12 Jun 2005 16:31:13 -0700, "GS" <guneet_it@yahoo.com> wrote:

Hello everyone,

I have been trying to use the white_noise and flicker_noise functions
of Verilog-AMS to include noise in my models and ended up facing some
problems.

Actually, I am using the Cadence LDV-5.1 toolset for my simulations and
the Verilog-AMS LRM that comes with the version that I have states that
white_noise, flicker_noise and noise_table have no effect when used in
models.

My question is: Is there a way I can make use of the noise functions to
include noise in my models and simulate using LDV-5.1??

So far, I have been writing my mixed signal models in verilog-ams and
compiling, elaborating and simulating using ncvlog, ncelab and ncsim
respectively.

Thanks,
GS

The reason they have no effect is because AMS Designer does not support noise
analysis - and these sources only have an effect in small signal noise
analyses. They are used in spectre's noise, sp, pnoise, psp, qpnoise.qpsp
analyses - and now in transient if you turn on the transient noise option -
but not yet in any AMS analyses. AMS can now do ac analysis (in IUS54), but
there's no noise analysis yet.

Regards,

Andrew.

 

[Andrew Beckett]

On Sun, 12 Jun 2005 23:19:16 -0400, "sunkey" <sunkey@263.net> wrote:

what i mean is , how to use the SKILL func of
hiInsertBannerMenu( (hiGetCIWindow) ...)
as well as the CIW window was loaded ??

for i put my skill in the .cdsinit, but it was
not correct work -- for the window was not build
yet when i the 'insert' func was load...

thx for help ~ !

You need to call:

ciwMenuInit()

before trying to add your menu. This tells the menus to initialise first from
the menus files before you try to add your menus using the SKILL mechanism.

Regards,

Andrew.

 

[Stefan Joeres]

analyses. They are used in spectre's noise, sp, pnoise, psp, qpnoise.qpsp
analyses - and now in transient if you turn on the transient noise option
- but not yet in any AMS analyses.

Where's that switch ?

Regards,

Stefan



--
Dipl.-Ing. Stefan Joeres
Lehrstuhl für Integrierte Analogschaltungen
RWTH Aachen
Sommerfeldstr. 24
D-52074 Aachen

Tel: +49-241-8027752 (office)
Fax: +49-241-8022199

 

[Andrew Beckett]

On Mon, 13 Jun 2005 09:45:52 +0200, Stefan Joeres <joeres@ias.rwth-aachen.de>
wrote:

analyses. They are used in spectre's noise, sp, pnoise, psp, qpnoise.qpsp
analyses - and now in transient if you turn on the transient noise option
- but not yet in any AMS analyses.

Where's that switch ?

Regards,

Stefan

Stefan,

You need to be using IC5141 USR1 together with spectre from MMSIM60.
When you do that, the switch appears on the main transient analysis form.

Regards,

Andrew.

 

[Andrew Beckett]

On 30 May 2005 13:35:50 -0700, "S" <sireeshakandula@gmail.com> wrote:

hi
Can anyone please tell me how to increase the blow up limit of a
quantity in a VHDL-AMS model.
Thanks in advance.
S

If you put a "quantity" statement in the analog control file for the AMS
simulation - for example:

//

defineVoltage quantity \
name="V" \
units="Volts" \
abstol=1e-6 blowup=1e10 huge=1e5 \
description="Voltage"

tran tran stop=5u

I tested this by setting the blowup limit to 1, and it fell over with my 5V
supplies.

Why are you getting signals at the blowup level anyway? It tends to suggest
you really have a different type of quantity which is scaled differently, and
you're using voltage (say) to represent that. Or is it that you've defined
your own quantity, and it's picking up the default blowup limit of 1e9?

You didn't give much information.

I'm not sure whether there is another way of doing this for VHDL-AMS. For
Verilog-AMS, the blowup limit is a Cadence-specific option in the files
such as:

<instdir>/tools/spectre/etc/ahdl/disciplines.vams

Regards,

Andrew.

 

[Andrew Beckett]

On 12 Jun 2005 08:17:16 -0700, wjcndyd@yahoo.com wrote:

Anyone?

Look at this current discussion thread:

http://www.designers-guide.org/Forum/?board=ms_design;action=display;num=1118555245

Regards,

Andrew.

 

[S. Badel]

I have a cell which is parameterized in the schematic. dfIIToVldb is
not happy about it. For example on a MOS transistor in the cell I have
something like:

w = pPar("Width")
l = pPar("Length")

sure, that will work only if the parameters are assigned a value in some way.
i.e., you have to instanciate your cell in another one and assign a value to
these parameters. is this what you do? or are you trying to verify this very cell?

stéphane

 

[Andrew Beckett]

On 24 Jun 2005 14:12:50 -0700, "jorobins" <wulf_homme@hotmail.com> wrote:

Here is a problem that I've kind of figured out recently. Whenever I
turn off the grid on any schematic (even small designs), schematic
refreshes or zooming operations on a vnc session (key: F6) become very
slow. Quite often every 5th or 6th zoom, I'll need to wait for about
10seconds for the cadence schematic to get updated. If I turn the grid
back on, it gets fast again. 24-bit & 8-bit color performance on the
server both show the same problem. I also changed # of colors on the
vnc viewer (Real VNC 4.1.1), different types of encoding etc. on
windows XP,and it didn't make any difference.

The slow performance happens when I increase the grid size to something
like 6.x + (I tried to do this to make the grid less obvious, since I
don't like looking at the grid), along with making the "multiple" field
1, performance starts to degrade.

Any ideas if this is something to do with cadence/vnc OR graphics
hardware? Any possible workarounds other than keeping the grid on? Just
a minor peeve! I hope others can replicate this. Cadence version 5.0.33
running on RHEL 2.1....

thanks,
Jose

Have you tried using:

icfb -noblink

Or typing in the CIW:

hiEnableBlink(nil)

Regards,

Andrew.

 

[Andrew Beckett]

On 25 Jun 2005 09:01:12 -0700, "jorobins" <wulf_homme@hotmail.com> wrote:

Yes, I tried both and made sure that blink was off using
hiIsBlinkEnabled(). Didn't make a difference. This along with the
experments with 8-bit color leads me to suspect that color depth is not
the issue here...

I suspect it's something to do with the driver you're using for your graphics
card - perhaps it does not have the settings optimized?

Anyway, I'd suggest contacting custmer support, with details of the graphics
card and driver that is being used (in the XF86Config file, if you're using
XFree86).

Andrew.

 

[Bernd Fischer]

Can you give the details how you implemt it
in your callback button and what kind of form it is?

Bernd

camelot wrote:

Hello,
When I call this procedure

procedure(MigEditRuleFile()
prog(()
edit "test.txt"
))

from a form button callback nothing append but when I type the command
edit "test.txt" in CIW the command is executed. Is there a solution to
this question?

Camelot

 

[S. Badel]

please try it with
edit("test.txt")
or
(edit "test.txt")

edit "test.txt" is not proper skill code.
it works within the CIW, and it happens to work in your prog()
by the result of a side effect: edit is taken to be a label,
(prog allows to label statements and to jump around with go())
and "test.txt" is a string constant. Thus, it does nothing.

cheers,
stéphane

camelot wrote:

Hello,
When I call this procedure

procedure(MigEditRuleFile()
prog(()
edit "test.txt"
))

from a form button callback nothing append but when I type the command
edit "test.txt" in CIW the command is executed. Is there a solution to
this question?

Camelot

 

[Bernd Fischer]

Surly it can, be used in SKILL code directly,
because it is a SKILL function!

I don't know what you mean with a command?

Bernd

edit(
S_object
[ g_loadFlag ]
)
=> x_childId

Description

Edits a file, function, or variable. This function only works if you are in
graphical mode. This is an nlambda function.

edit brings up an editor window in a separate process and thus doesn't lock up
the CIW. If the object being edited is a function that was loaded after debug
mode was turned on, then edit opens up the file that contains the function. If
the editor is vi or emacs it jumps to the start of the function. If g_loadFlag
is t the file is loaded into SKILL when the editor is exited. Be sure the editor
variable is set up properly if you are using an editor other than vi or emacs.
Arguments

S_object
If you are editing a file, the object you are editing must be a string. If
you are editing a function or variable, it must be an unquoted symbol.

g_loadFlag
Determines whether to load the file after the editor window is exited.
Valid values: t or nil
Default: nil.

Value Returned

x_childId
Integer identifying the process spawned for the editor.





Tom_Ding wrote:

edit is a command. It can not be used in SKILL code directly.

Tom

camelot wrote:

Hello,
When I call this procedure

procedure(MigEditRuleFile()
prog(()
edit "test.txt"
))

from a form button callback nothing append but when I type the command
edit "test.txt" in CIW the command is executed. Is there a solution to
this question?

Camelot

 

[Andrew Beckett]

Stephane is quite right. Omitting the parentheses when you have a
sub-expression leads to incorrect behaviour.

In this case it is referencing the value of a variable called "edit", and then
the return value of the function will be the literal string "test.txt". You
need to put in the parentheses.

I always describe missing out the parentheses as "lazy" mode (to
distinguish it from "C" mode - edit("test.txt") or "LISP" mode -
(edit "test.txt") . You should only ever use lazy mode when typing functions
in the CIW, and even then only when you don't want to do something with the
return value of the function.

Regards,

Andrew.

On Tue, 28 Jun 2005 11:15:55 +0200, "S. Badel"
<stephane.badel@REMOVETHISepfl.ch> wrote:

please try it with
edit("test.txt")
or
(edit "test.txt")

edit "test.txt" is not proper skill code.
it works within the CIW, and it happens to work in your prog()
by the result of a side effect: edit is taken to be a label,
(prog allows to label statements and to jump around with go())
and "test.txt" is a string constant. Thus, it does nothing.

cheers,
stéphane

camelot wrote:
Hello,
When I call this procedure

procedure(MigEditRuleFile()
prog(()
edit "test.txt"
))

from a form button callback nothing append but when I type the command
edit "test.txt" in CIW the command is executed. Is there a solution to
this question?

Camelot

 

[S. Badel]

Andrew,

this is detail, but

In this case it is referencing the value of a variable called "edit", and then
then it should complain that edit variable doesn't exists. It does it in fact if

you replace the prog with a let, but with the prog it doesn't.
I think what happens is, as i said, the symbol 'edit' is taken for a label.

the return value of the function will be the literal string "test.txt".
According to SKILL reference, prog always return nil, unless you specify a return

value with a call to return().
again, it would work with a let, since let returns the value of the last statement,
but not with a prog.

stéphane


procedure( test() let( () edit "test.txt" ) )
=> test
test()
=> *Error* eval: unbound variable - edit

procedure( test() prog( () edit "test.txt" ) )
=> test
test()
=> nil

procedure( test prog( ( x ) edit unless(x x=1) x++ when(x<n go(edit)) return(x) ) )
=> test
test(5)
=> 5

 

[S. Badel]

I've run accros this problem as well.
Deleting the file works if you are logged on the NFS server.
(I simply ssh and rm).

stéphane

Flash wrote:

Hello to all,

My problem:
I'm programming some stuff for Analog Design Environment (ADE)that
makes it necessary to remove old simulation results. If I'm doing so,
there is always a file left: ".nfsXXXX" (XXXX: running number).

What I already know (from several postings and others):
The ".nfsXXXX" file appears when I try to delete a file that is still
used by a process. Using "fuser" and "ps" I identified the "icfb.exe"
as that process holding the file as an "open file".
This happens as soon as I select the corresponding directory as result
directory (through menu: Results->Select... in ADE).
However, Results->Delete... or "deleteFile()" (SKILL) or "rm -r" (Unix)
can't handle this. (Who wrote the "PCR 427164" ?!? ;-) )
I know, the file will disappear at the moment of ending the "icfb.exe"
process, but this is a really bad option.

So:
How can I delete the file nevertheless? or
How can the ADE be stopped to hold the file open?
(how does the "Netlist and Run" button get rid of it?)

thanx for reading this massive text

 

[S. Badel]

<< Callback string for this field. If a callback is specified, the item
selected (either text or an icon) is appended to this callback
before being executed by SKILL. >>

thus you define the callback as a string like

?callback "myCallback arg1 ... argn"

and the callback gets one more argument which is the selected item

procedure( myCallback( arg1 ... argn item ) printf("selected %s\n" item) )

alternatively, include hiGetCurrentForm() in your callback and use the form handle
to retrieve the field value

?callback "myCallback hiGetCurrentForm()"
procedure( myCallback( form @rest args ) printf("selected %s\n" form->field->value) )


the trap is to define the callback as

?callback "myCallback( arg1 ... argn )" or ?callback "(myCallback arg1 ... argn)"

because then the evaluated skill is

myCallback( arg1 ... argn ) item

which calls the callback without the item argument, and evaluates the item upon return.



stéphane


kirscher_jerome@yahoo.fr wrote:

Hello,
I ve created a cyclic field in my form using the skill function
"hiCreateCyclicField" . But how can I launch a function depending on
the field I choose. I mean, I manage to create just one callback, which
will be activated whatever the field of my list is.
Thanks a lot
Jerome

 

[Andrew Beckett]

Stephane,

For some reason I had a temporary mind glitch and didn't read your previous
reply properly, nor indeed the original question. You are quite right.

I have a pathological hatred of the go() function, so that's probably what
threw me ;-)

Andrew.

On Wed, 29 Jun 2005 09:13:42 +0200, "S. Badel"
<stephane.badel@REMOVETHISepfl.ch> wrote:

Andrew,

this is detail, but

In this case it is referencing the value of a variable called "edit", and then
then it should complain that edit variable doesn't exists. It does it in fact if
you replace the prog with a let, but with the prog it doesn't.
I think what happens is, as i said, the symbol 'edit' is taken for a label.

the return value of the function will be the literal string "test.txt".
According to SKILL reference, prog always return nil, unless you specify a return
value with a call to return().
again, it would work with a let, since let returns the value of the last statement,
but not with a prog.

stéphane


procedure( test() let( () edit "test.txt" ) )
=> test
test()
=> *Error* eval: unbound variable - edit

procedure( test() prog( () edit "test.txt" ) )
=> test
test()
=> nil

procedure( test prog( ( x ) edit unless(x x=1) x++ when(x<n go(edit)) return(x) ) )
=> test
test(5)
=> 5