Re: ***[Possible UCE]*** Re: Generating noise in VHDL-AMS

From: John Shields <jshields_at_.....>
Date: Sun Mar 19 2006 - 16:28:47 PST
Hi,

Amro,  all of us are happy to listen and/or react to the discussion of your language and modeling issues.  It is appropriate.  I want to share with you that it is not appropriate to discuss product and vendor specific issues here.  IEEE rules obligate us to avoid it in the interest of fairness and to prevent any conflict of interest where criticism or promotion of products/vendors might be perceived.  The discussion you and Joachim have been having should be kept between you and not posted to vhdl-ams@eda.org.  I note that Joachim has been careful to avoid this, but you persist in replying with cc: vhdl-ams@eda.org.  Alain Vachoux also mentioned this with your first post, but I must presume you did not appreciate why or when it applies.  If you feel you have to refer to a particular product, tool, or vendor by name in order to make your point, assume that the email should not be posted here.

Regards,
John Shields

Amr Turk wrote:

Dear Joachim,
I have tried it and it worked fine. The problem was with the Systemvision on windows when I used it to try this file.
But here on the ADMS on linux it works good. I don't know why.
May be because I used VHDL on the top in the systemvision. Not sure but when I used here ELDO spice on top it worked fine,
Amro Tork
On 3/16/06, Joachim Haase <Joachim.Haase@eas.iis.fraunhofer.de> wrote:
Dear Amro,

It seems to run with your gaussian package.
I used ADVance MS 4.2_2.1.

Regards,
Joachim



Amr Turk wrote:
> Dear Joachim,
> I just used in the vnoisy the guassian distribution instead of the uniform.
> Here is the code,
> Amro
>
>
> On 3/15/06, *Joachim Haase* <Joachim.Haase@eas.iis.fraunhofer.de
> <mailto: Joachim.Haase@eas.iis.fraunhofer.de>> wrote:
>
>     Dear Amro,
>
>
>     it seems that this is a problem of the tool.
>     In the ADMS version I use the concurrent procedure
>     call is not supported for instance. To fix the problem,
>     I would need the model where the procedure
>     is called.
>
>
>     Best Regards,
>     Joachim
>
>
>
>     Amr Turk wrote:
>     > Dear Joachim,
>     > The code that I posted that implement the gaussian distribution.
>     When I
>     > compile it. It gives no error. But when I use it in the code, the
>     > following error comes out when running the a design :
>     >
>     > Session stopped on status VIF_OBSOLETE_UNIT with severity VIF_ERROR
>     > at line 305 of file V:/adms/src/hpi/hpi_library.c
>     > Error in front-end
>     > by the way, I don't have a drive named V:, and I went to the adms
>     folder
>     > and didn't find a folder named hpi. I don't know what is this
>     error it
>     > has no error number.
>     >
>     > Thank you anywhy,
>     > Amro
>     >
>     >
>     > On 3/15/06, *Amr Turk* < amrsfmt@gmail.com
>     <mailto:amrsfmt@gmail.com> <mailto: amrsfmt@gmail.com
>     <mailto: amrsfmt@gmail.com>>>
>     > wrote:
>     >
>     >     Dear Joachim,
>     >
>     >     I have finished the gaussian random number generator. I
>     included it
>     >     here.
>     >
>     >     I know that a VHDL process needs a sensitivity list or a wait
>     >     statement. But the problem, I didn't think of solving the
>     problem in
>     >     that way. Your solution was great.
>     >
>     >     best regards,
>     >     Amro
>     >
>     >
>     >     On 3/15/06, *Joachim Haase* <
>     Joachim.Haase@eas.iis.fraunhofer.de
>     <mailto:Joachim.Haase@eas.iis.fraunhofer.de >
>     >     <mailto:Joachim.Haase@eas.iis.fraunhofer.de
>     <mailto:Joachim.Haase@eas.iis.fraunhofer.de >>> wrote:
>     >
>     >         Dear Amr,
>     >
>     >
>     >         a VHDL process needs either a sensitivity list or a wait
>     >         statement. I have included a generic parameter TSAMPL,
>     >         a wait statement "wait for TSAMPL" and the argument TSAMPL
>     >         to the RAMP attributes. Thus,the random values are sampled
>     >         with a period TSAMPL. The consequence is that the
>     autocorrelation
>     >         function of the voltage should be 0.0 for arguments greater
>     >         than TSAMPL but unequal 0.0 for smaller arguments.
>     >
>     >         The remaining code I did not check and change.
>     >
>     >
>     >         Regards
>     >         Joachim
>     >
>     >
>     >
>     >         Amr Turk wrote:
>     >>
>     >> Dear Joachim,
>     >> I have tried the way you told me.but it seems not working. No
>     >         error in
>     >> compilation. But I can't see the noise. I made a sin source and
>     >> superimposed on the signal. Pure sin wave.
>     >> Here is the code.
>     >>  If anyone could help, Thank you
>     >>
>     >>
>     >>
>     >> On 3/14/06, *Amr Turk* < amrsfmt@gmail.com <mailto:amrsfmt@gmail.com>
>     >         <mailto:amrsfmt@gmail.com <mailto: amrsfmt@gmail.com>>
>     <mailto:amrsfmt@gmail.com <mailto:amrsfmt@gmail.com>
>     >         <mailto: amrsfmt@gmail.com <mailto:amrsfmt@gmail.com>>>>
>     >> wrote:
>     >>
>     >>
>     >>     Dear all,
>     >>     Thank you, Dr.Olivier Rolland, I went to your site and I
>     >         didn't find
>     >>     this noise generator application. Please if you can send
>     >         the link to
>     >>     it. And give me more info about it.
>     >>
>     >>     Joachim,
>     >>     I don't know how to thank you, This info that you have
>     >         provided me
>     >>     was very useful.
>     >>     Regarding this solution:"I am not quite sure if I
>     >         understand the
>     >>     problem. What you can do is to
>     >>
>     >>     update a real-valued signal in a process and assign the value
>     >>     to a quantity using for instance the 'RAMP attribute. The
>     >         signal has
>     >>     to be declared outside the process."
>     >>     I didn't try it,but I was concerned to thank because the
>     >         link that
>     >>     you gave me was very helpful for me. I will try it soon,
>     >         may be in
>     >>     the next 2 days. If it worked, I will send the code with it
>     >         so it
>     >>     can be like a sample for anyone else.
>     >>
>     >>     Thank you,
>     >>     Amro
>     >>
>     >>
>     >>     On 3/13/06, *Dr. Olivier Rolland* <
>     >         olivier.rolland@systemsvip.com
>     <mailto:olivier.rolland@systemsvip.com>
>     >         <mailto:olivier.rolland@systemsvip.com
>     <mailto:olivier.rolland@systemsvip.com >>
>     >>     <mailto: olivier.rolland@systemsvip.com
>     <mailto:olivier.rolland@systemsvip.com >
>     >         <mailto:olivier.rolland@systemsvip.com
>     <mailto:olivier.rolland@systemsvip.com >>>> wrote:
>     >>
>     >>         Amro,
>     >>
>     >>         We are developping a noise generator application which
>     >         creates
>     >>         automatically  the adequate VHDL-AMS code for temporal
>     >         discrete
>     >>         noise
>     >>         (graphically spectrally defined with a shape over any
>     >         branch
>     >>         desired of
>     >>         1/f**n asymptotic slop, n belonging to Z). This
>     >         application will be
>     >>         provided as an ASP on our web site.
>     >>         If you want  more information, please feel free to
>     >         contact us.
>     >>         Regards.
>     >>
>     >>         Olivier Rolland
>     >>
>     >>         Dr. Olivier Rolland
>     >>         Systems'ViP
>     >>         c/o SEMIA
>     >>         4, rue Boussingault
>     >>         F-67000 Strasbourg
>     >>
>     >>         Tel:   +33 671 128 130
>     >>         Email: olivier.rolland@systemsvip.com
>     <mailto: olivier.rolland@systemsvip.com>
>     >         <mailto:olivier.rolland@systemsvip.com
>     <mailto: olivier.rolland@systemsvip.com>>
>     >>         <mailto:olivier.rolland@systemsvip.com
>     <mailto: olivier.rolland@systemsvip.com>
>     >         <mailto:olivier.rolland@systemsvip.com
>     <mailto: olivier.rolland@systemsvip.com>>>
>     >>         Web:   http://www.systemsvip.com
>     >         < http://www.systemsvip.com/ >< http://www.systemsvip.com/>
>     >>
>     >>         Systems'ViP: Your innovation capitalization partner
>     >>
>     >>         This e-mail, including attachments, is intended for the
>     >>         person(s) or company named and may contain confidential
>     >         and/or
>     >>         legally privileged information. Unauthorized
>     >         disclosure, copying
>     >>         or use of this information may be unlawful and is
>     >         prohibited. If
>     >>         you are not the intended recipient,please delete this
>     >         message
>     >>         and notify the sender
>     >>
>     >>
>     >>
>     >>         Amr Turk wrote:
>     >>> Dear Joachim Haase,
>     >>> Thank you, for your support. But actually,I used this solution.
>     >>         But if
>     >>> there is a solution that is can be in the VHDL code without
>     >>         using eldo
>     >>> noise source, I would be grateful if someone point it out.
>     >>>
>     >>> By the way, there is a function : Uniform in IEEE.MATH_REAL that
>     >>> generates random number from 0.0 to 1.0 uniformly distributed.
>     >>         But the
>     >>> problem is :
>     >>> 1- This functions need variables, which mean needs process. I
>     >         don't
>     >>> know a way, to assign a quantity with a varaible through the
>     >>         process.
>     >>> Which make it finally, don't work for Analog.
>     >>> 2- The noise actually is gaussian distributed, But this one is
>     >>> uniform. But still can work.
>     >>>
>     >>> Finally, Thank you for your post, I was delighted to see it.
>     >>>
>     >>> Amro Tork
>     >>>
>     >>
>     >>
>     >>
>     >
>     >
>     >
>     >
>     >
>     >
>
>





Received on Sun Mar 19 16:28:54 2006

This archive was generated by hypermail 2.1.8 : Sun Mar 19 2006 - 16:29:37 PST