[SI-LIST] Re: small signal AC model for Current mode DC_DC converters, for full PDN simulation

hi
in a DCDC converter, there are lots of parameters, the 2 natural poles 
(mentioned previously: double pole at the half the switching 
frequency, and a single pole at the 1/(2*TT*R_load*C_out)), and the 
compensation network. All of these contribute into the behaviour of the 
converter, to the transient response, and to the output impedance vs frequency 
profile.
If we design a power distribution network in frequency domain, it has to be OK 
not only at the ceramic decoupling, or at the other elements separately, but 
they have to be OK together. Other elements can be modeled as RLC (capacitors) 
or touchstone models (planes), and simulated in AC simulation. 
How do we match the impedance of the DCDC to the bulk decaps, if we have just a 
very-very simplified model for the converter? matched distributed bypassing... 
I wanted to see all the possible antiresonances between elements, and to see, 
how flat the whole profile is...
The question was, how to make a model of a DCDC converter, compatible with the 
previous 2 models (planes, decaps). I had a solution, but I was not shure in 
the correctness, because of the strange way of operation of the current mode 
converter.

I know the Kemet spice, and that i has multi-element models for the caps. So, 
its nice that we can get a more accurate model of the caps from it.
Maybe I am wrong, but I have a feeling that to get similar accuracy from the 
capacitors and the DCDC, then we should start with a similar circuit I had, and 
have a  R-L-C-Lmount  model for the caps. As far as the DCDC is modeled with 
just an inductor, there is not much point in using more than these 4 elements 
for a cap.

Istvan Nagy
CCT
  ----- Original Message ----- 
  From: Allen Mayar/EXTON/USA/SALES/KEMET/US 
  To: Istvan Nagy 
  Cc: si-list@xxxxxxxxxxxxx ; si-list-bounce@xxxxxxxxxxxxx 
  Sent: Monday, March 23, 2009 2:48 PM
  Subject: Re: [SI-LIST] small signal AC model for Current mode DC_DC 
converters, for full PDN simulation



  Istvan:

  Regarding the RLC lumped models for decoupling caps, to avoid errors in 
Electrolytic capacitors (Tantalum/Aluminum and polymers of both) it is helpful 
to plot a more complex model that considers the RC ladder effect (to mimic the 
capacitance loss with the increased frequency). Also for low ESL<1.2nH it is 
important to include a RL network to mimic decaying ESL with increasing 
frequency.

  You can import the Electrolytic models that takes RC ladder effect into 
consideration from the site: 
  http://www.kemet.com/kemet/web/homepage/kechome.nsf/weben/kemsoft

  Instructions:
  
http://www.kemet.com/kemet/web/homepage/kechome.nsf/file/KEMETSpiceman361x.pdf/$file/KEMETSpiceman361x.pdf
   
         
        Kind Regards
        Allen Mayar

       
       
       
        Be gentle with your ecosystem, print only when necessary 







        "Istvan Nagy" <buenos@xxxxxxxxxxx> 
        Sent by: si-list-bounce@xxxxxxxxxxxxx
        03/22/2009 06:22 PM
       To <si-list@xxxxxxxxxxxxx> 
              cc  
              Subject [SI-LIST] small signal AC model for Current mode DC_DC 
converters, for full PDN simulation 

              

       



  Hi

  I would like to model a power distribution system in frequency domain with a 
  spice-like AC analysis, with all elements of it, together.
  For decoupling capacitors, I can use RLC lumped models, for IC pins 
  something similar, for power planes and package planes, I can use an 
  electromagnetic simulator to create a touchstone file that I can import into 
  the program (QUCS, or Agilent ADS).
  But, for the voltage regulator, I think I have to create a small signal 
  model to include in the simulation as a subcircuit. This model must have the 
  same output impedance versus frequency response, as the original DCDC 
  converter has, nothing else hast to be the same (switching circuits are not 
  needed in the model, nor correct voltage levels). I dont want to model the 
  DCDC converter with a single inductance or similar model, but put the whole 
  control loop equivalent circuit into the simulation. the whole thing is 
  described here: 
  http://www.buenos.extra.hu/download/PowerIntegrityDesign_prj.rar (there is a 
  pdf in it, and some circuit files) The "circuit" is all the models and 
  elements of the PDN together.

  For a voltage mode converter, I think it is quiet straightforward how to 
  make the equivalent circuit ( 
  http://www.buenos.extra.hu/download/voltagemode.jpg ), but for a current 
  mode converter, it is trickier. My assumption was this: I read somewhere 
  that a current mode converter has a double pole at the half the switching 
  frequency, and a single pole at the 1/(2*TT*R_load*C_out). So, based on 
  this, I made an equivalent circuit which has the same poles and no zeroes, 
  so the same transfer function: 
  http://www.buenos.extra.hu/download/currentmode.jpg The circuit element 
  parameters are automatically calculated based on the provided switching 
  frequency, load current, some other elements are coming from the original 
  schematics, like the compensation RC networks... Maybe this way of modelling 
  is not perfect for this purpose, but I think it's better than just using a 
  single inductor (or an RLC model) for representing the whole DC/DC, or than 
  guessing about the transient response.

  Is this model correct, or if not, how should I make it to be correct?
  I am not shure in this part.

  regards,
  Istvan Nagy
  CCT, UK

  ------------------------------------------------------------------
  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:
  http://www.freelists.org/webpage/si-list

  For help:
  si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field


  List technical documents are available at:
                 http://www.si-list.net

  List archives are viewable at:     
                                   http://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
   




  THIS CORRESPONDENCE CONTAINS CONFIDENTIAL INFORMATION OF KEMET ELECTRONICS 
CORPORATION AND ITS AFFILIATED COMPANIES. If you have received this e-mail and 
it was not intended for you, please let us know, and then delete it. We thank 
you for treating our confidential information in a courteous and professional 
manner.

  DISCLAIMER: Although we have taken reasonable steps to reduce risks against 
viruses, the reliability of this method of communication cannot be guaranteed. 
It can be intercepted, corrupted, or delayed, or it may arrive incomplete, 
contain viruses, or be affected by other interference.  The opening of any 
attachments to this e-mail indicates your agreement to hold us harmless and 
release us from liability for any damage sustained as a result of this 
transmission.


------------------------------------------------------------------
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:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field


List technical documents are available at:
                http://www.si-list.net

List archives are viewable at:     
                http://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: