The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Modeling >> Behavioral Models >> voltage limited current source
https://designers-guide.org/forum/YaBB.pl?num=1313587765

Message started by danmc on Aug 17^th, 2011, 6:29am

Title: voltage limited current source
Post by danmc on Aug 17^th, 2011, 6:29am

Sometimes it is useful to have a verilog-a or AMS model for something like a bias current generator but in those cases it is also good to provide voltage limiting. Is there a simple way to do this in Verilog-A or Verilog-AMS? The issue is that without voltage limiting, the current sources will simply force the current no matter what and sometimes that is really not what the rest of the circuit wants!

In the past I have done some stuff in spectre with hanging additional controlled sources that have limiting on my output node.

Thanks
-Dan

Title: Re: voltage limited current source
Post by boe on Aug 17^th, 2011, 12:39pm

Danmc,

you can do something like
Code:

if (V(ibias, vss) > 1)
I(ibias, vss) <+ CurrentValue;
else
I(ibias, vss) <+ 0;

However, for better convergence you should avoid the discontinuity of the current and use a soft transition.

B O E

Title: Re: voltage limited current source
Post by danmc on Aug 17^th, 2011, 1:17pm

you mean use something like like a tanh function instead of hard compare?

Title: Re: voltage limited current source
Post by Ken Kundert on Aug 17^th, 2011, 3:46pm

I don't recommend following BOE's suggestion. That is a recipe for convergence problems. I have seen two approaches to this problem that works well. One is to carefully model current/voltage limiting. You can use http://www.designers-guide.org/VerilogAMS/functional-blocks/regulator/regulator.va as a model. Alternately, you can model a two terminal clamp that you place in parallel with the current source. It would have infinite impedance within the compliance range of the current source, but then would look like a small resistor outside that range. It would be designed so that the IV characteristics were piecewise continuous.

-Ken

Title: Re: voltage limited current source
Post by boe on Aug 18^th, 2011, 3:25am

danmc wrote on Aug 17^th, 2011, 1:17pm:

you mean use something like like a tanh function instead of hard compare?

Yes; and that seems to work quite well, too.
B O E

Title: Re: voltage limited current source
Post by danmc on Aug 19^th, 2011, 8:36am

Ken Kundert wrote on Aug 17^th, 2011, 3:46pm:

Thanks Ken. I read those 2 pages from your book (which was only about arms length away) and implemented 2 boundaries instead of 1 and it seems to work like a champ.

Thanks again.

-Dan