The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> I would like to ask one question about Charge pump loop stability
https://designers-guide.org/forum/YaBB.pl?num=1326767979

Message started by yixiusky on Jan 16^th, 2012, 6:39pm

Title: I would like to ask one question about Charge pump loop stability
Post by yixiusky on Jan 16^th, 2012, 6:39pm

Dear all, i would like to ask one question:

As we can see from schematic, there is one loop existed from output to FBP (feedback point) of the error amp.

I would like to run AC simulation to check the loop stability. However, the part in white box is discrete signal, so i need modeling it, and then i can run AC simulation.

The white part is two diode with one cap, and input to cap is pulse ^^

Do any of you have experience to do it? Or do you have any book to recommend in order to modeling it?

Thank you so much ^^

Title: Re: I would like to ask one question about Charge pump loop stability
Post by Dan Clement on Jan 16^th, 2012, 7:32pm

I don't necessarily understand this circuit very well...

But I do know that switching regulators, which are similar to this charge pump, can be linearized and the ac response simulated.

Look on google books for a power management textbook. They talk about doing AC simulations on switching regulators.

I have never done a switching regulator so I'm afraid I am not much practical help.

Another option is to stick with transient simulations. You can inject a large transient pulse into the system and make sure it settles out. This method can be good since you get large signal effects but it will be dependent on the pulse characteristics.

Title: Re: I would like to ask one question about Charge pump loop stability
Post by yixiusky on Jan 16^th, 2012, 8:53pm

thank you so much, i will try to find the book what you mentioned ^^

Title: Re: I would like to ask one question about Charge pump loop stability
Post by raja.cedt on Jan 17^th, 2012, 12:05am

try for PAC in spectre RF.

Thanks,
raj.

Title: Re: I would like to ask one question about Charge pump loop stability
Post by yixiusky on Jan 17^th, 2012, 12:37am

thank you so much, but i do not know how to set up the PAC simulation. Can this simulation replace AC simulation to check loop stability?

Title: Re: I would like to ask one question about Charge pump loop stability
Post by RobG on Jan 17^th, 2012, 6:45pm

I'd just write the difference equations for Vreg(n) and use z-domain analysis so that you know what affects stability. Then verify the stability with transient simulations (pulse the load current).

That resistor/pnp combination is going to make the gain very non-linearly dependent on the output load. I bet it will be difficult to stabilize if the load varies a lot.

Watch that pnp... if it saturates you might turn on a parasitic npn which probably is not modeled.

Title: Re: I would like to ask one question about Charge pump loop stability
Post by RobG on Jan 17^th, 2012, 6:54pm

One more thought if you REALLY want to do w/ AC... you will have to include the the pnp circuit as in your discrete time model because the amount that Vreg changes with each cycle depends on the clocking frequency. The AC analysis will miss that.

I've never used PAC so it might take care of all this for you.

Title: Re: I would like to ask one question about Charge pump loop stability
Post by Frank Wiedmann on Jan 18^th, 2012, 6:14am

I suggest using the pstb analysis of SpectreRF (see http://www.designers-guide.org/Forum/YaBB.pl?num=1107716368 for some background information).

Title: Re: I would like to ask one question about Charge pump loop stability
Post by yixiusky on Jan 18^th, 2012, 4:50pm

thank you so much for all of you, i will try to run with pstb. ^^