The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> low power current starved ring oscillator
https://designers-guide.org/forum/YaBB.pl?num=1341047246

Message started by mixed_signal on Jun 30^th, 2012, 2:07am

Title: low power current starved ring oscillator
Post by mixed_signal on Jun 30^th, 2012, 2:07am

Hi,
I am designing a low power current starved ring oscillator as shown in attachment. The bias current for each inverters is in the range of 50n to 100nA.
M1 M4 are in strong inversion saturation and act as current sources. i have made w/l very small like 1/40 to put them in strong inversion.

How do i size M2 M3? If i size for general inverter like 4/1 and 2/1 then they will be in weak inversion during switching at such low currents. Is that OK?
OR should i size them with w/L small so that the are in strong inversion?

Title: Re: low power current starved ring oscillator
Post by AnalogDE on Jun 30^th, 2012, 11:43am

M2 and M3 are just switches. They will be in linear operation region of operation for the most part. Start with min-L and wide enough so the voltage drop across them is small enough to keep the outer devices in saturation.

Title: Re: low power current starved ring oscillator
Post by mixed_signal on Jun 30^th, 2012, 8:14pm

If w/l of M2, M3 is made greater than 1 then at current =40nA they will be in weak inversion. Isnt it a problem? To put them in strong inversion w/l of all transistors have to be <<1

Title: Re: low power current starved ring oscillator
Post by AnalogDE on Jul 1^st, 2012, 10:09am

The goal in sizing the switches is to minimize Vds. You may also want to minimize the gate cap of the switches (minimize loading). All this necessitates a min L device with the smallest W that works for you.

Title: Re: low power current starved ring oscillator
Post by loose-electron on Jul 3^rd, 2012, 2:57pm

what is this for?

THis type of architecture is very noise sensitive (power and ground) and full of jitter problems in a VCO ring oscillator PLL.

Title: Re: low power current starved ring oscillator
Post by mixed_signal on Jul 3^rd, 2012, 10:16pm

Hi loose-electron,

This is for an ultra low power temperature sensor. I found this vco frequenc to be highly linear wrt bias current. I like it!!!!

I want to do a transient simulation for the vco and at the same time sweep temperature during the transient. How can it be done in cadence?

Title: Re: low power current starved ring oscillator
Post by ywguo on Jul 4^th, 2012, 7:36am

Hi mixed-signal,

Please look at
http://www.designers-guide.org/Forum/YaBB.pl?num=1275773203/1#1
for changing temperature during a transient analysis.

Yawei

Title: Re: low power current starved ring oscillator
Post by loose-electron on Jul 7^th, 2012, 8:58pm

mixed_signal wrote on Jul 3^rd, 2012, 10:16pm:

Put some impedance in the power and ground, and put some noise on both.

You may not like it so much after that.

Title: Re: low power current starved ring oscillator
Post by mixed_signal on Jul 8^th, 2012, 9:02am

Hi loose electron,
I m interested in the simulations suggested. But how do u put impedance & noise in cadence? Can u explain ? :)