Hello,

I have an application where I need an on-chip MOS capacitor in my
device.  The electrodes consist of doped N(100) silicon [1-10 Ohms-sq] for
the bottom electrode and Cr-Cu-Au for the top electrode.  The dielectric is
approximately 1um of wet thermally grown SiO2.  After fabrication, the
capacitor characteristics are measured by plotting the C-V curve, which I
would expect to be consistent over time.  However, there appears to be a
transient effect where the C-V data has to be taken several times in order
to get repeatable data.  I wonder if this is some sort of charging
effect?  Since this will greatly complicate the operation of the completed
sensor, does anyone have any comments or suggestions about this phenomena?

Sincerely,

Robert Dean

Research Associate IV
Center for Advanced Vehicle Electronics
Auburn University
200 Broun Hall
Auburn, AL 36849

Voice: 334-844-1838
Fax:   334-844-1898
Email: [email protected]
Web: http://www.eng.auburn.edu/~rdean