Browse Prior Art Database

Imprinting model and imprinting compensation

IP.com Disclosure Number: IPCOM000126117D
Publication Date: 2005-Jul-01
Document File: 8 page(s) / 395K

Publishing Venue

The IP.com Prior Art Database

Abstract

Imprinting model and imprinting compensation Objectives I) Describe in a simplest way the “memory” effect of SLM mirror deflection D(t) caused by force E(τ), τ

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 64% of the total text.

Imprinting model and imprinting compensation

Objectives
I) Describe in a simplest way the memory effect of SLM
mirror deflection D(t) caused by force E(t), t<t. 

Requirement to the model:
Model is linear;
No explicit dependence from time;
Easy fit to experimental data (logscaling, for instance);
Small number of parameters for history storage;
Easy update of the history during consequent loads.

This results in D(t) = .E(t) G(t-t) dt, (-8<t<t), where the
main term in the kernel G has the form of (t-t)-1 (which
causes log scaling naturally), other terms (momentums)
have finite influence and are used both for fitting the
experimental data at finite time and for the main moment
update.  The finite number of moments is only required to
story whole history of loading.

II) Describe the algorithm for compensated loading(the
force E(t) correction) to have the desired deflection, i.e. to
eliminate the imprinting.

Objectives (cont.)

III) Memory restore (demagnetization):
a) to reduce the number of history parameters and
b) to remove the influence of unknown pre-history.

Imprinting compensation is simple computationally, but it requires
 large number of momentums to store the history of mirror loading
 during long period of time.
The number of parameters (the momentums) scales logarithmically
 in time. The initial history of loading is unknown. The reproducibility
 of the work session  depends on how large were the changes of the momentums
 during the session. The demagnetization...