Sunday, 20 October 2013

Week 4


1.       Progress Made:
This week I have tried to find more literatures about my project. Firstly, I searched for some materials about DMD in the IEEE website. Meanwhile, I wanted to find some resources about the application of DMD. Luckily, I found 4 articles about DMD. After reading the DMD background theories in them, I have a deep understanding of the work principle of DMD. However, the applications in them like volumetric 3D display do not have a close relationship with my FYP. In fact, I wanted to learn how the authors investigate their project, but they refer to too much technical knowledge. In my opinion, they may not give big favour to my project. Hence, eventually, I took notes about DMD work principle in my log book. This work principle is quite in detail.
DMD Principle:

l  The DMD pixel, which is an integrated micro-electronic mechanical system (MEMS) structure, is manufactured on a CMOS static random array memory (SRAM) cell.
l  The mirror (aluminia) is linked to an underlying yoke.
l  The yoke is connected by two thin, mechanically compliant torsion hinges (also aluminia).
l  The hinges are held by the posts which stick on the underlying substrate.
 (Explanation for some items:
Hinge: A kind of bearing linking two objects which allows a restricted angle of rotation between them.
Yoke: A beam applied between a pair of objects to capacitate them to pull together on a load. )
l  There is an electrostatic field between memory cell, yoke and mirror, which produces an electrostatic torque. This kind of torque works opposed to the restoring torque of the hinges to make the mirror rotate in the positive or negative direction.
l  The mirror and yoke rotate until the yoke comes to rest against mechanical stops which have the same potential as the yoke.
l  The address electrodes under the yokes are linked to SRAM cell.
l  The yoke and mirror are linked to a bias bus which interconnects the yoke and mirrors of each pixel to a bond pad at the chip perimeter
l  The yoke is attracted to one or another address electrode relying on which one is activated.
l  Which of the electrodes is activated relies on the state of the SRAM cell.
l  A 1 stored in the cell will lead the mirror to rotate to the degree between +10 and +12 degrees to an on state.
l  A 0 is stored in the cell will cause the mirror to move to the degrees from -12 to -10 degrees to an off state.
l  If the memory cell is nether 1 or 0, there is no electrostatic force on the mirror. The torsion hinges will cause the mirror to 0 degree.
l  If the mirror is tilted in either direction, a bias current will keep it in the current position even if a new bit of data is being loaded into the SRAM cell.



Electronic working principle:
1)        Memory ready-All SRAM cells have been loaded with the new address states.
2)        Rest-All mirrors are reset in parallel position (voltage pulse are applied to the bias bus).
3)        Unlatch-The bias is removed to unlatch mirrors and permit them to release and begin to rotate to flat state
4)        Differentiate-Under the retarding fields, the mirrors which keep in the same state from those that are to cross over to a new state are separated.
5)        Land and latch-The bias is restored in order to capture the separated mirrors and help them to rotate to the address states, then settle and latch.
6)        Update memory array-The bias keep turning on to make the mirrors latched in order to stop them from responding to changes in the memory, while the SRAM cell is written with new video data.
7)        Repeat sequence starting at step 1.


Furthermore, I searched the merits of DMD technology which mainly reflect in four aspects.
*  Digital advantage
Since DMD applies the digital signal, it is able to save the time and cost for the analogue digital conversion and have lower noise. Nevertheless, other display techniques still use analogue signal, like LCD.
*  Reflective advantage
DMD is a reflective device which can achieve the light efficiency larger than 60%. However, LCDs are polarization-dependent devices so that it cannot get the 50% of the lamp light.
*  Seamless picture advantage
DMD has a quite high fill factor up to 90%. The reason is that DMD mirrors are 16um square and the interval between them is just only 1um. Therefore, DMD can achieve higher perceived resolution due to high fill factor, finally which produces natural and lifelike image. By contrast, LCDs can just reach 70% fill factor.
*  Scale advantage
DMD systems are high integration so that the DMD projectors can be produced much lighter and smaller.
Then I read three articles about the application of DLP LightCrafter. As a matter of fact, I wanted to know how those researchers develop a prototype by DLP LightCrafter.  Nevertheless, DLP LightCrafter is not their main equipment. Those articles do not introduce how they use this equipment to implement their objectives.
2.       Problems and Challenges:
At present, I merely know the operation principle of DMD. There are following problems needed to figure out. To begin with, I do not know what kind of field in clinical eye testing will be applied by the DLP LightCrafter. Furthermore, it is difficult to find other resources about the using methods for DLP LightCrafter. Maybe I can understand the equipment better when I get it.   
3.       Plans for Next Week:
In the next week, maybe I can get DLP LightCrafter. I can do the project in the 4th floor of EEE building. On Tuesday, I will learn to use the equipment in the lab room. I will do my best to be familiar with all the connectors in the DLP LightCrafter.

Reference:
L J. Hombeck, Current Status of the Digital Micromirror Device (DMD) for Projection Television Applications, http://ieeexplore.ieee.org.ezproxy.liv.ac.uk/stamp/stamp.jsp?tp=&arnumber=347329

No comments:

Post a Comment