FFD is a technique developed by Advanced Display Inc & Mitsubishi Electric to reduce the response time of AMLCDs.
If you want to know all about FFD you can look at the paper of the inventors here:
http://www.melco.co.jp/service/tft_tech/image/sid_2001_29_03.pdf
Also you can look around the melco site and you will find more. So far there is only one panel known to me made by Chunghwa which has FFD featured. See my other posts for the link.
What is FFD:
FFD stands for Feed Forward Driving. All major AMLCD manufacturers (eg LG/Philips, Samsung, Hitachi, Fujitsu, IBM ...) are working on special driving techniques for AMLCDs to make them faster. In general these techniques are called "over and under drive" techniques. FFD is only a special version of it. The primary application area is LCD-TV where fast moving pictures have to be displayed without ghosting.
The key idea behind FFD and others is to optimize the driving voltage for each pixel picture for shortest response time. These days the individual pixel of a AMLCD are driven by a voltage which is only dependent on the desired gray level to which a specific pixel has to change. For example, a high driving voltage is applied to change a pixel to a bright grey level and a low voltage is applied to make it a dark gray pixel. On the other hand, the response time of a pixel, i.e. how fast an individual pixel can change its state, depends on the height of the applied driving voltage: The higher the voltage change the faster the state change (simply speaking).
Now, the idea of FFD is to drive the pixel always at an extra high or low level , a level which is much higher or lower than the required voltage for the target gray level . By making the driving voltage 'extra high' or 'extra low' and dependent on the current and target state the change from one grey level to another can be significantly accelerated. Additionally, since the amount of extra voltage is made dependent on the current pixel state and the target gray level different response times for different grey lavels can be avoided. The picture quality is improved.
However, it takes extra effort in the control ciruitry, e.g. additional RAM and extra processing power. Also, the grey resolution is reduced: If there is 8 bit reserved for the coding of grey levels, the effective solution decreases with over and under drive because now the 8bits must be used to code the extra 'over and under drive levels' as well.
It is not clear (to me) if these special driving techniques will come or if they are already there (except for the one Chungwha panel). A lot of papers and patents are already known from various companies. Originally, this technique was invented to make LCDs based on existing LC modes (like TN, IPS and MVA) faster. But there are limits of course. The most promising LC mode is OCB which is ten times faster than those we have now. I have no idea when OCB is ready for mass production. Maybe FFD will come as an intermediate step but the reportet response time of <20ms for *all* grey levels is not fast enough to get rid of ghosting. I think flashing back light will come first with the 'old' techniques and then OCB. Maybe even combined, because OCB alone needs high frame rates to get rid of ghosting. Combined with flashing backlight its cheaper.
"what means driving a pixel"
In order to understand that and how it affects the response time of AMLCDs it is necessary to look a bit more deeper :
An active matrix liquid crystal display- which AMLCD stands for - has simply speaking 2 layers :
one LCD layer, this is where the liquid crystals are and
a second layer with an array of switches (transistors)
Behind these two layers there is a light source: the back light.
It is not necessary to go into details about the LCD layer (the terms MVA, PVA, IPS, FFS, TN have something to do with this layer)
The switches of the second layer are transistors and typically realized in thin film technique. This is why this layer is called TFT, which stands for thin film transistor. These TFTs are used to move the liquid crystals. Why do we want that?
Well, the basic principle of an AMLCD is that the light which is emmited from the back light is filtered by a matrix of liquid crystal cells. Via the TFTs the properties of a single cell can be changed in such a way that more or less light can pass through it. Simpliy speaking, the 'transparency' of each cell can be changed electrically via the transistors. For our eys it looks as if each cell or pixel is a small lamp which can be dimmed. In this way the display controller can affect the transparency or grey level for each pixel. The panel controller itself is not directly connected to the display. Instead it talks to an electronic circuitry which generates the necessary electrical signals to affect the properties of each individual cells. This circuitry is called driver electronics.
The size of this pixel matrix corresponds to the resolution of the display, e.g. 1280x1024. Since we are dealing with color displays the number is in fact 3 times higher because the resolution is needed for each color red, blue and green.
Now, it takes some time to change the grey level. If a pixel is completely off and the panel controller wants to turn it on, it sends a signal to the TFT driver electronics which generates a specifc voltage for the addressed pixel to turn it on. Once the voltage is applied, the liquid crystals are moving around under the influence of the electrical field so that more light can pass through the cell. This elapsed time is called "response time". It depends - and this is important - on the initial state and the end state of a pixel. The response time is the shortest if it was 100% off and has to turn 100% on or vice versa.
Facebook
Twitter
Instagram