Understanding Physical and Electrical Principles of LCD Displays
Liquid crystal display (LCD) is a technology being widely used across many end equipments such as smartphones, TVs, automotive infotainment systems, and in smart home or factory environments.
Start with this video to understand the key physical and electrical principles behind a single LCD pixel before moving onto the next video linked below.
Resources
[MUSIC PLAYING]
Hello. My name is Andreas. I'm with Texas Instruments, in charge of product marketing for LCD and AMOLED Display Power Products. I have created two short videos you can watch to learn about the functionality of LCD displays. The first video explains the physical fundamentals. The second video, how these work together as a complete LCD display solution.
Same as me, you might have a general curiosity for technology and how things work or if you're designing a display solution for your company, this video will provide you with the knowledge to design the display towards your target, such as lowest power consumption or best picture quality and to budget.
I will describe the basic functionality of LCD displays as you know it from various end equipment such as smartphones, TVs, automotive, and industrial display applications. So stay with me watching this first video. One of the reasons for LCD displays being widely used in high volume and across many end equipment is that this is a mature technology. This translates into a well-known technology at reasonable costs, good quality, and lifetime.
There were actually three fundamental discoveries or developments in physics needed to make an LCD display work. First, I will introduce these in this video. Let's think of them as ingredients. Later, I will explain how these ingredients work together in a recipe.
From a physical perspective, light is fascinating. It is both a particle as well as an electromagnetic wave. As an EM wave, the vectors have a certain orientation. Most natural and artificial light sources emit light in an undefined or mixed orientation. The plane of the EM vectors can be at any angle. A polarizer filters for a specific EM vector orientation. This is the first important ingredient to our display.
Let's shine light of mixed EM orientations on the polarizer. How does it work? Maybe you have seen a dog trying to cross a narrow doorway with a stick in his mouth. Trying to walk through with a stick held horizontally, the doorway will block the stick and the carrying dog. If the stick is oriented more vertically, both stick and dog can pass through.
A polarizer works similar to the doorway. Light with proper EM orientation can pass through. Light of other angles will be blocked. The light that has passed the filter is polarized so the EM vectors are oriented to fixed planes. I did not explain yet why we need polarizers, but we can see already that for a polarizer, the light passing has only 50% of its original light power, the luminous flux. Some types of sunglasses use this effect. Let's keep the function of a polarizer in mind for now and the image of the dog that wants to pass through the narrow doorway. That's ingredient number one.
Liquid crystals are the next fascinating ingredient to our display recipe. Liquid crystal, as the name suggests, is a crystal with liquid characteristics. The ones we need have two essential physical properties. They are electrically bipolar and have light polarizing characteristics. If put between two substrates which are very close to each other, where the liquid crystals are given the preferred orientation set off by 90 degrees to each other, they will arrange like a helix between the two substrates. This is called twisted nematic.
The preferred orientation needed for the twist can be achieved by brushing the glass substrates. This setup is almost magic. It can turn the orientation of polarized light by 90 degrees. It is basically telling the dog to turn his head to pass through the doorway.
Let's look now on the bipolar characteristics of the liquid crystals. If an electric field is applied from the outside, the liquid crystals start to turn. They might leave their helix structure and lose the ability to twist the light vector orientation by 90 degrees. Depending on the strengths of the electric field or voltage applied, this can regulate the orientation of the EM vector planes. With that, an externally applied voltage can control the angle by which polarized light is twisted.
Without going into details, from this twisted nematic, there has been a lot of development on these cells to super twisted nematic, double super twisted nematic, triple super twisted nematic, to in-plane switching, and so on and so on. This is an evolution to improve contra-switching time, light transmittance, and viewing angles. These are important parameters to specify the optical performance of a display. These can be found in display specifications.
Now we get to the last important ingredient. The voltage needed to twist the liquid crystals must be applied somehow. It also needs to be stored somewhere until the next refresh happens. Probably these structures need to be very tiny to achieve high density but also to have little structures to increase the transmittance of light. Thin film transistors, TFT, are the solution. These FETs are built by depositing semiconducting materials on a substrate made of amorphous or polysilicon. It can be also metal oxides or organic materials. To a certain extent, these materials can be transparent. Why we need this to be transparent will become clear soon.
Now we know where the acronym TFT LCD display comes from. It is short for thin film transistor liquid crystal display. Strangely, the polarizers did not make it to be part of the acronym. By the way, one should correctly say LC display, but LCD display became colloquial. So no need to be picky about that.
So far, we looked on the ingredients. But how does the recipe work together? Let's have a look on how one pixel cell is built or actually we should rather say subpixel for now. We need a strong light source, the backlight. It emits light in undefined EM planes.
Next, we look on the sandwich cookie arrangement. The cookie wafers are polarizers but turn by 90 degrees to each other. So no light would be able to pass through this array. The dog carrying the stick vertically can pass the first narrow doorway, but it will be blocked by the second right after, which is too low now for the stick to pass vertically.
Something needs to help the dog to turn the stick into a horizontal position to pass the second doorway. This is what the sandwich filling does. The filling is the liquid crystal fluid, as described earlier, building the helix structure, turning the polarization plane by 90 degrees. With that, the polarized light by the first polarizer is turned by 90 degrees and passes the second polarizer without additional luminous flux losses. The light shines through.
You might find information if light would pass through a plot without an external voltage applied, called normally white and normally black modes. In our example, this is a normally white arrangement. Also in our sandwich cookie is the TFT structure to control the light twisting angle of the liquid crystal included. If a voltage is applied on the isolating substrates, this is charging a capacitor, creating an electric field. The liquid crystals would twist almost proportionally to the magnitude of the electric field.
In this case, they would not turn the light vector by 90 degrees but less. The subpixel would become less bright since the light travels through only with reduced luminance. Making the field even stronger by applying a higher voltage, the pixel can be driven black. We might leave the image with the dark at this point, as not half or one-third of the dark could pass the gate.
But we have seen, depending on the electric field, we can regulate the amount of light passing through the setup in an analog way. This is a dynamic light valve controlled by the external voltage applied. That's quite fascinating. With an externally applied voltage, we can regulate the amount of light that shines from the back light through the sandwich cookie filled with liquid crystals, which is regulating the brightness of a subpixel.
I did not explain yet how to get to a color picture presentation. That's done by using a red-green-blue filter grid. Then three subpixels control the brightness of each red, green, and blue content. The human eye will mix the three colors to one pixel with an RGB representation. In the color room, the display can cover the color gamut.
So each pixel is represented by three RGB LC subpixel cells. Each cell works like a capacitor where externally applied voltage loss turning the brightness and color of each pixel to reach any coordinate in the addressable color gamut.
So far, so ideal. As described above, each cell could keep its brightness level for eternity. Power would only be consumed if the image content changes to twist the LC liquid into a different angle. But no power consumption for as long as the picture does not change. This is only the case for EL displays. For LCD, we need to stop looking on each cell as being physically ideal. There are free carriers in the liquid crystal layer. And this has bad consequences.
Also for the same brightness of a subpixel, each cell needs continuously refreshes of the voltage applied on the capacitor to keep the charge on the plates constant. And since the capacitance of each cell is low, we need a storage capacitor in parallel to get to an acceptable self-discharge time until the next refresh happens. As a result, there are not only dynamic but also static power losses by leakage currents.
When I mentioned free carriers earlier, you might already have thought about another challenge. If the voltage applied was unipolar, these free carriers would migrate to the positive or negative side. Not a good thing, as this would destroy the functionality long term and result in image sticking. Maybe you recall that image sticking was a problem in early LCDs. But no worries. There is a solution to it. Instead of applying a DC voltage on this cell, let's introduce a virtual ground named VCOM somewhere in the middle. The cell is charged to a voltage within ground or VCOM in one cycle and within VCOM to maximum voltage in the other.
Since the VCOM is the common virtual ground, this is like applying an AC voltage on the cell. The voltage or electric field would alter its sine from cycle to cycle. This would make the liquid crystals twist by 180 degrees. It has no effect on the optical functionality of the cell. The liquid crystal still twists proportionally to the voltage applied. But the problem of free carriers gathering up is gone and with that, the problem of image sticking. This technique is called pixel inversion and happens usually every frame or in certain patterns.
We can already have a look on the electric representation of a subpixel. The thin film transistor triggers the refresh of a subpixel. If triggered on, the voltage from the data line is applied on the cell and storage capacitor. Once charged, the thin film transistor would turn off until the next subpixel refresh happens. The IC turning the gate on and off is referred to as gate driver. The IC setting the voltage on the liquid crystal cell is named source driver.
A word on the TFT. The TFT can only be a compromise between good electrical performance such as low resistive losses, little parasitic gate charge, fast switching, and optical characteristics, so being as transparent as possible. To fully turn the transistor on and off, the magnitude at the gate must be quite high compared to the maximum voltage at the drain. The voltage or currency is less a concern. Since the TFT also has a high temperature coefficient, you'll sometimes see that the gate driver supply has a temperature compensation block to adopt the voltage to the actual temperature to save power but also to increase lifetime of the display.
We have seen already that the voltage applied on the cell turns the liquid crystal to trigger each subpixel from white to black and each step in between. If a pixel has a color depth of 16.7 million colors, each subpixel must be able to represent 256 gray scales or 8-bit. The source driver sets the voltage to 1 out of 256 discrete steps. Actually, it is 2 times 256 steps. Remember the pixel inversion. Sounds like a deck. So probably the source driver supply needs to be very tight in output voltage accuracy. Actually, the source driver I see is an array of parallel decks. I'll explain soon how it works.
Thanks for watching the first video. If you would like to understand how these basic components work together as an LCD display, stay tuned and watch the second video.
[MUSIC PLAYING]