Hi everybody:
Recently i bought an 1.8" LCD TFT from ebay. They have web store but they sell on ebay also. It is called INESUN and they have a lot usernames on ebay.
The item they have is LILLIPUT 1.8" TFT.
I bought it planing to do a Diyprojector using a slide proyector. You would have heard about doing it.
The problem is that INESUM says it is 116160 pixel resolution, lilliput also says that. But it is not true. The resolution is measured (V*H) that means Vertical x Horizontal.
This monitor has 220 lines (vertical=220) lines. Then if it is 116160 pixel, then there should be 528 rows. There is something else called RGB strip that means each pixel is one color displaying and next one would be next collor and so on.
So RGB (Red Green Blue) a line would be RGBRGBRGBRGB.....
then we divide teoric 528 rows by 3 and we have 176 comlete pixels. I mean there are 176 Red ones, 176 Green ones and 176 Blue ones every line.
Actually it is not true. There are not such pixels. I counted them and found there are 93 red 93 green and of course 93 blue pixel every line. So 93 x 3 x 220=61380 pixel at all. Half the resolution they claim.
It looks to be hard to count the pixels (you would think microscope is need) but there are several ways to count them. the easiest would be taking a photo with a hi resolution camera (macro option, to take close photos you know) mine is 4 Megapixel and works just fine. I can count every pixel.
Other way could be as simple as removing the lcd from the lilliput and put it at the slide proyector. Then the proyected image can be large enough to count every pixel. (you just need to count 1 line (rows) and 1 row (lines). Then multiply these two terms and you'll have the true resolution of the LCD.
I sent my photos to INESUN and they didn't get convinced at all allthought they changed their website specs during two weeks of discussion. Finally they told me each pixel is count twice (if I multiply by 2 my measured resolution it is not 116160 anyway) then they sayd it there were some dark pixels between the ones i could count in my photo. (Really amazing)
I would like to show you my photos. I have some of them on my webspace. I'll drop a link here.
those are projected images (far to close)
http://www.telefonica.net/web2/roxilandia/pro1.JPG
http://www.telefonica.net/web2/roxilandia/pro2.JPG
http://www.telefonica.net/web2/roxilandia/pro3.JPG
http://www.telefonica.net/web2/roxilandia/pro4.JPG
this one is large file 800K
(hi resolution lcd photo so you can zoom in)
http://www.telefonica.net/web2/roxilandia/complete3.JPG
If someone just have measured also the resolution, could get in contact with me to interchange information.
Here are the Seller (INESUN)
http://www.inesun.com/
and lilliput:
http://www.lilliputweb.net/212GL-20NPTFT.html
Thanks for this forum
Recently i bought an 1.8" LCD TFT from ebay. They have web store but they sell on ebay also. It is called INESUN and they have a lot usernames on ebay.
The item they have is LILLIPUT 1.8" TFT.
I bought it planing to do a Diyprojector using a slide proyector. You would have heard about doing it.
The problem is that INESUM says it is 116160 pixel resolution, lilliput also says that. But it is not true. The resolution is measured (V*H) that means Vertical x Horizontal.
This monitor has 220 lines (vertical=220) lines. Then if it is 116160 pixel, then there should be 528 rows. There is something else called RGB strip that means each pixel is one color displaying and next one would be next collor and so on.
So RGB (Red Green Blue) a line would be RGBRGBRGBRGB.....
then we divide teoric 528 rows by 3 and we have 176 comlete pixels. I mean there are 176 Red ones, 176 Green ones and 176 Blue ones every line.
Actually it is not true. There are not such pixels. I counted them and found there are 93 red 93 green and of course 93 blue pixel every line. So 93 x 3 x 220=61380 pixel at all. Half the resolution they claim.
It looks to be hard to count the pixels (you would think microscope is need) but there are several ways to count them. the easiest would be taking a photo with a hi resolution camera (macro option, to take close photos you know) mine is 4 Megapixel and works just fine. I can count every pixel.
Other way could be as simple as removing the lcd from the lilliput and put it at the slide proyector. Then the proyected image can be large enough to count every pixel. (you just need to count 1 line (rows) and 1 row (lines). Then multiply these two terms and you'll have the true resolution of the LCD.
I sent my photos to INESUN and they didn't get convinced at all allthought they changed their website specs during two weeks of discussion. Finally they told me each pixel is count twice (if I multiply by 2 my measured resolution it is not 116160 anyway) then they sayd it there were some dark pixels between the ones i could count in my photo. (Really amazing)
I would like to show you my photos. I have some of them on my webspace. I'll drop a link here.
those are projected images (far to close)
http://www.telefonica.net/web2/roxilandia/pro1.JPG
http://www.telefonica.net/web2/roxilandia/pro2.JPG
http://www.telefonica.net/web2/roxilandia/pro3.JPG
http://www.telefonica.net/web2/roxilandia/pro4.JPG
this one is large file 800K
(hi resolution lcd photo so you can zoom in)
http://www.telefonica.net/web2/roxilandia/complete3.JPG
If someone just have measured also the resolution, could get in contact with me to interchange information.
Here are the Seller (INESUN)
http://www.inesun.com/
and lilliput:
http://www.lilliputweb.net/212GL-20NPTFT.html
Thanks for this forum
Um...
Correct me if I'm wrong but...
1) you have 220 vertical lines
2) you have 3 pixels (1 for each color, R/G/B)
3) stated res was 116160
4) you are trying to determin the horizontal lines
Formula:
Take V for vertical lines
3P for pixels
R for resolution
H for horizontal line
H=R/V(3P)
H=116160/220(3)
H=116160/660
H=176 horizontal lines
So the screen size is 220x176 which seem right for a screen that size. Now if you times H by V you only get 38720....is that wrong...no you must then multiply by 3 for each RGB pixel, which would give you the stated 116160 res.
Anyone...feel free to correct my math if it is wrong.
Correct me if I'm wrong but...
1) you have 220 vertical lines
2) you have 3 pixels (1 for each color, R/G/B)
3) stated res was 116160
4) you are trying to determin the horizontal lines
Formula:
Take V for vertical lines
3P for pixels
R for resolution
H for horizontal line
H=R/V(3P)
H=116160/220(3)
H=116160/660
H=176 horizontal lines
So the screen size is 220x176 which seem right for a screen that size. Now if you times H by V you only get 38720....is that wrong...no you must then multiply by 3 for each RGB pixel, which would give you the stated 116160 res.
Anyone...feel free to correct my math if it is wrong.
I can´t understand you,
116160 pixel claimed
220 lines at all (claimed also)
then 116160 pixel/220 =528 RGB strip pixel each line
actually 279 RGB strip pixel measured each line
So 279 x 220=61380 pixel REAL RESOLUTION
116160 pixel claimed
220 lines at all (claimed also)
then 116160 pixel/220 =528 RGB strip pixel each line
actually 279 RGB strip pixel measured each line
So 279 x 220=61380 pixel REAL RESOLUTION
Heya Rox, ill, try to make things abit clearer here, we have whats known as dot resolution and pixel resolution. Dot resolution is the resolution of the lcd's individual dots, ie: each colour counted as a dot. Pixel resolution is 3 combined dots to make up one pixel, one pixel comprises of a red, green and a blue dot.
Example:
We have a lcd that states its 960x234 dot resolution. To find out the pixel resolution all you have to do is this:
960/3 = 320 pixels horizontal
So the lcd's resolution is 320x234
And if we want to find out the pixel count 320x234 = 74880 pixels
To get the dot resolution you just multiply the horizontal pixel count by 3, (reverse the forumula).
So 320x3 = 960
960x234 = 224640 dots
Becarful of some of the lcd's you buy, quite often the resellers quote the dot resolution as the pixel resolution and therfore trick people into thinking they have a higher resolution then they realy have. Its always best to go to the lcd's manufactures websites ( sharp, panasonic, hitachi ect) to see the specs on the lcd used in the unit, they wont lie.
Trev🙂
Example:
We have a lcd that states its 960x234 dot resolution. To find out the pixel resolution all you have to do is this:
960/3 = 320 pixels horizontal
So the lcd's resolution is 320x234
And if we want to find out the pixel count 320x234 = 74880 pixels
To get the dot resolution you just multiply the horizontal pixel count by 3, (reverse the forumula).
So 320x3 = 960
960x234 = 224640 dots
Becarful of some of the lcd's you buy, quite often the resellers quote the dot resolution as the pixel resolution and therfore trick people into thinking they have a higher resolution then they realy have. Its always best to go to the lcd's manufactures websites ( sharp, panasonic, hitachi ect) to see the specs on the lcd used in the unit, they wont lie.
Trev🙂
yes the pixel and the dot is just 3 times the other.
But The only thing i was triyng to say is that they say it is
116160 dot resolution, or 38720 pixel resolution whatever you want to see. But If you check it (i mean if you count them)
you'll find they are 61380 dot or 20460 pixel at all.
check the last photo at the beggining of the post.
I mean, why do they say it is 116160 dot resolution?
they lie or am I doing something wrong?
But The only thing i was triyng to say is that they say it is
116160 dot resolution, or 38720 pixel resolution whatever you want to see. But If you check it (i mean if you count them)
you'll find they are 61380 dot or 20460 pixel at all.
check the last photo at the beggining of the post.
I mean, why do they say it is 116160 dot resolution?
they lie or am I doing something wrong?
I can tell ya 93x220 isnt a reso made, its certainly not a 4:3 ratio. From looking at you pics it's not a stripe colour config, its more of a star wich is common on those small lcds, could be just the cam angle though.
A stripe config is where the colours run up and down the screen in lines, like a red line, a blue line and a green ect.
Trev🙂
A stripe config is where the colours run up and down the screen in lines, like a red line, a blue line and a green ect.
Trev🙂
one thing is how many pixels are there at the screen and other thing is how are they distributed. allthought it could be posible that the pixels are wider than height. Then 93 * 220 could be 3/4 ratio as well.
for instance you said at your example 960 * 234 (it is not 3/4 ratio even) but the size of the pixels are made so total 3/4 ratio is made.
mine is RGB strip, but each line is shifted one pixel to get a litle more difused color display i guess.
for instance you said at your example 960 * 234 (it is not 3/4 ratio even) but the size of the pixels are made so total 3/4 ratio is made.
mine is RGB strip, but each line is shifted one pixel to get a litle more difused color display i guess.
just drive it with a diagonal line
It may be that the color filters each cover two row's LCD dots. You should be able to see that if you drive the LCD panel:
Use Paint or some other drawing program to make a VGA image of a diagonal line in white on black. Then make some more nearby diagonal lines in red, green, and blue. project it or look at it with a strong magnifier and you will see the pixel, dot, stripe, etc. geometry.
Or maybe they are counting interlaced frame rows (like NTSC video), even though the LCD displays them in the same row of dots!
It may be that the color filters each cover two row's LCD dots. You should be able to see that if you drive the LCD panel:
Use Paint or some other drawing program to make a VGA image of a diagonal line in white on black. Then make some more nearby diagonal lines in red, green, and blue. project it or look at it with a strong magnifier and you will see the pixel, dot, stripe, etc. geometry.
Or maybe they are counting interlaced frame rows (like NTSC video), even though the LCD displays them in the same row of dots!
Re: just drive it with a diagonal line
I thought that too but when i multiply 61380(my measured resolution) by 2=122760 dots. It isn´t 116160 claimed dots anyway.
Recently i took it under 600 power microscope and saw clearly a few dots. I can´t see a union at the center of a dot. I mean there should be a darker line i guess. Where i see dark bands is between every diferent color dots. (like 1/4 of dot wide)
Any other idea?
Guy Grotke said:
I thought that too but when i multiply 61380(my measured resolution) by 2=122760 dots. It isn´t 116160 claimed dots anyway.
Recently i took it under 600 power microscope and saw clearly a few dots. I can´t see a union at the center of a dot. I mean there should be a darker line i guess. Where i see dark bands is between every diferent color dots. (like 1/4 of dot wide)
Any other idea?
Hello rox 😀 😀 , que estas en todos lados tiu, juassss.
Por cierto, How much is the lcd? can you post pics with the proyection results? i think that this lcd it,s a good choice to use with an dvd, cheap and small (the proyection screen too).
Un saludo chaval.
Por cierto, How much is the lcd? can you post pics with the proyection results? i think that this lcd it,s a good choice to use with an dvd, cheap and small (the proyection screen too).
Un saludo chaval.
metal is transparent!
The conductors that go to each cell are so thin, they are transparent. You are seeing the boundaries of the color filter layer. Hook up the controller and drive a single red, green, or blue pixel. Then you will be able to see under the microscope what the pixel and dot geometry is. in relation to the color filter layer geometry.
The conductors that go to each cell are so thin, they are transparent. You are seeing the boundaries of the color filter layer. Hook up the controller and drive a single red, green, or blue pixel. Then you will be able to see under the microscope what the pixel and dot geometry is. in relation to the color filter layer geometry.
Re: metal is transparent!
actually i removed the polarizer film and then the color layer.
Would love to drop photos here but don´t know how to take them through the mricoscope. I can see very beautifull images on the microscope. The cell behind the color is same size as color cell. Then i could see there are numbered lines at left (220) and upside (243... go on till 279 i guess, cos i removed color layer at the very corner but not at the edge)
Then i could see the TFT (thin film transistor) that is at the corner of each dot. If there were 2 for each color cell, i would see them, but only 1 is for cell so still don´t understand the resolution measurement.
If somebody knows how to do a homemade videomicroscope or something like this, let me know and i´ll drop them here.
Guy Grotke said:
actually i removed the polarizer film and then the color layer.
Would love to drop photos here but don´t know how to take them through the mricoscope. I can see very beautifull images on the microscope. The cell behind the color is same size as color cell. Then i could see there are numbered lines at left (220) and upside (243... go on till 279 i guess, cos i removed color layer at the very corner but not at the edge)
Then i could see the TFT (thin film transistor) that is at the corner of each dot. If there were 2 for each color cell, i would see them, but only 1 is for cell so still don´t understand the resolution measurement.
If somebody knows how to do a homemade videomicroscope or something like this, let me know and i´ll drop them here.
CCD on the end of a eye veiwer.
There is 1 transistor per dot as its just a switch to turn the dot on and off.
Trev🙂
1 transistor per dot?
You sure about that Trev?
I would think an active matrix would need one transistor to pull up and another transistor to pull down at each cell. LCD cells are essentially tiny capacitors. If one transistor pulled the charge high, there would be no reason for it to go low when the transistor was turned off. That would take either a resistor (slow) or another transistor (fast). You can see how slow it would be with no resistor or second transistor, if you just touch the inputs of a raw passive matrix LCD: Cells go dark just from integrating the potential between one of your hands and the other. When you remove your fingers, it takes about 5 seconds for the dark cells to return to normal.
Also, last time I looked at raw LCD drive specs, you had to drive them in the positive polarity and the negative polarity at the same intensity and for the same length of time, or they would be damaged by physical changes in the liquid crystal material. Most panels have a clock input that has to run at twice the dot clock rate, just to switch the cell drives between positive and negative polarity during each dot's drive period.
You sure about that Trev?
I would think an active matrix would need one transistor to pull up and another transistor to pull down at each cell. LCD cells are essentially tiny capacitors. If one transistor pulled the charge high, there would be no reason for it to go low when the transistor was turned off. That would take either a resistor (slow) or another transistor (fast). You can see how slow it would be with no resistor or second transistor, if you just touch the inputs of a raw passive matrix LCD: Cells go dark just from integrating the potential between one of your hands and the other. When you remove your fingers, it takes about 5 seconds for the dark cells to return to normal.
Also, last time I looked at raw LCD drive specs, you had to drive them in the positive polarity and the negative polarity at the same intensity and for the same length of time, or they would be damaged by physical changes in the liquid crystal material. Most panels have a clock input that has to run at twice the dot clock rate, just to switch the cell drives between positive and negative polarity during each dot's drive period.
Re: 1 transistor per dot?
Very sure, read the pages posted in the links, its one transistor per sub pixel, ie: for each dot. It should also have its own capacitor.
http://electronics.howstuffworks.com/lcd8.htm
http://electronics.howstuffworks.com/lcd9.htm
1024x768x3 = 2359296 subpixels = 2359296 transistors as stated in the article.
Trev🙂
Guy Grotke said:
Very sure, read the pages posted in the links, its one transistor per sub pixel, ie: for each dot. It should also have its own capacitor.
http://electronics.howstuffworks.com/lcd8.htm
http://electronics.howstuffworks.com/lcd9.htm
1024x768x3 = 2359296 subpixels = 2359296 transistors as stated in the article.
Trev🙂
1 Transistor per cell
Right you are.
I was thinking in terms of regular bipolar transistors, but in LCD panels they use MOSFETs. They can conduct current in either direction. The control signal operates the gate, so it connects one end of the MOSFET to the other end. Since one end is connected to the desired drive voltage, the cell capacitor at the other end is charged (or discharged) to the drive voltage. It does not matter if that voltage is positive or negative. When the gate signal is shut off, the capacitor holds the charge.
Right you are.
I was thinking in terms of regular bipolar transistors, but in LCD panels they use MOSFETs. They can conduct current in either direction. The control signal operates the gate, so it connects one end of the MOSFET to the other end. Since one end is connected to the desired drive voltage, the cell capacitor at the other end is charged (or discharged) to the drive voltage. It does not matter if that voltage is positive or negative. When the gate signal is shut off, the capacitor holds the charge.
There is one corner of each dot with some kind of circuity. I guess it is the TFT. There are writings at the wirings also, I guess they are for reference for enginers. They have numbers each row and each column. 220 rows at all (as stated) and 280 columns at all (240 that i can see and other 40 if do extrapolation.
Still don´t understand why they say it is 116160 dot
Still don´t understand why they say it is 116160 dot
real specs vrs marketing specs
116160 = 220 * 264 * 2, so I would guess there are really only 264 columns. And the 2 probably is a bit of marketing hype justified because they assume you will be displaying an interlaced image. (Even though this LCD will show the pairs of interlace rows in a single row.)
There is a lot of such hyperbole in low-end LCD specs. Look at all the LCDs sold as having 3 times as many pixels, because they count each red, green, or blue dot as a pixel.
Instead of just reading the marketing literature, get a copy of the actual data sheet. It should show you the driving waveforms, so you can tell how many dots per row, etc. Then fire it up and verify it under the microscope.
116160 = 220 * 264 * 2, so I would guess there are really only 264 columns. And the 2 probably is a bit of marketing hype justified because they assume you will be displaying an interlaced image. (Even though this LCD will show the pairs of interlace rows in a single row.)
There is a lot of such hyperbole in low-end LCD specs. Look at all the LCDs sold as having 3 times as many pixels, because they count each red, green, or blue dot as a pixel.
Instead of just reading the marketing literature, get a copy of the actual data sheet. It should show you the driving waveforms, so you can tell how many dots per row, etc. Then fire it up and verify it under the microscope.
if there they were 264 dots per line, it would have sense. But i´m sure there are 279 dots per line. So try it once again.
Anyway, i don´t see what does it to do the refresh rate and the resolution. If i doble the refresh, then i could say it has doble resolution?
That way, i can use 3 leds and refresh them at a very hi rate (1Mhz, then i have 3 Megapixel display!!!! sounds stupit but still can´t understand why manufacturers say that. I´m getting mad with this....
Anyway, i don´t see what does it to do the refresh rate and the resolution. If i doble the refresh, then i could say it has doble resolution?
That way, i can use 3 leds and refresh them at a very hi rate (1Mhz, then i have 3 Megapixel display!!!! sounds stupit but still can´t understand why manufacturers say that. I´m getting mad with this....
data sheet will tell all
Okay, who is the LCD manufacturer and what is the exact part number? Let's find an engineering data sheet for it.
I've looked at a lot of LCD data sheets: They can be tricky. Sometimes you have to look at the drive waveforms to figure out the actual geometry. Sometimes you have to divide the active display area by the individual dot size to figure it out.
Okay, who is the LCD manufacturer and what is the exact part number? Let's find an engineering data sheet for it.
I've looked at a lot of LCD data sheets: They can be tricky. Sometimes you have to look at the drive waveforms to figure out the actual geometry. Sometimes you have to divide the active display area by the individual dot size to figure it out.
- Status
- Not open for further replies.