Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Graphical LCD dilemma

Status
Not open for further replies.
Does any one know of any TFT color LCD module with an onboard controller. lcd size preferred would be 100 x 90 mm. information regarding cost be be helpful. links to such product is also welcome
 
You will not be able to store enough data in an 8 bit to do it. LCDs of that caliber do not have on board controllers of the sort you need to do what you are asking since it is to expensive to do it with out restricting the application of the module to much.

You really have not thought it through. At 8 bit QVGA you need to send out 80K of data to define what is on the screen. The controller then keeps all that in local memory to refresh the display at 60-100Hz.

At 640x480x24 bits you are talking about close to a MB sent to the display memory for each image. In other words you would need a 20 bit address bus to talk to the display memory. So if you use a controller with enough memory to keep 2 display planes you end up talking 0.1 seconds to send the data for the next static screen to the display memory using a 10MHz bus speed. That does not include the processing time to generate the image on the fly since you are not going to hook up multiple megabytes of memory to an 8 bitter.

I would recommend backing off to a monochrome module with the controller on board. You do not seem to have the capability to handle 160 pin ICs with pins 0.5mm apart that are LCD controllers.
 
Thanks for the reply, ubergeek... right now i've settled for the 320 x 240 TFT LCD, and i'm not going for the 640 x 480...

As for the controller.. i'm going for an 8 bit controller (as i'll be using the LCD in 8 bit mode), and yes ram is around 80K per screen. However, calculations reveal that using this controller will give me enough juice to only read from RAM and write to the screen at the desired refresh rate.. Could any one answer these points for me :

1. will a 2 (8 bit) controller approach do the job satisfactorily... here one will refresh the screen (60Hz) and the other will do the on-the-fly computations communicate with the outside world and edit the RAM.

2. Going for one controller(8bit) and using a DMA sort of circuit that will generate timimng and refresh the RAM, whereas the controller will handle the other tasks(as mentioned above)
 
Just a heads up. LCD displays must be driven with a psudeo square wave or else it's possible to damage them. Driving LCD glass is not as simple as LED.
 
If you really want to do this in one IC and the ARMs are too much, think about the Propeller CPU. It has eight 20mip CUPs in one package. The RAM is short but that brings up the second subject. I do not know what type of graphics you need. Normally there is a RAM location for every pixel. If you graphics is simple meters and simple text you can build the image and send it at video speed without RAM. There is an application note on the propeller where they make video (text) without storing the graphics in RAM.

Blueroomelectronics. You are thinking about very small LCDs where you drive the panel direct. The ¼ VGA we use do not require a square wave.
 
The propeller ics look good, i'll just take a look at them. As far as the graphix are concerned, for the initial stages my graphix are going to be real simple, a simple rectangle and line, moving in different directions. Text will also be displayed for debugging sake(and for other info).... the tougher graphix if required will come only later on.. (once we are well settled with the concept).
 
Which is better to be used as a graphical controller the s1d1305 (epson) or the ssd1905 (solomon). I will be using a 320 x 240 TFT LCD, and i think i'll be running it in an 8 bit mode.

Then again, in the future if i have to go for more, i.e. 16 bpp then, what are my options

Thanx
 
Which is better to be used as a graphical controller the s1d1305 (epson) or the ssd1905 (solomon). I will be using a 320 x 240 TFT LCD, and i think i'll be running it in an 8 bit mode.

Then again, in the future if i have to go for more, i.e. 16 bpp then, what are my options

Thanx
The s1D13705 is maxed out at 8 bit QVGA with 80K, though with enough horsepower you can get some animation with it (we do so with it and an ARM).

The s1d13A04 has 160K of ram, which makes things easier since you have a display page and a build page or if you do not need a whole page refresh it has a bitblit engine to help you out.

After that, from Epson, you need to jump up the complexity to external memory devices.
 
I was planning to use an ARM9 controller in a system that will be gathering info from a set of sensors processing the data an displaying this info onto a graphical TFT LCD 320(RGB) x 240.

Should i burdern this controller with all this or should i shift the LCD workload onto a dedicated graphical controllers like the epson, etc. I know that the LCD refreshing is resource hungry, so can all this be done using a single arm9 controller. I am considering the AT91SAM9263 or the AT91SAM9261 (monster) controller
 
NO!!! We have tried to be gentle with you and failed miserably. You can not use a chip that does not explicitly say it has a built in LCD controller, and even then you have displayed the complete inability to get it working!

And yes I DID look. I was thinking you, had you not been a completely imbecilic noob, might have gotten away with it. The problem was that the SSC in that chip has a 1024 bit frame limit. In order to have a chance you need to be able to use DMA to at the very least do an unassisted horizontal scan line, which at 3204x240x8 is 2560 bits.

Now since all of the above was above you, I can only hope that the information might just help someone else here that actually has a chance of implementing something!
 
Last edited:
My opinion is the same as übergeek63. What are you trying to build anyway?
he already told us... he wants us to do his job for him-to get the graphics on a custom industrial controller working ... sounds to me a lot like one of the "lie for the job" types bit off more than he could chew and now is panicking.
 
Last edited:
It's sounds more like an epic fail school project. It's like those fellows who want massive 640x480 LED displays yet have zero clue about refresh rate, power requirements and so on. This project is much more complex and we might be able to offer a better solution if only the OP would say what it's for. (Yes its for a car dash but what does it do?)
How does he even plan to wire up an ARM to the display? Good luck on ARM hobby forums with code examples.
 
Last edited:
It's sounds more like an epic fail school project. It's like those fellows who want massive 640x480 LED displays yet have zero clue about refresh rate, power requirements and so on. This project is much more complex and we might be able to offer a better solution if only the OP would say what it's for. (Yes its for a car dash but what does it do?)
How does he even plan to wire up an ARM to the display? Good luck on ARM hobby forums with code examples.
so I see... in one post he says it's his "work project" and in the very next post he says it is for a car dash.

So what do you think then, auto maker's VP of personnel's son fresh out of a electronics certificate mill?
 
Last edited:
Who can say, he's not said what it's for.

I recall one poster said that in some countries rich kids whose grades were too low for doctors or lawyers were sent off to electronics school :(

The mystery continues.
 
Thanks for the sarcasm guys. I was just trying to look at aaaaaalllllll the options before taking the plunge. The thing is, people like you are very hard to find (people who know to get a "color LCD" working, and the ones i find here don't actually want to tell much.

Anyways, i have mentioned what the project is about in one of the threads. It was originally using an 8 bit controller (a very bad one too), so now i have to "upgrade" it to the one using an LCD.

Btw, fyi i'll be using an solomon ssd1906 or ssd1926 to refresh the LCD and as for the MCU, i've not exactly made up my mind, it's either an infineon(xc16xx) or a freescale (mc9s12x) 16 bit controllers. freescale has a lot of info and app notes on their site and their tech support is also a good one. Any more advice (or 'sacarsm') ....

My only problem is the IDE, i've been using the KEIL uV3 but there are upgrade issues over it. Any views over other free IDEs (like eclipse, etc.) that have a good c compiler out there...

ThanX
 
You still haven't told us what it does. Is it a TV? Message center? GPS? Why not use a small VGA monitor? Why do you want to attempt to interface it yourself?

That solomon ssd1906 is a very complex IC, how are you going to test it? Do you plan on doing PCB artwork before a prototype?
 
Last edited:
It's an automotive project, We use a set of sensors to "map" the surroundings while on the move or steady. Using this info from the sensors we process the coordinates of the vehicle. These coordinates are obtained using the 'speed' along with 'steering angle' of the vehicle. This is to be further processed and displayed on the screen as objects (preferrably rectangles for the vehicle) and some text. Currently we are using an 8 bit controller and the a single scan of the sensor + processing takes 30 - 40 ms. earlier we were using a set of LEDs and 7seg to display this information. Now we require an LCD. Using the current 8 bit controller is not possible. The system will enable a person to view an obstacle around his vehicle. FYI it's an isolated system, so there is no other inputs comming from outside.

The selected LCD is 5.7in 320x240 TFT-G320240DTSW-50W-E (i have posted the datasheet is some previous thread, i guess). I will be using this LCD in 16 bit mode.


Michael Rodrigues
 
Thanks for the sarcasm guys. I was just trying to look at aaaaaalllllll the options before taking the plunge. The thing is, people like you are very hard to find (people who know to get a "color LCD" working, and the ones i find here don't actually want to tell much.

Anyways, i have mentioned what the project is about in one of the threads. It was originally using an 8 bit controller (a very bad one too), so now i have to "upgrade" it to the one using an LCD.
The sarcasm comes from YEARS of hearing school children (college) and wannabe engineers asking for us to do their work for them, it is not our fault that the caliber of your questions hits a nerve made raw by them. Particularly when you, like them, do not seem to take a hint!

There are a few facts involved here:

1. It is impossible to get the kind of performance you obviously want to see without dedicated graphics memory and controller.

2. The controllers only come in 0.5mm lead spacing or BGAs with over 100 connections.

3. I am a design engineer with a full production line at my beck and call and can not work with BGAs since they require xray verification and border on impossible to rework if something is wrong.

4. To even consider doing yourself what you are proposing stinks of wishful thinking of a production product, and to use us to do all your work for you is despicable.

5. If it is indeed a college project, you do not have the capability to mount those parts and should not be considering them anyhow, in which case you should just get a damned SBC for $300-$500.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top