TFT LCD controller

I need some idea to realize a structure like this: MCU LCD CONTROLLER TFT LCD

I'm using a 32 bit MCU (ARM7 or Cortex-M3), supposed to drive external FLASH and SRAM. The TFT LCD has to be a QVGA-sized panel with no internal dedicated controller, has the company does not want to be be tied to a specific display model. We have this model to try with:

formatting link

We have discarded the NXP2378 MCU and (obviuosly) old discontinued Sharp models. The graphic wuold be static: some 16bpp images and icons, with no smooth animations (max 2-3 frames per second): the company doesn't want to move to an ARM9 controller (or similar). So the choice will be as described above.

Thanks

Reply to
news.tin.it
Loading thread data ...

news.tin.it scriveva il 07/08/2009 :

Sorry for the typo... it's the NXP2478 MCU.

Reply to
news.tin.it

Have you considered an FPGA solution?

Reply to
larwe

Why? It's probably the best fit for what you want to do

Why not? I think NXP have a pretty cheap ARM9 with TFT controller

Reply to
Mike Harrison

Mike Harrison ha pensato forte :

I know, but the company has got some serious EMC troubles with NCP2xxx family.

No ARM9 because of MCU packaging: they don't want BGA on those boards.

Reply to
news.tin.it

Yes, but we're still looking at it. Any idea?

Thanks

Reply to
news.tin.it

formatting link

In having just been evaluating all of this sort of stuff, I think you'll find that most of the ARM7s and all of the Cortexes will really let you down on pin count. If you're talking about an SRAM, a parallel flash, and a QVGA, that's a whole lot of pins. I found few Cortexes with more than 100 pins, and none with more than 144.

--
Rob Gaddi, Highland Technology
Email address is currently out of order
Reply to
Rob Gaddi

Nel suo scritto precedente, Rob Gaddi ha sostenuto :

Hi, we have recently used a STM32F103ZE Cortex-M3, 144-pin with external static memory interface (look at STM32 eval board for reference design, usign NOR/NAND FLASH and SRAM).

But the problem is the LCD controller... we need something that's going to be available for a few years (3 to 5) and most controllers are too specific to be managed.

Reply to
news.tin.it

Why not glue an LCD controller produced by one of these companies to the EBI? Chances are that the LCD controller will do what you want, and there is always technical support available to help you interface with your processor.

Epson Research:

formatting link

STMicroelectronics:

formatting link

Reply to
Nicholas Kinar

controller/lcd/...

While Epson show numerous QVGA capable controllers, STM have none (just VGA -> LCD convertors). Siliconmotion are the only ones to my knowledge to make somewhat larger controllers since the b690x0 by Chips (intel, asiliant) were slaughtered.

formatting link
. Mind you, the old b69030 can do SXGA on an LCD and on an analog output with independent sync rates - all this at < 1W, framebuffer memory (4Mbytes) included on chip.

Dimiter

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

formatting link

------------------------------------------------------

formatting link

Reply to
Dimiter Popoff

Without a controller, you need to bring out 500+ pins to the MCU. Perhaps you can try a programmable controller with your own interface spec?

Reply to
linnix

Well, what are you asking? No I don't have a ready-rolled design in my back pocket, though likely there is something at opencores that will serve.

The reason I suggest it is because a simple LCD driver is not a complex project, and if you do it "soft" then you'll be insulated as far as possible against parts availability issues.

Reply to
larwe

Can you get FPGA in COG die? Otherwise, 500+ wires between PCB and LCD would be very expensive, in addition to the expensive FPGA.

Reply to
linnix

We're evaluating Lattice FPGA, can you give me some links please?

Thanks

Reply to
news.tin.it

with:

formatting link

You and the OP have different concepts of controller. The panel linked above has a 40 pin connector. It's a pretty typical 6 bit/color interface. 500+ pins would be a bit of overkill.

Bob

Reply to
Bob

Scriveva Bob martedì, 11/08/2009:

I think so. I don't need to tie 500+ pins, as that display has a sort of digital interface, hiding all that transistors driving stuff.

The problem is that I can't buy an "intelligent display", one with a smart digital interface (I've seen some driven by SPI too): I need a general purpose display with that kind of timed interface.

Thanks

Reply to
news.tin.it

above

0+

But you can do a QVGA controller in a CPLD+RAM, you only need 153600 bytes. A 128 cell or sort of coolrunner + an SRAM will be plenty. A mostrous FPGA for that would be way beyond overkill, sounds more like a madness to me.

------------------------------------------------------ Dimiter Popoff Transgalactic Instruments

formatting link

------------------------------------------------------

formatting link

Reply to
Dimiter Popoff

Sembra che Dimiter Popoff abbia detto :

Hi Dimiter, we have some external consultant that could do FPGA development on Lattice FPGA, as we don't want to begin that kind of work internally. I've thougth even to a Freescale X-Gate solution, but it would be nice only if the company used Freescale MCUs: that's not our case.

Thanks

--
http://www.grgmeda.it
Reply to
news.tin.it

in a CPLD+RAM, you only need 153600

What I am saying is that an FPGA is insanely huge for a QVGA display. The coolrunner is a Xilinx (originally Philips) CPLD and a QVGA controller should fit comfortably inside. It will take

Reply to
Dimiter Popoff

The SM502 is really nice, once you've fixed some bugs in the sample code they provide, e.g. in the line drawing function. Currently I'm using this chip in a project. You can get it with integrated RAM and it has fast 2D graphics acceleration functions. But of course it is BGA, so the OP could not use it. I think the "no BGA" requirement is a big problem, because most modern chips with fast LCD controllers are BGA.

Regarding CPLD for a LCD controller: Maybe it is possible to fit something which scans a framebuffer in 64 cells of a Coolrunner, but graphics update would be really slow, because I think 2D graphics acceleration are impossible with 64 cells. But this depends on the application of the OP. A simple status monitor, where it doesn't matter if the update of the whole graphics contents needs some 100 ms, should be possible.

Someone in de.sci.electronics has developed a nice new DIL board with a Xilinx FPGA. I'll design a simple display controller with 2D accleration function, with SPI interface for small microcontrollers, when I got the free sample. This is the idea of my design:

formatting link

(the automatic Google translation to English of my posting has not too many flaws:

formatting link
)

--
Frank Buss, fb@frank-buss.de
http://www.frank-buss.de, http://www.it4-systems.de
Reply to
Frank Buss

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.