LPC1850 / Touch-TFT / SPIFI Flash, Assessment

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

LPC1850 / Touch-TFT / SPIFI Flash, Assessment

281 Views
lpcware
NXP Employee
NXP Employee
Content originally posted in LPCWare by kai.becker on Tue Feb 10 07:39:40 MST 2015
Hi all,

this is my first post in this forum. I hope I don´t break any forum rules, by
asking this without first thoroughly checking if a similar question has
been posted before.
I have done a search, but I haven´t found something comparable.

We want to develop a control unit for our devices. The control
unit should consist of a 4.3'' 16bit TFT Display with a
resisitve touch screeen.
Resisitve touch, because of the costs and because of the
fact, that our devices are always close to water.

So I´m looking for an assessment from people who have experience
using the LPC1850 processor with the following or similar peripherals.

LPC1850 @180MHz
parallel 512KB SRAM (no SDRAM)
Quad SPI Flash (Spansion S25FL164K, 108 MHz)
RGB 16bit TFT with 480x272 pixels
resistive touchscreen
UART communication every few seconds with 38400Baud.

My question is:
Does anyone "know", if the above configuration
will be responsive enough, so that a user will not see spurious
or even contious "lags" when using the control unit?


I mean is the Quad SPI Flash and Cortex-M3 fast enough
to satisfyingly solve this task?
The resisitve touch must also be tightly polled, because
a slider functionality should be realized.

We´re also thinking if it is better to use cheaper and bigger
SDRAM and to load the code from SPIFI Flash to RAM to
retrive a faster execution speed.

We have a done similar configuration with a
"computer on module" solution. The module had a
ARM9 (Toshiba TMPA900 @192MHz)
parallel NAND Flash
parallel SDRAM
and we connected a RGB 16bit TFT with 480x272 pixels (no touch)

This was fast enough for the control unit, but I think the ARM9
is considerably stronger than the Cortex-M3.
Also the code was loaded from Flash into SDRAM and
executed from SDRAM.

I would be very thankful if someone could post a plausible
assessment to the mentioned configuration.

Many many thanks in advantage to everyone who takes the
time to read my post.

Best regards
Kai
Labels (1)
0 Kudos
2 Replies

249 Views
lpcware
NXP Employee
NXP Employee
Content originally posted in LPCWare by kai.becker on Tue Feb 17 00:36:37 MST 2015
Many many thanks for your very helpful answer!

This truly makes me confident, that our chosen hardware
setup will serve us well.

My good impression of the NXP people and the NXP
hardware is confirmed again.

Thank you!

Best regards
Kai
0 Kudos

249 Views
lpcware
NXP Employee
NXP Employee
Content originally posted in LPCWare by bavarian on Thu Feb 12 08:20:59 MST 2015
Hello Kai,

here are some considerations:

[list]
  [*]  Summary right at the beginning:  yes, the LPC1800 is powerful enough for this application
  [*]  The SPIFI has an internal accelerator (small cache) which helps to speed up things. Depending on the code you end up with 35 - 50% execution performance compared to internal RAM
  [*]  Replacing the SRAM with SDRAM is OK, only a pretty fast SRAM would give some advantage, but that's expensive.
  [*]  Execution from SDRAM increases the performance compared to execution from SPIFI, but this depends on the refresh rate of the TFT. If the LCD DMA shuffles more data from SDRAM into the TFThte periods for code fetching logically decrease. Consider to execute from internal SRAM for heavily used parts of the code.
  [*]  The ARM9 has an internal cache, therefore the execution from SDRAM was pretty good, in the LPC1800 the SDRAM interface fetches always 128 bits in one read cycle, that it.
  [*]  With this display size you can easily achieve good sliding effects, our partner Draupner has done a good demo with their graphic library (see link below, first video, please ignore the demos they have done with the STM32  8-)  )
[/list]

http://touchgfx.com/product-details/demos/

Regards,
NXP Support Team

http://touchgfx.com/product-details/demos/
0 Kudos