On Thu, 14 May 2009 15:05:17 -0700 (PDT), -jg <[email protected]> wrote:

>On May 14, 9:20*pm, Mike Harrison <m...@whitewing.co.uk> wrote:
>> What I'm looking at is the cheapest way to drive a large number (potentially hundreds) of
>> distributed TFT (PSP) displays with local SDRAM and/or NAND flash storage for a few tens to hundreds
>> of frames, with a relatively a low bandwidth network of some sort to update content in non-realtime,
>> and switch the display between stored frames realtime. The aim is to minimise the cost per node as
>> much as possible.
>
>Ah, so you need many slaves, and the smarts can be elsewhere.
>
>I found some info on TFT(PSP), [9MHz, 8+8+8 Colour] and it seems the
>simplest solution would be a 32 bit wide SDRAM
>and a CPLD, and 128MC should be plenty.
> - I don't see you need a line-ram in a SDRAM playback scheme (tho the
>Qimonda SDRAM data I had was poor ).

One possible solution I considered was SDRAM using page-reads, one per displayed line. However the
page size means it would need 2x 8-wide SDRAMS as the x16 device page size on a 64M SDRAM is only
256 bytes ( I've decided that the difference between 16 bit 5:6:5 and 24 bit display is negligible -
content can be pre-dithered if necessary).
However although this would work OK for reads, writes to SDRAM would be limited to the vertical
blanking interval (12 lines), so the sending of data would need to be synced to this, and bandwidth
would be limited - without a local buffer you'd need high peak bandwidth which would be idle most of
the time.
Having a couple of 512-pixel buffers (one for read, one for write) means data can be read and
written much more freely, as you can do a page read from SDRAM to the buffer at, say, twice the
pixel clock, leaving 50% of RAM bandwidth available for writes, and no sync constraints on write
data - the only constraint is the rate of data transmission, which is easily regulated. It would
also remove the 1 line/1 page constraint so RAM use could be more efficient, although wasting 32 out
of 512 words may be worthwhile to simplify the control logic.
Given a suitably fast network link, updates at full-frame rate would be possible, and the whole
system becomes a lot more flexible.