Content originally posted in LPCWare by tvink on Mon Feb 23 08:18:35 MST 2015
Hi,
I've been away from this for a while and what I think I understood had to be relearned today.
Anyway would you review the attached JPEG to sanity check my SDRAM connection to the LPC1837?
I am using a single 16 bit SDRAM device and am patterning the connection after AN11508, figure 6. ( Best performance using a single device and CCLK Div2 )
I am booting from internal flash so I do not have any of the boot option resistors connected.
I have the two EMC_DQMOUT data mask lines connected to the SDRAM upper and lower byte mask pins.
I am treating the SDRAM bank selects as address lines. ( BA0 and BA1 on the SDRAM )
I am assuming that address lines EMC_A12 and EMC_A13 are not ROW / COL muxed and can be used as I have shown.
I chose EMC_DYCS0 as my chip select and am using EMC_CKEOUT0 to go with it.
I am using CLK1 to drive the SDRAM and am leaving CLK0, CLK2 and CLK23 unconnected. ( Guess I could use CLK23 for another pin function but there isn't an alternate that I need )
Some more neurons are firing but I am not there yet! So I am updating this posting with some additional questions....
I am trying to figure out how the banking works and how interleaving is programmed. The app note uses EMC_A14:13 for bank select ( table 1 ) but I think that would leave holes in my memory map. ( I am using a 64Mb "16 bits x 4 banks" device ) I am also reading the users manual UM10430 chapter 21. Table 364 and 365 look like they have the answers to my banking questions but I am not sure how to read the tables. Table 365 has two mappings for my 64Mb device. One is for Roe, Bank, column and the other is for Bank, Row, Column. To the left there are 4 groupings of addresses. Dont know how to read this. Do they map to the three items ROW, COL, Bank? but how to map 4 things to 3? And there is no A13 in this mapping. So I think my answers are here but need some additional help figuring it out.
Thanks
Did I misunderstand anything?
Thanks,
Tony
Edit: I had the byte masks reversed so I uploaded an updated JPEG