Mixing addres lines of SDRAM.

ear all,

I have to boot-up a device based on Hitachi uP with external SDRAM
(2x256Mb). The device was designed by somebody else and I only have to
get it working. However there is something that bothers me very much.
Probably to easen PCB layout the designer has mixed data and address
lines for SDRAM: i.e. pin D0 of SDRAM is connected to pin D7 of uP. All
other interface signals (RAS#,CAS#,CLK,CKE,UDQM,LDQM,WE#,CE#) are
connected correctly. While mixing data lines seems ok to me (let me know
if I'm wrong) mixing address lines looks unacceptable.
Here are the connections:

SDRAM uP
------------------
A0 A5
A1 A4
A2 A3
A3 A2
A4 A14
A5 A13
A6 A12
A7 A11
A8 A10
A9 A9
A10 A6
A11 A8
A12 A7

BA0 A15 (bank select)
BA1 A16 (bank select)

Will this connection work? For the time being I know that I must forget
burst transfers. Are there any obstacles for SDRAM in this configuration
to work with 1-byte burst transfers?

Thanks
Wlad

 

Hello

If this were SRAM then there will be no problem with the random connections
as long as address is connected to address and data to data.

If think that mixing up the data and address lines is a bad thing. I don't
know much about SDRAM and therefore I read a data sheet about SDRAM.
It looks like you only will be able to get one address at the time when the
address lines are mixed.

Mixing up the signals is a bad thing because if there were more items
connected to the memory bus then it could give a lot of problems with the
exact locations of the data structures.

Could you tell which uP and which SDRAM chips there are used?

If the uP can make burst transfer I will have asked the designer to redesign
the circuit board because it's a bad design.

Cheers
Rune

 

While it might work, I'd be concerned about refresh...

 

Even if it was the boss?

 

Actually, as has been discussed several times here, address lines don't
have to be in any particular order either. Now if you mixed data with
address lines...


Bob

 

In the case of an SDRAM I beg to differ. The address lines are used to
issue commands to the chip, and they aren't just randomly treated as
addresses (at least on the one that I checked). Specifically the one I
looked at assigns a special meaning to A10 which has nothing to do with
where you want to read memory. See
http://download.micron.com/pdf/datasheets/dram/sdram/512MbSDRAM.pdf.

And even if it does basically work, as the OP pointed out it totally
screws up burst reads, which on an SDRAM slows you down to a terrific
degree (a factor of 8 or 10, IIRC).

 

Oops! That's what I get for not reading the OP carefully.

Bob

 

the hardware can not work! fire that somebody else if you can!
addr bus *must* be connected properly to get the SDRAM to work.

Antti

 

I have to boot-up a device based on Hitachi uP with external SDRAM

You have to read the fine print in the data sheets - both the
memory chip and the CPU.

Modern SDRAMs usually need some magic initialization commands.
The "data" for that command is typically encoded on the address
lines. Usually there is some magic firmware that does that.
You may have to modify that - that is use a different address.

I'd expect a CPU to read a whole cache block in burst mode.
If you can't get it to avoid that you are probably out of luck.

 

There is nothing wrong with mixing either type of lines. All it means is
that addresses will not physically be at the expected addresses in RAM. It
is more difficult to justify on a ROM, where programming in an external
programmer will become 'hard'...
Your address lines as listed, show A0, and A1 are not present from the
micro. This means that four consecutive addresses, will all refer to the
same memory cell. I'd assume from this, that there is actually a 4byte
'latch'/multiplexor, and this is being used to select the individual
'bytes' for the processor.

Best Wishes