Additional config info for Camintonn CMV-250, CMV-500, CMV-254, CMV-504
jnc at mercury.lcs.mit.edu
Thu Sep 17 11:34:05 CDT 2015
So I have a couple of these Camintonn boards (a -500 and a -254, to be
exact), both using 256Kx1 DRAM's. I wanted to upgrade them both (by adding
memory chips) to be -504's (2MB), and I noticed that the -254 had a couple of
jumper wires that the -500 did not, so I needed to know what those jumpers
did. I looked online, and although there is a little bit of info, it doesn't
cover those jumpers.
My first thought was that they might be timing-related; one board used -12
parts, the other -15. However, after some poking around, I think (with 98%
certainty, although I haven't traced etches to be 100.000% certain) that they
actually allow the boards to be used with both 64Kx1 and 256Kx1 memory parts.
I hereby offer up all the details in case anyone's interested:
I found a document which described them as "Starting and ending address
boundary" (alas, without giving any detail, but which confirmed they aren't
timing-related). The clincher as to their function was the capacities of the
various board versions:
CMV-504 2 MB Memory Module
CMV-254 1 MB Memory Module
CMV-500 512 KB Memory Module
CMV-250 256 KB Memory Module
How do you get a 256KB board using 256K devices on a memory board that has to
produce 16-bit wide words? Clearly, the board was first produced with 64Kx1
chips, and so it likely (like the similar NS23C) that it can be configured to
use either 64Kx1 or 256Kx1 chips.
Here are the details of how to do that: down near the fingers, there are a
block of 6 solder pads, denoted thus:
(Note that there is _another_ 'S' on the board, at the top.)
On board #1, the CMV-254, it has jumpers on M-N, P-R (apparently the
configuration for use of 256K chips), and etch cuts on R-S, N-O (likely the
configuration for 64K chips). On board #2, the CMV-500, it has a slightly
different PCB (likely a later rev), and has no jumpers, and has etch
connections M-N, P-R (note - the same as the #1 board has jumpers).
Hope this is useful to someone!
Note that the board was normally sold with 1, 2, or 4 rows of chips.
(Interestingly, there must be two ways to produce a CMV-500 - 1 row of 256K
parts, or 4 rows of 64K parts. I've never seen one of the latter, but would be
interested to know if anyone has.)
I plan to verify that the board actually works OK with _3_ rows of chips (i.e.
as a 1.5M board), with the appropriate settings - will update when I try that.
More information about the cctalk