3 Dec
2005
3 Dec
'05
9:39 p.m.
The pinout between the two chips is slightly
different, thanks to the
fact that the 2532 has only one enable line and the 2732 has two. I've
come up with an idea for conversion, but I'd really like to get a sanity
check on it, preferably from someone with an original TI data book that
covers the 2532.
Here's what I've got so far, based on comparing the pinouts of the
two devices.
(1) Move the trace originally connected to pin 18 (A11) on the 2532
up to pin 21, which appears to be A11 on the 2732.
(2) Tie pin 18 to pin 20 so that, on the replacement 2732, both /CE
and /OE are asserted simultaneously.
Yes... That should work. Alternatives for (2) are to ground one of the
enable lines and connect the other one to the enable line from your
address decoder. Watch out, though, that many EPROMs take a lot more
supply current when enavbled (CE/ low) than when they're disabled, so if
the PSU is marginal-ish, you might have problems if the EPROM is held
enabled all the time. Linking OE/ and CE/ together avoids this, but note
that the CE/ assetyed - to - data - valid time is often longer than the
OE/ asserted - to - data - vaolid time, which might be a problem in
timing-critical systems.
I've done this sort of thing many times, normally to replace a masked ROM
with an EPROM containing a modified version of the firmware. What I
normally do is put a turned pin socket in place of the EPROM on the PCB.
Then make up a little daughterboard using stripboard, another socket, and
some turned pin header pins. Cut and jumper as appropriate. Of course you
put the socket sltightly offset from the pins (but not totally outside
them) so as to be able to use the stripboard tracks for the signals you
don't want to move around.
This way I can easily go back to the original set-up if I want to.
-tony