On Sat, 4 Apr 2009 18:33:22 -0700 (PDT), "
put finger to keyboard and composed:

I'm restoring a mistreated DX7 to working condition.

Re this schematic:
http://www.bobdbob.com/~deneb/ftp/dxmemory.png

The battery is a lithium coin cell, like a CR2032. D4 on my board has
been replaced with a gigantic power rectifier. I'd like to re-replace
it with a part more appropriate for the application (or at least that
fits through the holes on the circuit board).

/IC is an "initial clear" signal that comes from the power supply.
From staring at the schematic for a few hours, it looks like during
power-on, the RAM is kept powered by the battery, and then when the
power has stabilized, Tr1 switches the 5V rail to the RAM (stopping
current from the battery via D4). Also, during power-on, it looks like
Tr3 prevents the /CE2 line from asserting (there are four RAM IC's,
and /CE2 doesn't go anywhere else on the schematic.) I guess this
prevents garbage from getting written if /WE wobbles during power-on?

So what I'm wondering is, what kind of diode can I replace this with?
Do I want to go with a Ge diode to keep the available voltage at a
maximum, or doesn't it matter? Is current an issue? I wish I had
another DX7 to examine but I don't. I can't find a datasheet for the
M5M5118P so I don't know what kind of Vdd it wants.

http://www.datasheetarchive.com/pdf-...DSAP002745.pdf

The operating voltage is 5V +/- 0.5V.

"Data can be held with 2V supply voltage".

It seems to me that a general purpose silicon diode should suffice.
The current draw of each RAM in standby mode is only 15uA.

Thanks as always..

(BTW the schematic came from this page:
http://www.maths.abdn.ac.uk/~bensond...ml#diagnostics )

- Franc Zabkar
--
Please remove one 'i' from my address when replying by email.