Korg M1 SRAM to FRAM upgrade, battery replacement no longer needed

[Chipguy]

Hi !

The obviously dead Zero Power RAM on Dave's Prema 6047 gave me the idea to upgrade some of my stuff with FRAM instead of replacing the battery. First on was my Korg M1 Synthesizer, due for battery replacement. Not actually a piece of EE related gear but they did use the same technology as testgear used back then, either those MOSTEK or DALLAS zero power RAMs or just battery backup circuits. Mostly the solution with a replaceable battery.

Since my M1 demands a battery change every 4-5 years I though that's a good idea.
Beware, when your stuff (like an old Tek DSO from the 90s) has also a RTC built in, a FRAM upgrade not a good idea to do, because you would like the internal clock still working.

So I took out the two 32Kx8 SRAMs and replaced them with FRAM.
However, FRAMs need the /CS signal to go low for each cycle. It's used to latch the address into an internal register.
FRAMs are not transparent like SRAMs are. So I needed a circuit which makes the FRAM compatible to SRAM.
In the Korg M1 for instance the /CS signal is actually the RESET signal. Once the system is up and running only /RD and /WE are used to access the

SRAM. That means the /CS signal to the chip must be high whenever both /RD and /WE are high and whenever the incoming /CS (RESET) is high.
That's easy to achieve with 2 single gate logic gates.



Since SMD is involved and the ciruit needs to fitt into a 28 pin DIP socket I did layout the board and got a bunch of the made by Itead.
I basically desoldered the old SRAMS, soldered nice machined DIL sockets into the board and plugged the little modules into them.



They work great, but now the Synth came up with the "Low Battery" message everytime.
So I thought: "Ok, 10K to +5V should do". Well they do well but that japanese software designer from the end 80s was very aware that Lithium cells can't do 4.5V. So the display now showed "Battery error".

Ok, so after bit playing around with Ohm's law I soldered a 22K resistor in parallel to the battery.
The the voltage then is about 3.1V which is in the range of a new Lihium cell.



Yay, no battery error messages anymore. So it works
And yes, I have tested it by turning it on and off. And of course I reprogrammed it with the factory settings using a MIDI SysEx tool.

The point is: Since this works great it should work for the majority off every equipment using battery backed calibration data etc.

Here is the gallery:

[cloudscapes]

Great job! I love hacks like these.

Makes me want to upgrade my juno 106's battery backed SRAM as well.

[ovnr]

Very nice hack, but in the future could you please use an image host which isn't completely horrifically bad? Giant in-page pop-ups and shit is not nice. Try something like imgur.com.

[Chipguy]

Very nice hack, but in the future could you please use an image host which isn't completely horrifically bad? Giant in-page pop-ups and shit is not nice. Try something like imgur.com.

Oops, since I use a lot of adblocking stuff nothing pops up for me. Gonna deactivate that temporarily and see what the problem is. I also realized that one picture link seems broken. I hope I can edit the original article and fix all these issues.

Cheers,
Chipguy

[DavidDLC]

This deserves to be on Hackaday

mmm, what do you need to do to have your post on Hackaday ? I have no idea

David.

[andersm]

mmm, what do you need to do to have your post on Hackaday ? I have no idea

An Arduino.

[Trigger58]

Hello Chipguy,

i read about your great Mod for the Korg1.

I have the same problem with an 3DO Console. It use the same SRAM-Logic with reset signals. i tried to change the sram with fram but it ends in a permanent reset circle.

Here is the shematic of the savesystem.

[ Specified attachment is not available ]

Do you have an idea how i could use fram with the 3DO?

best regards from germany