486 EISA system

On 10/15/06, Teo Zenios <teoz at neo.rr.com> wrote: If I understand what you are describing, it sounds like the same style as Sun "NVRAMs" - you are looking at a 24-pin part with no "missing pins" bent up into the battery area, but with the attachments between the chip and the battery at the ends. You might even be able to see some light between the battery and the chip portions of the package. If that sounds like what you have, the battery is in the top, and the IC is in the bottom, I don't recally off the top of my head if the battery leads are in the front or the back, but embedded in that potting compound are two flat IC leads close together on the same end. You can scrape off the potting compound to expose the leads, then check them with a meter. If you are close to 3V, the battery is probably OK. If you are under 2V (ISTR seeing 1.7 from essentially dead Li cells), you need a new battery. What I have done with a Sun NVRAM (ST MK48T2) is to break the leads at the end of the chip, disconnecting the battery, then soldering on two wires - I happened to use a lead from a 9V battery-powered device - then attaching a battery to those leads. In my rig, I had a spare solder-on motherboard Li coin cell with factory-bonded solderable leads - I soldered it to the snap-top of a dead 9V battery. These days, I'd probably glue a coin-cell holder to the top of the NVRAM if there was vertical clearance in the case to install it that way. Whatever technique you use, don't solder directly to the battery casing. OTOH, if you just want to replace it, the successor product is the 1225AB http://www.maxim-ic.com/quick_view2.cfm/qv_pk/2646 I think you should be OK with an MK48Z08 as well - it's just an 8Kbyte SRAM, no clock (unlike an MK48T08 as used in SPARC2s, etc). No matter what you use, the battery life is approx 10 years whether it's in a powered-on device or not - the leakage current is higher than the current used by the SRAM, so the RAM chip isn't what drains the battery - it's time. -ethan