Hi..

There is plenty of stuff on the interweb, showing how other people managed to remove the battery "Hat" of a NOVRAM, and wired in an external cell.


Check these links out...

NVRAM to RAMTRON FM18W08 FRAM replacement.


New external battery for NVRAM


Reworking the DS1287 / DS1387 RTC chip


Dallas Semiconductor DS1387 Module Rework/Battery Replacement



All the above links checked and are working.

"Have Fun"!

73.

Dave 'KBV



--
Sent from a Micro$oft free zone:

¿ªÔÆÌåÓý

I appreciate the link to the article on the FRAM modification. I have a 54501A headed this direction DOA and once the power supply is sorted I suspect that NVRAM will be an issue. That article saves me a bunch of time chasing data sheets.?

A quick update.

After having built a new battery hat for the M48Z08 (and verifying that the battery makes contact) I installed the contraption in the 54620C, powered it up and... got the same result as before: just a horizontal bar. Battery or no battery, same result.

Removed the modified M48Z08 alltogether: it powers up and reports a wrong CRC...
Conclusion: the M48Z08 is either damaged in the process, or must contain some critical data at power-up after all.

Next step is to try this with a simple 8kx8 SRAM, just ordered some from a local supplier.
If that works, I might convert to FRAM (ordered a FM16W08 at the same time), will be offering the surplus PCBs.
If it doesn't and I cannot find the correct data to ut into that RAM, I just might sell it for parts...

I have 4 more boards with battery clips available for those that want to give this mod a try: €1.50 a set + P&P

It looks to me like there's some foreign goo bridging the IC legs in that photo.? Have you tried scrubbing it with some kind of effective solvent and a toothbrush?

That goo is hardened PUR, the remainder of the original battery hat (and non-conductive, I checked).
That stuff does not dissolve in anything AFAIK, soaking it in paint thinner for a day or so lets it swell and become brittle.

I had to exert quite a bit of force and heat to get the original hat off, maybe that pushed the chip over the edge...
I expect the parts I ordered to come to a solution of sorts friday, will post the results (if any).

Well, it took a little longer to get results, but here it goes:
After constructing an adapter board for the FM16W08 (the readily available ones from China do not fit the socket) the patient seems fully cured: all things work as expected, and the settings are remembered.
"Look ma! No batteries!" ;)

For those interested: I have 1 fully assembled FRAM adapter and three bare PCBs left over for sale...