Image Map Image Map
Results 1 to 8 of 8

Thread: How to test a Dallas DS12887 RTC chip

  1. #1
    Join Date
    Mar 2009
    Location
    Pleasant Hill, CA USA
    Posts
    1,798
    Blog Entries
    1

    Default How to test a Dallas DS12887 RTC chip

    I just some new ones, they appear to have been made in 2010 (date stamp 1029A2) and I installed one into my 386 and everything seemed fine till I powered it on today and discovered all the CMOS settings are gone. Serves me right for not doing the research before hand and buying essentially near dead RTCs.

    So my question is this, how can one test this thing to verify that it in fact the battery that is dead? Surely the thing would be supplying some sort of voltage to the motherboard for it to retain it's settings in BIOS? If so, on what pins?

    IBM 5160 - 360k, 1.44Mb Floppies, NEC V20, 8087-3, 45MB MFM Hard Drive, Vega 7 Graphics, IBM 5154 Monitor running MS-DOS 5.00
    IBM PCJr Model 48360 640kb RAM, NEC V20,, jrIDE Side Cart, 360kb Floppy drives running MS-DOS 5.00
    Evergreen Am5x86-133 64Mb Ram, 8gb HDD, SB16 in a modified ATX case running IBM PC-DOS 7.10

  2. #2

    Default

    Hopefully someone would know this better then I, but IIRC, The cmos data is actually stored on the RTC itself. It's not much, maybe only a few bytes or so, but that does allow it to store some basic configuration data. Also, the batteries on some of these are the rechargeable type - you may have to leave the system on for as long as a day before they are fully charged. Is everything else on the board fine? Clear cmos jumper not shorted?

    if they were really made in 2010, then they should have plenty of life left. Hopefully they aren't rebadged old ones. I'm assuming it held the configuration data when you first tried it? IE, after install and setup, you powered off the system, waited a minute or two and then checked and everything else was ok. How long was the system on for when you first installed the chip? How long was the system powered down? Also, was the part number the exact same? Some of the dallas RTC's have differant pinouts.

    I hope you can get it to work. Had to replace one a number of years ago and getting some of the "hidden" (IE, saved in the RTC, but not accessable via CMOS) data back into the chip was a pain.

    As far as i know, the only way to check for a dead battery is to literally cut it open....not an option if its returnable.
    Last edited by salamontagne; January 2nd, 2017 at 10:47 PM.

  3. #3
    Join Date
    Jan 2014
    Location
    Western North Carolina, USA
    Posts
    687

    Default

    As a point of reference, I purchased three new DS12887A's from Mouser a couple of weeks ago for my CPU280 project, and they all had 2016 date codes. These are new, not 'new-old-stock' ones.
    --
    Bughlt: Sckmud
    Shut her down Scotty, she's sucking mud again!

  4. #4
    Join Date
    Mar 2009
    Location
    Pleasant Hill, CA USA
    Posts
    1,798
    Blog Entries
    1

    Default

    Quote Originally Posted by lowen View Post
    As a point of reference, I purchased three new DS12887A's from Mouser a couple of weeks ago for my CPU280 project, and they all had 2016 date codes. These are new, not 'new-old-stock' ones.
    That begs the question, how does one read the date code? And which number on the top is the date code (there are two possibilities on mine).

    IBM 5160 - 360k, 1.44Mb Floppies, NEC V20, 8087-3, 45MB MFM Hard Drive, Vega 7 Graphics, IBM 5154 Monitor running MS-DOS 5.00
    IBM PCJr Model 48360 640kb RAM, NEC V20,, jrIDE Side Cart, 360kb Floppy drives running MS-DOS 5.00
    Evergreen Am5x86-133 64Mb Ram, 8gb HDD, SB16 in a modified ATX case running IBM PC-DOS 7.10

  5. #5
    Join Date
    Jan 2007
    Location
    Pacific Northwest, USA
    Posts
    22,514
    Blog Entries
    20

    Default

    If the chips are DS12887 and not the -A or -+ parts, the 2010 date code is probably close, as the original non-suffix version hasn't been made in years.

    If you're wondering about the internal lithium, register D (address 13 decimal) in the DS12887 clears bit 7 if the internal lithium cell has failed. So you can access it from BASIC thus:

    OUT &H70,13
    PRINT HEX$(INP(&H71))

    The register should display 80 (hex) or greater if the lithium cell is still good.
    Last edited by Chuck(G); January 3rd, 2017 at 10:46 AM.

  6. #6
    Join Date
    Mar 2009
    Location
    Pleasant Hill, CA USA
    Posts
    1,798
    Blog Entries
    1

    Default

    Yep, the ones I have all the DS12887 (no A or +/- on them). So yay for Jameco selling old dead stuff. Ah, well at least they were cheap ($6 each). I may still complain to them about it though.

    The output of that basic code in Qbasic is exactly 80. So that would seem to imply that I have the one case where an DS12887 is *NOT* a drop in replacement for the original DS1287 that was in there. Given that this is a Dell, I really should have seen that one coming. I tossed the original 1287 since the previous owner had chopped the hell out of it trying to add a battery to it, and I am pretty sure that was what was causing all manner of odd behaviour.

    So I guess I'll need to jump on ebay and buy a N.O.S DS1287 and hack it with a battery.

    IBM 5160 - 360k, 1.44Mb Floppies, NEC V20, 8087-3, 45MB MFM Hard Drive, Vega 7 Graphics, IBM 5154 Monitor running MS-DOS 5.00
    IBM PCJr Model 48360 640kb RAM, NEC V20,, jrIDE Side Cart, 360kb Floppy drives running MS-DOS 5.00
    Evergreen Am5x86-133 64Mb Ram, 8gb HDD, SB16 in a modified ATX case running IBM PC-DOS 7.10

  7. #7
    Join Date
    Feb 2015
    Location
    Cleveland, OH, USA
    Posts
    183

    Default

    I have been trying to solve this problem with a '386 VLB board, but had thought the DS1286 was no longer available. I didn't know until reading this post that there was a replacement.

    I checked Mouser and found there are several variants, some with an internal battery and some designed to have the battery and crystal connected externally. Coin cell battery seems like the way to go.
    Link to PDF datasheet: http://www.mouser.com/ds/2/256/DS128...887A-86657.pdf

    Thanks to Chuck G for the hint on reading the CMOS data in the clock; the battery died on my clock and now the board will not boot because the CMOS is not backed up. I tried hacking a battery, unsuccessfully.

    Mouser, here we come...

    -CH-

  8. #8
    Join Date
    Feb 2015
    Location
    Cleveland, OH, USA
    Posts
    183

    Default

    More on this topic:

    App notes from Maxim, dated 2002, (http://dgmag.in/N1/Online/advanced/48629.pdf) say difference between DS12887 and DS12887A is that the A version chips have a "RAM Clear" pin; which functions as a BIOS Reset. If your mobo contains a jumper for that purpose you probably want that version. Also notes that a non-A will probably replace an A in functionality, but if you set the BIOS password and forget it you will have to replace the chip as there is no way to reset memory. Final note of significance is that replacing the clock is not a guarantee that the board will now boot, for two reasons: One is that the system may attempt to access the wrong partition on the clock. The other is that some BIOS expect the clock to be partially configured before installation, and only the manufacturer knows what that configuration would be.

    In my case my board is a very early BIOSTAR and the clock chip was an ODIN, no longer available. The BIOS is Phoenix. I may be in the last category, but more research is required to find out for sure.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •