• Please review our updated Terms and Rules here

A possibility for A Load-Diode/Resistor version of the Dallas RTC replacement?

T-Squared

Veteran Member
Joined
May 29, 2011
Messages
657
Location
San Antonio, TX
I have a Compaq Portable III, and unfortunately, I wouldn't know where to get the special battery that powers the real-time clock.

I figured that using Glitchworks' RTC replacement was easy enough, until I started having problems with it keeping time and settings, or even just starting up in-general.

I learned that any battery hooked up to the Compaq RTC needs to have a resistor and/or diode to regulate something. I don't know of any other computer that does that. Would it be possible to make a version that has these components already attached?
 
A bit of googling told me that Compaq uses an MC146818 RTC.
I'd use three alkaline batteries with a 1N4148 diode and a 220ohm resistor in series, as per the battery packs used in a BBC Micro.

The diode prevents accidental charging, and the resistor keeps the current draw down in the event of a fault.
 
THAT'S why I have problems! I have the wrong freaking chip! (I do have the MC146818, but the Glitchworks module doesn't have a limiting resistor or charge-prevention diode. I may get a DS12887 replacement instead.)
 
Neither, as far as I can tell, does the DS12887. Like the GW model, (which is nothing more than the base chip of a DS12887 (a DS12885) with an external battery and crystal) rather than all being potted in black epoxy. The battery is completely isolated from the charging circuit.
 
Take a look at the datasheet for the DS12885. here, which is what Glitch uses. Might it possibly be that the MOT pin somehow gets driven the wrong way? I can't really think of anything else, offhand.
 
One year later, and it seems, with cautious optimism, I have figured out the battery problem in my Compaq, thanks to @modem7.
It was the chip type and the support circuitry around it. I got rid of it and the Motorola CMOS chip that I used, and swapped over to the Necroware DS12887 replacement you can find on eBay.

I will say this, Compaq built their computers with really high quality. I lost a few surface through-hole pad-rings while trying to desolder the original socket, but the through-holes themselves are still in-place, meaning I could solder the provided machined socket in with no problem. Of course, this means that this is the final time I can solder a socket in that footprint.

I had a few hiccups when trying to get the CMOS assembly to work, but when I used the setup program the second time, I intentionally left the system unplugged while preparing for a get-together this evening, just to see if the contents would be battery-retained. I turned it on twice afterwards without errors. (Normally it would complain with a "102-System Board Failure" error.)

Now to work on the problem of using CF cards in the system... (I'm hoping it's a matter of using the extra pair of ROM sockets, and splitting an XT-IDE ROM into high and low bytes.)
 
Last edited:
When working on very old gear, I like to use a solution like Chip-Quik to lower the temperature of the solder joint to less than 100C. You can gently heat the whole area without damaging any other solder joints. It's really a godsend when working with fine-pitch SMT packages.
 
Back
Top