• Please review our updated Terms and Rules here

4xAA BIOS battery replacement not working

Bill-kun

Veteran Member
Joined
Nov 2, 2020
Messages
688
Location
Michigan
I have a tall tower 386 I got over summer 2020. It came with a barrel BIOS battery soldered to the motherboard. The battery was completely dead and the motherboard was not holding any BIOS data after shutdown (date, time, floppy drive settings, and hard drive type). So I unsoldered the battery and put 4 AA batteries in the battery case that is on the floor of the computer case. I got the proper wire to reach from the battery case to the pin header on the motherboard for the external BIOS battery(ies). The AA batteries are brand new, non-rechargeable. I verified with a multimeter that the contacts that touch the pins have the correct voltage (1.5 V. * 4 = 6 V.).

Problem is, the BIOS still does not hold any data after shutdown. Why?
 
Quick question: Did you use a diode in series with the battery pack? You need something to block the charging current.

But that shouldn't have let the smoke out of anything. So next question: What does your 386 board use for a clock/memory chip? (e.g. DS1285, MC14686, etc.)
 
Quick question: Did you use a diode in series with the battery pack? You need something to block the charging current.

No.

I do??

Good grief
, the more I get into fixing my vintage computers, the more I realize how clueless I am on even the simplest things about it, even after having used computers since 1990, done some programming, knowing how to solder, and and being an automotive engineer. :|

Can you please give me a step-by-step guide to follow (or link) of how to properly connect the batteries, then? Thanks.

But that shouldn't have let the smoke out of anything.

Do you mean that it is safe to assume the 6 V. at whatever current would not have broken any components in the computer? I have been using fully charged Duracell non-rechargeable batteries.

I have disconnected the batteries from the motherboard. I was having a repeated symptom of argumentative booting, where I would often have to restart the booting because the hard drive buzzed, or the floppy drives were not accessed, etc. Could such a voltage and current cause that sort of thing?

So next question: What does your 386 board use for a clock/memory chip? (e.g. DS1285, MC14686, etc.)

The clock chip on the motherboard says "DOC-20NA, 14.31818 MHz., KDS-9M, JAPAN."
 
I mean the so-called "CMOS" chip (it's really the RTC with a bit of memory). I think you just read off the legend on a crystal or XCO. Photos would help.

The idea behind a blocking diode is to prevent the computer from charging non-rechargeable (i.e. primary) batteries.

highdiode.png
 
I've powered 386 systems (have two so configured now) with 6V AA supplies, so probably not a big issue--however, they all use the 4-pin external battery header on the CPU board.
 
The OP said he is using the external batt header. He doesn't need an extra diode.

The actual problem could be one of the diodes on the mainboard failed though. I had that happen recently, board used two 1N4148s in series and one had failed open.
 
One thing that it is worth watching out for when on board pcb batteries get replaced by a battery holder with batteries and flying wires is the type of plug or connector used and whether or not the pins on the flying wires are exposed.

I saw a case where the battery pack, still had its negative wire connected and the + terminal got unplugged and got dragged across PCB connections. An external battery pack especially with Nicads or AAA or AA sized batteries can source a large short circuit current and can destroy IC's with an accident. At least most times there are diodes on the pcb which will prevent a reverse polarity accident.

But since very little current is taken from the battery, I have become inclined to put a series 1k resistor with one of the battery wires inside some heatshrink sleeving. At least this way if the battery pack wires gets shorted out or the is an accidental contact elsewhere the current is limited to a very safe value.
 
The OP said he is using the external batt header. He doesn't need an extra diode.

The actual problem could be one of the diodes on the mainboard failed though. I had that happen recently, board used two 1N4148s in series and one had failed open.

Missed that--I got all wrapped up in the "I clipped the old battery out" Okay, Emily Litella.

However, I have run into a board where the battery (obviously failing) was left intact and a DS1287 was substituted for the MC146818. An external battery would have accomplished nothing.

You never know...
 
Last edited:
That's actually a good approach if you can make it work. I tried to swap my already socketed MC146818 with a dallas DS12887 in lieu of repairing the diode, but the BIOS would just halt with a CMOS bad error. I know the dallas module worked on another system so it must have just been something incompatible. A similar ODIN RTC module did pass POST on the board, but I ended up just repairing the diode
 
Could this be a motherboard where a jumper needs to be changed to effect operation of the external battery connector? An example pictured at [here].

What is the make and model of the motherboard ?
 
You need to check that the battery voltage is actually powering the CMOS chip. Probably a broken trace or via around the battery terminals if leakage is visible.
If it has never worked, then you need to check all the jumpers are correct and not in reset position for the CMOS. 6V is perfectly fine as the 4.8V rechargable barrel
would be charged at around 6V anyway.
 
That's actually a good approach if you can make it work. I tried to swap my already socketed MC146818 with a dallas DS12887 in lieu of repairing the diode, but the BIOS would just halt with a CMOS bad error. I know the dallas module worked on another system so it must have just been something incompatible. A similar ODIN RTC module did pass POST on the board, but I ended up just repairing the diode

I did the DS1287 (not the 887) dodge on a 286 motherboard years ago--worked fine.
 
It came with a barrel BIOS battery soldered to the motherboard.
[...]
Problem is, the BIOS still does not hold any data after shutdown. Why?
Maybe damaged traces due to the old battery leaking. Show a picture of the area where the old battery once was.
 
I have a somewhat relevant question: older Compaq systems (such as Deskpro286) use a 7.2v battery for backup. It's plugged into a motherboard jumper. Do I need to include anything to use an external AA battery unit? Also, since these suckers are 7.2 volts, would 6 volts be enough? Also also I've found quite a few 7.2v RC car batteries. Are those a good idea?
 
If it's a 7.2V rechargable then only use a rechargable battery. I expect 5V would be the lowest voltage required to power the chips as the 7.2V would discharge down to around 4.5V. RC battert would be fine, I use them all the time but the 3.7V versions. You could also use conventional alkaline batteries but a diode and resistor charging blocking circuit would be required or breaking the positive voltage track from the charging component. I expect there is a 3.3V regulator somewhere which is keeping the backup RAM powered.
 
Last edited:
I have a somewhat relevant question: older Compaq systems (such as Deskpro286) use a 7.2v battery for backup. It's plugged into a motherboard jumper. Do I need to include anything to use an external AA battery unit?
If the backup battery is charged by the system, then you should put a diode in series to your battery pack to prevent it from charging the batteries. Otherwise, you should be fine.

If there is a voltage regulator on the board, then the battery pack voltage doesn't matter too much; but 6V should be sufficient in any case.
 
Alternatively use a super capacitor if cost isn't an issue and you power the system on at least every month. These work great for replacing rechargable backup cells
and can charge to full capacity within minutes. You would have to use 2 or 3 in series to reach the 7.2V charging circuit. 2 x 5V or 3 x 2.7V.
I'm talking about these if you are curious:
https://www.ebay.co.uk/itm/3x-Eaton-Powerstor-0-33F-5-5V-Super-Capacitors-cap-electronic-memory-back-up/153461558999

and

https://www.ebay.co.uk/itm/2-Pcs-4-0F-5-5-V-Super-Capacitor-H-Type-Button-Smart-Capacitance-Universal/272412952404
 
Last edited:
Could this be a motherboard where a jumper needs to be changed to effect operation of the external battery connector? An example pictured at [here].

What is the make and model of the motherboard ?

Happy New Year!!

The motherboard is a Micronics 386 model 09-00021. Here is the link at Stason.org.
https://stason.org/TULARC/pc/motherboards/M/MICRONICS-COMPUTERS-INC-386-MICRONICS-386-VER-5-0.html
Before you click that, however, I need to note a jumper that exists on the board but not on the diagram of the board. It is labeled W30 on the board itself. It is right above the system BIOS chips. I have rotated the diagram of the board to the orientation in my computer case and added the 3 pins in red in the picture below.

09-0021 motherboard upside down with added 3 jumper pins.png

Here is someone else's photograph of their entire motherboard. Note this photographer had also unsoldered his BIOS battery (the square labeled B1, near the upper left corner of the board). J2 below it is where the external batteries connect.
https://www.vogons.org/download/file.php?id=69167&mode=view

According to the 09-00021 specs at this webpage
https://www.infania.net/misc/moboarchive/Micronics/00021.txt
the jumper pin set W30 is "Reserved" and should be set to 2-3. So I don't know what it does. There are 3 pins, and there is presently a jumper connecting pins 2 and 3. I have not moved the jumper since I got the computer about July 2020.

Interestingly, not to complicate this thread, but I've had the same phenomenon on my Dell System 200 286 from 1988: last month I unplugged the dead BIOS battery from the motherboard and connected a pack of 3 AA batteries to the same pins (the original battery said 4.5 V.) but that BIOS also does not keep setup data after shutdown.
 
Back
Top