• Please review our updated Terms and Rules here

RTC Stopped Ticking on my 286 clone board when power is off

maxtherabbit

Veteran Member
Joined
Apr 23, 2019
Messages
2,153
Location
VA, USA
No, the battery is not dead.

When the power is on, the RTC keeps good time (tested with checkit). When the power is off, the clock just freezes and doesn't advance. CMOS settings are never lost. I realize that these are the exact symptoms of a weak battery, but the 4xAA external battery pack I'm using still shows 6.5V open circuit voltage.

The board is a Shuttle 12MHz 286 board based on the VLSI 100 chipset. It is almost identical in layout and design to this comically overpriced Biostar:
https://www.ebay.com/itm/BIOSTAR-MB-1212V-VLSI-AT-MOTHERBOARD-W-286-12MHZ-CPU-1MB-RAM/123543534170

When I first acquired this system a couple of years ago, I removed the onboard rechargeable battery before it leaked, so battery damage is not a factor. I then fitted it with the external battery box and it has been working reliably ever since, keeping good time. About a month ago I noticed the RTC had stopped working during power off. I am using 4x Energizer lithium AA batteries, which I recently learned have an open circuit voltage slightly above 1.5V. I am slightly concerned that 4 of them in series was just enough above 6V to over volt something and damage it over time. I don't think the RTC chip itself is damaged, since it still keeps perfect time on main power. I measured the Vdd pin of the RTC when the system power was off and the battery was connected and got 2.4V, which explains why the thing isn't ticking since it requires a minimum of 3V.

Any thoughts?
 
I don't think the RTC chip itself is damaged, since it still keeps perfect time on main power.
The RTC isn't really in use when an OS like DOS/Win9x is running. The OS is taking care of the time in that case. I have some systems with the same issue, because they have their 3.6V rechargeable battery replaced by a 3V coin cell, which has enough voltage to power the CMOS, but not for the RTC. When the system is running, the time advances, but is frozen otherwise. This is easy to see in the BIOS. If the clock is frozen there as well, the voltage is too low or the RTC chip is dead. If it was working before in your case with the 4 AA batteries, it's probably the latter.
 
The RTC isn't really in use when an OS like DOS/Win9x is running. The OS is taking care of the time in that case. I have some systems with the same issue, because they have their 3.6V rechargeable battery replaced by a 3V coin cell, which has enough voltage to power the CMOS, but not for the RTC. When the system is running, the time advances, but is frozen otherwise. This is easy to see in the BIOS. If the clock is frozen there as well, the voltage is too low or the RTC chip is dead. If it was working before in your case with the 4 AA batteries, it's probably the latter.

As I said in the OP, I used checkit to verify that the RTC is ticking while the system is on
 
And you know for 100% that checkit really tests the hardware RTC and not just the time reported by the OS? The RTC gets its power from the battery all the time, no matter if the system is on or off. So it's either working all the time or not at all.

And what about the BIOS? Does it work there or not? That's the only place where you can see the RTC doing its thing.
 
And you know for 100% that checkit really tests the hardware RTC and not just the time reported by the OS? The RTC gets its power from the battery all the time, no matter if the system is on or off. So it's either working all the time or not at all.

And what about the BIOS? Does it work there or not? That's the only place where you can see the RTC doing its thing.

Yes, checkit has a specific "RTC test" where it shows the real time clock output along side the DOS clock to compare them. It ticks in the BIOS as well. I believe I have located the problem - the diode immediately downstream from the external battery header is failing a diode test in circuit
 
You replaced the rechargeable battery with a 6V battery pack just like that? No diode between the pack and the board?
Next question: what is the voltage of the pack when in use?
 
You replaced the rechargeable battery with a 6V battery pack just like that? No diode between the pack and the board?
Next question: what is the voltage of the pack when in use?

No lol the board came with an external battery header
 
OK, then still my last question stands. What about replacing the RTC IC? IMHO it is a common available MC146818A. If the problems stays after replacing the IC, the problem are the batteries IMHO.
 
I am using 4x Energizer lithium AA batteries, ...
I measured the Vdd pin of the RTC when the system power was off and the battery was connected and got 2.4V, which explains why the thing isn't ticking since it requires a minimum of 3V.
I believe I have located the problem - the diode immediately downstream from the external battery header is failing a diode test in circuit
Like as shown at [here], when the RTC chip is powered by battery, the voltage drop across the diode (a diode that is forward biased) is quite small (actual voltage dependent on diode type and curent). If you are seeing a voltage drop that is significantly larger, then the diode is suspect.
 
Like as shown at [here], when the RTC chip is powered by battery, the voltage drop across the diode (a diode that is forward biased) is quite small (actual voltage dependent on diode type and curent). If you are seeing a voltage drop that is significantly larger, then the diode is suspect.

Thanks for the schematic, that pretty much eliminates anything but the diode
 
I'm recently learned that in most cases a Dallas 12887 can be used as a drop in replacement for the MC146818A. Just so happens my board has the RTC socketed. I think I'm going to try swapping it out with a brand new Dallas from Digikey, because IMO an encapsulated DIP with battery inside is the most reliable option. Not to mention the RTC accuracy is generally better since they include the oscillator and are pretrimmed at the factory

I'm still of the belief that replacing the diode would solve the problem, but that would require a complete teardown of the machine, which I'd prefer to avoid.
 
Those Dallas chips aren't so great when the battery dies. They aren't cheap to replace.

Although that is one way to solve this particular problem. I'd be interested in hearing the results.
 
$10 every 10 years for an encapsulated unit which can never leak is well worth it to me. Honestly the lithium AA batteries are damn near as expensive. 3 of them even at bulk pricing is $4, and if you only buy a 4 pack it's closer to $8
 
Back
Top