RTC int. error - Toshiba T3100

[gleegum]

Got this machine that was working fine, but after a reboot it gave me this error: "RTC int. error".
It stays on this message and doesn't boot.

Checking the maintenance manual it appears to be a "RTC INTERRUPT ERROR". I've no idea what this mean, but checking the manual it says just replace the entire board.

Any hints If this is repairable/ what should I check?
Thanks

 

[modem7]

Checking the maintenance manual it appears to be a "RTC INTERRUPT ERROR". I've no idea what this mean, but checking the manual it says just replace the entire board.

Somewhere on the system board is a chip that contains real-time clock (RTC) functionality. If desired, software can configure the RTC to generate a hardware interrupt in certain conditions. Normally, the RTC is not configured to generate an interrupt.

The power-on self test (POST ) will be testing the RTC functionality, including the ability of the RTC to generate an interrupt.

If this was an IBM AT of mine, I know that the IBM AT motherboard has a dedicated RTC chip, the MC146818, and the interrupt from that, IRQ8, heads off to one of the two 8259A interrupt controller chips. Using suitable test equipment, I could see whether the interrupt failure problem is because:
- The MC146818 is not generating the interrupt; or
- The MC146818 is generating the interrupt, but the interrupt is not reaching the 8259A; or
- The interrupt is reaching the 8259A, but the 8259A is not processing it.

I do not see mention of the RTC in the maintenance manual for the T3100. The maintenance manual for the T3100/e is a different story; it states that there is a MC146818 chip. Is there a MC146818 chip on your system board? Look for 146818 (or similar) because the chip may have a different prefix than MC.

You can see where this is going. You may need to replace chip on the system board.

Another possibility that I can think of is system board damage created by a leaking RTC/CMOS battery. Looking at the maintenance manual for the T3100, the battery connector (PJ1) on the system board is right next to one of the two 82C59 interrupt controller chips. Quite suspicious. Have you inspected the insides of your T3100 yet?

 

[gleegum]

Somewhere on the system board is a chip that contains real-time clock (RTC) functionality. If desired, software can configure the RTC to generate a hardware interrupt in certain conditions. Normally, the RTC is not configured to generate an interrupt.

The power-on self test (POST ) will be testing the RTC functionality, including the ability of the RTC to generate an interrupt.

If this was an IBM AT of mine, I know that the IBM AT motherboard has a dedicated RTC chip, the MC146818, and the interrupt from that, IRQ8, heads off to one of the two 8259A interrupt controller chips. Using suitable test equipment, I could see whether the interrupt failure problem is because:
- The MC146818 is not generating the interrupt; or
- The MC146818 is generating the interrupt, but the interrupt is not reaching the 8259A; or
- The interrupt is reaching the 8259A, but the 8259A is not processing it.

I do not see mention of the RTC in the maintenance manual for the T3100. The maintenance manual for the T3100/e is a different story; it states that there is a MC146818 chip. Is there a MC146818 chip on your system board? Look for 146818 (or similar) because the chip may have a different prefix than MC.

You can see where this is going. You may need to replace chip on the system board.

Another possibility that I can think of is system board damage created by a leaking RTC/CMOS battery. Looking at the maintenance manual for the T3100, the battery connector (PJ1) on the system board is right next to one of the two 82C59 interrupt controller chips. Quite suspicious. Have you inspected the insides of your T3100 yet?

Thanks for your very helpful reply.
Today I inspected the battery area as suggested and I found some corrosion on the RTC chip, which is indeed an MC146818.
I've removed it to inspect the traces underneath, and they seem a little suspicious, but upon testing them for continuity they checked ok.
I added an improvised socket and put the RTC chip back on.

Problem persist, bad chip??

I've also removed the battery and checked for damage on the underside of the pcb, but it was perfect.
the solder joints of the little crystal next to the RTC is corroded, could this cause the problem? or the crystal is for the battery?
Could be another one of the smd chips next to the RTC the culprit?

PS: When you say that the RTC is software programmable, do you mean I need to program it before adding it to the machine? (just in case I get a new one) Sorry for my ignorance on this topic.
Also I see that there's the 82C59A chip you mention, I've searched for it on ebay and it says that is a programmable chip, do I have to program it too?

I've attached some pics.
Thanks a lot!

 

[modem7]

Problem persist, bad chip??

One possibility of a few. They do fail. You can see some example failures at [here]. At [here] is a Toshiba T3200SX where the chip failed.

I've also removed the battery and checked for damage on the underside of the pcb, but it was perfect.
the solder joints of the little crystal next to the RTC is corroded, could this cause the problem? or the crystal is for the battery?

That small crystal, with some other circuitry, is sure to be generating the 32.768 kHz clock signal for the MC146818.
I'm sure, because of the combination of crystal size and its proximity to the MC146818.
The equivalent circuit in the IBM AT is shown at [here].
Do you have a logic probe or oscilloscope so that you can look for the 32.768 kHz clock signal going to pin 2 of the MC146818 ?

( The diagram that I pointed to also shows that pin 19 is the pin that the interrupt leaves the MC146818. )

PS: When you say that the RTC is software programmable, do you mean I need to program it before adding it to the machine? (just in case I get a new one) Sorry for my ignorance on this topic.

No.

Also I see that there's the 82C59A chip you mention, I've searched for it on ebay and it says that is a programmable chip, do I have to program it too?

No. Routines in the power-on self test will program it after power-on.

( As expected, 'Figure 1-4 System PCB' in the maintenance manual, shows two of those chips. )

 

[gleegum]

Do you have a logic probe or oscilloscope so that you can look for the 32.768 kHz clock signal going to pin 2 of the MC146818 ?

No, I don't have one.
Anyway, I think I'm gonna source that RTC chip + the crystal as it's really cheap. I'm gonna report back when it gets replaced.
To reach the RTC in this machine, you need to almost completely tear it apart, including the screen.
Thanksss

 

[gleegum]

After some months waiting for the RTC chip (I bought a failed one on ebay), the machine is up and running again.

Thanks