Is this motherboard no longer usable?

[alejack12001]

I have an issue with a IBM PC-XT/286 that uses a Baby AT motherboard that seems to be hung up within POST. First, this is not an actual IBM PC-XT/286 motherboard. The motherboard is made by Informtech International Inc as a Baby AT 80286 10-MHz with 1-MB RAM. The motherboard uses Chips and Technologies with Award BIOS dated from 1984-1986. I have not been able to get this motherboard beyond POST.

I have a POST analyzer, which I have inserted in a 8-bit ISA slot. When I turn on the POST Analyzer the LED’s all light except for the reset; if I engage the reset button the light doesn’t come one. The digital display shows an error at 15 for the left digit and 14 for the right. Apparently, the right two digits precede the left two digits. According to the PC analyzer manual for Award BIOS, the error for code 14: test 8254 timer 0 counter 2 and error code 15 is “verify 8259 Channel 1 Interrupts by Turning off/on the interrupt links.

I found a chip on the motherboard that is NEC product PD71054, which according to its data sheet is a high-performance programmable timer/counter designed for timing control applications in microcomputer systems. This chip is a replacement for the 8254 by the maker of the board. In addition, I found the NEC D8259AC-2 Chip, which programmable interval timer.

If these chips are defective, is this motherboard no longer usable?

 

[Ruud]

Define "usable" first. I would start replacing these chips and preferably ones that have been tested on another board. If possible, let the board run at 6 MHz to avoid possible speed problems when running at 10 MHz. If one of the chips was indeed the culprit, you'll have a usable board again.

Also: there are Landmark diagnostic ROMs available for the AT>

 

[modem7]

Apparently, the right two digits precede the left two digits.

It sounds like you have the 'Example of a modern POST card' that is shown at [here].

According to the PC analyzer manual for Award BIOS, the error for code 14: test 8254 timer 0 counter 2 and error code 15 is “verify 8259 Channel 1 Interrupts by Turning off/on the interrupt links.

It is misleading to think of these as 'error' codes. Think of them as progress indicators.

It is kind of like:
You take your car to a mechanic to get checked out. You tell the mechanic to stop the examination/test as soon as something is found wrong. You are off in a nearby coffee shop, but can see your car being worked on.
The mechanic signals to you that he/she is about to test the doors, by momentarily holding up a '1' sign. The doors are successfully tested. Successful, so on to next test.
The mechanic signals to you that he/she is about to test the lights, by momentarily holding up a '2' sign. The lights are successfully tested. Successful, so on to next test.
The mechanic signals to you that he/she is about to test the tyres, by momentarily holding up a '3' sign. The tyres are successfully tested. Successful, so on to next test.
And so on.

So you need to ignore the "14". Whatever that test/check signifies, passed.

According to the PC analyzer manual for Award BIOS, ... error code 15 is “verify 8259 Channel 1 Interrupts by Turning off/on the interrupt links.

You need to be cautious there. Have a read of what I have at [here].

In that regard, Ruud's suggestion of use of the SuperSoft/Landmark diagnostic ROM is good, because in effect, it is 'another set of eyes'. If it reports an 8259 issue, then you have confirmation that the issue relates to the 8259.

You will notice that I used, "relates to". That is because sometimes, the problem cause is not the chip pointed to. E.g. Chip good, but has a poor connection to its socket. E.g. Faulty chip that supports the target chip.

 

[alejack12001]

I received your message. My apologies the internet is down again where I live. Verizon Fios cable was cut when neighbor installed a fence. Temporary cable fails a lot. When cable is restored I will respond.

 

[daver2]

Note that the 8259 is a programmable interrupt controller (PIC).

Dave

 

[alejack12001]

Internet restored.

The recommendation of Ruud’s Landmark diagnostics has a couple of problems for me. The first is that Landmark doesn’t support VGA card, according to Chuck(G), which I would be using in the test. I do have CGA card, but, I don’t have a CGA monitor. Now, because I don’t have a CGA monitor, I don’t know if the CGA card works. So, with that in mind I do have a CGA to VGA cable, however, I read that these generally don’t work because frequency of the horizontal sync pulse is too low for VGA monitors to receive. I know that there is some arcade game card that demodulates the frequency pulse to a level that VGA monitors can receive. Acquiring the card is not difficult, the question is it worth it for just a test? The second question is what is displayed when the Landmark discovered a failure such as the suggested 8259 chip by the POST analyzer? What if the issue is a blown capacitor or defective RAM chip or whatever. Will the Landmark discover that possibility?

 

[alejack12001]

Note that the 8259 is a programmable interrupt controller (PIC).

Dave

If this is programmable. The chip[ is soldered to the motherboard. So, getting it off may be a challenge especially with my Parkinson's directed tremors. Is there a way to re-program it?

 

[daver2]

It is programmed by the BIOS via registers.

It is not a ROM, EPROM or anything like that.

The 8259 was originally identified as a programmable interval timer (PIT) whereas it is a programmable interrupt controller (PIC). A different device.

Dave

 

[modem7]

I do have CGA card, but, I don’t have a CGA monitor.

I know that there is some arcade game card that demodulates the frequency pulse to a level that VGA monitors can receive.

I suggest that you be careful there. 'Arcade' CGA is not the same as IBM CGA. Discussed by tomxp411 at [here].

Maybe you have a TV that can accept the composite video from the CGA card's RCA jack.

The second question is what is displayed when the Landmark discovered a failure such as the suggested 8259 chip by the POST analyzer?

An example screenshot of the AT version of the diagnostic is at [here].

If I put a faulty 8259 onto my fully functional IBM AT motherboard, one of the two '8259 INT CONTROLLER' tests would fail, obviously which one depending on which one of the two 8259's I replaced.

However, if one day, I discovered that one of my IBM AT motherboards had failed, and the Supersoft/Landmark diagnostic indicated failure of the 'U125 8259 INT CONTROLLER 2' test, I do have to allow for the possibility that the pointed-to chip is good, and that something else failed, e.g. the chip that generates the 'chip select' signal for the pointed-to chip. A bit like the 'battery warning' light in my car turning on. Maybe it is the battery, but instead, maybe I have faulty alternator.

What if the issue is a blown capacitor or defective RAM chip or whatever. Will the Landmark discover that possibility?

If the capacitor is 'blown' open-circuit, no, in that the software cannot inform you that there is an open-circuit capacitor somewhere.
If instead, the capacitor has gone short-circuit, that overloads the power supply, resulting in the power supply shutting down its outputs.

Defective RAM chip. Usually. But like other diagnostic software, which includes IBM's diagnostics, it cannot definitively state that a RAM chip/s is faulty. It tests the RAM subsystem. In that test, if the diagnostic writes FF to address 0, then reads back FD, maybe the bit 1 chip in bank 0 is faulty, but there can be other causes.

 

[alejack12001]

I suggest that you be careful there. 'Arcade' CGA is not the same as IBM CGA. Discussed by tomxp411 at [here].

Maybe you have a TV that can accept the composite video from the CGA card's RCA jack.


An example screenshot of the AT version of the diagnostic is at [here].

If I put a faulty 8259 onto my fully functional IBM AT motherboard, one of the two '8259 INT CONTROLLER' tests would fail, obviously which one depending on which one of the two 8259's I replaced.

However, if one day, I discovered that one of my IBM AT motherboards had failed, and the Supersoft/Landmark diagnostic indicated failure of the 'U125 8259 INT CONTROLLER 2' test, I do have to allow for the possibility that the pointed-to chip is good, and that something else failed, e.g. the chip that generates the 'chip select' signal for the pointed-to chip. A bit like the 'battery warning' light in my car turning on. Maybe it is the battery, but instead, maybe I have faulty alternator.


If the capacitor is 'blown' open-circuit, no, in that the software cannot inform you that there is an open-circuit capacitor somewhere.
If instead, the capacitor has gone short-circuit, that overloads the power supply, resulting in the power supply shutting down its outputs.

Defective RAM chip. Usually. But like other diagnostic software, which includes IBM's diagnostics, it cannot definitively state that a RAM chip/s is faulty. It tests the RAM subsystem. In that test, if the diagnostic writes FF to address 0, then reads back FD, maybe the bit 1 chip in bank 0 is faulty, but there can be other causes.

Well the CGA/Composite video didn't produce even a flicker on the Panasonic TV that I had using the composite connection. I am not sure the video card is good or not.

I removed the ROM's from the Award BIOS motherboard and substituted the Landmark ROM's. I made sure that they were correctly seated and followed the notched drawing on the motherboard. When I powered up the machine nothing seemed to happen except for the fan coming on the power supply.

 

[modem7]

I removed the ROM's from the Award BIOS motherboard and substituted the Landmark ROM's. I made sure that they were correctly seated and followed the notched drawing on the motherboard. When I powered up the machine nothing seemed to happen except for the fan coming on the power supply.

Some possibilities that I can think of:
1. Bad/corrupted Supersoft/Landmark ROM's.
2. Odd ROM went into even socket, even ROM went into odd socket.
3. The motherboard is advanced enough that its chipset requires some chipset-unique configuration, something that the Supersoft/Landmark diagnostic does not do.
4. See the 'Important Notes' section of [here]. You did not wait 20 seconds.

Did you have a speaker connected to the motherboard? See the 'Speaker' section of [here].

 

[alejack12001]

Some possibilities that I can think of:
1. Bad/corrupted Supersoft/Landmark ROM's.
2. Odd ROM went into even socket, even ROM went into odd socket.
3. The motherboard is advanced enough that its chipset requires some chipset-unique configuration, something that the Supersoft/Landmark diagnostic does not do.
4. See the 'Important Notes' section of [here]. You did not wait 20 seconds.

Did you have a speaker connected to the motherboard? See the 'Speaker' section of [here].

Well number 1 is possible. I downloaded the Landmark software version 1.2 from Archive.org. I then flashed the binaries odd/even on to 82128 EPROMS. The Award BIOS were labeled HI/Low. I found a correlation of odd/even with hi/low in the minuszerodegrees website.Odd is Hi and Even is Lo.

Nope on number 2. I have pictures of the Award ROM seated on the motherboard and I re-seated the Landmark the same way.

On number 3, I have very little information on this motherboard. I have a Stason.org diagram, which I believe may be incorrect as to the jumper/switch settings.

On number 4, its possible even though I thought I waited about 1 minute.

The TV might be another issue. While there is a yellow composite port on the TV there is also YPBPR port. On screen there was a comp.1 and comp.2 and I kept hopping from one to the other to see anything. Another possibility is that the Landmark may itself got hung on something, which prevented me from seeing anything.

 

[alejack12001]

Here is one that will make you laugh. I was using the wrong Landmark tool. I had an XT model instead of the AT model. So, I will try again.