IBM 5162(XT 286) not booting up

[xyz378537]

Hello, I have a problem with my IBM XT 286. It doesn't want to boot up, although there is activity on both data and address buses.
When I plug in a CGA card, nothing is displayed. The computer doesn't beep either. Is there anything else that I need to plug in in order to test the PC, or is the speaker, PSU and the motherboard enough? I also have all 640K RAM on the motherboard, and it doesn't have any parity errors. The ROM is also good.
Thanks in advance for help.

[modem7]

Hello, I have a problem with my IBM XT 286. It doesn't want to boot up, ...

Out of interest, is it one that you have just acquired, or one that was working for you but no longer working ?
If the previous, it opens up other possibilities, e.g. wrong type of SIMMs fitted by a previous owner.

Any visible damage caused by a leaking battery ?

When I plug in a CGA card, nothing is displayed. The computer doesn't beep either. Is there anything else that I need to plug in in order to test the PC, or is the speaker, PSU and the motherboard enough?

I do not have a 5162, but I have a few 5170's. The 5170 is very close in design to the 5162.

In a functional system, the speaker, PSU, and motherboard, are enough to expect a beep at some point (e.g. the POST in the IBM 5170 complains about no video card).

I also have all 640K RAM on the motherboard, and it doesn't have any parity errors. The ROM is also good.

So, the good RAM and ROM functionality must have been verified by you elsewhere.

The two SIMMs alone [512KB total] will be enough RAM for diagnostic purposes. Both need to be fitted. Your "all 640K RAM on the motherboard" informs us that both are fitted.

..., although there is activity on both data and address buses.

Does that mean that you have a logic probe and/or oscilloscope ?

Is the activity constant, or does it stop shortly after power-on ?

Per the diagram at [here], early in the the POST, the POST programs the 8254 timer chip to output a regularly occurring pulse. If you observe that, you know that the POST is starting and progressing at least partially.

You did not mention the use of a POST card. A modern example is shown at [here]. That is something worth trying.

Something else worth trying is a set of the Supersoft/Landmark Diagnostic ROM's. See [here]. You would use the ROM set for the IBM 5170, with the 5170 ROM created for socket U27 ('low') going into ROM socket U34, and the 5170 ROM created for socket U47 ('high') going into ROM socket U35.

[xyz378537]

I have just aquired the computer, and it was dead when I got it. The battery has been removed before I got it, and there is no signs that it leaked before removing. I compared the ROM images of my ROMs with those found on the internet, and they are the same. I have both a logic probe and an oscilloscope. I didn't use a POST card, because I don't have one, though I'll get it ASAP. I'll try the Diagnostic ROMs. Thank you very much for your reply.

[Ruud]

One thing you can do as minimum is to connect the IORDY line to GND when starting up. This will halt the CPU. Then use a voltmeter to measure all address and data lines. The result should be 0FFF0 for the adress lines and 5BEA for the data lines. 0FFF0 is the reset address for all 80x86 CPUs. At this address you will normally find the instruction "jmp F000:E05B" which will translate as "EA 5B E0 00 F0". If this is not the case, then there is a serious problem, ranging from a bad CPU to bad glue logic. (the ROMs were OK, you said).

If the address and data lines contain the correct values, I advise you to insert a set of diagnostic ROMs.

[xyz378537]

Thanks for all the replies. I will try to check the initial address and data lines, as well as the diagnositc ROMs.
Thank you very much.