• Please review our updated Terms and Rules here

IBm 5150 dead motherboard ...

fafonso

Member
Joined
Mar 6, 2015
Messages
11
Hello,

I have 5150 rev B that I am attempting to resurrect. The issue is the dead motherboard symptom.

Initially i suspected a problem in the first 16K of bank 0. Therefore I programed the supersoft diag rom on a 27256, created am pcb/dil adaptor and the outcome was puzzling. No error beyond the obvious. 1 in Keyboard, for not having a keyboard plugged in at the time, and an FFD reading error also negligible.

Automatically suspected a ROM corruption/malfunction, I replaced the ROM and the dead symptom persisted.

After several passes of the supersoft, it returned a error on memory location 20402(hex), memory bank 2. But being outside the bank 0 and 1º 16k it wouldn’t be a problem or "THE" problem.

An odd thing is that at CGA memory test (passed) the character set becomes strange, as seen in the pic.




I think that the problem affecting this MB is related to the ttl logic ic's. I will check the schematics for probable culprit. Meanwhile any info or comments are highly appreciated.

Thanks,
Filipe
 
On the screen bit 2 of the video memory seems to hold incorrect values for some character cells. Is this CGA card known good?
If yes, there might be a problem with the data line on the mainboard.
- something conductive inside the slots?
- any of the 74LS245 temperature sensitive?
 
On the screen bit 2 of the video memory seems to hold incorrect values for some character cells. Is this CGA card known good?
If yes, there might be a problem with the data line on the mainboard.
- something conductive inside the slots?
- any of the 74LS245 temperature sensitive?


Sorry for the absence, have been studding and working so the mighty 5150 was a bit on hold.

The artifacts at the video memory occurs during cga test. Attempting to discard the cga (the only I have) i tried one WD based VGA card (Expert TopVGA2300) that is known to work in 8 bit mode, but to no avail.

I did verify all 74ls245 using an 16 bit logic analyzer, and found no problem.

Is it possible to be a DMA problem ??? The CGA card only malfunctions after cycling to "graphic mode" and i do notice that the pattern at the CGA test appears a bit strange.

The diagnostic ROM (SuperSoft/Landmark) excludes DMA,IRQ,PPI and Timer.

Im starting to suspect address lines A19 to A16. Will test 74ls20(u64), 74ls670(u19) and 74ls373(u10).

Any tips to solve this problem is highly appreciated,


Thanks,
 
I think your cga passes because it tests video ram which means all address and data lines are functional. I think bit 2 is getting stuck in the encoding process internal to the card. More data encoding bits are required for graphics which is why it worsens. I think it's your cga card. Not sure why vga card not working. Inspect your cga card closely for physical damage.

Larry G
 
The artifacts at the video memory occurs during cga test. Attempting to discard the cga (the only I have) i tried one WD based VGA card (Expert TopVGA2300) that is known to work in 8 bit mode, but to no avail.
To me, your wording suggests that your use of the VGA card was with the Supersoft/Landmark ROMs. The Supersoft/Landmark ROMs are not compatible with VGA.
 
To me, your wording suggests that your use of the VGA card was with the Supersoft/Landmark ROMs. The Supersoft/Landmark ROMs are not compatible with VGA.

Sorry if i didn't explain myself properly, the VGA card was tested using the U33 rom (1501476).
 
I was referring to U33 on the motherboard, and only 1 card on expansions slots, the VGA. Did excluded a U33 malfunction because I tried also a 27c256 whit ibm original rom (using the same adpter that I made for the SuperSoft/Landmark rom) .
 
Good news, everyone!.........................

The problem is solved, I replaced all 4164 of bank 0, u37 to u45, fitted new DIL16 sockets(not original, but convenient). The failure was on the address side of one 4164, the symptom was terribly misleading..

thank you all for the help.
 
Back
Top