VOGONS


First post, by dataino.it

User metadata
Rank Member
Rank
Member

I have the twin of this motherboard http://www.win3x.org/uh19/motherboard/show/8225,
which I revived after countless problems caused by the loss of battery acid.

I am left with a mysterious start-up error "checking unexpected interrupt and stuck nmi" but I don't know if it is related
to the real problem of the card and that is and as if the keyboard is talking empty
the clock and data signals are present on the 8742 but nothing more.

I have already swapped the chips of the two cards, discovering the bad ones and replacing them.

Can anyone tell me which chip / signal I need to analyze?

Attachments

Last edited by dataino.it on 2021-06-22, 17:51. Edited 1 time in total.

Reply 1 of 9, by Deunan

User metadata
Rank Oldbie
Rank
Oldbie

The 8742 keyboard controller (KBC) also controls NMI line masking, and you seems to have problems with both NMI and keyboard. Also, there's CMOS RAM error being reported. I'm pretty sure the culprit is the 74ALS245 chip next to battery - either died or the spill broke some connections to it.

Data bus on KBC (pins 12-19) and on the RTC/CMOS chip should be connected to one side of the '245 (pins 11-18). The other side of the '245 (pins 2-9) should be connected to one of the mobo chips (but the pins might be somewhat random). Then finally pins 1 and 19 of the '245 should also go to another mobo chip. Check all that before you start doing anything else.

Reply 2 of 9, by dataino.it

User metadata
Rank Member
Rank
Member

Well I checked the connections from the 74ALS245 lead to the various chips and everything is ok.

Since I have two identical motherboards, I started a parallel measurement of the various pins with respect to the ground, and I immediately realized that the problem is much more complex, since the board is a 4-layer.

The measurement of the resistance between 5V and ground on the good one is 1200 Ohm while on the bad one it is 14.5 Ohm, I disassemble every single chip to find which is faulty, or which capacitor

Reply 3 of 9, by snufkin

User metadata
Rank Oldbie
Rank
Oldbie
dataino.it wrote on 2021-06-20, 19:40:

Well I checked the connections from the 74ALS245 lead to the various chips and everything is ok.

Since I have two identical motherboards, I started a parallel measurement of the various pins with respect to the ground, and I immediately realized that the problem is much more complex, since the board is a 4-layer.

The measurement of the resistance between 5V and ground on the good one is 1200 Ohm while on the bad one it is 14.5 Ohm, I disassemble every single chip to find which is faulty, or which capacitor

Given the difference in resistance, you might get lucky and be able to find a component that gets hotter on the bad board than on the good board.

Reply 4 of 9, by Deunan

User metadata
Rank Oldbie
Rank
Oldbie
dataino.it wrote on 2021-06-20, 19:40:

The measurement of the resistance between 5V and ground on the good one is 1200 Ohm while on the bad one it is 14.5 Ohm, I disassemble every single chip to find which is faulty, or which capacitor

This is not a pure resistance but MOS transistor leakage, so quite often the value you measure will be different depending on how you connect the meter leads (which goes to GND and which to 5V). Make sure you use the same connection on both mobos.

14 ohms is not a full short but with resistance that low I would expect something to get rather hot - it might be useful to let the mobo run for a minute and then touch the caps and chips, then maybe compare with the other one if nothing really stands out.

Reply 5 of 9, by dataino.it

User metadata
Rank Member
Rank
Member

At the and I fixed it but it was a struggle.
After removing all the chips the quartz and replacing a part of the capacitors.
I made other measurements and concluded that there was no short on the 5V.

Too bad that after having reassembled everything the situation got worse and the card started but stopped at post 17.
So I started to test the area affected by the battery acid again and found that other tracks were interrupted.
Provably in disassembling and reassembling the chips the thin copper that connected the pads to the tracks gave way.
Fortunately, the photos I take before mounting the socket have made my life easier in looking for breaks
The last strange thing apparently the two chips of the keyboard are not interchangeable between the two twin cards, or rather the first in my possession accepts both of them and the second does not want its original.

Thanks for the tips, they still saved me time and I learned something.

Attachments

Reply 6 of 9, by dataino.it

User metadata
Rank Member
Rank
Member

This was the first step, to fix and restore my 286 "ASEM THOR" which weighs about 20KG
The next step will test the power supply.
Who meanwhile took a bath ..

Attachments

  • PSU-ASEM.jpg
    Filename
    PSU-ASEM.jpg
    File size
    636.59 KiB
    Views
    360 views
    File license
    Public domain
  • Asem-THOR-8050-1.jpg
    Filename
    Asem-THOR-8050-1.jpg
    File size
    604.16 KiB
    Views
    360 views
    File license
    Public domain
  • Asem-THOR-8050-2.jpg
    Filename
    Asem-THOR-8050-2.jpg
    File size
    365.88 KiB
    Views
    360 views
    File license
    Public domain

Reply 8 of 9, by dataino.it

User metadata
Rank Member
Rank
Member
Doornkaat wrote on 2021-06-22, 10:47:

That PC looks seriously awesome!

It is fantastic apart from the frightening weight, it has a particular design with the power supply in view and the connections at the top ..

When I saw him in the picture in the midst of various computer junk it took me a while to understand what it was