VOGONS


First post, by lackofpatience

User metadata
Rank Newbie
Rank
Newbie

Hey everyone,
I'm hoping someone might have some pointers for 286 keyboard controllers.

I've got a DTK PTM-1010 v2, 286 Motherboard I saved from the recycling center.
https://theretroweb.com/motherboards/s/dtk-pt … 10v2-edit.-1.01

Like most of these motherboards, the battery had leaked and cut some traces. Overall the damage wasn't too bad. I only needed to bridge 4 traces with wire, from my testing.
I've got the machine to a point where it will POST past a memory count and won't show the previous keyboard fault code 0742. This code would show pre memory count before I did some repairs, not completing post. Now I get passed this error code and completes a memory count and continues to post.

So it seems now the keyboard controller is being detected at least.
Now I'm getting an error for "User to check keyboard" Aka a keyboard detection issue.

Where I'm kind of stuck is with two questions.

1. Do most 286 machines require a XT compatible keyboard?
2. Am I missing a signal to/from the keyboard controller or keyboard port? See continuity tests bellow

I don't know about question 1.
But if someone could eyeball the following testing to see if I'm missing something obvious. Hopefully I get some pointers.

For question 2, I have done the following continuity tests from the keyboard controller chip to the Keyboard port. Also have check that +5v and Grounds are connected.
Keyboard LED's do now flash when I power the motherboard up. No caps lock or num lock toggles work and the keyboard is still not detected. Before the trace repair, no LED's would light up at power up.

I tested the following by going over each socket pin and checking which pin has the lowest resistance in continuity mode. Please forgive my lack of skills, I'm not familiar with testing such IC's.

Screenshot-2024-02-12-115329.png

At this point I'm not sure if I'm missing a connection to the controller or if it's simply a AT vs XT keyboard compatibility issue?
Something odd was for ground, the VSS and Pin 7 EA pin are connected to some sort of ground. Perhaps that is correct, I'm not sure.

Anything that I'm missing from the above or things worth checking?
Cheers!

Reply 1 of 10, by lackofpatience

User metadata
Rank Newbie
Rank
Newbie

As a side note, I will dig out a AT/XT Keyboard from storage with a switch on the bottom.
As I was using a XT to PS2 adapter, I'm wondering if the following is related to my errors,
https://www.win.tue.nl/~aeb/linux/kbd/A20.html

Reply 4 of 10, by lackofpatience

User metadata
Rank Newbie
Rank
Newbie
Nexxen wrote on 2024-02-12, 01:34:

Could you post a pic of the board?

Yes, sure can. I have attached them bellow.

The traces near the keyboard socket have been bypassed/repaired on the underside of the motherboard.

Attachments

Reply 5 of 10, by Horun

User metadata
Rank l33t++
Rank
l33t++

8042's can be programmed for either XT or AT KBC, have never seen a 286 with XT required keyboard yet but suppose it is possible.
DId you pull the KBC out and check for corrosion near the old battery location ? It is possible that some via's or bottom of socket has corrosion....

Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. Stuff: https://archive.org/details/@horun

Reply 6 of 10, by lackofpatience

User metadata
Rank Newbie
Rank
Newbie
Horun wrote on 2024-02-12, 03:09:

8042's can be programmed for either XT or AT KBC, have never seen a 286 with XT required keyboard yet but suppose it is possible.
DId you pull the KBC out and check for corrosion near the old battery location ? It is possible that some via's or bottom of socket has corrosion....

I do have an XT compatible keyboard, I just need to pull it from storage. When I de soldered the socket, I saw only one suspect trace.
I tested it, and it's still connected, but decided to put some fresh solder over the exposed copper.

The other option is to remove the socket and solder the chip directly to the motherboard.

But yes, I did remove the keyboard connector and tested the continuity. I've had a look over it for a few days but haven't found any more open circuits. Thankfully the damage was mostly isolated and didn't wick too far into the traces.

Reply 8 of 10, by lackofpatience

User metadata
Rank Newbie
Rank
Newbie
rasz_pl wrote on 2024-02-12, 11:04:

it looks like there are still weak tracks on the pictures. You fixed thick power ones, what about those very thin?

Indeed there are some thin ones. I ended up scraping some of the corrosion away to the copper. Then put some fresh tin over it.
Anything that was beyond repair was done with bridge wires, mostly on the underside.

Then tested with the multi meter. The resistance looked okay as well.

I'm thinking I need to triple check, the last triple check as I could be missing something or as you've pointed out, a weak trace may still exist.

Reply 9 of 10, by rasz_pl

User metadata
Rank l33t
Rank
l33t
lackofpatience wrote on 2024-02-12, 21:34:

Indeed there are some thin ones. I ended up scraping some of the corrosion away to the copper. Then put some fresh tin over it.

I also see dark vias, that means corrosion got inside.

Open Source AT&T Globalyst/NCR/FIC 486-GAC-2 proprietary Cache Module reproduction

Reply 10 of 10, by lackofpatience

User metadata
Rank Newbie
Rank
Newbie

Good progress has been made!
I decided to remove the keyboard plug, and two controller IC's and re check my work.

Also took some nice high res photos of the damage, so I could zoom in. ( I don't own a microscope) My desk looked like a bomb site

20240213-215617.jpg

20240213-215631.jpg

There were only to traces missing. The keyboard connector has a 2nd set of traces going to some I guess gate IC's (u62). I didn't spot the 2nd path until creating the blow up images and tracing.

u62.jpg

It looks like a mess, but it works at at least.

20240213-215710.jpg

I managed to complete POST and even boot to MS-DOS 5

post2.jpg

post1.jpg

Thanks everyone for your help! Another 286 saved from the recycling center. Also pleased with myself as this is the first time I've ever repaired a battery damaged motherboard.