Image Map Image Map
Results 1 to 10 of 10

Thread: DEAD - NEC PowerMate 1 Plus (APC-H2010F) 80286

  1. #1
    Join Date
    Dec 2019
    Location
    Sydney, Australia
    Posts
    7

    Default DEAD - NEC PowerMate 1 Plus (APC-H2010F) 80286

    Hi all,

    Some of you may have already seen my posts on the OCAU forum. This thing died while I was in the process of restoring it.

    First I was getting "error in EXE file' while trying to run programs in DOS. However this may be a problem with my DOS version, drives or disks...

    But then I started getting random characters sometimes while pressing the keyboard. Then it stopped detecting disk drives, resetting would fix it, but then it stopped turning on intermittently or it'd shut down after a few seconds.

    I'm was getting beep codes 3-1-1, 1-2-3 and 1-2-2. Which are all DMA (Intel N82231-2) failure codes:
    http://www.bioscentral.com/beepcodes/phoenixbeep.htm

    The socket of the Intel N82231-2 did have a crack but I replaced it and the N82231-2 itself. But it didn't fix it. The replacement N82231-2 was from ebay, was meant to be NOS but looks scratched up...

    The last few times I got it to boot it also displayed "Keyboard failure". "Gate A20 Failure" and "Time-of-day clock stopped" errors (see attached pics).


    Now I'm not even getting beep codes, the power LED comes on and keyboard lights flash but that's it. I think something related to the N82231-2 or the keyboard controller might be the issue. Can anyone help me narrow down the fault a bit? I have test equipment including an oscilloscope etc but I don't understand computer architecture well enough to know what to probe.
    Attached Images Attached Images

  2. #2
    Join Date
    Dec 2019
    Location
    Sydney, Australia
    Posts
    7

    Default

    The keyboard controller is an Intel 8742 (programmed as YAN25D05). I've got an 8041A from another PC, would the two be compatible? Can I use the 8041A for troubleshooting?
    Attached Images Attached Images

  3. #3
    Join Date
    May 2006
    Location
    Melbourne, Australia
    Posts
    6,427

    Default

    Quote Originally Posted by SpaceDog707 View Post
    Hi all,
    Welcome to these forums.

    Quote Originally Posted by SpaceDog707 View Post
    Now I'm not even getting beep codes, the power LED comes on and keyboard lights flash but that's it.
    With the majority of AT-class keyboards that I have tried it on, simply connecting power only to the keyboard, results in the keyboard's LED's flashing on momentarily. Presumably, that is the controller in the keyboard flashing the LED's as part of its own power-on self test.

    ( In the IBM AT, the IBM keyboard's LED's flash on when the keyboard gets power, then again much later, when the motherboard's power-on self test (POST) issues a reset command to the keyboard. )

    Quote Originally Posted by SpaceDog707 View Post
    I think something related to the N82231-2 or the keyboard controller might be the issue.
    You have to be careful when interpreting error codes/messages. For example, the DMA failure codes that you heard earlier, may have been because the CPU, at the time, was unable to communicate with a fully functioning DMA chip.

    Before your computer got to its present state, the symptoms were 'all over the shop' and varying.

    Quote Originally Posted by SpaceDog707 View Post
    Can anyone help me narrow down the fault a bit?
    First, is the CMOS battery okay? At these forums, we have seen some strange symptoms caused by a depleted CMOS battery.

    Have you done basics, such as a thorough visual inspection, and re-seating all socketed chips on the motherboard ?

    I suggest that you ignore the symptoms that you have seen in the past, and diagnose based on the current symptom (motherboard appears not to be starting).

    If I had your machine, an early thing that I would do is to isolate the cause to the highest level: motherboard, or PSU, or other (card, battery, drive, ...)
    I expect that with only a 286 motherboard with speaker connected (no CMOS battery connected, no cards, no keyboard, no switches) and PSU, that the POST is going to beep the speaker, complaining about something (missing video card, no CMOS battery, etc.).
    If no beeping, I would know that the cause has been narrowed down to the motherboard and PSU, and I would then try another PSU (known compatible and known working).

    And maybe try the motherboard/speaker/PSU out of the case, to rule out a case problem (e.g. metal support post contacting a motherboard trace).

  4. #4
    Join Date
    May 2006
    Location
    Melbourne, Australia
    Posts
    6,427

    Default

    Your second post has appeared.

    Quote Originally Posted by SpaceDog707 View Post
    The keyboard controller is an Intel 8742 (programmed as YAN25D05). I've got an 8041A from another PC, would the two be compatible? Can I use the 8041A for troubleshooting?
    Looking at the datasheet, from an electrical perspective, I cannot see why not. There is though, a remote possibility that the code in the 8041 is not suitable for your motherboard.

    If your 8742 had failed, and is the only failure, and the failure codes that you earlier pointed to do apply to your motherboard's BIOS, then I would expect that your motherboard beep 3-2-4 (for 'Keyboard controller failure').
    However, it is possible for the 8742 to fail in such a way that it negatively affects the address or data buses. You could test for that by powering up the motherboard after the 8742 has been removed.

    Again, before 'diving into' the motherboard, I think you should first prove that the current symptom of 'motherboard appears not to be starting' is caused by the motherboard.

  5. #5
    Join Date
    Dec 2019
    Location
    Sydney, Australia
    Posts
    7

    Default

    Hey thanks for your replies modem7.

    I should have said, at the moment I do just have the motherboard, psu and speaker connected. The keyboard, monitor and cards are all disconnected. I've taken it our of the case as well to check, but it's still the same, no bios beeps.

    I think the PSU is okay, with it connected to the motherboard I'm getting +5.1V, +11.85V, -11.94V and -5.1V. Does that seem correct? It's not a standard AT psu, the connector is a little bit different.

    Quote Originally Posted by modem7 View Post
    First, is the CMOS battery okay? At these forums, we have seen some strange symptoms caused by a depleted CMOS battery.
    It's pretty dead, but I've unplugged it for the moment. It's not the button cell type, it's a larger 'Panasonic BR-E3'.

    Quote Originally Posted by modem7 View Post
    Have you done basics, such as a thorough visual inspection, and re-seating all socketed chips on the motherboard ?

    I suggest that you ignore the symptoms that you have seen in the past, and diagnose based on the current symptom (motherboard appears not to be starting).
    That's a good point. I did re-seat all the socket chips, but not since the bios beeps stopped. I should give it another inspection as well, the boards quite dirty so I may have missed something.

    Quote Originally Posted by modem7 View Post
    Your second post has appeared.
    You could test for that by powering up the motherboard after the 8742 has been removed.
    Thanks, good idea, I tried it with the 8742 removed and with the 8041 in its place, but both didn't change anything.

    Would you expect bios beeps with the Intel N82231-2 removed? I might try powering it on in between reseating it.

    There's a bunch of PAL chips on there, I really hope one of those haven't failed.

  6. #6
    Join Date
    Dec 2019
    Location
    Sydney, Australia
    Posts
    7

    Default

    Okay, I found a group of vias which are badly corroded, I can't get a continuity measurement between two which I can see are meant to be connected. I'll have to try and clean them up.

  7. #7
    Join Date
    May 2006
    Location
    Melbourne, Australia
    Posts
    6,427

    Default

    Quote Originally Posted by SpaceDog707 View Post
    I think the PSU is okay, with it connected to the motherboard I'm getting +5.1V, +11.85V, -11.94V and -5.1V. Does that seem correct?
    Yes.

    Quote Originally Posted by SpaceDog707 View Post
    It's not a standard AT psu, the connector is a little bit different.
    Do you know if your motherboard requires a POWER GOOD signal (or equivalent) from the PSU? In most AT-class motherboards, the motherboard requires the POWER GOOD signal to go high in order to start.

    Quote Originally Posted by SpaceDog707 View Post
    Okay, I found a group of vias which are badly corroded, I can't get a continuity measurement between two which I can see are meant to be connected. I'll have to try and clean them up.
    I recently finished repairing a motherboard that had been damaged by a leaking battery. Apart from some damaged chips (near where the battery would have been), there were many damaged vias (near where the battery would have been).
    With the vias, the solder within the via had become concrete like, but only at the component side of the PCB. I have seen the same with vias under leaking aluminium electrolytic capacitors. Of the affected vias, some were still fully functional, but with others, I did not measure continuity between the top and bottom of the via. I decided to bypass the faulty vias using wire-wrap wire. Photo at [here].

    ( For the concrete-like vias that I determined were functional, it will be interesting to see if they fail down the track. And if more chips fail. )

  8. #8
    Join Date
    Jan 2005
    Location
    Principality of Xeon the Great State of Central New Jerky
    Posts
    1,523

    Default

    I had thought every 8086 or later uP needed a power good signal to operate.

  9. #9
    Join Date
    May 2006
    Location
    Melbourne, Australia
    Posts
    6,427

    Default

    Quote Originally Posted by tipc View Post
    I had thought every 8086 or later uP needed a power good signal to operate.
    No.

    For example, the power supply in the IBM PCjr does not generate a POWER GOOD signal (or equivalent). Instead, motherboard reset at power-on time is controlled by a capacitor/resistor/diode arrangement (like as shown at the far left of the diagram at [here]).

  10. #10
    Join Date
    Dec 2019
    Location
    Sydney, Australia
    Posts
    7

    Default

    Okay so I fixed up the worst of the vias. I also found a crystal (I think?) which had one leg broken off the board (you can see it in the photos above the power connector, next to the green trim pot, its designator is 15B2). But fixing these did not fix the issue.


    Thanks for mentioning the POWER GOOD, it doesn't seem to come from the power supply. The larger connector is just GND and +5V, the smaller connector is GND, 12V, -12V and -5V.

    But the diagram you attached gave me an idea. I probed some pins on the CPU, VCC is +5V, CLK is present and running at 16MHz.

    But RESET (active low) is oscillating at ~60KHz. Sometimes straight away as the board is powered up, or sometimes it starts at some random time latrer. When its not oscillating it's at 0V.

    READY is going berserk! A whole bunch of pulses, would be 3.8 MHz if they were continuous, but the pulse trains are all different lengths.

    I'm going on holidays tomorrow, so this will need to be put on hold for two weeks. It's just starting to get interesting too.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •