"Dead" IBM PC 5160 (256k-640k) motherboard

[jesolo]

It shouldn't. If memory (sic) serves, the 280 only applies to the memory on the planar. Other cards must do their own memory parity check.

I'm assuming that you've set the configuration switches correctly for the display adapter you're using.

Yes, made very sure of my settings - any ideas what might cause the Supersoft Diagnostics to just freeze up at the MDA (or CGA) memory test?

 

[modem7]

Yes, made very sure of my settings - any ideas what might cause the Supersoft Diagnostics to just freeze up at the MDA (or CGA) memory test?

Are those IBM cards, or clones ? I have heard of some clone MDA/CGA cards causing a corrupted display (for all tests) when used with the Supersoft/Landmark diagnostics. I have such a card somewhere. But I have yet to hear of a clone card causing a lockup.

If there was bad video RAM, or a refresh issue (where video RAM is dynamic rather than static), then I would expect:

* If the video card does not contain parity generating/checking circuitry, data read back is different to that written, fail the video RAM test.

* If the video card does contain parity generating/checking circuitry, then per [here], the video card would take pin A1 low, resulting in an NMI to the CPU, resulting in a jump to the Supersoft/Landmark's NMI handler. Fail the video RAM test.

 

[jesolo]

Are those IBM cards, or clones ? I have heard of some clone MDA/CGA cards causing a corrupted display (for all tests) when used with the Supersoft/Landmark diagnostics. I have such a card somewhere. But I have yet to hear of a clone card causing a lockup.

If there was bad video RAM, or a refresh issue (where video RAM is dynamic rather than static), then I would expect:

* If the video card does not contain parity generating/checking circuitry, data read back is different to that written, fail the video RAM test.

* If the video card does contain parity generating/checking circuitry, then per [here], the video card would take pin A1 low, resulting in an NMI to the CPU, resulting in a jump to the Supersoft/Landmark's NMI handler. Fail the video RAM test.

The CGA card is the original IBM CGA card (later revision).
The MDA card is an HMC6311 clone card.

Both cards exhibit the same symptom - complete lockup of the system once it reaches either the MDA or CGA test.
In some cases (with the CGA card), I get a garbled screen and then it freezes. In other cases, the screen blacks out.
But, I always end up with a complete lockup - no messages are displayed at those tests because, the system locks up beforehand.

 

[Malc]

Just an update of my progress thus far:
I burned a copy of the "Supersoft Diagnostics" ROM to a 27C512 EEPROM (by making a "duplicate" copy of the ROM contents) and inserted this into U18....

Just a thought, Are you 100% sure that the 'EEPROM' you are using is compatible with the 5160 motherboard you have, My own experience is that not all 'Eproms' are compatible with the 5160 motherboards, The same goes for the 5162 and 5170, What works in one board may not work in another, Also you may find that the board may boot but display some strange errors, Bios Rom and Memory. I've had it before using various EPROMS from different manufactures.

 

[jesolo]

Just a thought, Are you 100% sure that the 'EEPROM' you are using is compatible with the 5160 motherboard you have, My own experience is that not all 'Eproms' are compatible with the 5160 motherboards, The same goes for the 5162 and 5170, What works in one board may not work in another, Also you may find that the board may boot but display some strange errors, Bios Rom and Memory. I've had it before using various EPROMS from different manufactures.

It is a possibility but, wouldn't this then exhibit other symptoms?
I'm using the Wimbond W27C512-45.
I do have 27C256 EPROM chips that I can also flash and try out but, I don't think that is the problem.

 

[Malc]

It is a possibility but, wouldn't this then exhibit other symptoms?
I'm using the Wimbond W27C512-45.
I do have 27C256 EPROM chips that I can also flash and try out but, I don't think that is the problem.

I've never tried using EEPROMS in my IBM Boards, I've always used UV Erasable EPROMS, Modem7 has a page on his website with some known to work Eproms in the 5160. If it were my board i'd be much happier knowing that i'm using a known compatible and working EPROM, I tend to use AMD or INTEL 27256 now.

 

[jesolo]

I've never tried using EEPROMS in my IBM Boards, I've always used UV Erasable EPROMS, Modem7 has a page on his website with some known to work Eproms in the 5160. If it were my board i'd be much happier knowing that i'm using a known compatible and working EPROM, I tend to use AMD or INTEL 27256 now.

Flashed an ST M27C256B EPROM chip with the Supersoft Diagnostics - unfortunately, I'm still getting the same result (complete lockup at the graphics card memory test).

 

[Malc]

Flashed an ST M27C256B EPROM chip with the Supersoft Diagnostics - unfortunately, I'm still getting the same result (complete lockup at the graphics card memory test).

Oh well it was worth a shot, I do remember having an IBM MDA card that did not work with the supersoft rom but i think all i got was a blank screen when using the SS rom, I've never had one lock up at memory test though, unfortunately can't think of anything else.

 

[jesolo]

Oh well it was worth a shot, I do remember having an IBM MDA card that did not work with the supersoft rom but i think all i got was a blank screen when using the SS rom, I've never had one lock up at memory test though, unfortunately can't think of anything else.

I think I should probably still look at the "16K Critical Memory Region", as per modem7's suggestion. The fact that it is failing randomly, tells me that one needs to resolve that issue first before moving onto diagnosing other problems on the motherboard.

 

[modem7]

... Modem7 has a page on his website with some known to work Eproms in the 5160.

Noting that the 'what EPROMs work/fail' information there is specific to the early 5160 motherboard. For example, the NEC uPD27C256 does not work in the early 5160 motherboard, but it works in the later 5160 motherboard

... The same goes for the 5162 and 5170, What works in one board may not work in another, ...

If you do encounter a particular make/model of 27256 (including 27C256) that does not work in the IBM 5162 or 5170, please let me know. I will verify the situation, then put the information on my website.

Oh well it was worth a shot, I do remember having an IBM MDA card that did not work with the supersoft rom but i think all i got was a blank screen when using the SS rom,

Per [here], I see that with the IBM 5170 version of the diagnostics.

 

[modem7]

I think I should probably still look at the "16K Critical Memory Region", as per modem7's suggestion. The fact that it is failing randomly, tells me that one needs to resolve that issue first before moving onto diagnosing other problems on the motherboard.

Yes. Maybe the diagnostics are using that area in some way (e.g. variables).

 

[jesolo]

Thought I'd post some pictures of my motherboard.
Strangely, on the board itself it's actually "printed" as being 64-256KB motherboard but, it's covered with a "sticker" (which has party come loose over the years).
On the 256-640KB motherboard, are the wires at the bottom of the motherboard supposed to be there?
What's the rating of the capacitor at C53 and the resistor at R6?

 

[DDS]

I'm going to throw my two cents worth in here. I'll let you all decide if it's even worth that. We sometimes forget that our digital devices aren't all clean and logical like in boolean algebra unless they get as much stable clean power as they want. Whenever one of mine starts acting quirky I start thinking about the power supply. This might be a good time to check yours for noise and ripple under load.

 

[jesolo]

I'm going to throw my two cents worth in here. I'll let you all decide if it's even worth that. We sometimes forget that our digital devices aren't all clean and logical like in boolean algebra unless they get as much stable clean power as they want. Whenever one of mine starts acting quirky I start thinking about the power supply. This might be a good time to check yours for noise and ripple under load.

That was one of the first things I tested - all levels within tolerance. To be sure, I've tested the motherboard on 3 different PSU's .

 

[Malc]

Thought I'd post some pictures of my motherboard.
Strangely, on the board itself it's actually "printed" as being 64-256KB motherboard but, it's covered with a "sticker" (which has party come loose over the years).
On the 256-640KB motherboard, are the wires at the bottom of the motherboard supposed to be there?

Yes, IBM changed the wiring of sockets U18 / U19 more info [ Here ]

What's the rating of the capacitor at C53 and the resistor at R6?

I just had a look at a couple of my 64 -256 boards and C53 is 100V DC .01 Uf or 10 nF and R6 is 33 Ohm, Orange - Orange - Black - Gold

 

[jesolo]

Yes, IBM changed the wiring of sockets U18 / U19 more info [ Here ]

I just had a look at a couple of my 64 -256 boards and C53 is 100V DC .01 Uf or 10 nF and R6 is 33 Ohm, Orange - Orange - Black - Gold

Thank you for the info - my eyes were deceiving me a bit and I couldn't quite make out the colour codes.

 

[jesolo]

If the D8284A clock generator is faulty, could this cause the problems that I'm experiencing? Random memory errors & the graphics card memory test causing the system to hang?
I've noticed that the chip is getting hot while the system is on.

 

[modem7]

If the D8284A clock generator is faulty, could this cause the problems that I'm experiencing? Random memory errors & the graphics card memory test causing the system to hang?

"Faulty" as in intermittent or unstable. I do not know.

I've noticed that the chip is getting hot while the system is on.

For the past couple of hours, I have had a 5160 motherboard running the Landmark/Supersoft diagnostic ROM. A few chips are warm, with the CPU and D8284A measuring the warmest. The D8284A showed, using a laser type thermometer, as about 35 degrees Celsius (measured at center of top).

 

[modem7]

I have done some further experimentation with the Supersoft/Landmark diagnostic ROM.

Per [here], if the 16K CRITICAL MEMORY REGION TEST (before it displays "FAILED") always reports address 05040, then the cause appears to be either:
* a faulty RAM chip in bank 0; or
* faulty parity generation/comparison circuitry; or
* faulty circuitry that always results in a failed write operation to RAM; or
* faulty circuitry that always results in a failed read operation from RAM.

* I do not believe that the 16K CRITICAL MEMORY REGION test depends on operational refreshing.

I tested that by tying pin 8 of U69 high. To my surprise, the 16K CRITICAL MEMORY REGION test always failed.
The results were not completely random, possibly because the RAM chips are holding their contents well.
In something like 30 motherboard power-ups:

* The reported address varied, but not always. The ones that I saw were 5040/5041/5042/5046/5049/5073. The address usually changed when I waited a long time (say, a few minutes) before powering up the motherboard again.

* Same behavior for the bits failed. And usually only one or 2 bits.