IBM 210 error codes wth expansion unit

[RetroSpector78]

Hi,

I've got an IBM 5150 with a 5161 expansion unit. Both work great together (got the 2 cards + cable). The 5161 contains the 10MB seagate mfm hard drive + Xebec controller card and the IBM 5150 boots off the hard drive fine ...
The 5150 has 256kb ram (all banks populated on the mainboard). There is no memory expansion card in the PC or expansion unit.

It worked fine until yesterday that is ... After the unit was turned on for about an hour I went back to check on it and noticed the screen was blank and there was no reaction from the keyboard.
Turns out the PSU died on me (no longer outputting any voltage on the 5v and 12v rails, despite the fact that the fan was still spinning that I thought was weird).
However, I put in a new PSU and the PC was booting again ... but with error messages.

- First I got "1010 210" followed by parity check 1 , so I replaced the 5th chip in bank 1. PC started again.
- A couple of reboots later "2001 210" followed by parity check 1, so I replace the 1st chip in bank 2. The PC started again
- When the expansion card was inserted but the cable disconnected I sometimes get "0400 201" followed by parity check 2

Now, the first 2 errors are "normal" and so are the fixes. But should I be worried that all of a sudden 2 started failing (before the PC worked fine, and there were only a couple of reboots between the first failure and the second.
The partity check 2 and 0400 201 seems strange, as there is no expanded memory installed (the 5150 has 256kb ram onboard, and no memory expansion cards in the 51761).
Could this somehow be caused by the fact that there is an extender card in the PC ?

Thx.

[Stone]

Have you checked the voltages since you replaced the PSU? Although it seems unlikely it's not totally impossible that there is a voltage issue stemming from the replacement PSU, i.e., an overvoltage. Perhaps this contributed to the ram failures. This is just a far out guess.

[RetroSpector78]

On the replacement PSU I got 5.09 and 12.7 when the machine was running. Hooked up the mainboard and the 2 floppies.

On the PSU that is presumed dead I got millivolt readings on the 5v and 12v rail (but the psu fan was spinning).

[Stone]

On the PSU that is presumed dead I got millivolt readings on the 5v and 12v rail (but the psu fan was spinning).

Is it an AC fan? That would explain it's running in an otherwise dead PSU.

[modem7]

When the expansion card was inserted but the cable disconnected I sometimes get "0400 201" followed by parity check 2

The partity check 2 and 0400 201 seems strange, as there is no expanded memory installed (the 5150 has 256kb ram onboard, and no memory expansion cards in the 51761).
Could this somehow be caused by the fact that there is an extender card in the PC ?

Random 0400 201

For the '64KB - 256KB' type of 5150 motherboard, the '04' in '0400 201' corresponds to address 16K, i.e. an address within RAM bank 0.
I do not know how an 'uncabled' extender card could cause a random '0400 201'.

Random PARITY CHECK 2

Per the NMI diagram at [here], PARITY CHECK 2 is triggered by a card (one containing RAM) driving the I/O CH CK pin on the ISA connector (ISA in 5150 or ISA in 5161) to a low. PARITY CHECK 2 can only occur after the POST has completed (i.e. NMI's are disabled during the POST).

I have seen that sometimes, issues with 5150 motherboard RAM (which includes wrong motherboard switch settings) can result in PARITY CHECK 1, but I never seen PARITY CHECK 2 in those situations.

The extender card can also drive the I/O CH CK pin low, the circuit diagram showing that, as expected, the extender card simply buffers (via an open-collector driver) the I/O CH CK signal that it sees on its 62-pin cable connector.

Pull-up resistors on pins 12 and 13 of U22 ensure that, in the case of an 'uncabled' extender card that the I/O CH CK signal put onto the 5150's ISA bus by the extender card, is high (a 'pulled high via a resistor').

I do not have an extender card to check if I also see the random symptom.

If you were curious, you could put tape onto the I/O CH CK pin of the extender card's edge connector, then see if the random PARITY CHECK 2 persists.