IBM 5100 Restoration in progress

Christian Corti cc at
Wed Mar 18 05:27:51 CDT 2015

On Tue, 17 Mar 2015, Santo Nucifora wrote:
> Update:  Unfortunately, I think I'm at a dead end.  After cleaning
> everything up, I am still at the "A" in the "Bring Up" program that isn't
> very far but it's better than garbage.  I removed both memory cards and get
> garbage so the memory cards might be working (I have two and tried both
> individually).  I think my problem is in the "Basic I/O" card as indicated
> in the Maintenance manual.  I am also not able to get into the Diagnostics
> screen because it must be failing too early.  Either that or my keyboard
> isn't working.  It's hard to tell.

What you don't say is whether the machine just halts, or stops with a 
machine check.

Here's the relevant part of the Bring Up test 'A' from the 5110. Since I 
still haven't got a 5100 I can't verify what the 5100 does, but I suspect 
that it must be quite similar at this point.

                         ; --- Test A ---
                         ; Bus In bit test

0078    251E            MOVE R5, $3C
007A    81C1            LBI R1, #'A'
007C    7150            MOVB (R5)+, R1
007E    8FFF            LBI R15, #$FF
0080    00FE            GETB R15, $0
0082    CF04            SS R15
0084    0000            HALT


So essentially, the routine reads the byte from I/O device 0 (the 
processor or display adapter card) and halts if it isn't $FF.
You should be able to check what R15 contains after the halt by switching 
to the register display.
According to the 5100 MIM, the 'A' routine is described as "Test bus in 
for the ability to turn all bits on". So apparently you have a stuck bit.
Are you shure that you don't have any external devices attached to the 
machine? And check the top connectors and all the jumpers.
Then you might find the stuck bus in bit with a scope/logic probe by 
probing the external connector A2.

> I'd be happy to try to replace some of the logic chips but IBM uses their
> own numbering scheme.  Does anyone know of a way to read them?  I assume
> they are normal logic chips but with IBM numbering in most cases?

The normal "black" TTL ICs, yes, they are standard ICs. For a (incomplete) 
list, have a look at our FTP server at /pub/cm/ibm/ibmparts.txt

> You can find new pictures here:

One note: the terminator box is only needed when you use the 5106 drive 
*without* the 5103 printer, because the printer has a terminator built in 
and thus must be the last device in the chain.


More information about the cctalk mailing list