IBM 5155 analogue display fault

Tony Duell ard at p850ug1.demon.co.uk
Thu Jun 2 17:26:32 CDT 2005


> > I would argue that history is lost if _any_ change is made, and 
> > certainly if any part is replaced.
> 
> Much like modifying PCBs... :-)

Absolutley. I do NOT receomend modifying PCBs for no good reason. I do do 
it, though, when there are significant benefits that make the machine 
more useable.

In fact I am going to recomend you do a mod to the mainboard in your 5155 
(all of them). I am going to assuem they;'ve got 256K RAM, almost all have.:

Remove the motherboard from the case (this is the worst bit of the job, 
you have to take out the drives first to get to the screw in one of the 
brackets). 

Remove the 4164 64K*1 DRAMs from banks 0 and 1 of the motherboard and 
replace them by 41256 256K*1 DRAMs.

Fit a 74F158 (or 74S158) multiplexer chip in the empty socket towards the 
front right of the board, pointing the same way as the TTL chips round it.

Solder a jumper wire between pads E1 and E2 at the RHS of the board, 
about half way down.

You now have a 640K machine. 

This mod is sort-of official in that IBM did sell XTs with 640K on the 
mainboard, and the revised schematic is printed in later TechRefs. I 
don't think any 5155s were shipped like that, though.

However, I recomend doing the mod because very little software will run 
in 256K (much more will run in 640K), memory expansion cards are hard to 
find, and slots are tight on the 5155 anyway (with the CGA card and FDC 
fitted, you only have one full-length slot left, the others can be very 
short indeed). 

> 
> Remember this is a machine that's going to be used by children at events
> (well, assuming the primary one is out of action); it's going to have a

What's that go to do with the choice of internal screws...


> very different life to static exhibits or operational machines that
> aren't left totally open for people to play with.
> 
> > Also, please remember that this machine is owned by a museum. I would 
> > expect museums to have higher standards for preserving originality than 
> > most enthusiasts. Pity this doesn't seem to be the case.
> 
> We have several 5155s, all in original state. Unifying the screws on one

Well, what will you do when a monitor fails on one of the others? Put the 
right screws back, or bodge it? If the former, you will have to get the 
Bristol Spline tool. So why not get it now? In fact you darn well ought 
to have a set anyway.

> If at some stage they *do* become rare then fitting original screws is
> possible if needs be as we have reference machines (and very likely the

Provided those have not been changed!

> relevant technical info on the shelf)

Well, it's not in the Guide to Operations (which does cover pulling the 
case, it does cover inserting expansion cards and the like). It's not in 
the TechRef. It may be in the HMS (Hardware Maintenance and Service) 
manual, I don't have that one, but the HMS manual for the PC/AT does 
_not_ describe what scews go where. It assumes you put back the screws 
you take out. DO you have the HMS manual?

One final point (I hope). I am, as I am sure you know, an active member 
of HPCC. A few years back, we looked at the HP12C calculator, which was 
the machine that HP had made for the longest time (I think we looked at 
it on the 21st aniversary or something). We got as many of the machines 
together as we coold, and looked at the changes that were made over the 
production run. 

Of course there were many. Early machines had a separate logic module 
containing the display and ICs that fitted into the case and connected to 
the keyboard PCB via a zebrastrip. Later machines had it all on one 
board. but still with 2 ICs (the NUT CPU and the R2D2 ROM/RAM/Display 
Driver). Then there was a single IC version. And so on.

Now one, and only one, of the machines we looked had had TX6 Torx screws 
holding the case together rather than the Pozidriv used in all other 
machines. I am pretty sure that was original, and we noted it as a 
separate variant. 

We do _NOT_ want a variant of the 5155 IBM Portable PC with Pozidriv and 
not Bristol Spline screws in the monitor suddently springing into 
existance to confuse averyone....

Incidentally, have you figured out what's wrong with this machine yet?

-tony


More information about the cctalk mailing list