Original IBM PC (was Re: Prices to pay for old

From: Russ/Alice Blakeman <rhblake_at_bbtel.com>
Date: Fri May 22 20:05:37 1998

Tony Duell wrote:

> >
> > > >I hope that "16K" means "16/64K model" as opposed to "actually equipped with
> >
> > > >16K of memory."
> >
> > Yup, came with 16k soldered down and room to add DRAM up to a whopping 64k.
>
> So what people are saying is that if you find one with all 4 rows of 16K
> chips fitted, it becomes more valuable if 3 of the rows are pulled?
> Strange...

You can get one with just the fixed DRAM and insert more all you want, and if you get
one populated and want it stripped to the bare 16k then you just get out your "chip
puller" (the can opener on my utility knife is mine) and remove the extra rows. No big
deal. Finding the boards in either configuration is the main idea here, you can always
get 4164 DRAMs.

> This is the sort of thing which will make me leave the computer
> collecting world, alas. I can remember when computer collecting was fun -
> you went to a radio rally, or to the local(ish) junk shop and came home
> with a pile of bits. Fiddled with them, got them working, hacked about a
> bit, modified things a bit, and learnt a lot. Now, if we're going to have
> high-priced machines for no good reason, then I'm outa here...

It still is! It always will be a tinkerer's heaven to get a machine you know
absolutely nothing about and get it to a level where you know enough to be dangerous
and maybe even get it to work. That's the challenge, whether it's collecting old
radios, model trains, pooters, cars, etc.

> > Likewise the 64/256 board is the same, 64k soldered on, and room up to 156k. The
> > AT 5160 only goes to 512k on the board....
> ^^^^
> You mean 5170. The 5160 is the XT.

The fingers and the brain just don't get together much anymore ;-o

> Talking of XTs, there are officially 2 versions of that motherboard - the
> 64K-256K board and the 256K-604K board. It's not generally known how to
> turn the former into the latter, so here goes (assuming you want 640K in
> the end):

Wouldn't that hurt, 640k in "the end"??

> Remove the board from the case.
>
> Remove the 64K DRAMs from banks 0 and 1. Put 256K *1 (41256 or
> equivalent) chips in these 2 banks. Put/leave 64K chips in banks 2 and 3.
>
> Put a 74S158 multiplexer chip in the empty socket near the front of the
> board, same way round as all the other chips near it.
>
> Solder a link between (I think) pads E1 and E2 near the right hand edge
> of the the board.
>
> That's it - you now have 640K on the motherboard. I've done this to
> serveral XT's (and to my 5155 portable), with no problems. The schematics
> in the techref are otherwise indentical.

I have done this now to at least 8 XT's and it works fine on all of them, and it
appears that maybe IBM incorporated the same when they started manufacturing the 640k
series. The boards look very similar otherwise. When I had this nasty little Sanyo
MBC-55x machine we did a similar hack that made it 768k with a switch to bring it down
to 640k (for programs that couldn't deal with the overdose). Cost about $15 in parts
besides the extra DRAM chips and ungodly hours of leaning over a soldering iron making
delicate heat sensitive stacks of DRAMS 3 high and hooking up a pin of each chip to a
wire that went to a separate 74 series chip. After doiing that mod to over 20 machines
of friends and those I had mailed to me, I permanently swore off on the little silver
wastebaskets.
 --------------------------------------------------------------------
 Russ Blakeman
     RB Custom Services / Rt. 1 Box 62E / Harned, KY USA 40144
        Phone: (502) 756-1749 Data/Fax:(502) 756-6991
         Email: rhblake_at_bbtel.com or rhblake_at_bigfoot.com
           Website: http://members.tripod.com/~RHBLAKE/
                           ICQ UIN #1714857
                AOL Instant Messenger "RHBLAKEMAN"
       * Parts/Service/Upgrades and more for MOST Computers*
 --------------------------------------------------------------------
Received on Fri May 22 1998 - 20:05:37 BST

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:31:13 BST