I agree on both points, but if the cable is not longer than a foot or so,
it's likely to work fine. Certainly it's likely that one could figure out a
better way, on a case-by-case basis, but I just wanted to throw a very
general solution at the problem, one which everyone would easily understand.
It won't be hard to improve on that one.
Dick
-----Original Message-----
From: Tony Duell <ard_at_p850ug1.demon.co.uk>
To: Discussion re-collecting of classic computers
<classiccmp_at_u.washington.edu>
Date: Monday, November 01, 1999 3:11 PM
Subject: Re: EPROM issues, who can burn?
>>
>> I agree, mostly, with what you've said here, Tony. The easy fix, in many
>> cases will be to program the code into several larger but readily
available
>> and easily programmed parts, then wire an adapter from the target board's
>> socket to the physical EPROM on the board on which the EPROM resides,
with a
>> pair of appropriately sized IDC ribbon cable headers crimped on the ends
of
>> the cable. That should work, even if you have to hang the adapter board
>
>I'd be a little careful about hanging lengths of ribbon cable off an
>EPROM socket (the signals on which may well not be buffered). I am not
>saying it won't work -- many times it will, but I'd not do it unless
>necessary.
>
>Many times there's enough space around the EPROM socket and/or between
>the boards in the cardcage to allow for a simpler replacement method. The
>fact that most EPROM pinouts are similar helps here. Tricks include :
>
>1) Bending out those pins of the EPROM that are different (high-order
>address lines, for example), plugging the rest into the socket and
>soldering wires to the ones you've bent out
>
>2) Replacing the socket on the board with a wire-wrap socket. Wrap wires
>around the pins that are different, and cut those pins short. Then solder
>the remaining pins to the board so that the 'different' ones don't touch
>the tracks (the socket is probably about 0.5" above the board). Solder
>the ends of the wires to appropriate points. Insert the EPROM.
>
>3) Make an adapter. In the UK you can get pin strips designed to plug
>into turned pin IC sockets. What I normally do is replace the EPROM
>socket on the PCB with a turned-pin one of the same size. Then take a
>piece of stripboard and solder a socket for the new EPROM to it (cutting
>the tracks down the middle). Then solder pin strips to the track-side of
>the stripboard with one spare hole between the socket pins and the pin
>strips. Then cut tracks for the pins you don't want to connect straight,
>and then solder wires (wire-wrap wire is good for this) on the track side
>of the stripboard to make the necessary connections. Plug the adapter
>into the socket on the PCB and plug the EPROM into the socket on the
adapter.
>
>4) Ditto, but etch a PCB rather than using stripboard.
>
>5) Read the tech manual for the machine (!). It's not uncommon for
>machines that use mask-programmed ROMs to have some way of using EPROMs
>instead. There may be links for this on the board, for example.
>
>> from a hanger in the rack. Packaging problems are what the REAL
engineers
>> work hardest at, while the youngsters conjure up the fancy circuits.
This
>> is mostly a packaging problem. The higher speeds of today's common and
>> cheap parts will compensate for the few nanoseconds lost in cables, even
if
>> some form of termination has to be introduced.
>
>I'd be more worried about the stray capacitance and noise pickup from the
>ribbon cable than the delay it introduces.
>
>-tony
>
Received on Mon Nov 01 1999 - 19:04:39 GMT