PDP-8 Straight 8 restoration - failure modes

[Bob Rosenbloom]

On 3/30/2020 12:02 PM, Josh Dersch via cctech wrote:
> On Sun, Mar 29, 2020 at 1:48 PM Brendan McNeill via cctech <
> cctech at classiccmp.org> wrote:
>
>> Here in NZ and around the world many of us are in lockdown and spending
>> more time on our computers, if that were possible.  I have just completed
>> the restoration of a PDP-8 Straight 8 which I believe is the only one in
>> New Zealand.  You can view the restoration story and find appropriate
>> resources here:  https://pdp-8.nz <https://pdp-8.nz/>
>>
> Very nicely done, and an excellent write-up.  That core memory repair was
> amazing. It's interesting to read about a PDP-8 with such a high failure
> rate -- I've personally worked on two straight-8 systems (one at LCM+L, and
> one in my personal collection) where relatively few component failures were
> found.  I wonder what accounts for the difference -- batches of diodes more
> prone to failure, the environment the machines were stored in, or the
> number of years of service...
>
>
>> While it plays Chess, it would be great if someone wanted to write (say) a
>> Prime Number Generator, or some other application and email it to me off
>> list.  I have Focal-69 and can probably source other languages for this
>> wonderful old machine with 4K of memory.
>>
> There's a 4K LISP as well, though it's a pseudo-LISP-1.5 dialect so it's a
> bit different.
>
> - Josh
>
>
In restoring my PDP-8/s, diodes were the problem. Turns out they have 
steel leads that rust. Easy to see with a microscope.
Badly rusted ones broke the glass body.

I had to replace more than 200. Most turned into ~600 Ohm resistors, 
some open circuit. I'm sure I'm going to have to
replace all of them over time. So how (and where) the machine was stored 
probably plays a big part in it's reliability.

Bob

-- 
Vintage computers and electronics
www.dvq.com
www.tekmuseum.com
www.decmuseum.org