pdp8/L restoration time, part 1: switches

[Adrian Stoness]

If u hunt around u can find similar switches in e-waste or even new. Then
make some baskets u can attach. Like they did or if u got patience u can
take those right apart and rebuild as the pins sit in a cardboard like
material the plastic switch slides on with metal inserts for making
contact.  Need to get around finishing rebuilding mine.

Mine had been ripped off but it u can see how they are assemble here

https://www.flickr.com/photos/1ajs/16586222513/in/album-72157651577592077/

On Sat., Dec. 4, 2021, 3:46 p.m. Chris Zach via cctalk, <
cctalk at classiccmp.org> wrote:

> I've decided it's time to fix my second pdp8/L. First one is missing (I
> think it's at the bottom of the closet) so I'll work on this one for
> awhile.
>
> First problem: Testing the lamps and switches. If the switches don't
> work or the lamps are out it's not going to be easy to troubleshoot.
> Currently I'm missing 5 of the plastic switches, and I have 2-3 switches
> with a broken pivot.
>
> Since I was going to need new plastic switch covers I fired up the 3d
> printer, downloaded the STL file for making them (thank you for putting
> that up!) and ran off a few for testing. It helps a lot to set the
> bridge angle value to 60% or so, that allows a nice solid bridge for the
> two "teeth" that engage the rocker switch. Also don't bother with PLA:
> The plastic pivots are too weak and break pretty quickly. Use PETG,
> which requires higher heat (240 instead of 200 at the head and 70
> instead of 60 at the base) but it is more than strong enough for the
> pivots to go in without issues.
>
> Testing the switches I found that one of the data/address switches was
> reading high ohmage (20) when closed in the down direction and two of
> the data entry (spring loaded) switches were intermittent. Sprayed in a
> bit of Deoxit, and they now switch cleanly and without bounce. That
> would probably have driven me crazy if I tried to test with that.
>
> Next step: Check out the power supply.....
>
> C
>