Don't take this article too serious: there are still many PDP-11/34a out there.
It's surely cheaper to get a new CPU backplane instead of repairing one.
But sometimes you miss the point where you have to let something go, so this story continues ...
Well, as repair strategies, the following ideas came up:
1. Manual Patching
Procedure: Jiggle all (1944!) pins strongly, prepare a wiring diagram, and measure all connections. Afterwards resolder all soldering joints that are accessible with the iron, rewrap the remaining connections.
Well, you could do that once in your life. However, this is a systematic error in the production of the backplane by DEC … It can be expected that other machines have the same problem yet.
2. Automatic Soldering
- Because the number of soldering joints is much too high and many soldering joints are not accessible with a regular soldering iron, manual resoldering has to be ruled out.
- Classical soldering baths are not practicable, because the wire-wrap structure would melt in the solder wave.
- Condensation soldering is an industrial process of heat up the soldering joints to melting temperature with condensation vapor. The condensate only precipitates on metal surfaces and therefore only heats up the wire-wrap pins and the soldering joints. The whole bottom of the backplane is exposed to the vapor.
- The question remains if the contact between the pins and the solder can be restored without adding flux. The pins are very resistive to mate with the solder.
3. Chemical Processes
The idea is this: The backplane is send through several baths that clean the corroded pins in a first step and restore the contacts afterwards. With only a few process steps all pins can be treated at once.
The following processes already have been tried:
3.1. A recipe for cleaning sterling cutlery: The backplane was laid into a batch with warm salt water together with aluminum foil for three hours. Something did actually happen: Blisters occurred at the solder joints, the aluminum foil became thinner and brittle, but else no changes could be seen.
 |
 |
Even a complete nerd sometimes uses his kitchen!
3.2. The backplane was treated with silver immersion bath, a common solution from the hardware store for cleaning tarnished silver. No changes. Apparently, the black deposits on the wire-wrap pins are not silver sulfide.
3.3. The backplane was cleaned with salt acid. While pure solder from a roll is becoming shiny, the soldering joints on the backplane became black! Attempt aborted.
3.4. The backplane was treated with "chemical tin" (Octamex). The backplane was set into solution once for 20 minutes and once for 10 hours. The right exposure period is probably reached if you don’t hear any outgassing sound anymore after shaking the wet backplane.
"Chemical tin" usually deposits only on clean copper, nevertheless, the bad connections on the NPG chain seemed to have closed. However, after mechanical movement of the pin on the top of the photo, the contact was open again. Nevertheless, I wouldn’t say this process is a total failure.
4. Galvanic Processes
For restoring the electrical contacts also galvanic processes can be taken into account. Metal is galvanically coated onto the soldering joints until the hairline cracks are closed again.
Fortunately, many metals a suitable for galvanic coating: gold, silver, zinc, nickel, copper, tin. All metals have sufficiently high conductivity in order to fill hairline cracks. Therefore, that process can be chosen which promises the best adhesion on polluted solder, and which is as inexpensive and nontoxic as possible.
Configuration:
- On the socket side, contact strips must be plugged that connect all slot pins among each other. These contact strips can be won from sawed-off contact fingers of old flip chip boards. The socket pins are connected among each other and then to the negative terminal of the current source.
- Only the wire-wrap pins and the soldering joints are in the solution with the metal electrolytics.
- As anode a sheet metal that is connected to the positive terminal of the current source is laid on the bottom of the electrolytic bath.
... perhaps to be continued ...