Novag Aquamarine RISC II 26.6MHz
First repair

On this computer, part of the sensory board was failing.

Columns B and H are found to fail.

It is disassembled:

We make sure that the problem is not in the motherboard. To do this, the motherboard is supplied with 9V DC, the display is positioned so that we can see it well, and rows are short-circuited with columns as the sensor membrane would do:

The motherboard is working properly.

I also check the maximum resistance between row and column at which the microcontroller still detects the signal. The maximum resistance is 10Kohm: above this value, no matter how much we connect a row with a column, the microcontroller does not detect it.

Now I proceed to examine the sensory board:


These 'black' areas are degraded and their resistance to the current flow is higher than normal.

After carrying out some tests on the ribbons, we came to the conclusion that there must also be some areas in bad condition on the inside, so we proceeded to disassemble the sensory board, with the help of hot air and patience!

There are signs of deterioration of the conductive tracks at many points:

If we don't want the film to accidentally stick back in place, it is a good idea to interleave a "laminated" piece of paper:

We prepare the tools and materials for the repair: silver paint, a 6/0 brush, a blade, blotting paper, etc.

The repair begins, bearing in mind that:

  1. We must try to cover the damaged area of the track, but also cover a little of the healthy area.

  2. The solvent in the silver paint dissolves the conductive paint on the track, so we must paint without pressing on the area.

  3. Once the silver paint has dried, it is brittle, so try not to bend the repaired area.

Detail of a painted area.

All affected areas are painted and the top film is glued back on:

Now I repair the connection area of the ribbon. The two easiest ways are:

  1. If the ribbon is long enough, the affected area is cut out and the tracks are conditioned to connect properly.

  2. Silver paint is used as before.

Option two is chosen in this case.

We use a cutter to remove, very carefully, a couple of millimetres of the track coating, to obtain a "healthy" piece of track.

This is what it looks like after painting. Try not to bend the ribbon, as this could crack the silver paint.

We assemble the computer and test it:

The result: One column is still failing (failed repair). Possibly because the solvent in the silver paint has "dissolved" the conductive paint at some point.

The whole process could be repeated, but I discard it because of the following:

  • The deteriorated areas are those most exposed to the air, as the tracks are much more blackened on the perimeter of the board, in the exposed connection area, etc. Now that we have disassembled and reassembled something that was glued, it is clear that the sensory board will be less airtight and it is only a matter of time before it fails again.

  • Silver paint is fine for use on copper or non-bending areas, but here we find that its solvent will remove the conductive paint from the connection ribbons. Also silver paint, once dry, will crack if the ribbon flexes.

Conclusions: This repair system might be effective when the damage is localised, but on a machine model that we have found to suffer from defective conductive tracks, all we achieve is to delay the problem. This does not mean that this method can be applied satisfactorily to other machines!

I decide to try another method: SECOND REPAIR