As you look at my last photo, the four switches on the left side are active. . .sic. . .(only one). . .at one time, and become unlatched as another
is engaged. The four switches on the right of the photo work independently. It is those switches that I think work as you explained.
Well after a bit of NASA enhancement technique along with , Digital-dithering-averaging-supposition, I now come up with your photo view, as is now provided.
I do see that the 4 switches on the left have been upgraded from Schadow's prior use of a blue styrene
plastic tab, up to a hardy white Delrin plastic part (it is used in place of the pivoting spring wire), as is now seen on the four LEFT switches.
Using the [RED] referencing lines.
The four on the RIGHT are using the exact pivoting spring wire part which I described,
( [YELLOW] referencing lines, of which, the [RED #1] seems to be the most visible ).
Also you say that they are all working O.K., whereas the [MAGENTA] referencing lines to the frontal springs reveal that areas frontal springs are the ones in the worst corrosive shape. . .particularly [A].
However, it’s the left side switches that are the problem.
Some of these buttons do not stay depressed when pushed.
Since it is solely the four on the LEFT side of the bank, and individually, when trying then one at a time,
does it seem to be that possibly one switch is the problem child, with the other three, repetitively latching and unlatching as they should ?
I can only access the switches in the position you see in the photo since the card they are mounted on I think is soldered in. I did clean the switches with a spray cleaner called Miller-Stephenson (MS-230) Contact Re-Nu that contains no lubricant. Because of the switches location, it is difficult to examine them closely. So, I can clean (or spray) and assume I am doing some good.
I can certainly see your cramped quarters that the switches are mounted in.
For the four on the left to be in a “only one latches in---three are popped out” situation, there certainly has to be a side to side lateral mechanical coupling medium, be it a very small metal plate /shim / wire.
In those cramped quarters I would have to be using my small dental mirror and that STRONG hi intensity light for inspecting that mechanical situation. It could be located in 4 possible lateral positions around the switches in my previous ”circle reference” area .
If you look at all of the switches you will see that there is a metal rail to the front and another one just behind it that runs the length of the switch bank and I remember it as being in a square U bend with the bend being at the bottom of the form. All of the compression springs are pressing into the front rail, whereas the rear rail is set back about ~1/4--- 3/8 inch for accommodating the normal side mounting ears that those switches use in single mount situations. In this case you can see tabs at the top that are skewed 45 degrees for holding the switches in a vertical mounting restraint.
If a micro mirror view will show it, see if the mystery interconnect mechanics might be hidden down within that channel, if not visible outside of the rails, on their other two sides.
I probably have tech info for that set in my SAMUELS 1473-4, 1574-3, 1583-4 or 1453-4, but I surely think that they will NOT be giving a photo of any high detail, such as we now have.
That old info is now storage archived at “The mule barn” and will not be out there ‘til next month and will check out, but don’t expect much relevant mechanical info / photos to be therein.
BTW on re-referring to the photo below:
With you then looking at the very extreme left switch and then looking downward-left to the series of solder connections,on that PCB. . . .or . . .how you say. . ."card". . . there is a cluster of 7 joints at extreme left, be sure to check the extreme left joint for a "floating" "cold solder" "ring joint", and possibly, the one just to its right. . . .the five to the right look Mil spec..
73's de Edd