915 terminal

[jbdigriz]

Ok, I see now what you mean about a possible conductive glue or adhesive holding the silicon die to the PCB.

 

I need to research that ceramic material and how the traces were made on it, the best way remove the chip for testing without damage, and the best way to remount it. If, in fact, the ICs are still good.

 

Your plan looks like the most direct and timely approach to get a working keyboard

[Jimhearne]

Well, I now know that pcb material reduces the sensitivity of the hall effect sensors.

The position that seemed to work fine with the sensor on the end of wires didn't work so well once it was mounted on the pcb.

So another batch of tiny pcb are on there way with the sensor location moved.

 

Jim

 

[+Ksarul]

4 hours ago, Jimhearne said:

Well, I now know that pcb material reduces the sensitivity of the hall effect sensors.

The position that seemed to work fine with the sensor on the end of wires didn't work so well once it was mounted on the pcb.

So another batch of tiny pcb are on there way with the sensor location moved.

 

Jim

 

Maybe that sensitivity also explains why they would have mounted things in raised-blob-mode BITD.

[jbdigriz]

7 hours ago, Jimhearne said:

Well, I now know that pcb material reduces the sensitivity of the hall effect sensors.

The position that seemed to work fine with the sensor on the end of wires didn't work so well once it was mounted on the pcb.

So another batch of tiny pcb are on there way with the sensor location moved.

 

Jim

 

I hope it doesn't take too many iterations to get it dialed in.

 

Trying to get back to the DS1 here.Been swamped with other things.

 

Edited October 26, 2022 by jbdigriz

[Jimhearne]

8 hours ago, Ksarul said:

Maybe that sensitivity also explains why they would have mounted things in raised-blob-mode BITD.

It could be, i guess the hall effect sensors probably weren't as sensitive back them.

Mind you, they were still sensing the magnet through the ceramic "PCB" as i am.

I have a couple of different sensitivity hall sensors here, one of the others triggers even before you push the button, it's really sensitive.

The one i'm using now does trigger on the existing pcb but only at almost full travel, maybe 0.5mm travel left.

This is a lot more that the original switches which triggers at about half travel.

And i was worried this long travel might make the keyboard awkward to use and also with tolerances in the sensors, magnets and housings that i might get some that never triggered.

So i have moved the sensor up the PCB by about 2mm which fingers crossed should make it similar to the original.

 

Jim

 

 

[Jimhearne]

Rev 2 of the pcb works nicely , operating at about the same travel as the original.

Some simple code on the PIC 10F200 to create the required 50uS pulses on 2 outputs to match the original switch output.

 

Now to buy parts and make up another 90 odd pcb, I'm not doing all those by hand.

A few switches need different code as the shift, control etc keys do have a continuous output when pressed.

 

I know there is solder blob on the Hall effect chip, 2 of the pins are used on that side so it didn't mater.

 

Jim

[jbdigriz]

Neat. Are you planning to market these?

[Jimhearne]

I'm not sure there will be a massive demand for them but i will be happy to share pcb's and code and maybe make up a few boards if anybody does need some.

I suspect that so many switches failed on this keyboard as it was a very early batch, the keyboard is serial number 0000019, dated 1977

I could just replace the failed ones but since it's so hard to get the switches and keyboard apart i will replace them all on this one.

I will keep the original pcb's with the terminal, working and non working.

 

I'm undecided what to do about the pins / legs for the pcb.

The originals had stamped pins soldered on, but i'm thinking to just use tinned copper wire (thats why i left a via in the pads).

If i use the original legs i will have to take them from all the still working pcb's as well which seems a shame.

 

Jim

 

[jbdigriz]

6 hours ago, Jimhearne said:

I'm not sure there will be a massive demand for them but i will be happy to share pcb's and code and maybe make up a few boards if anybody does need some.

Awesome. I feel like that's going to be very helpful here once i tear my keyboard down and do some testing. Maybe there won't be MASSIVE demand, but I do know there are people looking for Microswitch keyboards just for the switches, in many cases sending otherwise usable or repairable equipment to the scrapyard in the process. So if there is a viable way to repair the switches, that will  hopefully delay entropy a while longer and increase the chances we'll get more terminals like these restored and preserved. Thanks! 

6 hours ago, Jimhearne said:

I suspect that so many switches failed on this keyboard as it was a very early batch, the keyboard is serial number 0000019, dated 1977

I could just replace the failed ones but since it's so hard to get the switches and keyboard apart i will replace them all on this one.

I will keep the original pcb's with the terminal, working and non working.

Gogd deal.

6 hours ago, Jimhearne said:

 

I'm undecided what to do about the pins / legs for the pcb.

The originals had stamped pins soldered on, but i'm thinking to just use tinned copper wire (thats why i left a via in the pads).

If i use the original legs i will have to take them from all the still working pcb's as well which seems a shame.

Doesn't look too hard to make new pins from flat stock. Some kind of dies to cut them out would be helpful, and a simple jig to make the right bends. But, yeah, wire would work. Tinned mechanic's wire might give a little better support. Put a little j-hook in the end at the solder pad for a better mechanical connection via a thru hole, maybe.

 

Edited November 5, 2022 by jbdigriz