IBM 3277 Conversion (Updated w/ Completion Pics)
Posted: 27 Oct 2017, 01:43
After the beautiful work of both hansichen and ipreferpie this may be a bit redundant, but I converted my 3277 today and wanted to share some shots. The keycaps are drying now, but I'll update with completion shots tomorrow.
xwhatsit's conversion guide can be found here: http://downloads.cornall.co/ibm-capsens ... spring.pdf
Update:
A couple of notes for anyone else doing the conversion.
1) When I first brought up the Capsense GUI, the state grid showed that all keys were pressed. Lowering the current threshold enough showed the skipped columns (five on the far left, one near the middle) as pressed, but no keystrokes would register. Lowering the threshold further showed all keys as released; again, no keystrokes would register.
I believe I had tightened the screws holding the PCB, rubber mat, and steel plate to the main body too much. I tried to tighten all screws only to the point where I felt resistance, but I guess I still went a bit tight here. I took the PCB out and reinstalled it, this time really holding back on the screws, just enough so they shouldn't work themselves back out, and everything worked fine from that point on.
2) I was advised against soaking the keycaps, but, as you can see, I did. I left them to dry for more than 12 hours, but still some of the stems had water in them. I ended up folding small strips of toilet paper into little stakes and poking them into the stems to soak up any water. Wouldn't want to cause the mounts to rust.
The following are from day one:

I don't know the relevance of the yellow barrels (?) vs white. If anyone knows, I'd love to find out!

Before shots. The case and caps were wiped down with a damp cloth. No other work had been done at this point.




I was thrilled to see what nice condition the 3277 was in when it arrived. The next two photos show the minor rust on the board.






The foam shown in the next three pictures may need to be replaced at some point. A corner had detached from the main body, and when I picked it up it felt like molasses-soaked bread that didn't spring back to shape when released.



The next two pictures show the original PCB.


And here's the emdude PCB.


The best advice I received was to wash the caps in cold water, no soap. But I was impatient, and switched to warm water with Dawn dish soap--the same as I use for dyesubs. If I detect any damage I'll update this and post a warning thread.


The caps are drying now.

Update:
The caps went back on this morning. I am typing on her now! Here are the shots from day 2.















Thanks to:
xwhatsit's conversion guide can be found here: http://downloads.cornall.co/ibm-capsens ... spring.pdf
Update:
A couple of notes for anyone else doing the conversion.
1) When I first brought up the Capsense GUI, the state grid showed that all keys were pressed. Lowering the current threshold enough showed the skipped columns (five on the far left, one near the middle) as pressed, but no keystrokes would register. Lowering the threshold further showed all keys as released; again, no keystrokes would register.
I believe I had tightened the screws holding the PCB, rubber mat, and steel plate to the main body too much. I tried to tighten all screws only to the point where I felt resistance, but I guess I still went a bit tight here. I took the PCB out and reinstalled it, this time really holding back on the screws, just enough so they shouldn't work themselves back out, and everything worked fine from that point on.
2) I was advised against soaking the keycaps, but, as you can see, I did. I left them to dry for more than 12 hours, but still some of the stems had water in them. I ended up folding small strips of toilet paper into little stakes and poking them into the stems to soak up any water. Wouldn't want to cause the mounts to rust.
The following are from day one:

I don't know the relevance of the yellow barrels (?) vs white. If anyone knows, I'd love to find out!

Before shots. The case and caps were wiped down with a damp cloth. No other work had been done at this point.




I was thrilled to see what nice condition the 3277 was in when it arrived. The next two photos show the minor rust on the board.






The foam shown in the next three pictures may need to be replaced at some point. A corner had detached from the main body, and when I picked it up it felt like molasses-soaked bread that didn't spring back to shape when released.



The next two pictures show the original PCB.


And here's the emdude PCB.


The best advice I received was to wash the caps in cold water, no soap. But I was impatient, and switched to warm water with Dawn dish soap--the same as I use for dyesubs. If I detect any damage I'll update this and post a warning thread.


The caps are drying now.

Update:
The caps went back on this morning. I am typing on her now! Here are the shots from day 2.















Thanks to:
- cml for finding and posting the 3277's from the classifieds in Spain,
Scottex for buying and re-distributing them,
emdude for developing the PCBs,
xwhatsit for developing the beam spring controller,
orihalcon for running the PCB GB and supplying other beam spring parts,
And everyone in the Telegram channel for their advice and information!