There have been people posting about NS638 RAM boards in the past
couple of weeks.

Recently, a kind soul sent me an M7609 and an NS638.  (Thank you very
much, you know who you are!)  The M7609, of course, worked completely
painlessly.  But the NS638 didn't.  My first attempt produced

| Performing normal system tests.
| 
|   7..6..5..4..3..
| 
| Tests completed.
| 
| 
| >>> b
| 
|   2..1..0..
| 
| 
| ?50 SCB2NDINT, XQA0
| ?06 HLT INST
|     PC = 00000EE6
| >>> b xqa0
| 
|   2..1..0..
| 
| 
| ?50 SCB2NDINT, XQA0
| ?06 HLT INST
|     PC = 00000EE6
| >>> 

I then tried flipping the switch at the back.  Then instead of the
SCB2NDINT and HLT INST lines, I got

?05 DBL ERR.

which I assume means it took a double error halt.

On pulling the board out and inspecting it, I noted there was a
two-position jumper.  I moved it to the other position and then it
passed POST and booted (showing 9M) with that board.  So I added the
M7609 and booted - and it still showed 9M.  So I flipped the switch (I
forget whether I flipped it back in between), and got DBL ERR again.  I
flipped the switch back and swapped boards (so the M7609 was next to
the CPU, with the NS638 farther away) and now I have a happy 16M
system.  (Presumably it'd be 17M except for the 24-bit limit.)

I didn't make any tests with the jumper completely disconnected, and
the way the pins are laid out, only two positions are possible.

I'm tempted to drag out another KA630 and set myself up with a two-CPU
system with 9M each; perhaps someday I will....

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B