On 06/11/2020 02:29 AM, Mattis Lind via cctalk wrote:
>/If that would be the case I think the system would fail />/quite soon rather 
than on test 5. A guess is that this is />/a memory problem. /
That was a good guess, everyone ;) I got some new 4116's and piggybacked (dry, no solder) two of them atop my suspects at E3 & E4.

Didn't fix it. Of course :/

In the meantime I've acquired a nice HP 1630G logic analyzer complete with pods and cables. Setting it up was going to take quite a while since I'm not familiar with this model. So I decided to try a simple brute-force approach before the analyzer. I piggybacked another 4116 onto each soldered-in 4116, one at a time. Actually easy to do since with the leads properly formed, I didn't even have to solder it in place, just turn off the power and move it to the next chip.

On the 16th, the last one of course, the terminal booted normally and works again. :)

I confirmed the bad one by removing the piggyback and the failure returned. Now I need to desolder the bad one without ruining the board. I may just cut the leads off close to the bad chip, and solder the replacement to the stumps. (Normally I remove the legs and install a machine-pin DIP socket). Or just solder the piggyback and leave it there... thoughts?

Reply via email to