I made a comment about an AVX problem I was having in another thread, but it really deserves its own. It was somewhat off topic there, and this might be easier to find if someone else has the same problem.
The mount would sometimes start up correctly, but when commanded to slew, usually to begin alignment, it would slew correctly in RA, but just oscillate back and forth by 15° or so (that's a guess) in Dec. If I canceled the slew by pressing a direction button, it would immediately stop slewing in RA, but Dec changed from back and forth, to continuous slew, always - it seemed - in the CCW direction when viewed from the saddle end of the dec axis, without ever stopping. The only way to make it stop was to turn power off.
At first this happened only rarely, and would sometimes work correctly after cycling power a time or two. I originally, mistakenly, attributed it to a power supply problem because attaching it to a freshly-charged battery would seem to fix it. More recently, it became more frequent, and power cycles and changing power sources very seldom did the trick any more, although it would sometimes work the following day at least for a while. In the last couple of days, it was always be doing this. Fortunately, I was just testing some things that didn't absolutely require driving the dec axis, so I didn't lose any observation time, but it was exceedingly aggravating, and I was going to need it to be working very soon.
Yesterday morning after it again failed to work correctly on startup, and on a whim, I shut it off, returned it to the index positions, and swapped the dec cable end-for-end. It worked for the first time in more than a day! Aha... a hint! This morning, it was failing again. Suspecting a connection problem, while it was executing its initial slew in the mock alignment, I wiggled the dec cable's connection to the main controller, but no joy. When I moved the connector at the motor end of the cable, however, it stopped moving back and forth and moved toward the alignment star's position and stopped slewing in what looked like its general direction! Eureka!
I opened the dec motor housing, removed the small circuit board with the RJ45 jack, and inspected its internal connections with a bright flashlight. The wire that makes the Pin 1 connection was dislodged at the far end, and was angled off to the side. It was easy to pop it back into place with a dental pick. After this the mount was again working smoothly and correctly, and has been perfect after much use all day, including many power cycles, mock alignments and calibrations, go-tos all over the sky, and a solar observation session. I think it's fixed.
I meant to trace the circuit and take a picture of the board before buttoning it up after testing with the board still loose, but in my haste to put it all back together and really exercise it, I forgot to. I think Pin 1 is in the encoder circuit; if so that would be consistent with suggestions in the other thread that this was an encoder problem, just not the particular one expected!
Because the problem is clearly after the dec motor controller, that explains why there were wasn't a "No Response 17" type error, which I had been wondering about after concluding it was a connection problem, but before I knew what-all was inside that motor housing; it makes sense that both motor controllers are on the same board, but I didn't actually know for sure.
I have no clue how that connection got set awry, but it does seem to be the root cause of the problem. I hope so! Thanks for the suggestions, everybody.
Edited by SkipW, 19 September 2019 - 06:30 PM.