TL;DR: Wiggled stuff inside printer, accidentally fixed the issue. Not confident itās fully resolved, but it has completed a few dozen filament swaps successfully now.
Printer: X1C
Print hours when symptoms appeared: ~1500
Length of time problem was ignored: 6+ months
Did problem resolve itself: no
Symptoms: Print starts fine using filament from AMS and may change filaments 1-2 times, but four flashing lights appear regularly during a print job. Turning printer off/on usually, though not always, returns the AMS to a temporarily-working state. If using only one color, print usually (est 80+%) completes successfully even if connection with AMS is lost.
Preliminary troubleshooting steps taken:
1.) Applied DeoxIT D100L from bottle to 4-pin and 6-pin connectors on the back of the printer, buffer, and AMS.
2.) Replaced 6-pin cable after step 1 did not remedy issue
Result: Initial improvement likely placebo, issue remains
Story:
It got to the point where not being able to use support material was holding up a few projects and I needed to finally address this issue. I followed the troubleshooting guide (Linked above by mfetting. Iām unable to include links for some reason) and found that the resistance for GND-sigA and sigA-sigB were in spec, but GND-sigB was ~5.37 k-ohms on the port on the AMS and ~5.95 k-ohms on the port on the printer. I opened up the AMS and determined it was the AMS main board that was out of spec, rather than the AMS power board.
Not keen on throwing parts at it without understanding the root cause, I didnāt feel confident that the measurements were far enough out of spec to cause these issues and talked to one of the EEs at work. He said it was probably fine, and after describing the symptoms, suggested it was likely something in the connection chain, given the intermittent nature and how the issue somehow resolves itself by power cycling the printer. He suggested hitting the contacts on the board with a soldering iron to re-flow the solder, that it could be a cracked solder joint that wasnāt visible.
When I got home the red lights were flashing on the AMS, so I started wiggling the external connectors, starting at the AMS. When I got to the 4-pin going into the printer, the lights changed from red to white. Interesting.
I turned the printer off, took the back off, noted the connections on the two boards looked fully seated, took one of the spare 4 pin cables I had and cut it in half, stripped the wires and tinned the ends. I plugged it back into the port on the back of the printer and connected alligator clips to the wires. GND-sigB was open at that point, so I started methodically wiggling the external 4-pin plug, and the other two in the printer on the AMS interface board and the MC (machine controller) board.
I thought I had a pattern figured out of what made resistance go up and what made it go down, but nothing was really repeatable and eventually I couldnāt make it go up at all, it was solid at 4.34 k-ohms, unless I fully unplugged a connector, right in the expected range from 4-5 k-ohms. I left the back off so it was easy to access again and started a print with about 50 filament swaps and it ran all night without issue (aside from the usual issues with the spool being too light and causing feeding issues).
Conclusion:
Wiggling the connections on the printer side of the connection chain appears to have brought the resistance reading into spec. Iām not confident that the issue is fully resolved but itās of some comfort that my issue is likely a connection issue rather than a dead board. The printer so far has completed a few dozen filament changes, which is a few dozen-1 more than pre-wiggling.
Thanks to those who shared their experiences previously, some were helpful. Hope this helps someone else.
Picture from mid-wiggling session before resistance value had stabilized. I switched to alligator clips as I wasnāt certain the leads pictured werenāt contributing to the fluctuating readings.