Thursday, March 8 2018
The capacitor I ordered to fix temperature stability issues on the CR-10S arrived in the mail, yesterday. Today I took the CR-10S control unit apart and replaced the factory not-quite-up-to-spec capacitor with the unit I ordered. The difference:
Temperature stability
The above gif shows temperature stability on the unit just after it had been turned on before and after the fix. The rest of this post contains some details of the fix.
Need to remove this board
Most cables removed
Board freed from case
Removing the bottom of the control unit and the power supply gives access to the control board. I did not disconnect the power supply. I swung it out of the case and let it rest on a small cardboard box that was roughtly the same hight of the case. With it out of the way I remove the cables, noting where each belonged. I made some labels out of masking tape to mark the cables that weren’t otherwise marked.
Ready for mods
Replacement cap
One leg lifted
Capacitor removed
Once the board was out of the case I warmed up the soldering iron and made sure the tip was freshly tinned. I applied heat to the visible section of one of the pads and applied the slightest bit of pressure with some tweezers to lift the cap once the solder had melted. With that done I moved to the other pad and lifted the cap off the board.
New cap installed
Board installed and wired
I applied a dab of fresh solder to the pads on the board and the leads on the cap. The new cap was held in place with tweezers while applying heat to the pads. I must have moved as the cap is not 100% square with the pads. However, it is good enough. I have never been very good with a soldering iron. That said, I decided against removing and re-installing to lessing the chance of messing up the board.
I put the board back in the box and hooked up the wiring using the labeled wires and and my notes. The notes must have been good enough as there was never a “where does this go?” moment. I re-installed the power supply and buttened up the control unit. Time to check it out.
I mved the control unit back by the printer and hooked everything up. Lastly I plugged the control unit and switched it on. The above gif shows the results: the reported temperatures are not jumping around +/- 2 degrees as they were before the fix. I fired up octoprint and used the terminal tab to issue m303 s235 c9 to initiate PID Autotune using 9 cycles and a target of 235 °C. The temperature graph looked much better than that last time I tried a PID autotune.
PID Autotune (9 cycles)
The results also weren’t bouncing all over the place from cycle to cycle. The plotted temperature was not nearly as noisy as it was before the fix (see previous page). The final PID values were P=19.42 I=1.44 D=65.62. I loaded and saved the new values and tried a test print using PETG.
I just started playing with PETG and my slicer settings still needs tuning. Ignoring that the printer came to temperature in about 7-8 minutes; bed at 70, hot end at 235. I’ll call this repair a success.