Did you try to put the official Creality firmware instead of the Smith3D?
Creality controller boards are notorious for inherently bad serial (USB) communications. Very sensitive to the USB cable used, EMI, and power or ground issues. Blocking the 5V pin on the RPi side of the USB cable is very often necessary.
Printing from the SD card bypasses this notoriously bad serial communication issue. You can change the firmware, reimage the RPi operating system all you want but until you solve the communications problem, you won't be able to print reliably.
1 Like
I wouldn't wish this on anyone... The official firmware has some of the most problems we find anywhere. Vanilla Marlin or well-maintained forks usually work far better than official Creality stuff. If the Smith3D one is outdated, yes upgrade it. But don't use Creality, it is usually a downgrade.
1 Like
Hi @b-morgan,
Thanks for your tip. The first thing I've done when install Octoprint the first time was to cover the vcc pin of USB cable because of that.
But this problem I don't think it's communication related because it only happens when Leveling......
@Fil, You did not use the </> button when you posted your serial.log extract. Unfortunately, some of the characters in that log are interpreted as markdown by this forum so its hard to tell exactly what is happening. It would be better to just post the log file itself. Please delete the log you have, capture another failure, and post the log file.
@phenriques3000, if you have the same problem as @Fil, then we need the same information from you as we asked for from @Fil. Enable the serial.log, capture a failing case, and post the log here.
A probing sequence can take a while to complete, while the printer appears to support busy processing, OctoPrint still has timeouts associated with long running commands. The complete serial.log will help determine if this might be the problem.
1 Like
I took it one step further and built a cable which has worked fine and done 5hr prints with no communication errors.
Really?
So as @b-morgan says it's really a communication problem... Hell...
Can you tell us how did you do that cable? Thanks
I'm trying to capture the serial.log for him right now but of course the first print i did with the log enabled it of course worked! Picked a short print (20mins) so once that's done I'll reprint it and it should fail. Therefore the log should have a success and a failure. crossing fingers. As for the cable I took a working data/power micro USB cable cut it in half, soldered the ground and two data inputs, heatshrunk it and bobs your uncle. 
Oh but even with that cable you have this issue right? You solder direct in the board? Did you have the pinout for that? If we can have Octoprint working fine one of my objectives was to put the raspberry inside Ender 3 v2 (in the tools box).
Correct I have the same issue with this cable. No not direct to the board. I cut a regular MicroUSB cable in half and then reconnect it but without the +5v line connected. I guess you could just strip back the wire on the sheathing and cut the +5v line if you know which color it is but I confirmed the lines with a multimeter.
The best USB cable for connecting to a Creality printer is usually shielded with ferrite cores on each end and as short as possible. Putting tape on the 5V pin on the RPi end (USB A connector) is probably easier than cutting through the shielding to get to the actual wire.
Hmm. The one I'm using doesn't have the ferrite cores. I'll order one with those and give it a shot. Of course now i want it to fail it's not failing. Just cancelled the second print and trying again. 
So there is a success followed by a failure in the attached serial.log file. Are there any other logs you would like that could assist in troubleshooting?serial.zip (1.8 MB)
Comparing the successful probe and the failed probe:
2021-02-07 17:53:26,990 - Send: N10 G29*35
2021-02-07 17:55:50,324 - Recv: Bilinear Leveling Grid:
2021-02-07 17:55:50,337 - Recv: 0 1 2 3 4
2021-02-07 17:55:50,342 - Recv: 0 -0.232 -0.148 -0.067 +0.097 +0.273
2021-02-07 17:55:50,347 - Recv: 1 -0.218 -0.140 -0.043 +0.120 +0.299
2021-02-07 17:55:50,351 - Recv: 2 -0.226 -0.135 -0.045 +0.126 +0.291
2021-02-07 17:55:50,354 - Recv: 3 -0.280 -0.174 -0.058 +0.118 +0.289
2021-02-07 17:55:50,358 - Recv: 4 -0.277 -0.163 -0.060 +0.112 +0.286
2021-02-07 17:55:50,362 - Recv:
2021-02-07 17:55:50,365 - Recv: X:220.00 Y:220.00 Z:12.72 E:0.00 Count X:17600 Y:17600 Z:5272
2021-02-07 17:55:50,380 - Recv: ok
Success reported by firmware 144 seconds later.
2021-02-07 18:38:45,632 - Send: N10 G29*35
2021-02-07 18:40:36,219 - Recv: Error:Probing Failed
Failure reported by firmware 111 seconds later.
The failure is reported by the firmware. OctoPrint is just passing the failure on. The timeout values in OctoPrint are not effecting the outcome (i.e. the current values are sufficient).
I don't have (nor do I want) a Creality printer but this appears to be yet another issue with their firmware. Apparently, when printing from the SD card, the firmware seems to ignore errors and just keeps going.
A wild shot in the dark... Add an M420 S0 before the G29 in the start gcode.
1 Like
Thanks @b-morgan I'll give that a shot. So you think that those errors are also occurring when printing locally but it continues on. Wonder if there is a way of capturing that somehow. Maybe I should give klipper a try?
Thanks,
Fil
Flash vanilla Marlin firmware first, Klipper takes a lot more effort to setup (IMO, ready to get slaughtered by some klipper fan).
Unless the printer shows a message on the LCD then it is not possible to find more details, since there is nowhere for it to go.
2 Likes
Hi mates,
Here I send my serial.log but I thing it's exactly like the one of @Fil ....
@Charlie_Powell where do I get that firmware to test?
serial.log (41.5 KB)
Just in case @Charlie_Powell doesn't have a pointer in his back pocket, I did a google search for "creality ender 3 v2 firmware update" and found many links and videos that should answer that question.
1 Like
Yeah need to compile the base marlin for the ender 3 v2 with bltouch for it. Found a site where you can pay for the fw but haven't had the time to go through and compile it with the correct settings/options yet. If I get it working properly I'll post it somewhere.
Ok I will check that too.... Besides Smith3D I can only found another solution from th3dstudio.com, they provide a firmware for our printer but he way I see if changing the lcd firmware too.