[Fermé] "checksum mismatch" when printing via USB  

Re: "checksum mismatch" when printing via USB

Which firmware are you running?

Re: "checksum mismatch" when printing via USB

Ahh, forgot to tell that... the latest one (sadly) - 3.1.1-RC4 build 143

Re: "checksum mismatch" when printing via USB

I had the same issue with the same firmware. I gave up and printed via SD. Sorry not very helpful but at least you are not alone.

Re: "checksum mismatch" when printing via USB

Thanks for your reply. The "at least you're not alone" part is indeed helpful. I think I'll contact support about this, but it is still interesting to know, if anyone is printing via USB. Anyone? 😉

Re: "checksum mismatch" when printing via USB

Just successfully printed a 5 hour print with USB, using latest firmware (143) and latest Octoprint. The only catch I can think of would be that I'm using an older Raspberry Pi (the one with only 2 USB ports).

The Raspberry Pi 3 has some changes when compared to the older one (yes, I know this is not related to the USB used by Octoprint); wonder if it couldn't be somehow related, though.

I.e. those having trouble with OctoPi, try printing with an older Raspberry Pi if possible and see if there's a difference.

edit: dammit, I've jinxed it 🙂

Send: N33386 G1 X103.121 Y154.591 E0.00995*103
Recv: Error:checksum mismatch, Last Line: 33385
Recv: Resend: 33386

Also had a crash detection (but the nozzle did not seem to hit anything) and a move to the front of the bed and back, but no plastic extruded during it.

Re: "checksum mismatch" when printing via USB

Although my issue do not relate to OctoPrint (I'm printing from a PC), I decided to try out the OctoPi thing.

Interestingly, I see very little checksum errors, and the print seems to turn out OK ❓

So... Einsy seems to like talking to RPis and really dislike the "real" computers. I am at my wits end...

(I'm using an RPi3, so I guess it is not the RPi version that matters...)

@LeMotan: Did you only get that one checksum error? I do get some with OctiPi, but it doesn't seem to affect the print...

Re: "checksum mismatch" when printing via USB

No, there are several checksum errors. The print however appears to be fine so far. Got one more slide to the front of the bad and back, but it resumed OK.

Re: "checksum mismatch" when printing via USB

Had to go crazy with this and do my own tests to determine the cause of these checksum errors. I am now quite sure that this is a firmware problem and not a computer/cable/software/noise issue.

I made a GCode file moving each axis (extruder, x, y and z) very rapidly 0.5 mm back and forth to stress the gcode interpreter in Einsy. The forum system will not let me attach the file to this post, so I included it at the end of this post, if you want to try it out yourself. Just make sure that the extruder is located more or less in the center with regards to all axes. Also remove any filament from the extruder, and disable the filament sensor (mine just stopped working, so I'm just guessing).

Anyway... this is what I get back from Pronterface, when "printing" the script:


echo:Unknown command: "$2 - test extruder movements"(2)
Error:checksum mismatch, Last Line: 28
Resend: 29
Error:Line Number is not Last Line Number+1, Last Line: 34
Resend: 35
Error:No Checksum with line number, Last Line: 38
Resend: 39
Error:checksum mismatch, Last Line: 40
Resend: 41
Error:No Checksum with line number, Last Line: 43
Resend: 44
Error:checksum mismatch, Last Line: 47
Resend: 48
Error:No Checksum with line number, Last Line: 76
Resend: 77
Error:checksum mismatch, Last Line: 78
Resend: 79
Error:checksum mismatch, Last Line: 94
Resend: 95
Error:checksum mismatch, Last Line: 116
Resend: 117
Error:checksum mismatch, Last Line: 117
Resend: 118
Error:checksum mismatch, Last Line: 138
Resend: 139
Error:checksum mismatch, Last Line: 153
Resend: 154
Error:checksum mismatch, Last Line: 156
Resend: 157
Error:checksum mismatch, Last Line: 157
Resend: 158
Error:checksum mismatch, Last Line: 159
Resend: 160
Error:checksum mismatch, Last Line: 168
Resend: 169
Error:checksum mismatch, Last Line: 185
Resend: 186
Error:checksum mismatch, Last Line: 196
Resend: 197
echo:Unknown command: "$204 - test X axis movements"(2)
echo:Unknown command: "$406 - test Y axis movements"(2)
echo:Unknown command: "$608 - test Z axis movements"(2)


The "Unknown command" lines are my crude way of including "comments" in the output to be able to follow at what part of the script the errors occur. As you can see, ALL errors happen when moving the extruder. NO errors happens when moving the other axes. Therefore, this cannot be a communication problem. I have re-run this script numerous times with the same result. The number of errors vary, but they always happen when moving the extruder.

By the way, all checksum errors disappear, when I decrease the speed of the extrusion in the first G1 line from F6000 to F600. Of cause, it is not an option to slow down printing that much.

Here is the GCode, if you dare 😉 Just copy the contents of the box to a .gcode file and print from your computer using Pronterface.


G91

G90


Re: "checksum mismatch" when printing via USB

I got exact same issue printing over USB and it seems to be related to latest firmware update, as I had no issue with earlier firm ware version.

Re: "checksum mismatch" when printing via USB

What is the USB type on the MK3 and what type is on your computers? Just checking if connecting a USB 2.0 to a 3.0 or something different is the cause. There's been some errors with other electronics when you connect a USB 2 to the USB 3 when it came out. I haven't kept track of errors over time so I don't know if that's cleared up.

Re: "checksum mismatch" when printing via USB

Had to go crazy with this and do my own tests to determine the cause of these checksum errors. I am now quite sure that this is a firmware problem and not a computer/cable/software/noise issue.
...
As you can see, ALL errors happen when moving the extruder. NO errors happens when moving the other axes.

Yes, I think you are correct. My guess is that the new linear advance and crash detection code in the stepper interrupt routine is taking too long to run, and is causing the system to miss characters coming in on the serial line. This will only affect moves using the extruder. One way to check that would be to disable linear advance and see what happens. This is a tricky one.

It prints perfectly from the SD card so I don't think this is a fundamental printing issue. Rather it's the limited way that this processor has to deal with serial input.

Re: "checksum mismatch" when printing via USB

Could this be solution for you ?
https://shop.prusa3d.com/forum/software-f13/checksum-mismatch-t6214.html#p43543

One other thought (this was a firmware fault I reported over 6 months ago...) OctoPi settings/Serial Connection/Advanced Options/Simulate an additional OK for resend requests - make sure you have ticked the box.

Re: "checksum mismatch" when printing via USB

My guess is that the new linear advance and crash detection code in the stepper interrupt routine is taking too long to run, and is causing the system to miss characters coming in on the serial line. This will only affect moves using the extruder. One way to check that would be to disable linear advance and see what happens. This is a tricky one.

I tend to agree with you on the Atmel microcontroller having a hard time dealing with all these new tasks. Another thread in this forum discusses OctoPi and similar issues, and people seemed to conclude that disabling crash detection helps. I re-ran my GCode file to test it, and sure enough... the number of checksum errors dropped by 90 %, when crash detection was disabled.

I really, REALLY hope that Prusa Research was right when saying that it wasn't necessary to move to a 32 bit platform for the MK3 😕

Re: "checksum mismatch" when printing via USB

Could this be solution for you ?
https://shop.prusa3d.com/forum/software-f13/checksum-mismatch-t6214.html#p43543

One other thought (this was a firmware fault I reported over 6 months ago...) OctoPi settings/Serial Connection/Advanced Options/Simulate an additional OK for resend requests - make sure you have ticked the box.

Mmmm, as I understand the issue, I would say "no". This is not directly related to OctoPrint, and in the end this is more about preventing the resends to occur in the first place.

Re: "checksum mismatch" when printing via USB

Has anyone rec'd help from support on this issue? I just took my MK3 out of thebox and tried to print over USB and ran into this problem. Kind of frustrating :/

Re: "checksum mismatch" when printing via USB

Has anyone rec'd help from support on this issue? I just took my MK3 out of thebox and tried to print over USB and ran into this problem. Kind of frustrating :/

Prusa has a firmware out for initial testing here: https://shop.prusa3d.com/forum/original-prusa-i3-mk3-f30/octoprint-issues-and-tips-t13156.html

It seems to do the trick for me 🙂

Re: "checksum mismatch" when printing via USB

hi to all,

i have the same problem with latest firmware 3.4.1 and latest slic3r PU 1.41.1

i print via USB without problem with S3D but i receive the error if i try to print via USB with slic3r

"Error:No Line Number with checksum, Last Line: 0"