Thermal model for E3D REVO #4105
Comments
3d-gussner
added
enhancement
help wanted
report
volunteer needed
|
Here the plots From @kromeninja From @ulab From @JWvP From @snafu1282 From @matthiazzz From @sdh2 |
PS: I had a lot of crash detections with my Y axis during the benchy print. I'm not sure if I just never noticed it before since I mostly print in stealth mode, but I guess I have to look into that. |
|
Unfortunately, in my case, |
|
My latest logs. No anomalies encountered. |
|
@3d-gussner I tried the M301 W1.1.... 1.0.... 0.9..... etc. and got no warnings again. This happened before and you pointed out something I needed to do to get it to work. Unfortunately, I don't have a reference to what that was. Can you refresh my memory please? |
Thanks, Your log file is very short and I can't see everything. Please follow the steps above and report back. |
|
@alexiri @kromeninja @ulab @JWvP @snafu1282 @matthiazzz @sdh2 Everyone else who didn't provide complete log files please start with Task 1 to get more data. |
|
Here is my Task 2 log file. It's a long one. My basic result is that M310 W0.8 works for me. W0.7 almost works, but it's a tiny bit unstable. I could not get an anomaly by trying to create backflow from the fan. I was able to repeatably trigger an anomaly by using a small air compressor with a finger trigger nozzle from about 500mm away. What I noticed is that the anomaly gets triggered at between 227C and 229C. |
I hope it will be good at this time. BTW. There is something wrong with the thermistor because on hotend I have 13deg and on the bad I have 20deg. |
Sorry but this is not a complete log file. What concerns me more is that your thermistor reads 13°C while the bed is 20°C they should be quiet close when the printer is turned on after few hours. EVERYONE: Do you know what this could cause? |
|
@snafu1282 Thanks for the log files of Task2 Please check steps 8 and if that makes sense in your opinion to speed up finding the correct value. Tip for putty and sending gcode files:
Yes the REVO compared to the E3Dv6 catches up very fast with blow backs and drafts so it is hard to trigger. |
Yeah, PuTTY not having a command input box like those in Pronterface or OctoPrint is a bit of a pain. I had hoped that PuTTY would keep all the typed text together, so it shows up as a single command line, but nope. On another question, if I find the W value needs to be at 0.8, shouldn't the E value also be changed (i.e. lowered to something like 1.35)?? |
|
Task 2: I did several heat/cool cycles but unfortunately do not have time for a print at the moment. I will follow up with that later. Before the fan is turned on, W0.5 seems to work for heating and cooling. But, this is too tight when the fan cools it down. W0.6 & W0.7 almost works, but triggers until about 180C. W0.8 seems to do ok cooling with the fan on. I was able to trigger the anomaly with a can of compressed air, and during the latter half of the tests I had a large floor fan pointed at the printer to simulate a draft. @3d-gussner There are a couple typos in Task 2 above in lines 10 & 15. Took me a couple minutes to figure out why my values weren't changing, I'm not terribly fluent in this stuff! |
By |
|
Here are log files from Task 2; the initial step through of the procedure, plus 3 more logs with values of W0.7, W0.75, and W0.8. snafu1282_20230325_TASK-2_W0.75.log.zip Just finished a Cali Dragon in PETg (235C/75C) It came out pretty clean. Here are 3 shots of it. I'm including them as I think the Cali Dragon is a reasonable test of what could happen with blowback cooling, which didn't happen.
|
Here is my print log. With a setting of W0.8, I didn't have any issues on a four hour print. |
|
Task 1; steps 1-24. Thermal anomaly occurred at W0.4 Compressed air from a can took about 15 seconds from 15cm to cause a thermal anomaly |
|
My MK3S+ with E3D Revo isn't printing at all with the new firmware. It keeps failing on the wizard. When heating up to PLA, it aborts at 70-ish C and nothing is working. How can I help to provide more data? This isnt working for me now. |
|
@zuidwijk Hi, thanks that you want to help and contribute. Please read #4105 (comment) and start with Task 1, provide the Thermal Model Debug serial log, this helps to fine tune the REVO default values. |
|
@peschkaj Thanks for the Thermal model debug log and feedback of your findings of warning value. I will add your data to the calculated default values. |
Update I take that back, |
Not sure if it was a typo here or during testing: Please note that |
'Twas but a typo |
|
Hi, attached my Task 1 logs from my MK3S, freshly serviced and upgraded with MK3S+ upgrade kit and the Revo Hotend. Regards, Marcel |
|
My 1st logging... I couldnt print, when loading filament and therefore pre-heating I got "THERMAL ANOMALITY" Edit: I forced reheating after each error and now it starts printing (with z-layer at 0.000 hahahaha) Logging during printing: |
First we want to thank everyone here for providing feedback, testing and reporting log files in #3636.❤️ 👏 🎆
As the issue got very long I would like to continue collecting data log files here.
We have a new firmware release FW 3.12.2 which includes some PTC / REVO improvements see https://github.com/prusa3d/Prusa-Firmware/releases/tag/v3.12.2
It would be great if you could run the tests below and give us some feedback.
Please don't use Pronterface/printrun as it UPPER-CASES
D3commands which are case sensitive.Please use Putty, OctoPrint (with D3 as exception to be UPPER CASED), or other serial terminal scripts:
Octopi Blacklist is in
Settings->Features->Terminal Auto Uppercase Blacklistadd here theD3gcode.Here the steps our internal test team was following.
Task 1
M503to show current settingsM310 S0to temporary disable TM for PID calM310 S1to enable TMG1 X125 Y105 Z1to move the hotend to the middle of the bedM155 S1 C3to auto report temps and fansM310 P40 U-0.0014 V1.05 D0.15 L270to set REVO valuesD70 S1to enable tml debug codeM310 A F0to run Thermal model cal. without checking existing TM settingsM500to store TM valuesD3 Ax0ca6 C1to check calibration v2 statusD3 Ax0f5f X01to set Wizard activeCheck TM simulation vs real REVO
25. Lower the TM warning values in 0.1 steps with
M310 W1.1,M310 W1.0...M310 W0.4during a print. The default warning is set to1.2and error to1.7426. Set warning value back to default
M310 W1.227. Try to force TM error
28. Hold nozzle (without silicone ring) during preheat with a big wrench
29. Blow compressed air on nozzle / block (without / with silicone ring)
Task 2
This task is for users who have provided the log files and have a well calibrated REVO MK3/S printer.
Goal is to find the narrowest TM Warning settings to detect issues as fast as possible while preventing false positives.
To simulate an issue it is needed to remove heat from the system that is unusual:
G28 Wand move it to the middle of the bedG1 X125 Y105 Z1M155 S1 C3M310 W0.1M104 S230TM: error ....is serial log8.1. increase the TM warning in to the highest warning value you see in the log file
Example:
8.2 The highest value in this example is
TM: error |0.782815|>0.100000so set the new TM warning to next 0.x value here aM310 W0.88.3 Please watch your serial log and try to find the highest TM Warning value and use that one.
9. Let the printer cool down with
M104 S010. Verify the warning value with a full cycle of heat up
M109 230and cool downM104 S0if needed adjust the TM warning value, so it is not triggered.11. Save the value with
M500in case you have to restart the test at some point.12. Let the printer preheat with
M109 S23013. Turn on the fan with
M10612.1 Try to create a blow back of the part cooling fan. Placing something behind extruder or so. hard to describe who to force a blow back.
14. Try some cold air, if needed to have access with tools just move the Z up, but don't forget to go back to
Z1if you try it again.14.1 If you have a better way to "suck" out heat from the system try that
15. If it doesn't trigger lower the TM warning value by
0.1withM301 W<old value - 0.1>to check when it triggers16. Let the printer cool down with
M104 S0and full fan blowing atZ1height17. Stop fan with
M10718. Verify the value again with a full heat up
M109 S230and cool downM104 S0cycle19. Repeat verification with full on fan
M106followedM109 S230,M104 S0andM10720. Save the new warning value that fits you best with
M50021. Print multiple things to verify your TM warning values (ideally with serial log files)
22. Report these values and log files here.
Task 2 results:
The text was updated successfully, but these errors were encountered: