MFM (magnetic filament sensor) chart - plugin?
-
Hi,
I am looking for a way to better utilize magnetic filament sensor, and let it live up to the official "monitor" part of the name
I have bought MFM because of recurring problems with some prints - hotend clogging (all-metal), always at the same stage. Probably slicer&nozzle problem with restricted extrusion, resulting in overheating and eventual total clog.
I would love to have a chart similar to "Temperature chart", but displaying extrusion rate - both "theoretical" (from gcode) and measured by MFM:
- preferably every 3 mm (configurable?) and not averaged from the beginning of the print
- maybe also log file for later analysis
- it would be perfect as separate tab - plugin?
Probably the most useful way would be to display "theoretical" vs measured extrusion rate as separate lines on the same chart, with percentage thrown in as an option (or all three selectable?).
"Theoretical" chart would be also useful with simulation, to identify potential problems, even without MFM.
If there is any chance for official or community plugin, I will gladly give my input in planning and testing. Too complex for my mediocre programming skills
I am not sure if this post belongs to "DWC Wishlist" or "Filament Monitor", so I start in "General Discussion"
-
@treetopped said in MFM (magnetic filament sensor) chart - plugin?:
I have bought MFM because of recurring problems with some prints - hotend clogging (all-metal), always at the same stage. Probably slicer&nozzle problem with restricted extrusion, resulting in overheating and eventual total clog.
Such a chart is a nice-to-have, but won’t help with the problem you describe. When hotend clogging became a problem to me, I installed a MFM and tried to monitor the readings during print from the console.
What I observed was that, as long as the hotend worked, the measured range was pretty neat, +/- 2 percent or so. In the event of clogging and/or the extruder skipping steps (or loosing grip), the MFM readings dropped all of the sudden, sometimes even to negative values.
Sure, the MFM triggered a pause, thus saving the print, but even with a narrow tolerance range and 3 mm detection length, I was unable to catch the problem before it fully evolved.
So, I had to solve the problem the hard way: clean the hotend, try other filaments, change temperatures and tool cooling rate, lower print speeds … sigh!
However, the chart you propose can be helpful: to detect mechanical issues with the MFM - especially its 3d printed housing tends to be inaccurate.
-
undefined T3P3Tony moved this topic from General Discussion
-
@treetopped Thats an interesting idea. I have moved to the plugins category.
-
@infiniteloop, you are absolutely right - cleaning is the best way to go. just now I use two nozzles, one is printing, the other one bathing in acetone
on the other hand, just recently I have had one irritating print which clogged three times in a row at exactly the same spot - so maybe there was some additional problem with slicer - BTW, it was Cura 5 early beta, so...
the simulated extrusion chart should help with indicating such problems (to many thin walls etc.)
and who knows, maybe when we have the chart, there will be some indication of coming nozzle clog?
using M591manually definitely doesn't realize full potential of MFM
-
@treetopped said in MFM (magnetic filament sensor) chart - plugin?:
using M591manually definitely doesn't realize full potential of MFM
D'accord. The MFM, encased and mounted properly, is a tool I don't want to miss, and yes, putting the data into a chart can help to better spot some problems. But then, I want it to reside in vicinity of the DWC temperature charts, and on top, I would like to see an additional graph with the power consumption of the extruder stepper: this would give us the full picture.
-
I have just realized that in gcode there is no explicit "extrusion rate", like in "mm/s"
but it is perfectly possible to calculate it for each move separately (?)I am in the middle in diagnosing the cause of clogging in new hotend, and it would be extremely useful to have this information
with all-metal it is possible (at least i have read so) to experience heat creeping up the radiator, so low extrusion rate in some parts of the model could be problematic, and maybe also avoided by changing slicer settings
on the other hand, if it proves not to be the source of the problem, I can focus on other tests
hence the chart idea
BTW, maybe it could be also implemented in the slicer (Cura plugin?)
-
I have just realized that in gcode there is no explicit "extrusion rate", like in "mm/s"
You address a topic which has lately been labeled „volumetric extrusion“ - this term has come up in forum posts about thermal limits of hot ends, especially at high print speeds.
But each and every slicer tells the extruder how much filament to deliver for a given linear slope. It’s the art of the firmware to calculate a proper timing, that is: at what point of the way it has to extrude how much filament.
As long as slicer and firmware are d’accord about the filament’s diameter, volume (to provide) and length (to feed) are just two sides of the equation. The time to feed a given volume depends on the speed (x/y) of the print head, and that’s where things become complicated.
The head has to accelerate and to slow down, physics don’t allow him to always move with the proper speed for a constant flow rate. Not to talk about pressure advance, retractions and other obstacles …
—
The „MFM chart“ you propose might be helpful to spot problematic parts of a print, but TBH I think you should focus on parameters like print speed, hotend temperatures and filament brands in order to overcome the problems with the all-metal hotend. I know what you are talking about. Heat creeping alone has kept me busy for days … -
I came here looking for exactly what the OP asked for.
It shouldn't be necessary to calculate data from the gcode, since DWC already knows the requested movement of the extruder. A graph showing this and the measured extrusion rate would be a very cool addition. I imagine it would look something like race track telemetry data with a perfect lap overlayed with you are achieving.
I also think it would be really cool if we had a way to manually, but more directly calibrate the filament monitor. Imagine if you could take the moitor out of the filament path and push a length of filament into it, then have a series of messages like this:
Mark filament "X" mm from reference point and press OK (to set 0)
Slowly pull filament through filament monitor and stop with the mark on the reference point then press OK
Measured sensitivity is XX.XX mm/rev
It would be good to be able to choose the length you want to measure. Obviously longer will lead to improved precision.
This would remove any error from printer settings such as steps/mm on extruder or losing extrusion rate at high speeds. To go even further with this, we could then use this data by comparing extrusion rate with extrusion speed and giving us an M592 non-linear extrusion rate line to paste into config.g in the same way we can with a PID tune and M500 to save to config-override.g
I wish I had the skill to write plugins to do this myself, but I do not.
-
I am bumping it up (a little), as I have just had a couple of mysterious stops.
How about writing check results to CSV file?
Every 3 millimeters by default, only current 'percent' value.It can be then analyzed in Excel (eg.).
If it proves useful, we can think again about 'live' chart.
