First issue was that stepper motor gets really hot when is in sync with extruder, that’s why I used custom casing. So far no issues with filament length offset.
For next futures I will try to implement clog detection that works when stepper is in sync (but PTFE tube is needed), runout sensor works out of the box already.
The whole idea came to my mind when I used filament that is to rough for PTFE, so now I can use buffer with 2 configurations with and without PTFE. I will update this thread with project status
Thank you will try that, but concerning is that I tried to calibrate and both motors are calibrated, it seems that stepper from toolhead extruder extrude less during printing.
The rotation distance for your toolhead extruder should be measured while extruding.
Keeping 2 extruders synchronized over several meters of filament is nearly impossible. If you manage to get it done a different filament will probably not be the same. The TRUE rotation distance for both extruders changes with how deeply the teeth on the feed rollers bite into the filament.
I don’t have a Max but l’ve studied the CAD model. The solution Sovol came up with is good in theory but the execution is POOR.
Take your pusher apart and polish the areas where sliding occurs. Put a spacer on the spring or add weight by putting a few M8 nuts around the PTFE tube and let them slide down against the pusher. Make sure there is nothing touching the PTFE tube that would prevent it moving fully down as filament is used.
Update: I created top shelf mount with spool, in that orientation filament will use gravity instead of fighting it, also this shorten the PTFE path.
Right now extruder stepper is in sync with this additional stepper, as I said for some reason external feeder feed a little to much even if it is exact the same as the extruder one. My idea is that for next upgrade I will create macro that will force move stepper from external feeder a little bit in opposite direction when one of sensors from feeder detect movement.
Question for now is: Could someone provide some theory behind this moving tension sensor? I would like to now how it detect tension and what principles are behind it.
The pusher shoves extra filament into the tube, which causes the slide joint to extend, until the tab inside the pusher blocks the IR light beam of the upper switch. Then as filament is used the spring and gravity cause the tab to move down until it blocks the IR light of the lower switch. Goto start.
Meanwhile a counter adds up every extruder move. If too much filament is used before the lower switch triggers it sets an error. The counter resets every time the upper switch triggers.
I wasn’t asking about the software I’m FW engineer outside hobby so I do understand code and how it works(with issues I hate python :D). I was curious about mechanical/physical principles of this extender, but i wrote sketch with force vectors and understood how it works.
Btw length correction was easier to implement that expected
I will create github repo for this mod with stls for custom mounts and updated macros, because it works perfectly (at least for now).
I created algorithm that automatically recalculates rotation_distance for external feeder to make length correction as rare as possible. So yea @cardoc I don’t need to keep them synchronized, they will synchronize themselves
