Nice man! I have been running a custom filament cutter on my ERCFs for about a year. Works way better. I also have it purge the color changing into a side bucket then runs over a nozzle cleaning brush. No purge blocks or faulty carrot tips. Only issue I have with the ERCF now is the servo not always working right.
I am wondering, what is the reason to cut at the toolhead and not at the ercf. Wouldn't that result in the same amount of filament left in the extruder as without cutting ? I guess it is easier to cut at the toolhead, you used the motion system to perform the cut. At the ercf it needs another motor (+ stepperdriver) and possibly even a gearbox.
Friction and tangling caused by the shape of the filament end can occur. The problem of filament selection and movement getting stuck due to filament spider webs can also occur. The reason why BamBu lab introduced the cutting method is probably for the same reason. This problem can be solved.
@@TheHanutaXD Have you tried to run it in the cutting mode? I am not having any problems in the process of reinsertion. My biggest concern is the friction that occurs in the Carrot buffer.
Actually in my opinion it depends on the hotend and the filament. With a long meltzone, a lot of filament actually is molten, so the tip of the filament is quite creapy and actually can be a problem to unload (actually this a problem the new version of the Qidi X-Max 3 seem to have with PLA). So a cutter inside the toolhead is the right way to go. On the other hand, this slows down the filament change by a lot. If you have a revo hotend and use a normal-flow revo nozzle on multi colour, this might even result in shorter over all printing time, as with the right ramping profiles you get a nice filament tip (also of course make sure your filament is dry...). And with wipe2infill you might drasticly reduce waste. On the other hand, on multi-material (like PETG+PLA for easy to remove supports) a cutter inside the head is practicly a most2have for reliable filament change imho. Actually, after disassembling the bambulab filament buffer (I do not have a bambulab printer, but I wanted to understand the buffer-concept, so I ordered the replacement part) I think the ERCF is an unnessarily slow concept anyways. The filament buffer is quite genius. A slider with a magnet inside detects how much filament is pulled by the extruder. This information is used to controll the feeder in the AMS to deliver the amount of filament needed. So the feeder permernantly can stay attached to the filament. So in my opinion, the ERCF isn't the best way to do filament changes. Actually such a buffer could increase the time to change filaments by a lot, as each filament can use its own permernantly attached feeder. Such a buffer had to be attached directy after the bowden to the hotend leaves the enclosure of the printer. So shortest distance between printhead and buffer. in front of the buffer is a hub (4 inputs to one output - actually bambulabs hub includes an internal buffer). With runout sensors for each filament. So as soon one filament is unloaded and left the hub, loading of the next filament could be started. The unladed filament now is further retracted, though a switch (one input, two outputs) after passing the swich, this changes to the other position and the filament is pushed back though the switch to a cutter. After cutting, filament retracts again, swich swiches back and filament is pushed to the hub again (stopping directly before the hub). All this time the new filament is already loaded and printing. Yust some thought how to improve filament change even more. Actualy in my opinion both is needed. A cutter inside the hotend esp. if you use different materials, higher temperatures or high flow nozzles. Esp. if you only have a few filament changes for seperation layers on support. But for PLA-multicour-prints in my opinion using a normal flow nozzle is reliable enough for unloading every time. So the cutting (for perfect reload) afterwards in parallel to the other filamnet being already loaded and printing seems more time and material effective. By the way: Even without multi material such a buffer-concept with permanently attached feeder would be nice. At the moment the feeder in the extruder not only has to push filament through the nozzle, but also have to tow the filament from the spool. This can lead to inconsitient extrusion if the spool does not rotate freely (sometimes happened to me printing directly from dryboxes). Also if you use bigger spools, this might be helpfull. And a second thought: such a "sensor-buffer" also could detect retraction. If there is more than 5mm of retraction, it's obviously no retraction, but an unload. So an unloading procedure could be started automaticly. A "two-material-management unit" obviously had to load the other filament next. So if you know the time a change need, you could do a two material toolchanger without any electrical connection between printer and MMU. Toolchange g-code simply would be retract x mm (x= lenght to unload), wait y seconds, extrude filament x mm (to load the new. With any FDM-Printer independend of the firmware used. Yust some thought, as in my opinion just transfering the "cutting filament inside the printhead" to a voron is only half way. This "sensor-buffer" is the other great idea that need to be transfered to voron. Using two modes (cutting in the printhead or externaly) depending of the users choice would be the most flexible way to increase over all printing time.
maybe a bit but you should be able to tune the distance to extrude to get the new colour primed. in theory you should be able to continue for a bit before the gradient between colours starts too but not sure if you can get the slicer to do that, cut the filament, continue to extrude until the colours are mixing, purge, print in new colour. in which case there would be no diffrence in wastage
I feel like the answer is yes but what has been putting me off about doing an ERCF is tuning the tips of the filament. If it's not done correctly your magnetic gates won't work as intended. This method seems to solve that issue entirely
Wasted filament is better than wasting time fixing inconsistent filament tips. Plus after using the ERCF for months before making my own cutter, you get these super thin strings that clog up the gears. It is near impossible to tune the ERCF tips for every filament and every brand. After making my own cutter, I dont have any loading / unloading issues anymore. Not one.
