2x filament tubes to a Y-splice @ the hotend for much faster swaps? (Asynchronous retraction/extrusion for all but the short distance after the Y to the toolhead)

[michael-jennings]

Is this an original idea or has this already been discussed (and rejected or planned for the future?):

It looks like running 2 filament tubes to a Y splice at the hotend would significantly increase switching speed. You could have the next filament queued up near the hotend, and only have to wait for the last filament to retract past the Y splice before extruding the next one in.

While I'm here, my next idea is if we had a mechanism at the ERCF end that would then have another Y configuration that worked like how train tracks can be switched between a left or right split - It would switch the path from the hotend-bound tube, to an open-air cutter that drops a length off in a tray, ensuring a clean tip. Then it would continue extruding fully, switch the Y back to the normal airtight hotend path, and a rubber membrane could help keep outside air from flowing into the spool upline (for avoiding humidity).

Finally, for purging I am wondering if a brute force mechanism would work for blasting the heatbreak & nozzle clean, avoiding the extra purging required when a new filament is pushing out old filament left inside - For example, if the tube leading into the extruder had a mechanism where it could reliably be pulled out / moved out of the way, so a high-temp tube could do basically the same thing you do when you clean a clogged nozzle with a blow torch (appropriate for the temp tolerance of the nozzle/heatbreak of course), and maybe if that isnt enough, a combination of water mist & high pressure heated air, then cleared of water with a final air-blast.

Heck, you could it move the nozzle over a purge station holding both the entire blast-out and even a blast-in system to do a fast brute force push/pull/incineration process, and the station would then reseat the filament tube into the extruder and there'd be no extra weight on the tool head.

[RomRider]

It looks like running 2 filament tubes to a Y splice at the hotend would significantly increase switching speed. You could have the next filament queued up near the hotend, and only have to wait for the last filament to retract past the Y splice before extruding the next one in.

How would that work as there is only 1 cart and the cart can't move while a filament goes through?