Warning to anyone using a mixing hot end
-
I've just done an incredibly stupid thing (although it seemed perfectly sane and logical at the time).
I was playing around trying to tune firmware retraction with my latest 6 in 1 out hot end. (For those who are not familiar with mixing hot ends, it is necessary to retract all filaments concurrently, not just the one(s) that are in use at the time, otherwise filament is drawn from the unused inputs, rather than the nozzle. This is how firmware retraction works - it retracts all filaments associated with a tool regardless of the mixing ratio being used).
So, as my two tower test part progressed, I noticed that I seemed to be getting a small blob after the non-print move (at the start of the next tower) but without any stringing as the print head moved away from the previously printed tower. I know that this hot end is prone to oozing due to the high volume of molten plastic. So I had the "bright idea" of using asymmetric retraction. That is to say, retract 3mm but un-retract (say) 2.5 mm. My thinking being that if there was no stringing as the print head moved away from the part, but there was small blob when the print head reached the adjacent part, this could be due to oozing during that non-print move, so using a lower un-retract amount might fix it.
This didn't seem to work. In fact things started getting worse and I started to get stringing as well as blobs. I increased the amount of retraction and tried numerous other settings but always less un-retract than retract. As it was getting late in the day, I decided to sleep on it.
What I failed to realise at the time was that each of the other 5 filaments which were not in use for this particular test, were being retracted by (say) 3mm and un-retracted by only 2.5 mm but there was no forward movement of those filaments between retraction cycles. So effectively, each retract/un-retract cycle was pulling the filament out by 0.5mm. After numerous such cycles, the filament had become completely unloaded from the hot end (while hot filament has been drawn up through the heat breaks where it has cooled and solidified).
The moral of the story is do not use asymmetric retraction/un-retraction with a mixing hot end! It's also a good idea to use clear Bowden tubes as I do - that way one can easily see the position of the filament inside the tube making such faults easier to spot.
Now to remove, strip and clean the hot end. It looks like 'er in doors will get her oven cleaned again (a pyro cycle is the best way to get solidified filament out of stainless steel heat breaks)
................ -
@deckingman said in Warning to anyone using a mixing hot end:
After numerous such cycles, the filament had become completely unloaded from the hot end
Theoretically you could have individual retraction management for each channel. E.g. once it was moved 0.5mm a few times, not retracting it anymore.
I am not sure how much control do the slicer and RRF give you but can you do it yourself with a gcode post processing script?
-
@zapta said in Warning to anyone using a mixing hot end:
Theoretically you could have individual retraction management for each channel. E.g. once it was moved 0.5mm a few times, not retracting it anymore.
That won't work with a mixing hot end. ALL filaments MUST be retracted whenever retraction is required, throughout the entire print. And if there is an air gap anywhere in the filament path, then retractig the filament will have no effect.
-
@deckingman I guess what might work is to scale the "extra length on restart" value with the mix ratio.
Duet WiFi hardware designer and firmware engineer
Please do not ask me for Duet support via PM or email, use the forum
http://www.escher3d.com, https://miscsolutions.wordpress.com -
@dc42 said in Warning to anyone using a mixing hot end:
@deckingman I guess what might work is to scale the "extra length on restart" value with the mix ratio.
Possibly. But I don't think it's worth your time and effort. Although, if extra length on restart was proportional with the mixing ratio, then it might prevent someone in the future from suffering the same problems if they make the same stupid mistake that I did. But with mixing ratios of single digit percentages of the whole, then the extra length on restart would be single digit percentages of those single digit percentages which end up as extremely small amounts. So then we get into sub micro-step territory and the need to carry over those fractions of micro steps to prevent a cumulative error effect. I'm sure you have more important things to do than mess around with this.
I'm planning a bit of a re-design, which if it works, will reduce the oozing issue and (hopefully) reduce the overall amount of retraction needed. As ever, I favour the hardware fix
-
@deckingman said in Warning to anyone using a mixing hot end:
ALL filaments MUST be retracted
Do you retract them by current mixing ratio or equal lengths?
-
@zapta said in Warning to anyone using a mixing hot end:
@deckingman said in Warning to anyone using a mixing hot end:
ALL filaments MUST be retracted
Do you retract them by current mixing ratio or equal lengths?
Equal lengths. It has to be that way because, as in the case of that test print, the mixing ratio was 100% of one filament and zero for the other 5. So retracting by mixing ratio would have meant that only a single filament would have been retracted, which would simply pull filament from the other 5 inputs, rather than the nozzle.
