Hi anyone have tested something with prime towers in H2D for try to waste less filament in tower?
For example now i start to print one model of 139gr with 2 colors in h2d and the prime tower waste 63gr for me the prime tower is only for avoid the marks in model when change to another nozzle (with P1S printer i never use prime tower normally).
Reminder that prime is a thing that tries to condition the molten filament inside the nozzle tip, it isn’t about measuring waste. If you adjust it too heavily you might get bad priming, which can result in worse quality in lots of small ways (bad initial seam, ooze, etc)
BL has amazing article about prime tower and how to optimize it.
https://wiki.bambulab.com/en/software/bambu-studio/parameter/prime-tower
In short yes, you can reduce the volume of flushes to a bare minimum, but keep in mind that materials with bad adhesion to each other (i.e PLA and PETG) are better to flush properly in their own block and potentially keep rib wall to avoid prime tower getting knocked out and ruining the whole print.
Thanks i go to see this article and experiment a little.
I see, now the prime volume is in the filament settings! for this reason can change nothing 
I go to try some changes in prime volumes and try to reduce a little this waste (i hate waste things).
While experimenting with prime towers, be on the lookout for errors when uploading any models to MakerWorld that use them.
On multicolor models the prime tower grows with every color in the print and can be tripped up when the upload filter thinks it is interfering with the model itself.
In the “mark model photos as real” fire drill I had a bunch of multicolor prints have to get their towers disabled because of the bug in the upload filter. It all depended on how many colors were in the print. Just a couple of colors or lots of room on the build plate - no problem. Lots of colors where the prime tower had to be managed to be able to fit and work - problems.
It’s been a problem for two months. Filed a ticket but no idea if still an issue. I quit posting multicolor prints with prime towers. Pretty frustrating.
You can use the “flush into this object” feature to make a multi color item instead of a useless tower.
Not on a lightbox lens you can’t. 
Actually, I learned with slicing that - with basically identical slicing settings - the material consumption on my single-extruder MK4S is slightly lower than with the H2D which mainly comes down to this option at PrusaSlicer:
I haven’t found it at BambuSlicer yet and the benefit is only when there are very few color changes (otherwise the dual extruder H2D will always win) but with what I do mostly “multi color” prints, support interfaces with different material, so only few changes, this option saves a bit of filament, even more than the dual extruder Bambu.
It does come with a slight theoretical disadvantage which never was an issue for me in reality. The time between priming and continuing to print goes up slightly.
Maybe it’s still worth putting it in as feature request.
Yes thats the problem with that setting. I always place my prime tower as close to the model as possible to improve the look of the seam. With a long pause between tower and model the seam will not be ideal.
Yep. The prime tower bug affects model quality and can make the difference between a good print and a failed print.
What Bambu really needs to do instead of rolling out buggy updates unannounced that breaks things and compromises prints is run the tests against already approved models and see what gets rejected FIRST.
I wouldn’t have deleted all my prime towers (a royal pain) had the process of certifying photos were real that we were commanded to do not included running the checks that rejected my models.
It would be trivial to run the rules on existing models and verify they wouldn’t cause problems before implementing them. And anything that breaks things should never be in an unannounced update. Somebody really needs a sit down to discuss proper procedures when so much is at stake.
That option is also in Orca Slicer - but not currently in BS - I always used it when I used an MMU2 on a Prusa Mk3s and it worked quite well for me.
Downsides though are the time it takes between finishing the priming and raising the bed back down again to carry on printing,
Plus it has also got a lot of the same limitations as “slice by object” in terms of needing plenty of space around the object and prime tower, plus not being able to have too much a difference in heights - otherwise the object will hit the top glass or cable chain.
I think there are only two real solutions to this issue.
Flush into object works fairly well already - but in almost all cases it makes the prints take a fair bit longer - due to the extra material usually needed to complete the objects beyond what is strictly needed for purging.
I think all of this could be done in a Gcode post processor - as a demo - prior to it being added to slicers eventually.
That was my point when I wrote it is a theoretical disadvantage but I never had issues with it in reality.
It simply does work quite well. And when I’m using that together with scarf seems I get flawless surface quality on the MK4S.
There is still a small difference, at the single-extruder MK4S printer it’s both a purge tower and a prime tower where with the dual extruder H2D purging is not necessary (when sticking with two filaments in one print) so it’s only a prime tower. Meaning potentially even more filament can be saved with H2D but the tower would be even lower (more Z travel).
