4 Likes
Do we know if the eco purge setting was done? So seems only max 4 seconds by making input sorter but also the 4->1 was not installed and most will⊠Good to have a topic open on thisâŠ
Itâs a shame Bambu is so obviously chained their business to the ams and ams2
Both of those devices do wonderful things for single nozzle. Just doesnât perform as a premium product when wedged into multi nozzle/tool. Physical reality of that filament strategy is just slow. It is fantastic margin for their store though!
Imagine if the ams was able to feed each vortek with its own ptfe path. Users without vortek could still just slap a 4in1 hub somewhere. Hmm maybe thatâs just bondtech with extra stepsâŠ
3 Likes
Looking at the video why doesnât it start feeding the new filament at 14seconds after the old gone ? it can push it nearly all the way to the head already no? seems odd or am i missing something? If i see it correctly there is about 5 seconds it could gain there âŠ
3 Likes
I am still new here but maybe we should try to keep this thead about how to make the swap time shorter not if indx will be better at some point in the future enough other places already discussing that.
2 Likes
I wonder if its just this guys table or if this thing is tuned so hard to max speed for socialMediaBucks© that the violent shaking is expected.
That video gives me very spooky maintenance vibes. I dont like my machines shaking themselves to death. I know, input shaping, quality is fine, yadda yadda. IMO That thing should have another 50 lbs on its frame to arrest that kind of motion normals
Not sure, but it is sexy. Pictures and video dont do it justice. Somehow feels more premium than the h2d. Especially the toolhead and outer shell. Like the H2D is now a p1s series lol. The H2S is now the p1p. The swapping mechanism is much nicer than video shows too.
2 Likes
The shaking is part of its design like it or not its been proven stable for the H2 so far no reason to expect anything different on the H2C.
3 Likes
I think there is about 7 seconds to be gained in the purge and less pause moments esp. donât get why the feeding doesnât start once the retract is done seems to me it can be pushed much closer to the head already. Guess we will find out over time. I also prefer the outside color of the h2c over the h2d and them head now has a led that moves from left to right no?
The time saving mode will probably knock a few more seconds off, with a warning about possible lower quality. It still needs to purge until the cut line enters the melt zone. otherwise, it cant retract. Most times, that wont matter. Only if youre doing something small or something with alot of retractions right away.
I plan on using the longer purge and just being happy theres drastically less waste.
yeah 3 seconds he claims in his first review. but its unclear if in this case it was on or off. Also his other tests are a little iffy like on one model the U1/XL had no tower thats matters a lot (we have seen some no tower h2c prints already) the dino seems the most real compare in a few color changes per layer not all and then the times are much closer. I feel it will take a while and probable shared models to bring times down. I do hope under the pressure of âits slow compared to x,y,zâ bambu does respond a little and gives some hints how they see this evolve. I agree we should compare what we have gained in speed and less waste compared to h2s and h2d first. Still comparing is fun and trying to decode what is happening too.
Ams 2 has a faster and stronger motor as i recall. Iâm sure ams 3 will be slightly faster when it comes out
Both AMS 2 and AMS HT will occasionally yeet the filament into low Earth orbit if it gets even slightly tangled, so yes â theyâre absolutely stronger than the original units. 
But about the swap times in that video: is that remotely realistic AMS usage? If you want true numbers, test under real conditions, not âlaboratory speedrun modeâ.
To me it looks like some people bought the H2C just to chase speed records â exactly what I mentioned earlier. A lot of users will grab it simply because itâs the newest, shiniest toy, not because they actually need it or will ever recoup the higher cost.
I didnât buy it and I never will, for several reasons. One is that it brings zero real benefit to my work compared to my current machines. Another is how Bambu Lab has started treating their customers lately.
So sure, go full speed demon if you want â but at least do it in realistic scenarios. Otherwise we can âproveâ that any machine is amazing⊠as long as we pretend real-world use doesnât exist. 
3 Likes
Yeah I found that weird too. It should push the filament to buffer or toolhead
1 Like
I think heâs already printing in purge saving mode. The tower is pretty small.
Hopefully they take a break on new releases and optimize the last 4. They should be able to get the time down more. Especially if they had a smart hub with runout sensors
3 Likes
You are asking too much from people who want to defend the existence of H2C
(sarcasm of course)
The number of times AMS 2 yeeted out the filament full force made be giggle a bit. Still canât figure out the logic behind it, but occasionally it goes forward and backwards full thrust.
Right now, there is no sensor anywhere in the machine that can tell where the filament is between the AMS and the toolhead. Given that, it has no idea if you 1m or 2m between your AMS and printer. So when a retract happens the AMS doesnât stop rewinding until it senses the end of the filament. Same thing when loading, the AMS pushes until it hits the sensor in the toolhead.
What is needed is a distribution box for AMSs that can load the next filament up to the distribution box, while the previous color is going. When the swap happens, your filament is that much closer, mounting the box on the back of the printer, similar to the X1c hub. As soon as the filament is cut, retract as a process running in parallel w/the nozzle switch. By the time the nozzle parks, the previous filament is fully retracted, and the next filament is only a few seconds away from the distribution box to the nozzle.
The slicer then would have to figure out what colors to put in what AMS so that you can accomplish this; the limitation being here that there is only one exit point from any AMS, if, somehow, you could change the switching in the AMS to allow two exit points at once. Each connection could go to the d-box, essentially having the next color on standby from ANY ams, not just a different one.
Was in bed thinking how to solve this, this is what i came up withâŠ
1 Like
That kind of âownershipâ is fine too, but it is a bit hilarious to watch things like this:
âIf you use 1 cm of PTFE it loads SUPER FAST!â
Sure, great â in that one carefully staged lab scenario that never happens in real life.
Can we actually use that in a real-world setup? Probably not.
Does it matter anywhere outside that super-niche test? Also probably not.
What really amuses me is how people happily chase tweaks that might reduce print quality or solve a problem that didnât even exist to begin with. Pushing machines to their limits is fun, absolutely, but with the H2C it feels like a lot of people bought it purely for the speed â and are now desperately trying to overclock what was already sold as the âoverclockedâ version.
3 Likes