Usefulness of the Filament Track Switch for the H2C/D

Bambu Lab H2 Series
21

Oh absolutely agree. The price is free, and i was watching for when it became available, ordered without hesitation. It is very satisfying when you have it set up. I was making the point that the benefit for H2C machines is less than one might first think, and the reliability is a real downside that may actually outweigh that benefit for many users.

My judgement in that still pending. With redoing all my lines it now seems reliable if a bit more struggling than before. I think the new 4-in-1 devices have more resistance than the old ones, so swapping to old ones is my next fix if I get another loading problem.

But i am still enamored with the device and if anything I wrote to the original post to highlight how it could be even better. If you love the idea of it and tinkering, then get it:)

22

The new ones increased resistance is likely only due to that added rubber wiping pad that you can just remove. Try without it, I removed mine though I am not sure if I even had to.

23

I think itā€™s two things:

  1. the excessive friction causing a lot of problems for many.
  2. The discrepancy between expectation and reality. At least for me, when I first read about the FTS, I was very excited about the flexibility it allows. Step by step, I realised that in many scenarios, it will not help me. So I was quite disappointed by the actual implementation compared to what it could have been.
24

Iā€™ve been using my H2C with the track switch since the beta firmware for it was released. Iā€™m removing it and going back to my old setup (ams 2 pro & 3 ams-HTā€™s).

Some things Iā€™ve noticed:

  • severe resistance in the tube path. Iā€™m sure I can optimize it a tiny bit more, but not much. It works for the most part but the additional print pauses are annoying.
  • Prints with tree supports are failing more often. I have my profiles dialed in pretty good and in the few years Iā€™ve been operating Bambu machines Iā€™ve had maybe one or two print failures due to failed tree supports. And that was before I found out about reduce infill retraction. Itā€™s been at least a solid year and a half since I had a tree support failure and not itā€™s happening more times than not. It just seems like theyā€™re not printing properly which may be due to the next bullet point.
  • I had to recalibrate my filament flow rates. Most of them went up. Some of them significantly. I can only assume this is due to the extra resistance in the feed path. I know this is due to having the track switch because I can use the same filament in my other h2c w/o the track switch and it will over extrude until I put the flow rates back to where I had them before installing the switch on the first machine.

Iā€™ll be removing the track switch and going back to my old setup until something changes for the better. If it never does, oh well. The machine still works amazing without it.

25

Hmmm. Iā€™m not there yet, I want to make it work. If the feed is so bad that it affects extrusion, yeah thatā€™d be a huge problem. Itā€™s not supposed to happen - thatā€™s what the buffers are handling, I thought.

I wonder if the FTS has an impact on calibration. The machine does a bunch of automatic stuff, and if you do manual stuff itā€™s tied to nozzles. I just wonder out loud if calibration (auto or manual) is working as well with the FTS. I donā€™t have a good method of verifying this, but itā€™s something I think about.

One theory Iā€™m exploring a bit now is that the 4-in-1 (new model) is really draggy, in fact it seems to contribute more than the FTS, and Iā€™m pretty sure the wiper is not very meaningful. I will try the old one and some other models next. I think it also helps to put draggy mergers closer to the ams rather than close to the FTS. This is due to how friction works when pushing filament in a tube - resistance further away takes more push to overcome than resistance nearby.

Curious to hear any learnings or thoughts.

26
  1. I was referring to people argumenting against its core functionality before even buying it. Like the OP. But yes, fair point.
  2. I think this is where I donā€™t get what those expectations were, your argument earlier up is seriously that manually reassigning paths and trying to keep the slicer in sync is more convenient, which is never going to be the case in reality. But maybe you mean that you prefer yourself to keep the current filament paths in your head is easier, then all good if it works for you, it would confuse the ā– ā– ā– ā–  out of me when using the slicer though as I would never know what color would be where.

But whatever, best aaccessory I have ever bought myself, though I do see that many have problems with resistance, which is a bummer if you have it happen to you.

27

I wish I had read this thread before ordering my H2C FTS - the extra resistance is a concern - so I donā€™t think I will bother installing it for now.

My setup is all 4 way AMS - 1 on left and 3 on right - mostly for 10+ colour prints.

Was hoping the FTS would help with times during prints where the main ā€˜leftā€™ colour changes.

I guess spool splitting and copies of colours and manual selection during different heights of the print will continue to be the most efficient approach.

I agree too with the point about being able to prioritise print speed would be my preference too as the recommended placements sometimes suggest quite a lot longer print times for tiny savings. I now almost always select convenience mode.

1 Like
28

When I saw the FTS at first, I thought that it doesnā€™t matter at all anymore how you connect the AMSs, the switch takes care for you. Of course it does not work like that. But I didnā€™t expect the limitations I describe below.

I have an X2D with 2x AMS2 + 1x AMS HT. One AMS holds all PLA, the other holds PETG and the HT is my hot spare for engineering materials, extra colors, external spool holder for TPU and simply for drying.

My hope was that for any print with at most 2 filaments (2 colors, 1 color + support) from different AMS, the printer can keep both filaments loaded throughout the print.

When I went through my usage scenarios, I realised that however I group those 3 AMS, there are always some scenarios that donā€™t work out:

  • single filament from any AMS on the main nozzle
  • PETG from AMS-B on main nozzle, PLA as support from AMS-A
  • PLA from AMS-B + extra color from HT on other nozzle.
  • same for PETG
  • TPU from HT on main + PLA as support from AMS-B on aux
  • TPU from HT on main + PETG as rigid from AMS-A
  • PLA multicolor from AMS-A + HT

However I would group my 3 AMS onto the 2 inputs, half of my scenarios above donā€™t work out.
I found that having both AMS2pro on the main nozzle and the HT on the aux nozzle without the FTS actually means the same amount of manual spool swapping, so I have almost no benefit and potentially have to deal with the extra friction.
And on top of that when you have installed the FTS, you canā€™t use an external spool holder for loading TPU directly into the nozzle. So printing TPU with the FTS would be really cumbersome.

What is really missing for me in the FTS is an input for each AMS that can be assigned individually to both tracks. And then I found that exactly that exists in the model I linked above, although manual switching only. Still this is very fast. You just flip a switch, go into AMS configuration and dragā€™nā€™drop the flipped AMS to the other nozzle. Takes maybe 10 seconds total.
This allows me optimal configurations for all the scenarios above, I donā€™t have the extra drag, no cascade of 4in1-adapters dangling at the back of the printer and 50ā‚¬ saved.

Iā€™m sure with an H2C and more AMSs, the situation changes and the switch gains usefulness, but even then, a switch with one input per AMS would be much better and what many hoped for like me, when they first found out about the FTS.
So once the idea is in your mind, every lesser solution is a disappointment, even if it means an improvement over the situation without FTS as you pointed out.
And the FTS adds quite some complexity to deciding your setup.

One thing BBL could improve by software to achieve the same goal:
Allow to connect multiple FTS, using 4in1 adapters after the switches to join the outputs for each nozzle.
That way you could connect 8 AMS to 4 FTS and assigning each AMS individually to a nozzle.

And the golden solution would be an AMS that already has a switch integrated, so that each roll can individually go to 2 outputs, one for each nozzle. So all you need is to join all the outputs from your AMS into the two nozzles by 4in1s without any extra devices. This would make it completely irrelevant how you distribute filament across your AMSs.

1 Like
29

I do have a general insight to offer on how to avoid excessive resistance. I had printed one of the FTS holders that also holds 2x4-in-1 mergers right by the FTS. I did some basic testing and confirmed that this is actually the worst way to place your mergers.

You can replicate this test: Take a length of PTFE tube, say 1 meter, and connect a 4-in-1 either at the end of the tube, or at the beginning. If you try pushing a filament by hand through this assembly you will notice that there is vastly more resistance if the merger is at the end, even though youā€™re passing through the same components. The mechanical explanation for this is that having this extra resistance at the end increases the ā€œpressureā€ for the entire length of tube before that merger, which results in a bowing/buckling filament rubbing harder on the PTFE walls.

Having applied this learning and moved all my mergers closer to the AMS unit, my feeding is much smoother. I use the new high-drag ones with the wiper installed.

To confirm and build an intuition for how this works, I built a quick simulator: https://ai.studio/apps/9fe3d35a-6114-4878-97e9-44852f32acfc?fullscreenApplet=true

The numbers used in it are not the real numbers, but not crazy either. Play with them and see just how sensitive the overall system can be to relatively small changes, since it highly non-linear. It does explain well why soft TPU canā€™t be feed through the tubes.

I think I am now realizing that Bambuā€™s limit of 4+8 AMS units is probably primarily about limiting the cases where people run into these limits. With more units, they would be further away with more mergers, and even more casual users would run afoul of the limits.

30

Yeah this is very accurate. Itā€™s an upgrade but a disappointment since it could be more. I hadnā€™t thought of the software-only fix of just having more FTS units.. genius!

Agreed on the golden solution. And while theyā€™re at it, to amortize that hardware cost a bit more, they should make a larger 8 spool unit or somethingā€¦ and make it stackable :enraged_face:

31

It baffles me that well written posts with actual formatting are now regarded as less good than sloppy walls of text.

1 Like
32

Although I have the FTS, Iā€™m now tempted to just connect my 4 AMS Pro 2 units to the right for the Vortek system to use, and my single AMS HT to the left, for support materialā€¦..

33

People gravitate towards something like that, or keeping some common base material on the left. Or engineering materials. I keep my ABS-GF, black/white PLA, and support material on the left.

34

Yeahā€¦ this was my first contact with that attitude. This forum is great with a lot of good technical analysis and advice. Letā€™s keep it structured:)

35

I disagree on the ā€œmulti-color printingā€ and "resistance pathā€ issue.

Let the Bambu Labs slicer help with filament changes, and youā€™ll find in multi color prints you are actually seeing far less filament cutting and changing where not needed more often. A bit filament saver and time saver.

Also, the switch has the ability to push filament, giving the path more power to keep filament flowing with fewer problems. I saw a dramatic effect on filament feed problems after adding the filament switch to my X2D setup and 2 AMS 2 Pros setup.

36

Are you talking about the feed assist or the track switch because the switch is purely passive, thatā€™s just a ā– ā– ā– -in-2 adapter, it has no way to push anything and with the friction it has more ability to pull than to push :wink:

37

I think what we are missing is actually the ability to lock the track switch to one output. This way, we can then either load the left nozzle via external spool, or even better allow loading from AMS with a 4 to 1 after the track switch.

38

Honestly i have the same feeling and same logic

Since i put the FTS with my H2C i have many trouble and many major problems of extruder or filaments go to the wrong side,

I didnā€™t test it yet on my X2D but for the H2C Iā€™m going to remove it

For the AMS 3 yes especially for H2C user, weneed something like, we need AMS with 4 output or 2 output and then next to the machine filament selector, something to gain time, tension on filament and of course in the life of the ams, as actually each color changes AMS xork hardly

1 Like
39

I just purchased the FTS with a 2nd AMS 2 Pro for My H2C and I LOVE it. I only have 2 AMS2 Pro, not HT, but not only does it let me direct any filament to any nozzle, but what people are leaving out is the massive time saver of not having to load support filament to the left nozzle. Yes, I could have have each AMS dedicated to one nozzle, but the absolute convenience of just loading spools and letting the FTS do the work is game changing for me.

Right now Iā€™m printing some ASA parts and put HIPS in the 2nd AMS for support. Thatā€™s it. super easy.

But, tomorrow Iā€™m starting a 7 color print on the H2C and the FTS is a huge time saver.

2 Likes
40

I must be missing something, because the FTS is actually a net positive for me. I used to get up constantly to move rolls between AMS units, but now I donā€™t have to. Itā€™s possible thereā€™s some nonā€‘obvious time or efficiency loss, or that my expectations are lower. I do occasionally see the resistance warnings, but they donā€™t seem to interrupt my prints.