H2C FTS works fine on each input separately, but fails when using both

[user_1411648265]

I have an H2C with 3 AMS 2 Pro units, 2 AMS HT units, two 4-in-1 adapters, and an FTS

Multicolor prints are unreliable. Eventually, I get “AMS feed resistance is too high,” clicking from an AMS, failed loading or retraction, or filament that reaches the toolhead but will not load into the extruder

Sometimes Resume works, but another failure usually appears later, so I have to constantly babysit the printer

Both FTS inputs work fine when tested separately

The same test print completed through each input, and two AMS units also worked together through one 4-in-1 and one FTS input

The failures return when the same print alternates between both FTS inputs

I have tried different AMS units, slots and filaments, short PTFE paths, removing the 4-in-1 cleaning pad, loosening the FTS housing screws, the latest Studio beta, rebuilding my older FTS, and the AMS hub spring fix

This is already my second FTS, and both had the same issue

The FTS firmware is 00.00.16.62

Has anyone fixed this exact problem? I am considering buying a third FTS, but I am worried it will behave exactly the same

[CarbonForge]

You probably will need to show us your setup.

Have you printed and installed the PTFE guides that go on the back of your AMS units?

[just4memike]

There are many possibilities of why you are having the issues you are, but the simple fact is… they didn’t do a good job on this accessory. I don’t think you can get a very reliable setup with this accessory. I find CF/GF infused filaments are the worse followed by wood infused filaments. So if you can, stay away from those.

But the single take away is… its not a good accessory. Its super useful, and well intentioned, but not very functional. IMO, they should go back to the drawing board and stop selling it for the H series at the minimum.

[user_1411648265]

@CarbonForge

ou probably will need to show us your setup.

Have you printed and installed the PTFE guides that go on the back of your AMS units?

Yes, I have already installed PTFE guides behind the AMS units. I used the adjustable guides on the tubes, and on the lower AMS I also installed Bambu Lab’s original PTFE guide. Neither made any difference.

The important detail is that the system works when I use only one FTS input. I can even use multiple AMS units through the same splitter and the same FTS input without errors.

The feed-resistance warnings return when AMS units connected to both FTS inputs are used in the same print. Because the exact same AMS units and PTFE paths work when only one FTS input is active, this does not appear to be caused simply by the PTFE tube routing or bends behind the AMS units.

[CarbonForge]

Apologies, I missed that detail.

What order are the cables connected in? What’s first, what’s last, etc?

[user_1411648265]

On one side, two AMS 2 Pro units and one AMS HT are connected to a 4-in-1 adapter, which connects to one FTS input.

On the other side, one AMS 2 Pro and one AMS HT are connected to another 4-in-1 adapter, which connects to the second FTS input.

Both sides are correctly assigned as left and right in the printer.

I’m also thinking about adding a fourth AMS 2 Pro soon.

[CarbonForge]

Cables, not PTFE.

[user_1411648265]

Do you mean the 6-pin bus cables between the printer, FTS and AMS units?

The 6-pin cable connections are:

• The FTS is connected directly to the H2C.

• The other 6-pin port on the FTS is connected to the upper-right AMS 2 Pro.

• The upper-right AMS 2 Pro is connected to the upper-left AMS 2 Pro.

• The upper-left AMS 2 Pro is connected to the left AMS HT.

• The left AMS HT is connected to the right AMS HT.

• The right AMS HT is connected to the lower AMS 2 Pro.

• The lower AMS 2 Pro is the last unit in the chain, so it has only one 6-pin cable connected.
  Do you see a problem with this order?

All AMS units remain detected and work normally when I use only one FTS input. The resistance warnings mainly appear when AMS units connected through both FTS inputs are used in the same print.

[CarbonForge]

For curiosity, I would try a “stripped down” test - just the FTS and one AMS HT directly into each input.

Just to see what happens.

[JohnnyKn]

The accessory works as intended From the Bambu Wiki

The AMS has specific requirements for the physical properties of filaments. The following three types of filaments are not recommended:

• Too soft (e.g., TPU 95A/85A/83A/80A, damp PVA or BVOH): These cannot pass smoothly through the AMS / AMS lite channel, and can easily cause loading/unloading failures, clogging, and other issues.

• Too brittle (e.g., certain brands of PET-CF, PLA Wood, PA-CF/GF, PAHT-CF/GF): These can easily break inside the AMS / AMS lite channel.

• Too hard and rough (e.g., certain brands of PET-CF, PA-CF/GF, PAHT-CF/GF, PLA-CF/GF): These can easily wear down the internal feed channel of the AMS / AMS lite.

In addition, the AMS has an extra transmission gear that works together with the extruder gear to provide stronger feeding force. This helps overcome greater feeding resistance, so the AMS can handle some softer or rougher filaments.

These types of filaments were not intended for use in the AMS and hence the FTS. I have had very little problems with my FTS and long as I follow the specific requirements clearly stated by Bambu

[just4memike]

Its not a question of it not working as they “expect”, it simply doesn’t work well. That’s it… errors and all, it just wasn’t done well.

If you like it, great. I’m not going to ask you to not like it. Personally, I’ve been using it since its release and have taken off multiple times due to the errors and drag issues. I’ll tell others to be careful if they install it. It will create a significant drag on the filament path and force the extruders and AMS motors to prematurely wear. Maybe the premature wear will be insignificant… maybe not. Can’t say, but I will say it definitely isn’t ideal for the system.

So to make a long story short, to anyone putting it on, just be aware that you are likely shortening the life of some of the components. How much… I don’t know. Could be a lot, could be very little. Just know this product is not good for the system.

[CarbonForge]

@just4memike

I don’t think you can get a very reliable setup with this accessory. I find CF/GF infused filaments are the worse followed by wood infused filaments.

You would be wrong. I’m feeding AMS units into both sides of my FTS and once did an all-nylon print with five different nylon filaments, all of which were either glass or carbon filled. No issues.

[user_1411648265]

Thanks — I tried the stripped-down test and then added components back step by step.

First, I connected one AMS HT directly to each FTS input, with no 4-in-1 adapters and no other AMS units connected.
I ran color-change torture test , and it worked perfectly: no clicking, resistance warnings or loading problems.

I then tested one AMS HT and one AMS 2 Pro directly into the two FTS inputs, and that also worked. After that, I connected the same two units through the two 4-in-1 adapters, and it still worked normally.

The problem returned as soon as I used three active filaments. It happened first with three different AMS units, and then again with only two AMS units — one filament from the AMS HT and two slots from the same AMS 2 Pro. In both cases, the resistance warnings appeared at around layer 3 and were reported from different sources, including both the AMS HT and the AMS 2 Pro.

So far, two active filaments seem reliable, even through both FTS inputs and the adapters. With three active filaments, the warnings return almost immediately. In normal prints I have also experienced clicking and actual loading failures, not only harmless warnings.

[CarbonForge]

Very interesting results.

I suspect two active filaments results in the filaments being loaded only at the beginning of the print, and for whatever reason (either just luck or perhaps different system behavior) it loads successfully, but for >2 filaments, the need to repeatedly move filaments through the FTS is increasing the odds of a single “bad” loading event until the force limit is tripped? Or is the threshold different after the initial loading? Or something else is being “sensitized” after a few load/unloads?

So your activity would appear to be covered and you have a good chance of being in violation, too.

Here’s another test if you’re game:

Start a 3-filament print, and when the error comes up leave the print paused, disconnect the offending PTFE at the AMS, and then try to push a loose piece of the same filament through the disconnected PTFE all the way to the extruder, and see what you feel along the way, and particularly when passing through the FTS.

[just4memike]

Able to work, doesn’t make it reliable. So, I think you’re wrong, but we will just have to agree to disagree.

The FTS is just too draggy. Using one or two track paths… it didn’t make ANY difference to me. Personally, I’m not judging it based on the errors (even though there are many), I’m judging it based on the resistance measured. The resistance is about 3x what it is without the FTS. Just free the lines an pull a piece of filament through, the difference is significant.

[user_1411648265]

@CarbonForge

Very interesting results.

I suspect two active filaments results in the filaments being loaded only at the beginning of the print, and for whatever reason (either just luck or perhaps different system behavior) it loads successfully, but for >2 filaments, the need to repeatedly move filaments through the FTS is increasing the odds of a single “bad” loading event until the force limit is tripped? Or is the threshold different after the initial loading? Or something else is being “sensitized” after a few load/unloads?

So your activity would appear to be covered and you have a good chance of being in violation, too.

Here’s another test if you’re game:

Start a 3-filament print, and when the error comes up leave the print paused, disconnect the offending PTFE at the AMS, and then try to push a loose piece of the same filament through the disconnected PTFE all the way to the extruder, and see what you feel along the way, and particularly when passing through the FTS.

Just to clarify, in my two-filament tests the filaments were not loaded only once at the beginning. I assigned both filaments to the same nozzle, so the test performed about 98 repeated unload/load cycles between the two FTS inputs, and it remained completely smooth.

With three active filaments, the warnings appeared around layer 3, both when using three different AMS units and when using only two AMS units, with two slots from the same AMS 2 Pro.

I understand the purpose of the manual resistance test, and it might help identify where the drag is located. However, the practical failure is already clear: in normal prints I have had repeated clicking and spool stalls inside the AMS, as well as filaments that fail to reach or load into the toolhead.

I have also already noticed substantial resistance when manually feeding filament through both FTS paths. Additionally, during a paused loading error there may already be filament partially occupying the path, so pushing another loose piece through may not provide a very meaningful comparison.

@just4memike

Able to work, doesn’t make it reliable. So, I think you’re wrong, but we will just have to agree to disagree.

The FTS is just too draggy. Using one or two track paths… it didn’t make ANY difference to me. Personally, I’m not judging it based on the errors (even though there are many), I’m judging it based on the resistance measured. The resistance is about 3x what it is without the FTS. Just free the lines an pull a piece of filament through, the difference is significant.

That matches what I am starting to suspect. The system can work with two active filaments, but that does not mean the resistance is normal or that the setup is reliable.

With three active filaments, the warnings return almost immediately, and in normal prints I have also experienced clicking, spool stalls and actual loading failures where the filament could not reach or enter the toolhead.

It may simply be that the FTS adds enough drag to leave the system very close to the force limit, and more complex filament-change sequences are enough to push it over that limit.

[CarbonForge]

I’m really baffled then. Why would three filaments being unloaded and loaded fail almost immediately, but two can achieve 100% success over ~98 changes??

What is actually different between these two scenarios?

I re-read the Wiki for the install, I have a few questions about your install…

1. For the tubes between the outlet of the FTS to the printer, did you use the supplied 400mm PTFE tubes, or did you use different ones/change the length?

2. On the inlet side, did you use the supplied 700mm tubes, or did you use different ones/lengths? The Wiki is a bit conflicting, one place says “<700mm”, but another spot has an interesting comment:

   

3. Are you using the new style PTFE 4 into 1 adapters or the old ones?

4. It sounds like you’re using a mix of AMS 2 units and HTs, it sounds from the wiki like the HT is more sensitive than the AMS 2 - for your 3 filament test what happens if you put all three in your AMS 2?

5. I wonder if this is a wrong mapping issue? Have you tried resetting the mapping and putting only one filament in each AMS initially so it can re-learn the connections to each nozzle?

[CarbonForge]

@just4memike

Able to work, doesn’t make it reliable. So, I think you’re wrong, but we will just have to agree to disagree.

I may not have been clear if that’s your impression.

My bar for acceptable performance is “The filament load process has to be as reliable with the FTS as the filament load process is without it.”

I was one of the first people to use the FTS on my H2. I ordered it the day it was announced, and sat on it until the beta firmware that enabled it for the H2 was announced, and it’s always worked for me. The first production run of the hardware, the first beta firmware, and no issues since.

That has been my experience. I don’t know why some people struggle with it, and I’m really trying to develop a deeper understanding of it’s operation, that’s why I’m trying to help OP.