I have 3 AMS units and 2 printers. I just finished an A/B switch that allows me to simply twist a knob to switch an AMS between the two printers. When I flip the switch, the target printer automatically detects the newly attached AMS, and scans its spools.
I’ve run a couple of prints with and so far they have all run flawlessly.
If anyone is interested in making one, let me know.
Steve
How do you change the PTFE?
I’m assuming the knob in the middle toggles the single PTFE input to either output. Pretty ingenious design. Just needs a custom PCB to also simplify the switch wiring.
Ahhh I can see that now you’ve mentioned it, the PTFE blended in with the background 
Fair enough, interesting design.
HECK YES! Do tell, Please!
Yes, the knob switches the PTFE tube from the shared AMS to printer A or B. It also controls 12 SPDT micro-switches to switch the cable from the AMS to the two printers. I simply switch 6 pins from the AMS to the respective 6 pins of the printer cables. It circuit is trivial, but there are a lot of pins to switch.
I bought a 12PDT switch but it needed way too much torque to turn it, and there was no easy way to control the PTFE switch with it. So I used 12 micro-switches instead. But a PCB is the way to go, wiring it was a pain.
I also took grinder to an AMS power board to isolate the two 6-pin connectors.
It took several tries to get a low drag path for the filament switch, but it is working smoothly now. If I make two more I will be able you switch any AMS to either printer. The next step is to automate it so I can control it from my desk.
The one thing I’d be concerned about is the AMS isn’t a hot swap plug, switching while powered could cause board issues.
Can I assume there is a small gap in time, either physically due to the mechanism or electronically due to software to remove the potential for both being connected simultaneously?
This could cause a short or electronics issue in one or more connected devices.
My less developed attempt at same problem. I used 2 multi pole switches in parallel (I initially thought only 4 poles needed to be switched). I used Molex 3.0mm ready wired cables, but like the idea of using a power board.
Ultimately I would like to move the switch with a servo.
If you look closely at the 6-pin AMS cable, the ground pin extends slightly more than the rest of the pins, so ground is established before power when plugging in the cable. I need to adjust the cams on my knob to emulate the same effect when twisting the knob. e.g., engage the ground signal before the rest of the signals when switching between the AMS units. And reverse the process when switching the other direction. But I have switched it without this feature several dozen times and I haven’t seen any issues. So far…
Hi Ukdavewood, Here is the schematic that I used. However I didn’t add the termination resistors. I figured that if I signal integrity issues I could add them later. But so far I haven’t seen any. If they aren’t needed, then it could be reduced to a 6P2T switch.
I guess a custom PCB as mentioned earlier above would be the best solution ultimately, with mounting for the 3 connectors and appropriate tracks to form one half of the 6 pole switch. My biggest concern is the reliability/smoothness of the ptfe filament part of this type of switch.
Found a post with a bit more detail of my attempt if it is of any interest. I didn’t progress it any where near as far as @mcmaven
That’s exactly what I want too. And it would be really nice to support printers other than Bambu, but later…
I posted a replacement spool tray for the AMS that supports all the features of the Hydra, but where the dividers can be moved so a spool can span 2 slots and support spools up to 97mm wide.
I’ve also started working with a friend to use a single slot, 124 mm wide variant (for 2KG spools) of my current tray design, with an AMS Feeder Unit that can load or unload an External Spool via WiFi. Sort of a ‘poor man’s’ 5th spool. I will be able to load/unload it from my desk, but it won’t be able to be used for multi-color prints.
The plan is to also make a dry box enclosure for the external spool. Designing the side tray to fit all the AMS hardware was the hard part, but designing a larger enclosure that would support a single wide spool should be easy, just a LOT of printing.
Would love to see this evolve into a kit with a PCB! This may sound dumb, but with 4 AMS, this is what has prevented me from buying a second printer. If I could switch the AMS between printers, I’d grab another X1C or maybe a P1S.
Hi DruiD,
I need to put together some documentation for it, and I will post it to my MakerWorld page. It requires a small ball grinder (to cut traces and the ground plane on the AMS Power Board), a soldering iron, and a small spring and 6mm ball-bearing to act as a detent for the knob. I used my spring winder to make the spring.
Hi d_p,
I am currently using 12 micro-switches to act as a 12PDT switch. The micro-switches are arranged in groups of 3 around the knob. The weird angles of all the switches do not lend themselves to PCB, but I guess one could use a flex circuit. The problem is that I’ve never designed a flex circuit… I did print a special ‘jig’ version of the base with no sides that holds the micro-switches, AMS Power Board PCB and 6-wire AMS cable in their relative positions so they can be easily soldered. Once I have all the wires neatly cut to length and soldered, I then move the whole wiring harness over to the real base.
My long term goal is to control the switch wirelessly. This would mean adding a servo to rotate the knob between the A and B outputs, and some relays to switch the signals. Relays are much more PCB friendly. One thing that I really like about the Bambu is the fact that I only have to visit the printer once per job, i.e., to pick the parts off the build plate. If I can control my AMS switch from my desk I will be able to preserve this usage model.
A friend of mine and I are looking at how to make this into a kit, along with some other devices for filament management. Keep in touch.
I feel like there must be some clever way to use tiny magnets for the switching. You want a double-pole contact which will reliably engage either pole, and doesn’t take too much force, since there’ll need to be six of them …
Just haven’t thought of it yet.
Maybe reed switches? Do they make double throw versions?
I initially held out on buying an X1 Carbon as I had a deposit down on an XL, however, all the delays pissed me off so much that I redeemed my deposit but held out on buying the X1C and AMS as I thought it might be possible that Bambu would release a tool changer printer to compete with the XL, this never happened and so I bought the X1C combo.
I still think Bambu may release a tool changer and so I am not happy to buy a second AMS, even with calibration for wastage it still poops far more filament than I am happy with.
Nevertheless, this A/B switch is a very good thing, at first I wondered why Bambu Lab had not released one, and I came to the conclusion that it would reduce AMS sales.
Until I know what will succeed the X1C I am not thinking of buying a second.
